6.1 (open)

Statistics

Total	579
Fixed	483
Invalid	13
Works for me	6
Duplicate	10
Won't fix	23

Distribution

defect	376 / 408
enhancement	117 / 126
task	38 / 40
project	5 / 5

Summary

#550

fcntl() fails to put a non-blocking socket into blocking mode

#1318

libcpu/shared is not helpful

#1459

Add rtems_shell_main_monitor to cpukit/libmisc/shell/shell.h?

#1593

RTEMS Sequenced Initialization API

#1712

Add LWIP Support to RTEMS

#1728

Patch for RTEMS-BSP m68k/mcf5225x

#1815

_ISR_Nest_level should be protected with interrupts disabled in function _Watchdog_Insert

#1820

sptests/sp36: Improper buildsystem integration

#2000

Re-add removed LINK_OBJS to MP tests

#2071

The object-file suffixes for default-bsp-post-link in leaf.cfg don't work with make-exe...

#2111

AVR needs avr-libc to compile RTEMS.

#2114

statvfs for MS-DOS filesystem and 'df' coomand for shell

#2189

Insufficient documentation for rtems_clock_get_tod()

#2215

Duplicate code for TAR support

#2269

x86 Context Switch Synchronization and SMP

#2301

cpukit is built in BSP, and BSP custom, specific part of build tree but isnt built using BSP's spec file

#2330

nios2 missing atomics, fails to link C++

#2348

Timecounter: Add NTP support

#2349

Timecounter: Add PPS support

#2365

Task pre-emption disable is broken due to pseudo ISR tasks

#2778

Error building RTEMS Source Builder recipe for libxml

#2828

Remove libmisc/serdbg

#2869

Remove support for 16-bit object identifiers

#2871

Use bibtex references thoughout the documentation

#2872

Get rid of CONFIGURE_MAXIMUM_PTYS

#2902

Port RTEMS to Microblaze

#2971

Add fenv.h to newlib

#2985

Add documentation for printk()

#3143

Remove obsolete rtems_gxx_*() implementation

#3199

New test framework

#3230

RSB does not report --rsb-file for patches correctly.

#3250

Remove Use of bsp_specs

#3269

Make the IRQ extensions API a standard API

#3321

Improve console driver documentation

#3414

Remove Makefiles from rtems-examples?

#3420

Deprecate/obsolete rtems_io_lookup_name

#3456

Add support for CPU counter timestamps

#3566

Remove checks for Newlib features introduced during the RTEMS 5 development

#3584

Remove types which are only available if RTEMS_DEPRECATED_TYPES is defined

#3586

Remove deprecated proc_ptr definition

#3588

Remove deprecated rtems_context

#3590

Remove deprecated rtems_context_fp

#3592

Remove deprecated region_information_block

#3594

Remove deprecated rtems_thread_cpu_usage_t

#3596

Remove deprecated rtems_rate_monotonic_period_time_t

#3613

Remove m32c architecture port from upstream tools

#3615

Convert Tools_Used Wiki page into Support Tools Selection with Criteria

#3657

Add support for poll.h method - poll

#3659

LEON3 kernel entry point is overwritten - secondary processors may enter into spurious handler

#3689

bdbuf: Replace automatic read-ahead with rtems_bdbuf_peek()

#3698

libdl failure on many PowerPC BSPs

#3702

Space profile for RTEMS SMP

#3704

Review and update Doxygen recommendations

#3705

Software Design Document (SDD) for space profile

#3706

Create a hierarchy of RTEMS software components using Doxygen groups

#3707

Assign each code file to a Doxygen group

#3715

Add Requirements Engineering chapter to RTEMS Software Engineering Handbook

#3717

Add test guidelines chapter to RTEMS Software Engineering Handbook

#3718

Add support for test plans

#3719

Update libcrypt to latest FreeBSD to address Coverity Scan Issues

#3726

Select a requirements engineering tool

#3818

New build system

#3828

Switch over to new build system

#3846

Build system does not track the dependencies of start.o files

#3847

Test programs are not re-linked if a start.o or linker command file changes

#3853

Add and use project-wide glossary to documentation

#3866

Add support for GCC 10 noinit attribute

#3878

rtems06: All BSPs Fail to build with BSP Builder on rtems 6

#3883

Trace linker is broken on small-data area targets

#3886

Fail an RSB build with --mail option if a repo is not clean

#3897

Dhcpcd does not daemonize

#3910

Add BSP for STM32H7

#3924

Test the application configuration options

#3925

Optimize work area initialization

#3929

Document use of virtualenv

#3935

Remove use of RTEMS_INLINE_ROUTINE in the RTEMS sources

#3937

Move content of bsp_specs to GCC

#3941

Remove epiphany architecture support

#3942

multiple definition of `pci_indirect_functions'

#3947

fdt_rw.c: Unchecked return value (CID #1047324)

#3951

Remove obsolete powerpc BSPs

#3957

Review wiki/Developer for deletion

#3959

Add rtems_task_construct()

#3962

Licensing Requirements for Submissions Poorly Documented

#3977

Add unexplained-failure as option for Test Configuration Files

#3978

Move _RTEMS_version to implementation header file

#3979

Remove RTEMS_MAXIMUM_NAME_LENGTH

#3980

Remove deprecated <rtems/system.h>

#3981

Remove _Copyright_Notice from API header file

#3982

Test failures on BSPs with sbrk() support

#3988

Remove deprecated rtems_extension

#3989

Remove deprecated rtems_get_current_processor()

#3990

Remove deprecated rtems_get_processor_count()

#3991

Remove deprecated Thread typedef

#3993

Specify the RTEMS Classic API

#3994

Generate Doxygen markup for the application configuration options

#3996

fenv support is missing on powerpc

#3997

fenv support is missing on sparc

#3998

fenv support is missing on arm

#4007

Add rtems_message_queue_construct()

#4018

libcpusupport: Architecture defines are present outside CPU port

#4020

Change all version numbers to RTEMS 6 on the master branches

#4025

NFS Client is broken on 64-bit targets

#4029

i386: SMP-System hangs with non-consecutive APIC IDs

#4031

i386: ISR can overwrite its own stack during system initialization (cloned)

#4032

Make deferred free in malloc() support optional

#4034

Add rtems_interrupt_server_create() and rtems_interrupt_server_destroy() (cloned)

#4036

Track dependency command for the RSB

#4037

Python script distribution standardisation

#4039

arm/atsam/SC16IS752: Make interrupt server configurable (cloned)

#4040

coverhd.h remnants left to remove

#4043

Update software eng manual to rtems-central

#4044

Add RSB Track command

#4045

Remove RSB sb-builder command

#4046

Remove RSB sb-bootstrap command

#4048

Move RTEMS 5 BSP and package build sets and configs to RTEMS 6

#4052

[libbsd] e1000 driver not working for i386 based BSP in current rtems-libbsd master

#4054

RTEMS_BARRIER_AUTOMATIC_RELEASE and RTEMS_BINARY_SEMAPHORE options have the same value

#4055

bsps/xilinx-zynq: Flush TX-Buffer before initializing the zynq-uart

#4061

The reworked <rtems/confdefs.h> has a cyclic dependency with RTEMS_MULTIPROCESSING enabled (cloned)

#4063

Change to a simpler TFTP server for rtems-test

#4065

Typo in rtems-boot.ini "ubootdior" should be "ubootdir"

#4066

test running for too long

#4067

Libbsd build system does not separate kernel and user land include paths

#4069

dl06 does not link on RISCV

#4070

RISC-V Multiple Definition Error (many BSPs)

#4072

testsuite: Add ticket ids for expected-fail and indeterminate states, add annotations

#4074

Add CONFIGURE_MAXIMUM_THREAD_LOCAL_STORAGE_SIZE

#4076

i386: bad asm in smp mode

#4077

Implement LibBSD module dependencies in the build system

#4078

GCC C++ compiler for powerpc defines _GNU_SOURCE as a built-in

#4079

build: i386/pc686/appstart.o is created more than once

#4080

build: Add start.o dependency to the executable link step

#4081

Remove Autoconf/Automake based build system

#4084

throwing std::runtime on PC BSP fails

#4093

bsps/pc386: Add missing license header

#4095

Remove doc from rtems-tools

#4097

shell: CRTL-U sets the cursor to the wrong position (cloned)

#4098

All BSP bsets fail to build on master

#4100

arm/edb7312 Build Failure When SMP is Enabled

#4101

atsamv fails to build on master

#4103

BeagleBoard BSPs Fail to Build

#4105

Add RTEMS_PARTITION_ALIGNMENT

#4107

waf -j24 fails on powerpc/psim -- probable dependency issue

#4108

Review minimum and ensure it really is a minimum configuration

#4110

Ubuntu package list is broken

#4115

Cortex-A9 MPCore based BSPs should include the workaround for Errata 794072 and 845369

#4116

raspberrypi fails to build on waf and autoconf

#4117

lpc1768_mbed_ahb_ram_eth sp01 and others fail to link

#4118

Add clang variants for sparc BSPs

#4120

altcycv_devkit and imx7 fail to build all tests on master

#4121

lm3s3749 fails to build all tests on master

#4122

Change RTEMS_NO_RETURN to use C11 and C++11 standard means

#4123

rtems-bsps needs to be switched to look at waf info for BSPs

#4124

rtems-bsp-builder needs to be switched to waf builds

#4125

Update wiki pages to reference waf not autoconf based build

#4126

Remove legacy networking stack from rtems.git and place in a new repo

#4127

x86_64/amd64 fails to build

#4128

all griscv variants fail to build on master

#4129

x86_64 not listed as architecture in components in Trac

#4131

"fdisk DISK_NAME mount" in bdpart-mount.c uses non-existing file system type "msdos"

#4134

rtemsspec: Support user-defined ordering of directives in header files (Doxygen)

#4140

waf install of RTEMS breaks Makefile system in rtems-examples

#4142

Sparse disk destroys data when it grows

#4143

fileio - Stack too small for login with password

#4145

rtems-source-builder: Update RTEMS Kernel Recipe to Use waf for RTEMS

#4148

RFS bit map search buffer overflow

#4155

Add a workaround for the LEON3FT RETT Restart errata: TN-0018

#4157

pthread_spin_unlock() may corrupt the thread stack if RTEMS_PROFILING and RTEMS_SMP is enabled

#4158

rtems_interrupt_server_delete() does not destroy the ISR lock of the server control

#4164

Fix NVMe disk synchronization and media block handling

#4166

rtems-bsp-builder: Add way to select rtems version (6 vs 7)

#4167

Coverity Scan Does Not Run on rtems6

#4175

Resolve warning during AArch64 build

#4180

Add BSP for i.MXRT1050

#4181

Add CONFIGURE_INIT_TASK_STORAGE_SIZE

#4184

arm/bsps: Small MMU pages are rounded to 1 MiB

#4192

bsp/xilinx_zynq: Enable support for small MMU pages

#4193

devctl.h does not compile from C++ (cloned)

#4194

RSB failed with mail subject: Build Linux: PASSED 6/rtems-arm on x86_64-linux-gnu

#4195

Add gr740 support available in SIS 2.25 to the RTEMS Tester

#4196

Improve gcov support provided by GCC for RTEMS

#4197

testsuite: ts-validation-0 broken on AArch64 due to SIZE_MAX issues in the test design

#4198

arm: Add multilib for Cortex-R52 processor

#4200

Build system doco link in the User Guide is broken

#4201

waf building out of source tree does not find all configured bsps

#4202

Add BSP for ARM Fixed Virtual Platform with a Cortex-R52

#4204

Make Termination Handler which Prints Exit Information Less Scary

#4206

aarch64/xilinx_zynqmp_lp64 does not run with RTEMS_DEBUG enabled

#4207

PZERO define in RTEMS is wrong

#4209

STM32H743ZI Nucleo Consle support not working without workaround

#4212

libio leaks location clones

#4214

nios2: Add thread-local storage support

#4215

nios2: Add multilib with all custom floating-point instructions supported by GCC

#4217

Change status code for nested obtain of MrsP semaphores

#4218

aarch64 as internal error with spconfig01

#4219

gdb master does not compile on CentOS 7

#4221

dtc recipe issue -- dependency on yaml.h

#4222

Add a per mount flag to not create a node on O_CREAT in open

#4223

Missing "extern" in RTEMS_LINKER_ROSET_ITEM_ORDERED_DECLARE()

#4225

Cobham Gaisler NOEL-V RISC-V 32- and 64-bit processor BSP

#4226

Buildbot needs a tool chain update

#4228

rtems_build_name() depends on the signedness of char type

#4229

conf.py files need updated for sphinxcontrib-bibtex ver. 2.0.0 and greater

#4230

Timeout for automatic barriers is broken

#4231

MVME 2600/2700 has no console output

#4237

consolesimpletask.c: Two Unchecked Return Values Spotted by Coverity

#4238

Change motorola_powerpc bsp to support irq-generic

#4239

PowerPC shared ISA IRQ support is broken

#4240

rtems-debugger-target.c: Dereference before null check spotted by Coverity

#4241

rtems-debugger-threads.c: Three Dereference before null check errors spotted by Coverity

#4242

rtems-debugger-threads.c: Three Dereference before null check errors spotted by Coverity

#4244

Possible infinite recursion in Classic API Signal handling

#4245

Add PCI regions support to RTEMS nexus bus support

#4246

Add libbsd DC NIC support to mvme2700 board

#4250

b1553brm.c: Dereference before null check error spotted by Coverity

#4251

b1553rt.c: Dereference before null check error spotted by Coverity

#4252

grtc.c: Dereference before null check error spotted by Coverity

#4253

grspw_pkt.c: Dereference before null check error spotted by Coverity

#4254

shell.c: Dereference before null check error spotted by Coverity

#4258

printertask.c: Unchecked return value errors spotted by Coverity

#4259

grspw.c: Unchecked return value error spotted by Coverity

#4260

record-client.c: Unchecked return value error spotted by Coverity

#4261

getgrent.c: Unchecked return value error spotted by Coverity

#4262

rtems-debugger-threads.c: Unchecked return value error spotted by Coverity

#4264

Activate ehci_pci in rtems-libbsd (cloned)

#4265

Add rtems_get_build_hash()

#4267

Add rtems_get_target_hash()

#4269

Add rtems_get_build_label()

#4270

A failing task extension produces zombi objects and resource leaks

#4277

grtc.c: Missing break in switch errors spotted by Coverity

#4278

cpuusagetop.c: Missing break in switch errors spotted by Coverity

#4279

pci_cfg_read.c: Missing break in switch error spotted by Coverity

#4280

gen_uuid.c: Unchecked return value from library errors spotted by Coverity

#4281

main_chmod.c: Unchecked return value from library error spotted by Coverity

#4282

pwdgrp.c: Unchecked return value from library error spotted by Coverity

#4283

gr_rasta_spw_router.c: Unchecked return value from library error spotted by Coverity

#4284

gr_rasta_io.c: Unchecked return value from library error spotted by Coverity

#4285

gr_leon4_n2x.c: Unchecked return value error from library error spotted by Coverity

#4286

gr_rasta_tmtc.c: Unchecked return value from library error spotted by Coverity

#4287

gr_rasta_adcdac.c: Unchecked return value from library error spotted by Coverity

#4288

gr_701.c: Unchecked return value from library error spotted by Coverity

#4289

gr_tmtc_1553.c: Unchecked return value from library error spotted by Coverity

#4290

gr_cpci_gr740.c: Unchecked return value from library error spotted by Coverity

#4291

main_help.c: Unchecked return value from library error spotted by Coverity

#4292

rtems-rfs-bitmaps.c: Useless call (CID 1399749)

#4294

satcan.c: Unsigned compared against 0 error spotted by Coverity

#4295

b1553brm.c: Unsigned compared against 0 error spotted by Coverity

#4301

grcan.c: Unused value error spotted by Coverity

#4302

Obsolete powerpc/haleakala board

#4320

Add driver for cadence-spi device for xilinx based BSPs

#4321

Add spi driver for AXI SPI ip core from Xilinx

#4323

grlib,can: introduce a new common CAN baud-rate timing calculating functions

#4324

grlib,grcanfd: extend the GRCAN driver with GRCANFD support

#4326

objectextendinformation.c: Dereference after null check error spotted by Coverity

#4329

main_rtrace.c: Dereference after null check error spotted by Coverity

#4331

main_pci.c: Dereference after null check error spotted by Coverity

#4336

Improve the workaround for the LEON3FT store-store errata: TN-0009 on master

#4338

rtems_clock_set(): Cannot set future dates later than approximately 2105

#4339

main_cp.c: Unused value error spotted by Coverity

#4340

print-ls.c: Unused value error spotted by Coverity

#4341

fastlz.c: Unused value error spotted by Coverity

#4342

disp_hcms29xx.c: Unused value error spotted by Coverity

#4345

Nexus Driver for Marvell mv643xx Ethernet Adapter (powerpc/beatnik BSP, MVME6100 board)

#4346

Require RTEMS_PRIORITY for MrsP semaphores

#4355

Remove leon3 test configurations which assume sis is built into gdb from rtems-test

#4356

rtems_semaphore_set_priority() uses an invalid SMP lock

#4358

Priority discipline is broken for semaphores and message queues in SMP configurations

#4363

GDB does not build on Windows for MSYS2

#4364

rtems-test Fails to list failed tests to user

#4365

main_cp.c: Ignore return value from stat()

#4368

rtems--tools tool kit temp file constructor does nothing

#4374

Handle symbol sets in covoar

#4375

Add futimens() and utimensat()

#4376

covoar/ReportsBase Coverage::GenerateReports() uses raw pointer to std::list

#4377

covoar/Explanations Explanations::writeNotFound() NULL check of wrong variable

#4378

covoar/symbol-sets add libuuid to symbol-sets.ini

#4380

Update RTEMS POSIX Compliance Spreadsheet

#4382

mingw - add flex to packages needed

#4383

covoar keeps DWARF info in memory too long

#4384

Add SMTP login to rtems-tools mailer

#4386

covoar missing conditional branch instructions on ARM

#4387

covoar marks taken/not taken incorrectly for AArch64

#4388

Copy improved mailer.py from rtems-tools

#4389

Undefined behaviour if the area size calculation in calloc() and rtems_calloc() overflows

#4390

Make zero size allocation result consistent across directives

#4391

SPARC GCC no longer needs sysconf(515) (_SC_STACK_PROT)

#4392

rtems: Constify timer fire when directives

#4393

rtems: Change rtems_task_get_affinity() status

#4395

Workspace initialization is broken for arm/imx7 and arm/raspberrypi

#4396

Add futimens() and utimensat()

#4397

Make utime() use utimensat()

#4398

Make utimes() use utimensat()

#4399

Add tests for futimens() and utimensat()

#4400

Change filesystem utime_h to utimens_h to support nanosecond granularity

#4401

rtems: Change rtems_scheduler_get_processor_set() status

#4402

rtemstoolkit/mailer.py option added twice

#4403

rtems_timer_fire_when() returns wrong status code when wall_time argument is NULL

#4406

rtems: Constify rtems_task_wake_when()

#4410

rtems_task_start() does not check that the entry point is not equal to NULL (cloned)

#4411

rtems_task_restart() should set the real priority to the initial priority

#4412

Unexpected rtems_task_restart() behaviour if called from within interrrupt context

#4413

Allow pthread_cancel() from within interrupt context

#4414

Return RTEMS_CALLED_FROM_ISR in rtems_task_delete()

#4415

ReportsBase.cc: Division or modulo by float zero error spotted by Coverity

#4416

rtems-exeinfo.cpp: Division or modulo by zero errors spotted by Coverity

#4417

Explanations.cc: Resource leak error spotted by Coverity

#4418

GcovData.cc: Resource leak error spotted by Coverity

#4419

TraceReaderLogQEMU.cc: Resource leak error spotted by Coverity

#4420

TraceWriterQEMU.cc: Resource leak error spotted by Coverity

#4421

DesiredSymbols.cc: Resource leak error spotted by Coverity

#4422

ReportsBase.cc: Resource leak error spotted by Coverity

#4423

record-client.c: Unchecked return value error spotted by Coverity

#4424

rtems-bin2c.c: Copy into fixed size buffer spotted by Coverity

#4425

GcovFunctionData.cc: Copy into fixed size buffer errors spotted by Coverity

#4426

clock_nanosleep() may use wrong clock for relative times

#4427

GcovData.cc: Dereference after null check error spotted by Coverity

#4428

ReportsBase.cc: Dereference after null check error spotted by Coverity

#4430

DesiredSymbols.cc: Uninitialized pointer read error spotted by Coverity

#4432

GcovFunctionData.cc: Uninitialized pointer field error spotted by coverity

#4433

DesiredSymbols.h: Uninitialized pointer field error spotted by coverity

#4434

record-filter-zlib.cc: Uninitialized pointer field error spotted by Coverity

#4435

Thread cancellation may produce ready threads with an active thread timer

#4437

powerpc/shared/sbrk.c: Do not reference errno

#4438

powerpc/shared/irq/ppc-irq-legacy.c: Use rtems_malloc() not malloc()

#4439

Open of /dev/console should only be in SysInit List if Console Enabled

#4442

CID 1399721: Copy into fixed size buffer

#4444

CID 1399727: Copy into fixed size buffer

#4445

CID 1468684: Uninitialized scalar variable

#4448

covoar reports uncovered ranges of size 1

#4451

DesiredSymbols: Implement move assignment operator

#4453

Global vs Local Attribute Unclear for SMP/Distributed MP

#4454

bsps/i386: TSC calibration inaccurate

#4455

bsps/i386: TSC calibration inaccurate

#4458

Simplify trap table initialization

#4460

CID 1437647: Explicit null dereferenced

#4461

grlib: Register system console as "/dev/console"

#4462

CID 1437638: Logically dead code

#4463

rtems_condition_variable_signal() is implemented as a broadcast

#4468

waf build system does not generate the same bspopts.h

#4469

rtems-exeinfo.cpp: Not restoring ostream format errors spotted by Coverity

#4470

CoverageMapBase.cc: Not restoring ostream format error spotted by Coverity

#4471

ReportsHtml.cc: Not restoring ostream format error spotted by Coverity

#4472

ReportsText.cc: Not restoring ostream format error spotted by Coverity

#4473

AddressToLineMapper.h: Pointer to local outside scope error spotted by Coverity

#4475

Add NFSv4 client support to libbsd

#4482

Document kernel character I/O support in Classic API Guide

#4483

GcovData.cc: Not restoring ostream format error spotted by Coverity

#4484

GcovFunctionData.cc: Not restoring ostream format error spotted by Coverity

#4485

GcovData.cc: Out-of-bounds acces errors spotted by Coverity

#4486

TraceWriterQEMU.cc: String not null terminated issue spotted by Coverity

#4487

GcovData.cc: Uninitialized scalar field error spotted by Coverity

#4488

TraceWriterQEMU.cc: Uninitialized scalar variable issue spotted by Coverity

#4489

waf configure's --rtems-config and build create an error

#4490

rtems_partition_return_buffer() wrongly accepts buffers which are exactly at the buffer area end

#4491

TraceWriterQEMU.cc: Buffer not null terminated issue spotted by Coverity

#4492

ObjdumpProcessor.cc: Uninitialized pointer field

#4499

rtems-exeinfo.cpp: Uninitialized scalar field error spotted by Coverity

#4500

rld-dwarf.cpp: Uninitialized scalar field

#4501

TraceConverter.cc: Uncaught exception issue spotted by Coverity

#4502

rld-rap.cpp: Uncaught exception issue spotted by Coverity

#4506

posix_devctl() should return the errno directly not -1 and set errno

#4507

mount call retruns error number and it should return -1

#4508

rtems_message_queue_receive(): flush() does not release waiting tasks

#4509

Message queue priority dicipline is broken in SMP configurations

#4510

Fix ILP32 multilib variant in newlib for AArch64

#4511

Count of postponed jobs is not set to zero for a newly created rate-monotonic period object

#4513

Document parts of the Cache Manager in the Classic API Guide

#4517

Build failures for MacOS

#4518

RTEMS and LibBSD both provide competing pipe() interfaces

#4519

record-main-lttng.cc: Out-of-bounds access

#4522

Optimize red-black tree insert/extract

#4524

Re-add lost capability for custom stack allocator to allocate IDLE thread stacks

#4525

rtems-tld.cpp: Logically dead code error spotted by Coverity

#4527

Add new directives to get all time values provided by the timehands

#4528

rate monotonic: reset of CPU usage time not always detected

#4531

Data corruption in SMP schedulers

#4532

Priority inversion issues with MrsP locking protocol implementation

#4533

libdebugger should only build for archs with a backend

#4534

SMP EDF scheduler violates priority group ordering

#4543

cannot build multiprocessor application on eclipse

#4544

The last processor must not be removed if it is owned by a helping scheduler

#4545

The SMP EDF scheduler can only support more restricted affinity sets of a thread

#4546

A thread restart does not update the priority of related threads

#4547

dtc build failure on msys2 - all rtems6 target tools fail to build on Windows 10

#4548

Add RTEMS version option to waf configure

#4550

Add --rtems-version to rtems-test

#4551

Workaround for GRLIB-TN-0011 required for sparc/leon3 BSPs in SMP configuration

#4556

rtems_shell_main_mmove problem

#4557

medit problem in RTEMS6 (was 4.11/5.1)

#4560

Use thread-local storage for Newlib reentrancy objects

#4566

Atomic store does not use the order parameter for C++

#4569

aarch64 cache support does not provide rtems_cache_disable_data()

#4572

Improve Clarity for New Clock Manager Directives

#4581

Add AArch64 to Tier 1 BSPs

#4583

rtems-bsp-builder: Do not vary on network

#4585

Multiple Prototype Issues/Mismatches in rtems-examples from trace linker

#4588

RTEMS Stack Checker Report Bad Access When Not Enabled

#4607

support for SATA（AHCI） drives

#4613

Deprecate sparc64 port in rtems6 and remove in rtems7

#4614

Deprecate sh port in rtems6 and remove in rtems7

#4615

Deprecate v850 port in rtems6 and remove in rtems7

#4617

Potential non-monotonic CLOCK_MONOTONIC

#4619

riscv/rv32iac: minimum sample test fails to build

#4627

Multitasking start is broken on SMP targets which do not restore the interrupt state during context switching

#4629

shell: Potential stack corruption in rtems_shell_init()

#4631

RSB fails to build gdb with Python 3.10 (and newer)

#4632

Error in rtems/thread.hpp with GCC 12

#4634

Fix hash for glib-2.48

#4635

untar on JFFS slow down

#4639

Error in imfs memfile when device is full

#4641

x86_64 elf_machdep header file missing body

#4644

wkspace "failed allocations" increase with heap protection enabled (cloned)

#4646

Control buildsets using the --with-* command line option

#4648

Documentation Needs to Use Version Variables not Hard-Coded Numbers

#4650

tcpdump: Fix dumping to file and reading from file (cloned)

#4652

if_atsam: Fix checksum offload, add multicast and VLAN support (cloned)

#4654

pfctl: Fix global state initialization (cloned)

#4656

sync() whould synchronize all file descriptors (cloned)

#4658

Fix build issue with riscv sample test

#4661

GCC 12 generates wrong code

#4662

Fix GCC 12 warnings

#4666

TFTP: Implement block and window size options

#4667

include/sys/_bitset.h defines struct bitset, pollutes namespace

#4668

Use thread-local storage in libstdc++

#4669

clock_nanosleep() uses the wrong clock to determine the start time point

#4670

Add option to build RTEMS with coverage instrumentation

#4672

Thread-local storage should be usable once an idle thread exists

#4673

Runtime loader exported symbols address size not consistent

#4675

sleep should be a thread cancellation point

#4677

incorrect handling of "inactive_per_block" from "Objects_Information" structure (cloned)

#4678

System initialization time increased compared to RTEMS 4.11

#4679

Use priority inheritance for thread join

#4680

POSIX tasks cancelled through rtems_task_delete() should have an exit value of PTHREAD_CANCELED

#4682

Add AArch64 support to libdl

#4684

Support Versal DDRMC0 regions 0 and 1

#4686

Refactor shared code in Arm and AArch64

#4689

mips/jmr3904 Massive Number of Test Failures

#4690

CLOCK_REALTIME thread queue not updated as part of clock_settime call

#4691

Add support for FACE Technical Standard Restriction on POSIX Absolute Timers

#4693

Remove ILP32 aarch64/xilinx_versal_* BSPs

#4694

sptests/sp37 fails to compile if RTEMS_SMP is not defined

#4695

Automatic version numbers in the documentation

#4697

Documentation for libbsd Configure options

#4700

Fix instances of "is" operator used with literals

#4701

Relocatable write data is placed in read-only region on some BSPs

#4702

Fix warnings for indexing past array bounds (primarily validation tests)

#4703

NetSNMP 5.7.2.1 Patch

#4705

Installed header clean up

#4706

Installed header break C++

#4710

RSB does not expand dir type macros correctly

#4712

powerpc: Add AltiVec VRSAVE support

#4713

Make pthread_atfork follow FACE Technical Standard Behavior for Single Process Environments

#4714

Use waf for LwIP walk sources

#4718

Coverity issues in cpukit/libfs/src/ftpfs/tftpDriver.c

#4724

Git commit message format instructions

#4726

RSB decode exception stops build

#4729

Add 64bit address support to RTEMS FDT

#4730

rtems-source-builder doesn't generate tar archives for all packages any more

#4732

Set top in RSB version.py

#4736

RISC-V - ISR_stack allocation and usage by harts > CONFIGURE_MAXIMUM_PROCESSORS

#4741

Fix GCC PR 107248

#4744

libdebugger arm backend does not single step

#4746

Armv7-M SysTick clock driver uses off by one reload value

#4750

Linker discards RTEMS_SYSINIT_ITEM object

#4751

chmod problem in shell of RTEMS 6

#4756

Docs build system does not build singlehtml (cloned)

#4759

TLS support is broken for ARMv6

#4760

RSB fatal error on missing hash checksums

#4762

Add Versal SDHCI Arasan driver to libbsd

#4763

Detect terminal size

#4766

Classic API User Guide Index has Functions without Parentheses

#4767

Add rtems-lwip to release packages

#4768

Use tarballs for stable versions development tools

#4769

config: Add CONFIGURE_RECORD_INTERRUPTS_ENABLED

#4771

Versal UART issues

#4772

Remove use of interval from rtems_task_wake_after() documentation

#4775

RISC-V: Default RV64 BSPs to medany

#4778

Deployed RSB does not find release version config

#4782

Correct Issues in Rate Monotonic Manager Background Section

#4783

Fix issues when compiling dtc with Cygwin

#4803

Revert or Mark CAN API as experimental

#4818

Incomplete Statement in “cpu->heir”

#4820

Hazardous cast

#4821

Ubuntu Host Instructions Wrong

#4822

Function returning unchanged function input

#4826

Software Engineering Guide has broken Reference

#4828

Macro defined but magic number used instead

#4832

CLANG flagged error

#4833

grlib/pci/grpci2dma.c Warnings Require Insight to Resolve Properly

#4834

close_editor problem of RTEMS4.13/5.1 (cloned)

#4835

medit malloc problem of RTEMS (cloned)

#4838

Fix for UART interrupt on aarch64 Versal

#4842

SPARC/grlib - Registers definitions wrongly defined when there are reserved bits

#4843

Output value not assigned

#4844

Unchecked boundaries

#4845

Global variable declared/defined in the wrong file

#4847

Goto statements

#4850

LEON3 - Variable not initialized in any file / missing #ifdefs

#4855

Input validity unchecked

#4857

RSB Recipe Missing %hash Does Not Fail

#4858

GDB fails to build on any host with readline 8.2 or newer

#4859

Update gdb to 12.1

#4860

Waf Failure on CentOS 8 Stream

#4861

"Single HTML" links on docs.rtems.org lead to empty pages

#4862

score/ & bsps/: Unused input parameters

#4863

Operations evaluation order.

#4867

Clean up rtems-lwip uLan directory

#4869

Add QSPI Flash Device API

#4870

Add features for QSPI Flash on Xilinx Versal

#4872

GDB 13.1 does not build on FreeBSD 13.1

#4874

Undefined behaviour

#4875

LEON3FT - Power-down workaround

#4878

sptests/spsysinit01 sem_open call only has 3 parameters with O_CREAT flag (cloned)

#4881

powerpc gas and ld warnings

#4882

mvme2703 legacy network build failure

#4884

I2C default ioctl handler does not handle set clock

#4887

RTEMS Net Legacy does not build as a package in the RSB

#4889

Remove file scanning from rtems-lwip

#4892

MacOS tools build failures on Ventura and M series processors

#4896

Makefile.inc exported BSP include path not a system include path

#4899

BSP powerpc/qoriq_e6500_64 and powerpc/qoriq_e6500_32 fail to build

#4900

PowerPC score warnings in score headers

#4901

rtems_task_construct Documentation Missing Some Details

#4902

Update zlib to version 1.2.13

#4905

SuperCore _TOD_Adjust No Longer Referenced and Should be Removed

#4906

libbsd not working on nucleo-h743ZI

#4909

Multiple Protected Heap Functions are Ununused

#4910

RSB options for python selection

#4912

Collect Compression Library Source in Single Place

#4913

libbsd/ipsec-tools: Problems with more than 64 open files

#4915

Building sphinx 7.0.1 fails

#4920

libdl does not support TLS

#4921

Tool versions for RTEMS 6.1 release

#4923

FPU context init/switch not working well on more than 2 tasks on Cortex-Mx/ARMv7-M platform

#4924

Add Regulator Helper to RTEMS

#4925

Deal with GR740 errata: Level-2 Cache Issues H1 2023 (GRLIB-TN-0021)

#4929

RTEMS tools installs into $PREFIX/share/rtems and not $PREFIX/share/rtems6

#4931

User manual RSB BSP build examples need updating

#4934

Update STM32 H7 HAL/LL drivers code

#4935

7/rtems-mips get source fails

#4940

Fix cache support for ARM926EJ-S processor

#4942

Regulator warnings

#4943

ARM unwind register is not being hooked

#4944

libdl test dl09 fails on arm and aarch64

#4945

rtems-test SIS support is broken

#4947

Testsuite libtest tests not built with optimise or debug flags

#4950

libdl test dl07 fails on PowerPC

#4951

rtems-tools fail on FreeBSD to get number of cores

#4953

Use of symbol addresses for arbitrary values may not work for some code models

#4954

sparc: Simplify the clock driver and CPU counter support

#4955

sparc: Potential stack corruption in uniprocessor configurations during start multitasking

#4956

Deadlock caused by locking inversion in JFFS2

#4957

spstdc17 warnings building i386/pc686

#4959

rtems-test tftp does not a session timeout

#4960

User manual test documentation update

#4961

Document and test CONFIGURE_JFFS2_DELAYED_WRITE_TASK_PRIORITY

#4962

MicroBlaze BSPs modify carry flag before MSR is saved

#4967

strtof leaks memory

#4968

Python 3.12 fix escape sequences and remove SafeConfigParser

#4969

Update libibery sources in rtems-tools

#4970

Fix C++ 17 deprecated functionality in rtems-tools

#4971

Report that nested validation test cases did run

#4973

Add RSB options to control a specific python

#4974

rtems-tools buid failed with '--host=x86_64-w64-mingw32'

#4975

Terminal row/col probe timeout too fast

#4977

tmux bug swaps lines and columns

#4980

Build documentation using sphinx 7.2

#4982

Add support for TMS570LC4357 for the arm/tms570 BSP

#4986

Improve CONFIGURE_TICKS_PER_TIMESLICE configuration and documentation

#4991

RSB source and patch checksum updates

#4993

Increase FD_SETSIZE to 256

#4994

Documentation Sphinx theme broken for HTML

#4995

powerpc/qoriq_e6500_64 rtems-syms unknown machine type (cloned)

#4996

libdl: Add 64-bit PowerPC ELFv2 support

#4997

GMP not building when not installed

#4999

missing bsp's in rtems 6.1 rc2

#5002

Add API to get and set interrupt priorities for interrupt vectors

#5003

Make GCC ARM Cortex-A53 errata workarounds available

Details

Ticket	Created	Resolution	Component	Reporter	Owner	Modified
#550	20 years ago	wontfix	network/legacy	Thomas Rauscher	Joel Sherrill	17 months ago
Summary	fcntl() fails to put a non-blocking socket into blocking mode
Description	A socket cannot be put into blocking mode using fcntl() once it has been put into non-blocking mode. ioctl() is working fine, so the problem is not in the IP stack. The problem occurs in real life with the GoAhead? webserver on high-latency connections as its non-blocking write socket remains non-blocking and a web page can be truncated when websDone() is called. In newlib (1.10, 1.11) there are several non-blocking flags. The following is defined in sys/fcntl.h #define _FNBIO 0x1000 /* non blocking I/O (sys5 style) */ #define _FNONBLOCK 0x4000 /* non blocking I/O (POSIX style) */ ... #define _FNDELAY _FNONBLOCK /* non blocking I/O (4.2 style) */ /* O_NDELAY _FNDELAY set in include/fcntl.h */ /* O_NDELAY _FNBIO set in 5include/fcntl.h */ #define O_NONBLOCK _FNONBLOCK So O_NDELAY is never defined. sys/fcntl.h is not patched by the rtems-newlib patch. In RTEMS, libio.c and libio_.h both contain the following code #if ! defined(O_NDELAY) # if defined(solaris2) # define O_NDELAY O_NONBLOCK # elif defined(RTEMS_NEWLIB) # define O_NDELAY _FNBIO # endif #endif As O_NDELAY is undefined and RTEMS_NEWLIB is defined, O_NDELAY is defined to 0x1000. Thus, using O_NONBLOCK in fcntl() (this was suggested in the mailing list some time ago) doesn't seem to work. Release: RTEMS-4.5 Environment: rtems-4.5.0, newlib-1.10 or newlib-1.11
#1318	16 years ago	wontfix	bsps	Ralf Corsepius	Joel Sherrill	6 years ago
Summary	libcpu/shared is not helpful
Description	IMO, libcpu/shared is not helpful Either something is cpu-specific, then it should go into cpukit/score/cpu, general, then in should go into cpukit/ bsp-dependent, then it should go into libbsp. I would strongly suggest to reconsider libcpu/shared.
#1459	14 years ago	fixed	shell	Ralf Corsepius	Chris Johns	17 months ago
Summary	Add rtems_shell_main_monitor to cpukit/libmisc/shell/shell.h?
Description	rtems_shell_main_monitor is a private function of cpukit/libmisc/shell/cmds.c However, the way testsuites/libtests/monitor02/init.c uses it (as a public/exported function) to me reads as a strong indication that rtems_shell_main_monitor() should be "made public" by declaring/prototyping it in cpukit/libmisc/shell/shell.h
#1593	14 years ago	wontfix	score	zhangwenjie	Joel Sherrill	6 years ago
Summary	RTEMS Sequenced Initialization API
Description	This is a part of GSOC2010 project RTEMS Sequenced Initialization. The attachment is the implement of RTEMS Sequenced Initialization API. More details are placed on the wiki ​http://www.rtems.com/wiki/index.php/RTEMSSequencedInitialization.
#1712	13 years ago	fixed	network/legacy	Joel Sherrill	Joel Sherrill	2 years ago
Summary	Add LWIP Support to RTEMS
Description	This port was done by Marcello Presulli <m.presulli@…> and has been in my inbox for a while. I think I have merged most of their BSP specific work but the lwip has been pending. They managed to run LWIP+SNMP and an application in something like 128K RAM Here are some email fragments about it. For now we have just finished to port the latest lwip-1.3.2 based on the latest rtems 4.9.4 version successfully including a test_app for pings. We have stressed the stack with ping floods (packetsize 1400 Bytes interval ~150ms) about 250.000 sequences or more without packet losses and an average latency of 2ms. Additional we have adopted a new bsp target for the mcf5225x platform. To lwip now: we integrated it directly via the autoconfigure/automake process under the rtems cpukit/liblwip subtree and therefore it gets built with rtems when u configure it with --enable-networking=lwip appropriately. The driver on the other side we put it in libcpu/m68k/mcf5225x and it got built in the library context of the librtemsbsp.a. The driver has only 2 references on lwip headers ... To the bsp: The bsp we have adopted from mcf52223 and have created it as a own bsp-target=mcf5225x also via the autoconfig/automake mechanism. In the bsp itself i have modified the bsp_get_CPU_clock_speed() and some improvements in the console driver and PIT2 timer interrupt. The bsp_start() we have defined as weak symbol so we don't have to touch the "empty" bsp_start() of rtems coz it's platform-circuit specific you know.
#1728	13 years ago	invalid	bsps	Marcello Presulli	Joel Sherrill	6 years ago
Summary	Patch for RTEMS-BSP m68k/mcf5225x
Description	This patch is based on a "clean" RTEMS 4.9.4 and offers some modifications/enhancements for the BSP. Generic Exception Handler can be used instead of _uhoh Flash Driver for the onboard flash including flash security DMA initializations can be used to substitute the whole termios, so we saved alot of code/memory size Bugfixes in fec for the lwip-port, the fec driver works now "optimized" with the lwip-port RTC driver routines which configures the RTC clock and time generic software trap handler capable of return values Bugfix in "old" bsp with MASKALL_BIT when use RTEMS without console bugfixes and extensions in bsp header file "mcf5225x.h" generic dbg_printk() which prints debug outputs directly via a dma channel I hope someone have the time to merge it into the CVS head, it shouldn't be a real task i think.
#1815	13 years ago	invalid	score	Helder Silva	Joel Sherrill	6 years ago
Summary	_ISR_Nest_level should be protected with interrupts disabled in function _Watchdog_Insert
Description	_ISR_Nest_level should be protected with interrupts disabled in function _Watchdog_Insert, because another interrupt can occur after _ISR_Nest_level is saved, causing inconsistency in the global variable.
#1820	13 years ago	wontfix	unspecified	Ralf Corsepius	Joel Sherrill	6 years ago
Summary	sptests/sp36: Improper buildsystem integration
Description	sptests/sp36 is improperly integrated into the rtems buildsystem. sp36 is missing from SUBDIRS in sptests/Makefile.am, but is referenced in sptests/configure.ac, as well as the code is present in sp36/ sp36/ either should be activated (== added to SUBDIRS) or be fully removed.
#2000	12 years ago	wontfix	build	Marta Rybczynska	Ralf Corsepius	3 years ago
Summary	Re-add removed LINK_OBJS to MP tests
Description	PR1589 commit: 2011-12-08 Joel Sherrill <joel.sherrill@…> PR 1589/build base_mp/node1/Makefile.am, base_mp/node2/Makefile.am, base_sp/Makefile.am, capture/Makefile.am, cdtest/Makefile.am, fileio/Makefile.am, hello/Makefile.am, iostream/Makefile.am, loopback/Makefile.am, minimum/Makefile.am, nsecs/Makefile.am, paranoia/Makefile.am, pppd/Makefile.am, ticker/Makefile.am, unlimited/Makefile.am: Remove obsolete optional manager capability. and 2011-12-08 Joel Sherrill <joel.sherrill@…> PR 1589/build mp01/node1/Makefile.am, mp01/node2/Makefile.am, mp02/node1/Makefile.am, mp02/node2/Makefile.am, mp03/node1/Makefile.am, mp03/node2/Makefile.am, mp04/node1/Makefile.am, mp04/node2/Makefile.am, mp05/node1/Makefile.am, mp05/node2/Makefile.am, mp06/node1/Makefile.am, mp06/node2/Makefile.am, mp07/node1/Makefile.am, mp07/node2/Makefile.am, mp08/node1/Makefile.am, mp08/node2/Makefile.am, mp09/node1/Makefile.am, mp09/node2/Makefile.am, mp10/node1/Makefile.am, mp10/node2/Makefile.am, mp11/node1/Makefile.am, mp11/node2/Makefile.am, mp12/node1/Makefile.am, mp12/node2/Makefile.am, mp13/node1/Makefile.am, mp13/node2/Makefile.am, mp14/node1/Makefile.am, mp14/node2/Makefile.am: Remove obsolete optional manager capability. remove LINK_OBJS from the MP tests. They do no compile anymore. This patch re-adds them.
#2071	12 years ago	wontfix	build	cynt6007	Ralf Corsepius	6 years ago
Summary	The object-file suffixes for default-bsp-post-link in leaf.cfg don't work with make-exe...
Description	Would it be proper to change the suffix of the input in default-bsp-post-link in rtems/make/leaf.cfg ? Namely changing: define default-bsp-post-link $(NM) -g -n $@ > $(basename $@).num $(SIZE) $@ endef define default-bsp-post-link + $(NM) -g -n $@.o > $(basename $@).num + $(SIZE) $@.o endef I'm working on the application side of RTEMS (as opposed to the source), but here are the errors I encountered, what I did to find the problem, and the results. I kept getting errors when I ran the BuildTests?.sh... ~/rtems-addon-packages/examples/openldap$ ... i386-rtems4.11-nm o-optimize/slapd-addel i386-rtems4.11-nm: 'o-optimize/slapd-addel': No such file ... ~/rtems-addon-packages/examples/openldap$ ... i386-rtems4.11-size o-optimize/slapd-addel i386-rtems4.11-size: 'o-optimize/slapd-addel': No such file After looking at the Makefile it looks like they were called by: “make-exe”. Doing grep -r “make-exe” /opt revealed “make-exe” is defined in /opt/rtems-4.11/make/leaf.cfg, which calls bsp-post-link, (which is exactly the stage of the build process giving errors). According to man, nm and size work on object files. So, I changed the suffixes in /opt/rtems-4.11/make/leaf.cfg and the errors went away. After making the change, the following additional objects appeared in o-optimize: *.bin and *.ralf, with an approximate size of 570,000. The sizes vary slightly... If changing the suffixes in /opt/rtems-4.11/make/leaf.cfg makes sense, would it make sense to change them for rtems/make/leaf.cfg ? Also, there are other target.cfg's in the different bsp s that may have similar problems. Thanks, Cynthia Rempel
#2111	11 years ago		config	cynt6007	Joel Sherrill	6 years ago
Summary	AVR needs avr-libc to compile RTEMS.
Description	The avr io.h header gives a warning that if not doing avr-libc, that if not heeded leads to many errors. Need an avr-libc configuration file... should be a place-holding stub that states this is a problematic build. Anyway, these are just some cans of worms... FYI AVR has issues so deep it requires ~50 patches for binutils, ~60 patches for gcc, and has ~14 patches for avr-libc. I suspect getting ONE functional AVR-RTEMS a major undertaking... I wonder if it might be worth exploring using ATMEL's prepatched toolchain for non-Ada builds... For Ada-builds, that's going to be another major undertaking... Eric Weddington (at ATMEL) is recommending the use of their source for C, C++ builds... ​http://distribute.atmel.no/tools/opensource/Atmel-AVR-Toolchain-3.4.1/avr/ Although the Rolf Ebert (at AVR-Ada) have their own patch-set (using different tool versions) etc, which is making updating the AVR port so challenging... ​http://sourceforge.net/projects/avr-ada/files/avr-ada/Source%20Dist/ Their source-builder is located at: avr-ada/tools/build/build-avr-ada.sh
#2114	11 years ago	invalid	fs	Andrey Mozzhuhin	Chris Johns	6 years ago
Summary	statvfs for MS-DOS filesystem and 'df' coomand for shell
Description	I need statvfs for MS-DOS filesystem in my project. Share implementation and example shell command 'df' - like in GNU coreutils, but very simplified.
#2189	10 years ago	fixed	doc	Sebastian Huber	Needs Funding	3 years ago
Summary	Insufficient documentation for rtems_clock_get_tod()
Description	It is not clear which date and time is returned by rtems_clock_get_tod(). For example which time zone is used. Is there a relation to UTC?
#2215	9 years ago	duplicate	fs	Sebastian Huber		6 years ago
Summary	Duplicate code for TAR support
Description	We have similar code in rtems_tarfs_load(), Untar_FromMemory() and Untar_FromFile(). These functions should instead use a common TAR archive iterator with specific visitor functions.
#2269	9 years ago	duplicate	score	Joel Sherrill		6 years ago
Summary	x86 Context Switch Synchronization and SMP
Description	The x86 context switch code does not implement the "is executing" hand-off protocol used on the other ports. It should.
#2301	9 years ago	wontfix	build	Jeffrey Hill	Needs Funding	3 years ago
Summary	cpukit is built in BSP, and BSP custom, specific part of build tree but isnt built using BSP's spec file
Description	The nios2 soft-core processor is probably a bit of a different beast because it has many options when the FPGA programming is configured. It is probably therefore not a good context for the use of multilibs, although it is possible to enable them. For the nios2 RTEMS automated system configuration I must interrogate the Altera sopcinfo file to properly specify some of the nios2 cpu instance specific compiler flags such as {-mhw-div, -mhw-mul, -mhw-mulx}. An easy way to do this is to build some custom tools that need to run _only_on_the_host_ during the build. The source code for these BSP instance custom tools are easy to write if they include header files from the Altera generated HAL {system.h, linker.h}, when they are built. The Altera HAL needs to also be generated, but not installed, during the gnu build. It seems that this is the way that Altera intends for users to learn about system configuration based on how often their file formats evolve, and also based on lack of documentation for their sopcinfo file. Nevertheless, for many reasons we don't include Altera generated header files into any RTEMS BSP or cpukit source code. Currently this approach is difficult to fully implement because I must finish the interrogation for the CPU flags during the gnu build system configure phase when CFLAGS_OPTIMIZE_V and CPU_FLAGS are fetched from the custom .cfg file, and so I am unable to build some custom tools that need to run _only_on_the_host_ during the build at that time. However, during the preinstall phase it is easy to create a custom "bsp_specs_cpu" file like this. *[cpuspecs]: *[cpuspecs]: %(cpuspecs) -mhw-div *[cpuspecs]: %(cpuspecs) -mhw-mul *[cpuspecs]: %(cpuspecs) -mhw-mulx Which could be used in "bsp_specs" like this. %rename cc1 old_cc1 %rename cc1plus old_cc1plus %rename lib old_lib %rename endfile old_endfile %rename startfile old_startfile %rename link old_link %include <bsp_specs_cpu> *cc1: %{!qrtems: %{old_cc1} } \ %{qrtems*: %{old_cc1} %{cpuspecs} } *cc1plus: %{!qrtems: %{old_cc1plus} } \ %{qrtems*: %{old_cc1plus} %{cpuspecs} } *lib: %{!qrtems: %(old_lib)} \ %{!nostdlib: %{qrtems: --start-group -lrtemsbsp -lrtemscpu -lc -lgcc --end-group \ %{!qnolinkcmds: -T linkcmds%s}}} *startfile: %{!qrtems: %(old_startfile)} \ %{!nostdlib: %{qrtems: start0.o%s start1.o%s crti.o%s \ crtbegin.o%s -e start -u exception_vector }} *link: %{!qrtems: %(old_link)} %{qrtems: -dc -dp -N} *endfile: %{!qrtems: %(old_endfile)} %{qrtems: crtend.o%s crtn.o%s} This works very nicely excepting that the cpukit isn't built specifying a spec file and so it could use the wrong instructions when generating code. Therefore its tempting to modify the RTEMS build like this. cpukit/aclocal/gcc-specs.m4 | 20 ++++++++++++++++++++ cpukit/aclocal/prog-cc.m4 | 8 ++++++++ cpukit/automake/compile.am | 5 +++++ 3 files changed, 33 insertions(+) diff --git a/cpukit/aclocal/gcc-specs.m4 b/cpukit/aclocal/gcc-specs.m4 new file mode 100644 index 0000000..dd49bb9 --- /dev/null +++ b/cpukit/aclocal/gcc-specs.m4 @@ -0,0 +1,20 @@ +dnl +dnl $Id$ +dnl +dnl Check whether the target compiler accepts -specs +dnl + +AC_DEFUN([RTEMS_GCC_SPECS], +[AC_REQUIRE([RTEMS_PROG_CC]) +AC_CACHE_CHECK(whether $CC accepts -specs,rtems_cv_gcc_specs, +[ +rtems_cv_gcc_specs=no +if test x"$GCC" = x"yes"; then + touch confspec + echo 'void f(){}' >conftest.c + if test -z "${CC} -specs confspec -c conftest.c 2>&1";then + rtems_cv_gcc_specs=yes + fi +fi +rm -f confspec conftest* +])]) diff --git a/cpukit/aclocal/prog-cc.m4 b/cpukit/aclocal/prog-cc.m4 index e8b27b5..8dae5de 100644 --- a/cpukit/aclocal/prog-cc.m4 +++ b/cpukit/aclocal/prog-cc.m4 @@ -27,6 +27,14 @@ dnl check if the target compiler may use --pipe RTEMS_GCC_PIPE test "$rtems_cv_gcc_pipe" = "yes" && CC="$CC --pipe" +dnl check if the compiler supports --specs +RTEMS_GCC_SPECS + +AS_IF([test x"$rtems_cv_gcc_specs" = xyes],[ +GCCSPECS="-B\$(PROJECT_ROOT)/lib/ -B\$(PROJECT_ROOT)/$RTEMS_BSP/lib/" +GCCSPECS="${GCCSPECS} -specs bsp_specs -qrtems_cpukit"]) +AC_SUBST(GCCSPECS) + # Append warning flags if CFLAGS wasn't set. AS_IF([test "$GCC" = yes && test "$rtems_cv_CFLAGS_set" != set], [CFLAGS="$CFLAGS -Wall"]) diff --git a/cpukit/automake/compile.am b/cpukit/automake/compile.am index 3d81426..0e43494 100644 --- a/cpukit/automake/compile.am +++ b/cpukit/automake/compile.am @@ -2,6 +2,11 @@ ## $Id$ ## +CC = @CC@ $(GCCSPECS) +CXX = @CXX@ $(GCCSPECS) +CPP = @CPP@ $(GCCSPECS) +CCAS = @CCAS@ $(GCCSPECS) + AM_CPPFLAGS = @RTEMS_CPPFLAGS@ AM_CFLAGS = AM_CCASFLAGS = @RTEMS_CCASFLAGS@ I am thinking that hopefully this wont break any existing CPU architecture and or BSP because of several reasons. 1) Perhaps typical BSPs included in RTEMS are only modifying the way that executables are linked, and we probably don't create executables during the cpukit build. 2) Probably most of the spec files are noop implementation if qrtems isnt specified, and for the cpu kit I specify -qrtems_cpukit 3) I am specifying -qrtems_cpukit so its easy to write a spec file that does different things when it is and isnt a cpukit build. We could I suppose also define -qrtems_cpukit_multilib in that specific situation, if it were important to break that situation out. In any case this is what I currently have implemented and I will need to look for plan B before I can merge my nios2 support if this type of change looks like a bad idea. I am therefore writing to the relevant RTEMS wizards for some guidance. Thanks for your consideration.
#2330	9 years ago	fixed	unspecified	Joel Sherrill	Needs Funding	3 years ago
Summary	nios2 missing atomics, fails to link C++
Description	nios2-rtems4.11-g++ -B../../../../../nios2_iss/lib/ -specs bsp_specs -qrtems -mno-hw-mul -mno-hw-div -O0 -g -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -mno-hw-mul -mno-hw-div -o cxx_iostream.exe init.o /users/joel/rtems-4.11-work/tools/lib/gcc/nios2-rtems4.11/4.9.2/libstdc++.a(ios-inst.o): In function `atomic_add': /users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:53: undefined reference to `sync_fetch_and_add_4' /users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:53: undefined reference to `sync_fetch_and_add_4' /users/joel/rtems-4.11-work/tools/lib/gcc/nios2-rtems4.11/4.9.2/libstdc++.a(ios_init.o): In function `exchange_and_add_dispatch': /users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:49: undefined reference to `sync_fetch_and_add_4' /users/joel/rtems-4.11-work/tools/lib/gcc/nios2-rtems4.11/4.9.2/libstdc++.a(ios_init.o): In function `atomic_add_dispatch': /users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:53: undefined reference to `sync_fetch_and_add_4' /users/joel/rtems-4.11-work/tools/lib/gcc/nios2-rtems4.11/4.9.2/libstdc++.a(ios_init.o): In function `exchange_and_add_dispatch': /users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:49: undefined reference to `sync_fetch_and_add_4' /users/joel/rtems-4.11-work/tools/lib/gcc/nios2-rtems4.11/4.9.2/libstdc++.a(locale-inst.o):/users/joel/rtems-4.11-work/rtems-source-builder/rtems/build/nios2-rtems4.11-gcc-4.9.2-newlib-2.2.0.20150323-x86_64-linux-gnu-1/build/nios2-rtems4.11/libstdc++-v3/include/ext/atomicity.h:49: more undefined references to `sync_fetch_and_add_4' follow
#2348	9 years ago	fixed	score	Sebastian Huber		6 months ago
Summary	Timecounter: Add NTP support
Description	The FreeBSD timecounter implementation supports the NTP. This support is currently disabled in RTEMS.
#2349	9 years ago	fixed	score	Sebastian Huber	Sebastian Huber <sebastian.huber@…>	6 months ago
Summary	Timecounter: Add PPS support
Description	The FreeBSD timecounter implementation supports the RFC 2783 PPS-API. This support is currently disabled in RTEMS.
#2365	9 years ago	fixed	score	Sebastian Huber	Sebastian Huber	21 months ago
Summary	Task pre-emption disable is broken due to pseudo ISR tasks
Description	We have for example: Scheduler_Void_or_thread _Scheduler_priority_Unblock ( const Scheduler_Control *scheduler, Thread_Control *the_thread ) { Scheduler_priority_Context *context = _Scheduler_priority_Get_context( scheduler ); Scheduler_priority_Node *node = _Scheduler_priority_Thread_get_node( the_thread ); _Scheduler_priority_Ready_queue_enqueue( &the_thread->Object.Node, &node->Ready_queue, &context->Bit_map ); /* TODO: flash critical section? */ /* * If the thread that was unblocked is more important than the heir, * then we have a new heir. This may or may not result in a * context switch. * * Normal case: * If the current thread is preemptible, then we need to do * a context switch. * Pseudo-ISR case: * Even if the thread isn't preemptible, if the new heir is * a pseudo-ISR system task, we need to do a context switch. */ if ( the_thread->current_priority < _Thread_Heir->current_priority ) { _Scheduler_Update_heir( the_thread, the_thread->current_priority == PRIORITY_PSEUDO_ISR ); } SCHEDULER_RETURN_VOID_OR_NULL; } So the unblock event of a pseudo ISR task (MPCI or timer server) may pre-empt a task with disabled pre-emption. Suppose a higher priority task than the non-preemptible task is ready. Then this task gets scheduled once the pseudo ISR task blocks again.
#2778	8 years ago	worksforme	unspecified	Sambeet Panigrahi	Chris Johns	17 months ago
Summary	Error building RTEMS Source Builder recipe for libxml
Description	I wrote the three essential files for cross build of libxml for RTEMS, i.e. the .bset file (libxml.bset), the version specific configuration file (libxml2-2.7.8.cfg) and the build configuration file (libxml2-2.7.8.cfg).As per the FreeBSD ports I have placed it under textproc.However when the build configuration script is run, an error "config file is not found" occurs. I tried to cd into the directory by providing a bash script but that is also not working,an error "can't cd to ..<config directory>" appears.
#2828	7 years ago	fixed	lib	Joel Sherrill	Needs Funding	14 months ago
Summary	Remove libmisc/serdbg
Description	This is old code which should be replaced by the new debug server. Any hints or reusable in this for termios/serial port support can be retrieved from an old branch.
#2869	7 years ago	fixed	score	Sebastian Huber		2 years ago
Summary	Remove support for 16-bit object identifiers
Description	The only target using 16-bit object identifiers is M32C. This optimization is not worth the trouble. Self-contained objects are much more attractive for low end systems.
#2871	7 years ago	fixed	doc	Sebastian Huber		18 months ago
Summary	Use bibtex references thoughout the documentation
#2872	7 years ago	fixed	score	Sebastian Huber		2 years ago
Summary	Get rid of CONFIGURE_MAXIMUM_PTYS
Description	Change the Termios PTY driver for Telnet to use the new Termios device API. This why the CONFIGURE_MAXIMUM_PTYS gets superfluous.
#2902	7 years ago	fixed	bsps	Tanu Hari Dixit	Joel Sherrill	2 years ago
Summary	Port RTEMS to Microblaze
Description	Port RTEMS to Microblaze Students: Past, Present, and Potential Students Status: The tools build from RSB properly, but GDB is not compatible with current XDM (Xilinx Debugging Module), Xilinx GDB version is working fine. Introduction: A new architecture port, not just BSP. Include a BSP for GDB simulator. Also needs BSP for more complete HW on simulator. Goal: Update the preliminary Microblaze port, complete clock timer support, merge into RTEMS, and continue to improve the BSP/port. Some work has been initiated here [1] (By Joel Sherrill and Hesham ALMatary) to get hello world working. It has been tested on Atlys FPGA board [2]. The BSP can run virtually on every FPGA board the Xilinx tools support building MicroBlaze? on. The work will need to be updated against the current RTEMS version and tools and then completed. There are multiple boards supported by qemu-system-microblaze. One of these should be suitable for completing the port including interrupts. It will require investigation to know if qemu includes networking support for the Microblaze but this is likely. References [1] ​https://github.com/heshamelmatary/rtems-microblaze [2] www.digilentinc.com/atlys/
#2971	7 years ago	fixed	unspecified	Joel Sherrill	Needs Funding	2 years ago
Summary	Add fenv.h to newlib
Description	fenv.h is not present. ​http://pubs.opengroup.org/onlinepubs/9699919799/ feclearexcept, fegetenv, fegetexceptflag, fegetround, feholdexcept, feraiseexcept, fesetenv, fesetexceptflag, fesetround, fetestexcept, feupdateenv Cygwin has x86 implementation. FreeBSD should be source for ARM and some other architectures. The implementation is architecture specific. When completed, see psxtests/Makefile.am and enable the fenv.h POSIX API Compliance Tests.
#2985	7 years ago	fixed	doc	Joel Sherrill	chrisj@…	3 years ago
Summary	Add documentation for printk()
Description	printk() and its variants should be documented in the Classic API User's Guide. I am not sure where though.
#3143	7 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Remove obsolete rtems_gxx_*() implementation
Description	GCC versions prior to 6.1 used a RTEMS thread model based on rtems_gxx_*() functions. GCC version 6.1 or later uses the self-contained synchronization objects of Newlib <sys/lock.h> for the RTEMS thread model. Remove the obsolete implementation.
#3199	7 years ago	fixed	lib	Sebastian Huber	Sebastian Huber	3 years ago
Summary	New test framework
Description	Requirements: Support for RTEMS and general POSIX systems such as Linux or FreeBSD Test output functions printf() like No floating point support by default Optional floating point support Non-blocking Arbitrary context, e.g. during system start, before device drivers, in interrupt context Automatic test case registrations (constructors, linker set) Standard test case output, e.g. begin/end of test case message Ability to sort test cases at run-time Safe assert functions, e.g. assert(a == b) vs. assert(a = b) vs. assert_eq(a, b) Easy to write, e.g. avoid long namespace prefix rtems_test_* Support multiple threads No dependency on C++ Context validation, context invariant validation (e.g. thread dispatch disable level, interrupt state, heap state) Test pattern support (e.g. spintercritical*, parallel tests on SMP) No ready to use framework exists which fulfils these requirements. We have to write something for RTEMS. API proposal (header file <T.h>): T_eq(a, b) T_gt(a, b) T_printf(fmt, ...) T_output_printf(putchar, fmt, ...) T_output_string(putchar, s, n) T_output_add_format_handler(handler, arg) Example for T_output_add_format(): #include <stdio.h> #include <string.h> typedef void (*T_putchar)(char c); typedef size_t (*T_format_handler)(void *arg, T_putchar putchar, const char *fmt, const char *s, int *out); int T_printf(const char *fmt, ...); void T_output_printf(T_putchar putchar, const char *fmt, ...); void T_output_string(T_putchar putchar, const char *s, size_t n); void T_output_add_format_handler(T_format_handler handler, void *arg); size_t T_output_float_handler(void *arg, T_putchar putchar, const char *fmt, const char *s, int *out) { char buf[32]; if (strncmp(fmt, "%f", 2) != 0) { return 0; } *out = snprintf(buf, sizeof(buf), "%f", s); if (*out > 0) { T_output_string(putchar, s, (size_t) *out); } return 2; } void f(void) { T_output_add_format_handler(T_output_float_handler, NULL); }
#3250	6 years ago	fixed	build	Joel Sherrill	Sebastian Huber <sebastian.huber@…>	20 months ago
Summary	Remove Use of bsp_specs
Description	The bsp_specs file in each BSP directory is GCC specific and a barrier to supporting other compilers including clang/llvm. The bsp_specs file is used in a variety of ways including some odd build system dependencies and this ticket is to track eliminating those uses and the direct use of bsp_specs themselves for extending GCC options. The primary use of the bsp_specs file is to extend GCC's notion of what to link with. The bsp_specs itself commonly specifies: start.o file program entry point crt[in].o or crtbegin.o/crtend.o linker script Frequently, arguments which are implicitly passed to ld. This is typical and from a recently added BSP: %rename endfile old_endfile %rename startfile old_startfile %rename link old_link *startfile: %{!qrtems: %(old_startfile)} \ %{!nostdlib: %{qrtems: start.o%s crti.o%s crtbegin.o%s -e _start}} *link: %{!qrtems: %(old_link)} %{qrtems: -dc -dp -N} *endfile: %{!qrtems: %(old_endfiles)} %{qrtems: crtend.o%s crtn.o%s } Note that it has the implicit "-dc -dp -N" arguments which are passed to ld. The -dc and -dp options are equivalent and assign space to common symbols even with linking with -r. Why these two options are included is a mystery lost to time. Similarly the -N sets the .text and .data sections to be readable and writable and do NOT page align the data segment. This option is also a blast from the ancient past and likely should not be used. The "-qrtems" option is used with this pattern when compiling and linking RTEMS programs: -B${PREFIX}/${BSP}/lib/ -specs bsp_specs -qrtems The -B says that the following directory is a system directory. It is treated as a "lib" directory and is also assumed to have an include/ subdirectory. They are added to the linker and include search paths, respectfully. The -specs option references the bsp_specs and adds the -qrtems options which are then tripped by specifying -qrtems. Every part of this must be eliminated and performed in a manner that is more amenable to using a compiler other than GCC. In order to eliminate the bsp_specs, the arguments used in each BSP's version will have to be reviewed, made common, and then eliminated. A similar effort was performed when the BSP specific make-exe rules were eliminated from the "make/custom" files and a common pattern was put in place. In addition to using the bsp_specs to specify details about linking and include files, the build system has quite a few odd dependencies on the presence of bsp_specs: all BSP configure.ac use bsp_specs like this: AC_CONFIG_SRCDIR([bsp_specs]) The idea was that this was a unique file to ensure the script was in the right directory. This can be changed to use a specfic make/custom/*.cfg file and not lose any integrity checking. all BSP Makefile.am have this: dist_project_lib_DATA = bsp_specs That can just be removed when we don't use the file. ./aclocal/check-bsps.m4 uses it to see if it is BSP directory bootstrap uses it generate the list of BSP directories for the libbsp/*/acinclude.m4 files Other references are either comments or obvious uses of it in some invocation of GCC.
#3269	6 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	10 months ago
Summary	Make the IRQ extensions API a standard API
Description	The IRQ extensions API defined by <rtems/irq-extensions.h> is currently only implemented by some BSPs and not covered by test cases. It is not documented in the RTEMS Classic API Guide. Change the BSPs so that every BSP implements this API at least as a stub so that we can write tests for this API. Specify and document the API. Extend the API to enable/disable specific interrupt vectors. The proposed functions for this are rtems_interrupt_mask() and rtems_interrupt_unmask(). Extend the API to raise/clear specific interrupt vectors. This allows to test specific interrupts. The proposed functions are rtems_interrupt_raise() and rtems_interrupt_clear(). Extend the API to set/get the interrupt vector priority. Extend the API to get some information about a particular interrupt vector, e.g. if the raise/clear functions work, if it is possible to change the priority, etc.
#3321	6 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Improve console driver documentation
Description	We have now three different standard console drivers. Improve the console driver configuration documentation to assist the user to select the right one.
#3414	6 years ago	invalid	unspecified	Joel Sherrill		17 months ago
Summary	Remove Makefiles from rtems-examples?
Description	It seems to be time to remove Makefile build support from examples-v2
#3420	6 years ago	fixed	rtems	Joel Sherrill	joel	18 months ago
Summary	Deprecate/obsolete rtems_io_lookup_name
Description	I have noted in the RTEMS Class that this method is only kept for legacy purposes for many years now. There is no real use in the tree now. Only the following: $ grep -rl rtems_io_lookup_name . ./cpukit/include/adainclude/rtems-io.adb ./cpukit/include/rtems/io.h ./cpukit/libfs/src/imfs/ioman.c ./testsuites/sptests/sp21/sp21.scn ./testsuites/sptests/sp21/init.c Deprecate in 5.1 and remove in 6.x or obsolete and just remove now?
#3456	6 years ago	fixed	score	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Add support for CPU counter timestamps
Description	The original use case for the CPU counter was low level profiling of SMP synchronization primitives and busy wait loops for device drivers. For low overhead timestamps the CPU counter may be also used, e.g. for tracing. To support timestamps, the CPU counter must increment monotonically within the whole range of the CPU counter data type, e.g. 32-bit unsigned integer. In addition, it should not be affected by power saving states of the processor.
#3566	5 years ago	fixed	build	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Remove checks for Newlib features introduced during the RTEMS 5 development
Description	Remove checks for O_DIRECTORY, O_CLOEXEC and O_NOFOLLOW.
#3584	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove types which are only available if RTEMS_DEPRECATED_TYPES is defined
Description	The types boolean, single_precision, double_precision, rtems_single, rtems_double, and rtems_boolean. are only available if RTEMS_DEPRECATED_TYPES is defined. They are deprecated for a long time. Remove them in this release.
#3586	5 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated proc_ptr definition
#3588	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_context
#3590	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_context_fp
#3592	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated region_information_block
#3594	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_thread_cpu_usage_t
#3596	5 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_rate_monotonic_period_time_t
#3613	5 years ago	fixed	arch/m32c	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Remove m32c architecture port from upstream tools
Description	The m32c architecture port was removed in RTEMS 5.1. Remove RTEMS-specific support in the upstream tools, e.g. Binutils and GCC.
#3615	5 years ago	wontfix	admin	shashvat jain	Chris Johns	17 months ago
Summary	Convert Tools_Used Wiki page into Support Tools Selection with Criteria
Description	Originally, this ticket requested that ToolStatus? and Tools_Used wiki pages be addressed. ToolsStatus? was horribly out of date and removed. This leaves Tools_Used as a list of the tools used at a point in time. This list needs to be updated and grown into a Support Tools section of the Software Engineering Guide. Beyond a list of the tools and their use with RTEMS, it should discuss the attribute(s) that led to the tools selection. Original ticket description below: ​https://devel.rtems.org/wiki/TBR/Review/ToolStatus This page's data is very old and needs to be updated so that it can be added to the wiki with a link to the home page name "Development Tools". All the data about the tools on the page must be updated to the latest used by RTEMS 5 , this new page can either be added to the page ​https://devel.rtems.org/wiki/Tools_Used or be made into a separate page altogether . Advice on the placement of the new page is needed.
#3657	5 years ago	fixed	posix	Marçal Comajoan Cara		2 years ago
Summary	Add support for poll.h method - poll
Description	According to the RTEMS POSIX 1003.1 Compliance Guide, poll() from poll.h is not supported (​https://docs.rtems.org/branches/master/posix-compliance/posix-compliance.html#poll-h) PS: I wrote the test for POSIX API Signature Compliance Test and it compiled correctly but I suppose that some things are still not fully supported.
#3659	5 years ago	fixed	arch/sparc	Martin Aberg		3 years ago
Summary	LEON3 kernel entry point is overwritten - secondary processors may enter into spurious handler
Description	When linking a LEON3 RTEMS SMP application, the entry point in the ELF output file is set to the symbol "start". "start" is the first entry in the trap table and directly jumps to "hard_reset". The boot CPU does the following in the boot_card(): Release other CPUs from power-down (but does not wait here) Some other initializations Overwrite trap entry 0 with spurious interrupt handler The rest It means that the entry point is guaranteed to be valid for the first CPU entering the RTEMS kernel. But 1. and 3. above gives a race. non-first CPU:s will either enter the kernel properly or end up in the spurious interrupt handler depending on how quick it reaches the "start" label. One example where this has been an issue is when secondary processors run for some time (self-tests) in a ROM boot loader before entering the RTEMS (ELF) entry point. It is convenient to use the ELF file entry point for all processors. Possible solutions: a. Do not install spurious handler on trap table entry 0. For example by changing bsps/sparc/leon3/start/spurious.c: - for ( trap=1 ; trap<256 ; trap++ ) { + for ( trap=0 ; trap<256 ; trap++ ) { (This changes address 0 on GR740. So following a null function pointer may not end up in spurious trap anymore.) b. Change the SPARC ELF entry point. For example to the symbol hard_reset. c. Document that the entry point for boot processor is "start" and entry point for other processors is "hard_reset".
#3689	5 years ago	fixed	lib/block	Sebastian Huber	Sebastian Huber	3 years ago
Summary	bdbuf: Replace automatic read-ahead with rtems_bdbuf_peek()
Description	The block device cache (bdbuf) tries to deduce a read ahead opportunity from the read access pattern. Using real world FAT file systems showed that this is not very reliable and the read ahead seldom happens. One way to fix this issue is let file systems give a hint to the block device cache that they will likely read a certain block in the future, e.g. the next block in a sequential file read. Remove the existing read ahead logic and add a new function rtems_bdbuf_peek() instead.
#3698	5 years ago	worksforme	lib/dl	Joel Sherrill		2 years ago
Summary	libdl failure on many PowerPC BSPs
Description	The following BSPs do not build dl09: powerpc-beatnik powerpc-brs5l powerpc-brs6l powerpc-dp2 powerpc-haleakala powerpc-icecube powerpc-mcp750 powerpc-mpc8260ads powerpc-mtx603e powerpc-mvme2100 powerpc-mvme2307 powerpc-mvme3100 powerpc-mvme5500 powerpc-pm520_cr825 powerpc-pm520_ze30 powerpc-qemuprep-altivec powerpc-qemuprep powerpc-qoriq_e6500_64 powerpc-ss555 powerpc-virtex4 powerpc-virtex5 They fail with this: gmake[5]: Entering directory `/home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/testsuites/libtests' powerpc-rtems5-gcc -mcpu=powerpc -mmultiple -mstring -mstrict-align -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -Wl,--gc-sections -mcpu=powerpc -mmultiple -mstring -mstrict-align -B./../../lib/libbsp/powerpc/motorola_powerpc -B/home/joel/rtems-work/rtems-testing/rtems/rtems/bsps/powerpc/motorola_powerpc/start -specs bsp_specs -qrtems -L./../../cpukit -L/home/joel/rtems-work/rtems-testing/rtems/rtems/bsps/powerpc/shared/start -Wl,--wrap=printf -Wl,--wrap=puts -Wl,--wrap=putchar -o dl01.pre dl01/dl01-init.o dl01/dl01-dl-load.o dl01-dl01-tar.o ../../lib/libbsp/powerpc/motorola_powerpc/librtemsbsp.a ../../cpukit/librtemscpu.a /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: ../../cpukit/librtemscpu.a(rtl-mdreloc-powerpc.o): in function `get_sdata_libdl_size': /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:73: undefined reference to `bsp_section_sdata_libdl_begin' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:73: undefined reference to `bsp_section_sdata_libdl_begin' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:74: undefined reference to `bsp_section_sdata_libdl_end' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:74: undefined reference to `bsp_section_sdata_libdl_end' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: ../../cpukit/librtemscpu.a(rtl-mdreloc-powerpc.o): in function `get_sdata_sbss_size': /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:63: undefined reference to `bsp_section_sdata_begin' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:63: undefined reference to `bsp_section_sdata_begin' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:64: undefined reference to `bsp_section_sbss_end' /home/joel/rtems-work/tools/5/lib/gcc/powerpc-rtems5/7.4.0/../../../../powerpc-rtems5/bin/ld: /home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qemuprep-rtems/powerpc-rtems5/c/qemuprep/cpukit/../../../../../rtems/c/src/../../cpukit/libdl/rtl-mdreloc-powerpc.c:64: undefined reference to `bsp_section_sbss_end' collect2: error: ld returned 1 exit status
#3702	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Space profile for RTEMS SMP
Description	Define a subset of RTEMS SMP functionality which is useful for space applications, and can be pre-qualified by project #3701. To figure out the feature set carry out an online survey. In addition, do a workshop at ESA / ESTEC premises (also accessible on-line via Skype for Business). The space profile needs to be documented.
#3704	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	9 months ago
Summary	Review and update Doxygen recommendations
Description	Review and update the ​Doxygen recommendations. Due to the removal of the pre-install build step we can now use Doxygen directly with header files in the source tree. It should be possible to generate documentation for all architectures and BSPs. All code files (headers, C source, assembler) should have an @file entry belong to at least one Doxygen group (@ingroup). There are a lot of source file in RTEMS (excluding test code and legacy network stack): find bsps cpukit -name '*.[chsS]' | grep -v libnetworking | wc 4898 4898 187167 Creating the groups and categorizing all code file is a labour intensive work package. Therefore it should be discussed if the @brief and descriptions for files should be removed. The file content will be covered by groups and individual documentation. The use of @param should be clarified. It is not clear if the [in], [out] or [in,out] should be used. If they should be used, what they mean exactly.
#3705	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Software Design Document (SDD) for space profile
Description	A Software Design Document (SDD) is required by ​ECSS-E-ST-40C specified by Annex F. The plan is to use Doxygen for the software architecture, and detailed design. How should links to from a software component (Doxygen group) to requirements be handled?
#3706	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	9 months ago
Summary	Create a hierarchy of RTEMS software components using Doxygen groups
Description	Create a hierarchy of RTEMS software components using Doxygen groups. Review the existing Doxygen groups (software components) first. ​https://docs.rtems.org/doxygen/branches/master/modules.html Grouping should be done by at BSP level architecture and then by BSP. There should be a device driver group with subgroups for each device class and specific device drivers, e.g. a BSP-specific device driver belongs to a device class and BSP. There should be groups for APIs, file systems, support libraries, super core, etc. Board Support Packages Device Drivers Block Devices Block Device Buffer Management Block Device Disk Management Block Device Management XYZ Block Device Block Device Partition Management Cache XYZ Cache Support Console Termios XYZ Driver Framebuffer XYZ Driver I2C Bus Driver XYZ Driver Device Driver XYZ Driver Legacy I2C Legacy Network RTC XYZ Driver Serial Mouse XYZ Driver SPI Bus Driver XYZ Driver Device Driver XYZ Driver API Classic Tasks Dynamic Loading File Systems Memory Management Shell Tracing Event Recording Capture Engine Internal C Library Support POSIX Shell Super Core Thread Handler CPU ARM
#3707	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 months ago
Summary	Assign each code file to a Doxygen group
Description	Assign each code file (header, C source, assembler) to a Doxygen group. Exclude the legacy network stack and test suites. For third-party code decide case by case. Use this template and place it at the top of the file: /** * @file * * @ingroup RTEMSClassicTasks */ Do not use @brief and a file description.
#3715	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	11 months ago
Summary	Add Requirements Engineering chapter to RTEMS Software Engineering Handbook
Description	The chapter should cover the following topics: Overview and introduction Evaluation of tools for requirements management Selected tool for requirements management Evaluation of data models and formats for the requirements Definition of data model and format actually used for the project (may overlap with tool) Requirements management workflow used in the project Requirements on requirements, e.g. derived from standards such as ​ECSS-E-ST-10-06C Requirement identifiers used to ensure traceability
#3717	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Add test guidelines chapter to RTEMS Software Engineering Handbook
Description	The test guidelines chapter should cover the following topics: tools involved in testing test classification test organization how to plan tests how to write tests how to run tests how to report tests
#3718	5 years ago	fixed	test	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add support for test plans
Description	Test plans are required by ​ECSS-Q-ST-40C: 5.5.3.2, Software unit testing, a. 5.5.4.1, Software integration test plan development 5.6.3.1, Development and documentation of a software validation specification with respect to the technical specification, a. The test plans are a content of several documentation sets. Test plans are subject to verification activities: 5.8.3.6, Verification of software unit testing (plan and results) 5.8.3.7, Verification of software integration 5.8.3.8, Verification of software validation with respect to the technical specifications and the requirements baseline It should be possible to create documents from individual test plan fragments for a particular test case. There are two approaches possible: Add the test plan directly to the test code in form of a specially formatted comment. Use separate files with a specific format for test plans. The test plans should include a test identifier, test classification (unit, integration, validation), test steps, links to requirements, design, architecture, unit, component, etc. It should be possible to verify that a test is executed according to the plan from the test output. For example the test plan could specify test steps. The test output could print the test steps. A tool could check that the test steps in the output and the plan are consistent.
#3719	5 years ago	fixed	lib	Joel Sherrill	Sebastian Huber	6 months ago
Summary	Update libcrypt to latest FreeBSD to address Coverity Scan Issues
Description	Coverity issues 1018087, 1018088, 1018089, and 1018090 all appear to be addressed in the latest source code for libcrypt from FreeBSD. This code needs to be updated. This is one of the issues (1018087) in crypt-sha512.c from Scan but the same issues are in sha256.c 275 if (copied_key != NULL) CID 1018088: &#39;Constant&#39; variable guards dead code (DEADCODE) [select issue] 276 memset(copied_key, '\0', key_len); null: At condition copied_salt != NULL, the value of copied_salt must be NULL. dead_error_condition: The condition copied_salt != NULL cannot be true. 277 if (copied_salt != NULL) CID 1018087 (#1 of 1): 'Constant' variable guards dead code (DEADCODE) dead_error_line: Execution cannot reach this statement: memset(copied_salt, 0, salt.... Local variable copied_salt is assigned only once, to a constant value, making it effectively constant throughout its scope. If this is not the intent, examine the logic to see if there is a missing assignment that would make copied_salt not remain constant. 278 memset(copied_salt, '\0', salt_len);
#3726	5 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Select a requirements engineering tool
Description	In order to ease the requirements engineering process (such a process doesn't exist currently, see #3715) select a supporting tool (or tools). Define tool selection criteria. Do a market survey. Use the following wiki page for this work: Developer/RequirementsEngineering
#3818	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	3 years ago
Summary	New build system
Description	Replace the existing Autoconf and Automake based build system with something new. Goals: Easier configuration of BSPs (BSP options) Easier to maintain (nobody understands the details of the existing build system) Faster builds (especially on Windows) Reusable build specifications (e.g. generate documentation for BSP options for the user manual) Validation of built artefacts (e.g. ensure that the objects are built as specified using the DWARF debug information) Support building of BSPs external to the project Customization of the build (e.g. build only a subset of the RTEMS functions) Support alternative compilers such as clang instead of GCC Proposed solution: Use waf for the build system Specify the build through build specification items maintained by Doorstop
#3828	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Switch over to new build system
Description	During a four month period every RTEMS user is encouraged to test its BSPs of interest with the new build system. The user should use the following check list. Issues should be reported on devel@… or this ticket. Review the build system specification in the user manual. If something is unclear, please complain. Use "./waf bsp_defaults" to get a list of options with default values for the BSPs of interest. Check that all options are present. Check the type of the options in the build specification (e.g. actions get-integer, get-boolean, get-string). Review the description of the options. Review the default values. Add sanity checks to options (e.g. actions assert-*). Build the BSPs of interest with all tests enabled (BUILD_TESTS = True). Review the ABI and linker options. Review the linker command files. Run the test on the boards. Build the BSPs of interest with different permutations of build options. Make sure a sample set of tests runs for each permutation. Check that all artefacts (libraries, header files, start files, linker command files, etc.) are properly installed. Check that you can build your applications and third-party libraries with the installed BSPs of interest.
#3846	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Build system does not track the dependencies of start.o files
Description	This defect relates to the new waf based build system. Most start.o are generated from start.S files which use the C preprocessor. The start.o files are not re-generated if header files (e.g. cpuopts.h, bspopts.h) change. The problem is that the automatic dependency generation via the GCC -MMD flags doesn't work with the current asm rule: def asm(self, bld, bic, source, target=None, deps=[], cppflags=[]): if target is None: target = os.path.splitext(source)[0] + ".o" bld( asflags=self.data["asflags"], before=["cstlib"], cppflags=cppflags + self.data["cppflags"], features="asm c", includes=bic.includes + self.data["includes"], rule="${CC} ${ASFLAGS} ${CPPFLAGS} ${DEFINES_ST:DEFINES} ${CPPPATH_ST:INCPATHS} -c ${SRC[0]} -o ${TGT}", source=[source] + deps, target=target, ) return target
#3847	4 years ago	duplicate	build	Sebastian Huber	Sebastian Huber	21 months ago
Summary	Test programs are not re-linked if a start.o or linker command file changes
Description	The test programs are not re-linked if a start.o or linker command file changes. The dependency tracking doesn't know anything about the linker command files or GCC specification files. The start files are normally included through the linker command file START() directive. Somehow these dependencies must be manually added to the test programs.
#3853	4 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add and use project-wide glossary to documentation
Description	See discussion: ​https://lists.rtems.org/pipermail/devel/2020-January/056728.html Approach A Use c-user/glossary.rst as the project-wide glossary. Generate document-specific glossaries from this file and the use of :term: in the specific document. Add new waf target "./waf regenerate" to carry out the glossary generation. Approach B Move the glossary terms to the RTEMS specification (e.g. in the RTEMS sources "spec/glossary/*.yml") and generate the glossary.rst for the documents with a script. This gives us more flexibility and removes the need for the special parser code, see: ​https://lists.rtems.org/pipermail/devel/2020-January/056811.html The AV-2 mentioned of the DoD Architecture Framework (DoDAF) wants the glossary terms in categories. We could add categories to the glossary specification items. This would be difficult with a master glossary in reST.
#3866	4 years ago	fixed	bsps	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add support for GCC 10 noinit attribute
Description	In GCC 10 there is a new common variable attribute noinit: ​https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-noinit-variable-attribute ​https://gcc.gnu.org/git/gitweb.cgi?p=gcc.git;a=commit;h=f0033821c1c9ba386a1983499a666d5759cdd943 To support this, a change in the linker command files is necessary.
#3878	4 years ago	worksforme	rtems	Joel Sherrill		2 years ago
Summary	rtems06: All BSPs Fail to build with BSP Builder on rtems 6
Description	I think there are a couple of places in an rtems6 build of RTEMS itself where rtems5 is being used from some hard-coded. Not sure what's up but here's the message to the build@ list. ​https://lists.rtems.org/pipermail/build/2020-February/010431.html From the post above, you only see that all failed. Looking at the build log from a single BSP, I see is early which doesn't cause a visible failure gmake[4]: Entering directory `/home/joel/rtems-cron-6/b-jmr3904/mips-rtems6/c/jmr3904/make' sed \ -e "s|[@]prefix[@]|/home/joel/rtems-cron-6/tools/6/bsp-install|" \ -e "s|[@]exec_prefix[@]|/home/joel/rtems-cron-6/tools/6/bsp-install/mips-rtems6|" \ -e "s|[@]pkgdatadir[@]|/home/joel/rtems-cron-6/tools/6/bsp-install/share/rtems5|" \ and this one which is error inducing. : error trying to exec 'mips-rtems5-gcc': execvp: No such file or directory error: compiling wrapper: Compiler error gmake[5]: * [dl08-sym.o] Error 10 The first looks like a typo that isn't breaking anything but needs to be fixed. The latter is some issue during the dl tests using rtems5 or it is hard-coded in a tool somewhere.
#3883	4 years ago	fixed	tool	Sebastian Huber	Sebastian Huber	4 years ago
Summary	Trace linker is broken on small-data area targets
Description	The trace linker generates some data, e.g. __rtld_trace_names. The arrays are not declared in a header file as arrays of unspecified size. So, in rtems-trace-buffer-vars.c the compiler assumes that the size one weakly defined defaults reside in the small-data area. This leads to linker errors with the strong definitions: Waf: Entering directory `/scratch/git-rtems-examples/build/powerpc-rtems5-qoriq_e6500_32' [22/29] Processing rtrace: build/powerpc-rtems5-qoriq_e6500_32/filesystem/fat_ramdisk/init.c.6.o build/powerpc-rtems5-qoriq_e6500_32/filesystem/fat_ramdisk/fs-root-tar.c.6.o -> build/powerpc-rtems5-qoriq_e6500_32/filesystem/fat_ramdisk/fat_ramdisk.texe /opt/rtems/5/bin/powerpc-rtems5-gcc: /opt/rtems/5/lib64/gcc/powerpc-rtems5/7.5.0/../../../../powerpc-rtems5/bin/ld: /opt/rtems/5/powerpc-rtems5/qoriq_e6500_32/lib/librtemscpu.a(rtems-trace-buffer-vars.o): the target (__rtld_trace_names) of a R_PPC_SDAREL16 relocation is in the wrong output section (.data) /opt/rtems/5/bin/powerpc-rtems5-gcc: /opt/rtems/5/powerpc-rtems5/qoriq_e6500_32/lib/librtemscpu.a(rtems-trace-buffer-vars.o): in function `rtems_trace_names': /opt/rtems/5/bin/powerpc-rtems5-gcc: /home/EB/sebastian_h/git-rtems-5/c/src/../../cpukit/libmisc/capture/rtems-trace-buffer-vars.c:51:(.text.rtems_trace_names+0x2): relocation truncated to fit: R_PPC_SDAREL16 against symbol `__rtld_trace_names' defined in .data.__rtld_trace_names section in /tmp/ccu6uaaa.o /opt/rtems/5/bin/powerpc-rtems5-gcc: /opt/rtems/5/lib64/gcc/powerpc-rtems5/7.5.0/../../../../powerpc-rtems5/bin/ld: /opt/rtems/5/powerpc-rtems5/qoriq_e6500_32/lib/librtemscpu.a(rtems-trace-buffer-vars.o): the target (__rtld_trace_signatures) of a R_PPC_SDAREL16 relocation is in the wrong output section (.rodata) /opt/rtems/5/bin/powerpc-rtems5-gcc: /opt/rtems/5/powerpc-rtems5/qoriq_e6500_32/lib/librtemscpu.a(rtems-trace-buffer-vars.o): in function `rtems_trace_signatures': /opt/rtems/5/bin/powerpc-rtems5-gcc: /home/EB/sebastian_h/git-rtems-5/c/src/../../cpukit/libmisc/capture/rtems-trace-buffer-vars.c:81:(.text.rtems_trace_signatures+0x2): relocation truncated to fit: R_PPC_SDAREL16 against symbol `__rtld_trace_signatures' defined in .rodata.__rtld_trace_signatures section in /tmp/ccu6uaaa.o
#3886	4 years ago	fixed	tool/rsb	Chris Johns	Chris Johns	4 years ago
Summary	Fail an RSB build with --mail option if a repo is not clean
Description	Fail a build if a user has the --mail option to post the build results and repo is not clean. Public postings of results with local changes has no public value. Users cannot replicate the results if changes are not publicaly available. What about a repo that is clean but has changes that are not pushed?
#3897	4 years ago	wontfix	network/libbsd	Ionut Vilsan	Sebastian Huber	20 months ago
Summary	Dhcpcd does not daemonize
Description	Background: the DCHP client is started using the rc.conf script. In the background a new task dedicated to the dhcp client is created and then the dhcp client takes over this task. This works fine, we can get IP form the dhcp server and then take part in network communication based on IPv4. Problem: The dhcp client cannot be stopped any time. Any shell commands will never reach the dhcpcd task, but rather it will instantiate a new dhcpd task which will take over the current task. E.g: in rtems console if you enter dhcpd -x, instead of killing the dhcpcd task a new dhcpcd task will be created which will run from the console. This means that the console cannot be used anymore. There seems to be noway to communicate with the already created dhcpcd task. I took a look over the dhcpcd in libbsd code and some of the features are turned off. Can someone give more details regarding this ?
#3910	4 years ago	fixed	bsps	Sebastian Huber	Sebastian Huber	9 months ago
Summary	Add BSP for STM32H7
Description	Add a BSP for the ​STM32H7 series. Import the HAL from ​https://github.com/STMicroelectronics/stm32h7xx_hal_driver. Add a VERSION file to allow future updates similar to ​https://git.rtems.org/rtems/tree/cpukit/dtc/VERSION. Add standard clock and console drivers.
#3924	4 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	20 months ago
Summary	Test the application configuration options
Description	There exists a specification of the application configuration options (work in progress) here: ​https://git.rtems.org/sebh/rtems-qual.git/tree/spec/acfg/opt/ For the RTEMS Qualification project we have to test the implementation according to the specification (validation and unit tests). A test plan needs to be developed. The tests should use the RTEMS Test Framework. Add C preprocessor error tests to the new build system, e.g. test #error and #warning directives.
#3925	4 years ago	fixed	bsps	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Optimize work area initialization
Description	The new work area initialization introduced by #3838 introduces a dependency on _Heap_Extend() for all BSPs. The BSP knows how many memory areas it delivers to the system. It should provide also the RTEMS Workspace and C Program Heap initialization functions to the system so that an unnecessary pull in of _Heap_Extend() can be avoided.
#3929	4 years ago	fixed	doc	Gedare Bloom	Needs Funding	17 months ago
Summary	Document use of virtualenv
Description	The use of python virtualenv to build docs or use RSB has been suggested a few times. We should provide a brief sketch in our documentation about that utility. For example, in modern versions of Arch Linux, python is now a shortcut to python3 from the python package. For tools that do not yet function in Python 3, users can set up a virtual environment with Python 2 as follows: At a root terminal, install python2-virtualenv with pacman. # pacman -S python2-virtualenv Examples of other architectures and uses for virtualenv would round out a reasonable section in the user manual for Preparation.
#3935	4 years ago	fixed	unspecified	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Remove use of RTEMS_INLINE_ROUTINE in the RTEMS sources
Description	Replace all uses of RTEMS_INLINE_ROUTINE with "static inline". This makes it a bit easier for third-party tools to parse the RTEMS sources.
#3937	4 years ago	fixed	tool/gcc	Sebastian Huber	Sebastian Huber <sebastian.huber@…>	2 years ago
Summary	Move content of bsp_specs to GCC
Description	Move content of the bsp_specs to GCC 10. Once this is done, update the tools and delete the bsp_specs. On powerpc this means we have to move the content of rtems_crti.S to GCC. It may not be possible to back port this to GCC 10 and we may have to maintain this as a patch until GCC 11 is released. Add support for ​nostartfiles and ​nodefaultlibs.
#3941	4 years ago	fixed	arch/epiphany	Sebastian Huber	Hesham Almatary	4 years ago
Summary	Remove epiphany architecture support
Description	I tried to build the epiphany BSP with GCC 10 recently and a lot internal compiler errors appeared. If it is not possible to build the BSP with a released GCC 10.1, then I suggest to remove this architecture. The GDB was never supported by the FSF upstream. To me this looks like an unmaintained architecture.
#3942	4 years ago	fixed	arch/mips	Sebastian Huber	Joel Sherrill	17 months ago
Summary	multiple definition of `pci_indirect_functions'
Description	Building with GCC 10 showed a multiple definition error: /build/rtems/6/lib/gcc/mips-rtems6/10.0.1/../../../../mips-rtems6/bin/ld: ./librtemsbsp.a(i8259.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: multiple definition of `pci_indirect_functions'; ./librtemsbsp.a(vectorisrs.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: first defined here /build/rtems/6/lib/gcc/mips-rtems6/10.0.1/../../../../mips-rtems6/bin/ld: ./librtemsbsp.a(simple_access.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: multiple definition of `pci_indirect_functions'; ./librtemsbsp.a(vectorisrs.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: first defined here /build/rtems/6/lib/gcc/mips-rtems6/10.0.1/../../../../mips-rtems6/bin/ld: ./librtemsbsp.a(pci.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: multiple definition of `pci_indirect_functions'; ./librtemsbsp.a(vectorisrs.c.19.o):/ramdisk/git-rtems-5/build/mips/malta/../../../bsps/mips/malta/include/bsp/pci.h:113: first defined here
#3947	4 years ago	fixed	lib	Joel Sherrill	Sebastian Huber	21 months ago
Summary	fdt_rw.c: Unchecked return value (CID #1047324)
Description	git blams says Sebastian. *** CID 1047324: Error handling issues (CHECKED_RETURN) /home/joel/rtems-work/rtems/cpukit/dtc/libfdt/fdt_rw.c: 351 in fdt_add_subnode_namelen() 345 if (offset >= 0) 346 return -FDT_ERR_EXISTS; 347 else if (offset != -FDT_ERR_NOTFOUND) 348 return offset; 349 350 /* Try to place the new node after the parent's properties */ >>> CID 1047324: Error handling issues (CHECKED_RETURN) >>> Calling "fdt_next_tag" without checking return value (as is done elsewhere 7 out of 8 times). 351 fdt_next_tag(fdt, parentoffset, &nextoffset); /* skip the BEGIN_NODE */ 352 do { 353 offset = nextoffset; 354 tag = fdt_next_tag(fdt, offset, &nextoffset); 355 } while ((tag == FDT_PROP) || (tag == FDT_NOP)); 356
#3951	4 years ago	fixed	arch/powerpc	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Remove obsolete powerpc BSPs
Description	Remove powerpc BSPs for unsupported ISA variants (e.g. SPE and the embedded FPU). Remove powerpc BSPs for dead projects. This includes: brs5l brs6l dp2 mpc5566evb_spe mpc5643l_dpu mpc5643l_evb mpc5674f_ecu508_app mpc5674f_ecu508_boot mpc5674fevb_spe mpc5674f_rsm6
#3957	4 years ago	fixed	doc	Joel Sherrill		16 months ago
Summary	Review wiki/Developer for deletion
Description	​https://devel.rtems.org/wiki/Developer has minimal content. All should likely point to somewhere in the Users Guide now rather than the Wiki. Also this page is a candidate for deletion. Verify the references and delete it. Check rtems.org also.
#3959	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add rtems_task_construct()
Description	In RTEMS 5 a lot of development was done to allow a static allocation of resources provided by RTEMS and avoid the use of a heap. One missing piece is the support for statically allocated RTEMS tasks. The rtems_task_create() function allocates the task stack from a heap. We need an alternative function which allocates an RTEMS tasks with a user provided stack (similar to the POSIX threads). There does need to be documentation on the allocation strategies, when to use, limitations, and particularly the use of rtems-exeinfo to get the TLS size. This is one case for sure where an RTEMS Tool needs explicit mention in both the API and configure option sections. Proposed API: /** * @brief This structure defines the configuration of a task constructed by * rtems_task_construct(). */ typedef struct { /** * @brief This member defines the name of the task. */ rtems_name name; /** * @brief This member defines the initial priority of the task. */ rtems_task_priority initial_priority; /** * @brief This member shall point to the task storage area begin. * * The task storage area will contain the task stack, the thread-local storage, * and the floating-point context on architectures with a separate * floating-point context. * * The task storage area begin address and size should be aligned by * #RTEMS_TASK_STORAGE_ALIGNMENT. To avoid memory waste, use RTEMS_ALIGNED() * and #RTEMS_TASK_STORAGE_ALIGNMENT to enforce the recommended alignment of a * statically allocated task storage area. */ void *storage_area; /** * @brief This member defines size of the task storage area in bytes. * * Use the RTEMS_TASK_STORAGE_SIZE() macro to determine the recommended task * storage area size. */ size_t storage_size; /** * @brief This member defines the maximum thread-local storage size supported * by the task storage area. * * Use RTEMS_ALIGN_UP() and #RTEMS_TASK_STORAGE_ALIGNMENT to adjust the size to * meet the minimum alignment requirement of a thread-local storage area used * to construct a task. * * If the value is less than the actual thread-local storage size, then the * task construction by rtems_task_construct() fails. * * If the is less than the task storage area size, then the task construction * by rtems_task_construct() fails. * * The actual thread-local storage size is determined when the application * executable is linked. The ``rtems-exeinfo`` command line tool included in * the RTEMS Tools can be used to obtain the thread-local storage size and * alignment of an application executable. * * The application may configure the maximum thread-local storage size for all * threads explicitly through the #CONFIGURE_MAXIMUM_THREAD_LOCAL_STORAGE_SIZE * configuration option. */ size_t maximum_thread_local_storage_size; /** * @brief This member defines the optional handler to free the task storage * area. * * It is called on exactly two mutually exclusive occasions. Firstly, when the * task construction aborts due to a failed task create extension, or secondly, * when the task is deleted. It is called from task context under protection * of the object allocator lock. It is allowed to call free() in this handler. * If handler is NULL, then no action will be performed. */ void ( *storage_free )( void * ); /** * @brief This member defines the initial modes of the task. */ rtems_mode initial_modes; /** * @brief This member defines the attributes of the task. */ rtems_attribute attributes; } rtems_task_config; /** * @brief Constructs a task from the specified the task configuration. * * In contrast to tasks created by rtems_task_create(), the tasks constructed * by this directive use a user-provided task storage area. The task storage * area contains the task stack, the thread-local storage, and the * floating-point context on architectures with a separate floating-point * context. * * This directive is intended for applications which do not want to use the * RTEMS Workspace and instead statically allocate all operating system * resources. It is not recommended to use rtems_task_create() and * rtems_task_construct() together in an application. It is also not * recommended to use rtems_task_construct() for drivers or general purpose * libraries. The reason for these recommendations is that the task * configuration needs settings which can be only given with a through * knowledge of the application resources. * * An application based solely on static allocation can avoid any runtime * memory allocators. This can simplfiy the application architecture as well * as any analysis that may be required. * * The stack space estimate done by <rtems/confdefs.h> assumes that all tasks * are created by rtems_task_create(). The estimate can be adjusted to take * user-provided task storage areas into account through the * #CONFIGURE_MINIMUM_TASKS_WITH_USER_PROVIDED_STORAGE application * configuration option. * * The #CONFIGURE_MAXIMUM_TASKS should include tasks constructed by * rtems_task_construct(). * * @param config is the task configuration. * * @param[out] id is the pointer to an object identifier variable. The * identifier of the constructed task object will be stored in this variable, * in case of a successful operation. * * @retval ::RTEMS_SUCCESSFUL The requested operation was successful. * * @retval ::RTEMS_INVALID_ADDRESS The id parameter was NULL. * * @retval ::RTEMS_INVALID_NAME The task name was invalid. * * @retval ::RTEMS_INVALID_PRIORITY The initial task priority was invalid. * * @retval ::RTEMS_INVALID_SIZE The thread-local storage size is greater than * the maximum thread-local storage size specified in the task configuration. * The thread-local storage size is determined by the thread-local variables * used by the application and #CONFIGURE_MAXIMUM_THREAD_LOCAL_STORAGE_SIZE. * * @retval ::RTEMS_INVALID_SIZE The task storage area was too small to provide * a task stack of the configured minimum size, see * #CONFIGURE_MINIMUM_TASK_STACK_SIZE. The task storage area contains the * task stack, the thread-local storage, and the floating-point context on * architectures with a separate floating-point context. * * @retval ::RTEMS_TOO_MANY There was no inactive task object available to * construct a task. * * @retval ::RTEMS_TOO_MANY In multiprocessing configurations, there was no * inactive global object available to construct a global task. * * @retval ::RTEMS_UNSATISFIED One of the task create extensions failed during * the task construction. * * @retval ::RTEMS_UNSATISFIED In SMP configurations, the non-preemption mode * was not supported. * * @retval ::RTEMS_UNSATISFIED In SMP configurations, the interrupt level mode * was not supported. */ rtems_status_code rtems_task_construct( const rtems_task_config *config, rtems_id *id );
#3962	4 years ago	fixed	doc	Joel Sherrill	Gedare Bloom	4 years ago
Summary	Licensing Requirements for Submissions Poorly Documented
Description	The discussion of licensing for code/docs in RTEMS is a bit of a mess right now. After posting this issue to the mailing list (​https://lists.rtems.org/pipermail/devel/2020-April/059489.html), Sebastian pointed out where the actual recommended text for code/docs was located. SW Engineering Guide sections 6.3.1.2 (Licenses) and 11 (Licensing Requirements) are somewhat overlapping and duplicative. In addition, Gedare's old blog ​http://gedare-csphd.blogspot.com/2013/05/software-licenses-with-rtems.html is still a decent discussion of the underlying issues and goals even though we have moved to BSD-2 and Creative Commons for documentation since he wrote that in 2013. The above references just describe the rationale and cite the licenses. For an example of what to put in the code, you have to visit ​https://docs.rtems.org/branches/master/eng/coding-file-hdr.html#copyright-and-license-block I THINK a good solution would be to reference the Chapter 11 Licensing Requirements from the Coding Conventions chapter. And incorporate an updated version of Gedare's blog text.
#3977	4 years ago	invalid	test	Joel Sherrill	joel@…	17 months ago
Summary	Add unexplained-failure as option for Test Configuration Files
Description	RTEMS has many BSPs and it is a large amount of work to investigate each test execution that has an unexpected result. The tcfg files were initially created to just avoid trying build tests where the executable did not fit into target memory. That has grown to incorporate knowledge about which tests require human input and are not automated. This ticket is to add a new test state for "unexplained failure". If a test has been investigated and has an explanation, it is a known failure. This could be associated with a ticket (e.g. dynamic loading or fenv not supported on an architecture) or just a comment explaining it (e.g. mips jmr3904 simulator trapping invalid memory accesses before the trap handler is tripped). But many tests are currently just marked as failures because they do fail but no one has investigated and explained it. The addition of "unexplained failure" is to capture those. This will make it possible for someone looking to know they need investigating but are known to fail. Marking these are "known failure" permanently puts them in a category where they will not get looked at again. This has two side-effects. The first is that we can have test results with no "unexpected failures." They are either known and explained or known and need investigation. This helps anyone looking at test results since there should be no unexpected failures. The second side-effect is that it hopefully captures the set of repeatedly failing tests and encourages us to investigate and file tickets or explain. The milestone for this is NOT 5.x because this is moving the goalpost for test results. Release 5.x is the first with the body of test results data we have. The build@ mailing list has them and can be compared by any user interested. We also do not have a way to machine check results yet which would ease checking the test results for deviations. Note: As an odd case, there are multiple BSPs where the executables can run on hardware or 1+ simulators. Sometimes the results vary across the target environment even for the same executable. This is at least something to consider as we want to have a trustworthy and checkable set of test results available.
#3978	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Move _RTEMS_version to implementation header file
Description	The global variable _RTEMS_version was declared via <rtems.h> and may be used directly by application code. Applications should use rtems_get_version_string() instead. Move it to an implementation header file in RTEMS 6.
#3979	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove RTEMS_MAXIMUM_NAME_LENGTH
Description	Remove this deprecated define.
#3980	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated <rtems/system.h>
Description	This header file is included by <rtems.h> and contains only deprecated or internal content. Remove this deprecated header file.
#3981	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove _Copyright_Notice from API header file
Description	Remove _Copyright_Notice from API header file. Applications should use rtems_get_copyright_notice() which is available since RTEMS 5.
#3982	4 years ago	fixed	test	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Test failures on BSPs with sbrk() support
Description	There are some test failures on BSPs with sbrk() support such as psim, for example psxpipe01: #0 rtems_heap_extend_via_sbrk (heap=0x30468 <_Malloc_Heap>, alloc_size=5) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libcsupport/src/rtems_heap_extend_via_sbrk.c:41 #1 0x00005b48 in rtems_heap_allocate_aligned_with_boundary (size=size@entry=5, alignment=alignment@entry=0, boundary=boundary@entry=0) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libcsupport/src/malloc_deferred.c:116 #2 0x00005930 in malloc (size=size@entry=5) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libcsupport/src/malloc.c:39 #3 0x0001b720 in _malloc_r (ignored=ignored@entry=0x2fa68 <_RTEMS_tasks_Objects+736>, size=size@entry=5) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libcsupport/src/_malloc_r.c:26 #4 0x00015ff4 in _strdup_r (reent_ptr=0x2fa68 <_RTEMS_tasks_Objects+736>, str=str@entry=0x1ce94 "/tmp") at ../../../../../../gcc-7.5.0/newlib/libc/string/strdup_r.c:10 #5 0x00015fbc in strdup (str=str@entry=0x1ce94 "/tmp") at ../../../../../../gcc-7.5.0/newlib/libc/string/strdup.c:10 #6 0x00011470 in rtems_mkdir (path=path@entry=0x1ce94 "/tmp", mode=mode@entry=511) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libcsupport/src/rtems_mkdir.c:133 #7 0x0000a7c8 in pipe (filsdes=filsdes@entry=0x36ce8) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/libfs/src/pipe/pipe.c:42 #8 0x000004a0 in Init (ignored=<optimized out>) at /home/EB/sebastian_h/src/rtems/c/src/../../testsuites/psxtests/psxpipe01/init.c:65 #9 0x0000c828 in _Thread_Handler () at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/score/src/threadhandler.c:139 #10 0x0000c77c in _Thread_Get (id=<optimized out>, lock_context=<optimized out>) at /home/EB/sebastian_h/src/rtems/c/src/../../cpukit/score/src/threadget.c:43 The problem is that an rtems_heap_greedy_allocate() is ineffective due to the automatic heap extension.
#3988	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_extension
Description
#3989	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_get_current_processor()
#3990	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated rtems_get_processor_count()
#3991	4 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Remove deprecated Thread typedef
#3993	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	19 months ago
Summary	Specify the RTEMS Classic API
Description	For the RTEMS Qualification project we need a specification of the RTEMS Classic API. This specification can be used to generate an Interface Control Document (ICD), header files with Doxygen markup and content for the RTEMS Classic API Guide. The RTEMS Classic API is mainly defined in the following header files: <rtems.h> <rtems/config.h> <rtems/extension.h> <rtems/fatal.h> <rtems/init.h> <rtems/io.h> <rtems/rtems/barrier.h> <rtems/rtems/cache.h> <rtems/rtems/clock.h> <rtems/rtems/config.h> <rtems/rtems/dpmem.h> <rtems/rtems/event.h> <rtems/rtems/intr.h> <rtems/rtems/message.h> <rtems/rtems/modes.h> <rtems/rtems/mp.h> <rtems/rtems/object.h> <rtems/rtems/options.h> <rtems/rtems/part.h> <rtems/rtems/ratemon.h> <rtems/rtems/region.h> <rtems/rtems/sem.h> <rtems/rtems/signal.h> <rtems/rtems/status.h> <rtems/rtems/support.h> <rtems/rtems/tasks.h> <rtems/rtems/timer.h> <rtems/rtems/types.h> <rtems/score/stack.h> Examples for interface specification items are available here: ​https://git.rtems.org/sebh/rtems-qual.git/tree/spec/if/rtems/status You find an example for a generated header file here: ​https://git.rtems.org/sebh/rtems.git/tree/cpukit/include/rtems/rtems/status.h?h=qual The header files are generated by this script: ​https://git.rtems.org/sebh/rtems-qual.git/tree/spec2rtems.py As preliminary work, remove all RTEMS APIs deprecated in RTEMS 5.1. Use and improve the ​Interface Specification How-To during the work on this ticket. Carry out the following tasks for each header file: Add a interface specification item for each variable, typedef, enum, enumerator, define, macro, function, struct, and union. For each item analyse the change history using git log -p --follow to determine the copyright holders of the Doxygen markup. Use only Doxygen markup for the specification items from contributors which agreed on a dual-licensing (CC-BY-SA-4.0 OR BSD-2-Clause) in #3899. Use this information to provide copyright statements for the items. The API parts alone (e.g. function names, parameter names, type names, etc.) are not considered to be copyrightable information. Use the RTEMS documentation sources to document the API. Use only content from contributors which agreed on a dual-licensing (CC-BY-SA-4.0 OR BSD-2-Clause) in #3899. Add unspecified interface items for interfaces to the implementation, C language, compiler, user provided defines, build system provided defines, etc. Run the RTEMS test suite with all tests enabled and check for regressions (in particular new warnings). Create a commit for each header file and send it to the RTEMS mailing list for review.
#3994	4 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Generate Doxygen markup for the application configuration options
Description	The application configuration options are now available as specification items. From these items, the documentation in the RTEMS Classic API Guide is generated. The same approach can be used to generate Doxygen markup.
#3996	4 years ago	fixed	arch/powerpc	Sebastian Huber	Joel Sherrill	17 months ago
Summary	fenv support is missing on powerpc
Description	The test program psxfenv01 fails on powerpc.
#3997	4 years ago	fixed	arch/sparc	Sebastian Huber	Joel Sherrill	17 months ago
Summary	fenv support is missing on sparc
Description	The test program psxfenv01 fails on sparc.
#3998	4 years ago	fixed	arch/arm	Sebastian Huber	Joel Sherrill	17 months ago
Summary	fenv support is missing on arm
Description	The test program psxfenv01 fails on arm.
#4007	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add rtems_message_queue_construct()
Description	In RTEMS 5 a lot of development was done to allow a static allocation of resources provided by RTEMS and avoid the use of a heap. Similar to #3959 a new directive to build Classic API message queues without using the workspace is proposed: /** * @brief The message queue configuration to construct a message queue. */ typedef struct { /** * @brief The name of the message queue. */ rtems_name name; /** * @brief The count of the provided buffers. * * This value defines the maximum number of pending messages. */ size_t buffer_count; /** * @brief The size of the provided buffers in bytes. */ size_t buffer_size; /** * @brief The array of count message buffers. * * Each buffer shall have the size of the buffer_size member. The maximum * message size is buffer_size minus sizeof( rtems_message_queue_buffer ). */ rtems_message_queue_buffer *buffers; /** * @brief The attribute set of the message queue. */ rtems_attribute attributes; } rtems_message_queue_config; /** * @brief Constructs a message queue according to the specified configuration. * * @param config The message queue configuration. * @param[out] id The message queue identifier of the new message queue. * * @retval RTEMS_SUCCESSFUL Successful operation. * @retval RTEMS_INVALID_ADDRESS The id parameter is @c NULL. * @retval RTEMS_INVALID_NAME The message queue name is invalid. * @retval RTEMS_TOO_MANY No message queue is available. */ rtems_status_code rtems_message_queue_construct( const rtems_message_queue_config *config, rtems_id *id );
#4018	4 years ago	fixed	admin	Kinsey Moore	Kinsey Moore <kinsey.moore@…>	4 years ago
Summary	libcpusupport: Architecture defines are present outside CPU port
Description	The handler code for global constructors (threadhandler.c) and destructors (newlibc_exit.c) uses compiler-provided architecture defines which should be limited to the CPU port as they are in other locations.
#4020	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Change all version numbers to RTEMS 6 on the master branches
#4025	4 years ago	fixed	network/libbsd	Sebastian Huber	Sebastian Huber	2 years ago
Summary	NFS Client is broken on 64-bit targets
Description	Cloned from #4024.
#4029	4 years ago	fixed	arch/i386	Jan Sommer	Jan Sommer <jan.sommer@…>	4 years ago
Summary	i386: SMP-System hangs with non-consecutive APIC IDs
Description	If a processor enumerates its cores non-consecutively (e.g. 0,2,4,8 for a tested Intel Atom) the target cpu for IPIs is mapped wrongly.
#4031	4 years ago	fixed	arch/i386	Jan Sommer	Jan Sommer <jan.sommer@…>	4 years ago
Summary	i386: ISR can overwrite its own stack during system initialization (cloned)
Description	Cloned from #4030: During testing the rtems-libbsd examples, we experienced GP exceptions from within the ISR from time to time during initalization. When the init task is restored for the first time and the a pending interrupt is available, an ISR could overwrite its own return address if it is spawned between restoring the eflags register and restoring the esp register.
#4032	4 years ago	fixed	lib	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Make deferred free in malloc() support optional
Description	The free() implementation in RTEMS supports that allocated memory is freed in interrupt context. Since the heap is protected by a mutex, the frees issued in interrupt context cannot immediately be freed to the heap, instead they are placed on a deferred free list. This list is emptied by the core allocation function rtems_heap_allocate_aligned_with_boundary(). This adds a dependency to free() in rtems_heap_allocate_aligned_with_boundary(). In order to better support applications which only allocate memory an never free it, this dependency should be avoided. This can be done with a weak default implementation of _Malloc_Process_deferred_frees(). In the free() module a strong implementation can be provided.
#4034	4 years ago	fixed	lib	Sebastian Huber	Sebastian Huber	4 years ago
Summary	Add rtems_interrupt_server_create() and rtems_interrupt_server_destroy() (cloned)
Description	Cloned from #4033. In addition to the implementation add also unit tests.
#4036	4 years ago	fixed	tool/rsb	Chris Johns	Chris Johns	4 years ago
Summary	Track dependency command for the RSB
Description	Add a command to track dependencies in the RSB. Scan all the bset files and see what they include. Build a list of dependencies and also a list of those configs that are not referenced. The configs not references can be removed. Errors in build sets can result in an invalid list so the scan needs to be clean. This is difficult when some build sets like rtems-defaults.bset are designed to be included and not used from the command line. As a result it generates an error. A list of these could be maintained or a tag added to the bset file to indicate it is not a top level build set file. Removal of unreferenced configs can occur in the RBS because we use a branch model for releases.
#4037	4 years ago	fixed	tool	Chris Johns	Chris Johns	17 months ago
Summary	Python script distribution standardisation
Description	The RSB and Tools wrap python scripts in a shell script based wrapper that detects the installed version and uses it. Other python scripts in RTEMS do not. I would like to have a consistent model for all scripts. ​ PEP0394 discusses the issues and says a distro can decide not to provide a python command by default. It also says a virtualenv can be used to provide a consistent default environment. Do we assume #! /usr/bin/env python will always work in our script and we make sure we are python2 and python3 clean for user installed commands? If a user does not have a python command installed by default do we document how to install one or do we document using a virtualenv? A simple shebang is cleaner for us to maintain however it added an extra dependency we need to document. Note, development script and tools should follow the same model we decide on however they only need to support python3.
#4039	4 years ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	4 years ago
Summary	arm/atsam/SC16IS752: Make interrupt server configurable (cloned)
Description	Cloned from #4038.
#4040	4 years ago	fixed	test	Joel Sherrill	joel@…	4 years ago
Summary	coverhd.h remnants left to remove
Description	coverhd.h has been removed but there are as least two remnants to address: tmoverhd is not needed and still is present. (done) m68k/mcf52235/README references coverhd.h $ grep -r coverhd .. ../rtems/bsps/m68k/mcf52235/README:*) Update the coverhd.h (calling overheads) page 21 of the BSP guide ../rtems/bsps/m68k/mcf5225x/README:*) Update the coverhd.h (calling overheads) page 21 of the BSP guide
#4043	4 years ago	fixed	doc	Chris Johns	Sebastian Huber	4 years ago
Summary	Update software eng manual to rtems-central
Description	​https://docs.rtems.org/branches/master/eng/req/howto.html Update to rtems-central
#4044	4 years ago	duplicate	tool/rsb	Chris Johns	Chris Johns	4 years ago
Summary	Add RSB Track command
Description	Add a comment to track the RSB build set and configuration files.
#4045	4 years ago	fixed	tool/rsb	Chris Johns	Chris Johns	4 years ago
Summary	Remove RSB sb-builder command
Description	This command is now always wrapped bu the set builder. There is no need to use the command line with configuration files.
#4046	4 years ago	fixed	tool/rsb	Chris Johns	Chris Johns	4 years ago
Summary	Remove RSB sb-bootstrap command
Description	Bootstrapping is going away with the new build system and until then the command in rtems.git should be used.
#4048	4 years ago	wontfix	tool/rsb	Chris Johns	Chris Johns	17 months ago
Summary	Move RTEMS 5 BSP and package build sets and configs to RTEMS 6
Description	Move the BSP build sets with the related packages to RTEMS 6. Update the kernel and libbsd to point to RTEMS 6 commits. Drop the tier build sets as these are not being used. They can be brought back if we need them.
#4052	4 years ago	wontfix	network/libbsd	Jan Sommer		17 months ago
Summary	[libbsd] e1000 driver not working for i386 based BSP in current rtems-libbsd master
Description	Current rtems-libbsd master branch fails to compile the e1000 driver for i386 based BSPs. The corresponding patches from the 5-freebsd-12 branch need to be ported to master.
#4054	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	RTEMS_BARRIER_AUTOMATIC_RELEASE and RTEMS_BINARY_SEMAPHORE options have the same value
Description	The RTEMS_BARRIER_AUTOMATIC_RELEASE and RTEMS_BINARY_SEMAPHORE options have the same value. In order to better detect a misuse of option values (for example using RTEMS_BINARY_SEMAPHORE for rtems_barrier_create()), the options should have unique values.
#4055	4 years ago	fixed	bsps	Jan Sommer	Jan Sommer <jan.sommer@…>	4 years ago
Summary	bsps/xilinx-zynq: Flush TX-Buffer before initializing the zynq-uart
Description	We experienced that u-boot (at least from Xilinx repositories) does not wait until all its output has left the TX Buffer of the stdout uart, before handing over to the RTEMS application This causes some garbage output at the begin of the RTEMS application in some cases and corrupts the test begin marker prohibiting rtems-test to recognize the test correctly. Proposed fix is to let RTEMS wait until the TX Buffer of the uart is empty before configuring and using it.
#4061	4 years ago	fixed	config	Sebastian Huber	Sebastian Huber	4 years ago
Summary	The reworked <rtems/confdefs.h> has a cyclic dependency with RTEMS_MULTIPROCESSING enabled (cloned)
Description	Cloned from #4060: The changes for #3875 introduced a cyclic dependency between <rtems/confdefs/mpci.h> and <rtems/confdefs/threads.h>.
#4063	4 years ago	fixed	test	Chris Johns	Chris Johns	4 years ago
Summary	Change to a simpler TFTP server for rtems-test
Description	Add a simpler TFTP server that can have multiple instances on different ports running at once. Allowing more than one TFTP server to run at once is the first step to letting rtems-test run parallel hardware test jobs if the user has more one piece of hardware available for testing. Parallel hardware testing will speed up the time it takes to run the tests on target hardware. This feature combined with the rtems-tftp-proxy lets a standard bootloader configuration be integrated into a parallel hardware test set up.
#4065	4 years ago	fixed	tool	Heinz Junkes	Joel Sherrill <joel@…>	17 months ago
Summary	Typo in rtems-boot.ini "ubootdior" should be "ubootdir"
Description	Typo in rtems-tools misc/tools/config/rtems-boot.ini diff rtems-boot.ini ./tools/misc/tools/config/rtems-boot.ini 55c55 < first_stage = %{ubootdir}/spl/boot.bin --- first_stage = %{ubootdior}/spl/boot.bin
#4066	4 years ago	fixed	test	Chris Johns	Chris Johns	4 years ago
Summary	test running for too long
Description	The spcache01 test on my PC test hardware is not finishing. It is looping generating output. Add a Test too long error count and end the test after 5 minutes.
#4067	4 years ago	fixed	network/libbsd	Chris Johns	Chris Johns	4 years ago
Summary	Libbsd build system does not separate kernel and user land include paths
Description	FreeBSD has duplicates kernel and user land header files. LibBSD build system currently supports only one set of header include and the kernel and user land include paths.
#4069	4 years ago	fixed	tool	Chris Johns	Hesham Almatary <Hesham.Almatary@…>	17 months ago
Summary	dl06 does not link on RISCV
Description	rtems-ld -r /home/jiri/src/rtems/riscvmp/riscv-rtems6/c/griscv \ -C riscv-rtems6-gcc -c "-march=rv32imafd -mabi=ilp32d" \ -O rap -b dl06.pre -e rtems_main -s \ -o dl06.rap dl06-o1.o dl06-o2.o -lm error: rap::object: Section index '0' not found: dl06-o1.o Makefile:8528: recipe for target 'dl06.rap' failed make[5]: *** [dl06.rap] Error 10
#4070	4 years ago	fixed	arch/riscv	Joel Sherrill		2 years ago
Summary	RISC-V Multiple Definition Error (many BSPs)
Description	This error appears to occur on all RISC-V BSPs which do not fail at the dynamic loading rap section 0 error. I assume that eventually they may all hit this. List of BSPs failing below message. /home/joel/rtems-cron-6/tools/6/lib/gcc/riscv-rtems6/10.2.1/../../../../riscv-rtems6/bin/ld: ./../../lib/libbsp/riscv/riscv/librtemsbsp.a(htif.o):/home/joel/rtems-cron-6/rtems/bsps/riscv/riscv/include/dev/serial/htif.h:48: multiple definition of `htif_console_handler'; ./../../lib/libbsp/riscv/riscv/librtemsbsp.a(console-config.o):/home/joel/rtems-cron-6/rtems/bsps/riscv/riscv/include/dev/serial/htif.h:48: first defined here /home/joel/rtems-cron-6/tools/6/lib/gcc/riscv-rtems6/10.2.1/../../../../riscv-rtems6/bin/ld: ./../../lib/libbsp/riscv/riscv/librtemsbsp.a(htif.o):/home/joel/rtems-cron-6/rtems/bsps/riscv/riscv/include/dev/serial/htif.h:48: multiple definition of `htif_console_handler'; ./../../lib/libbsp/riscv/riscv/librtemsbsp.a(console-config.o):/home/joel/rtems-cron-6/rtems/bsps/riscv/riscv/include/dev/serial/htif.h:48: first defined here mv -f cdtest/.deps/cdtest-init.Tpo cdtest/.deps/cdtest-init.Po rv32iac_spike rv32imac_spike rv32imafc_spike rv32imafdc_spike rv32imafd_spike rv32im_spike rv32i_spike rv64imac_medany_spike rv64imac_spike rv64imafdc_medany_spike rv64imafdc_spike rv64imafd_medany_spike rv64imafd_spike
#4074	4 years ago	fixed	config	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add CONFIGURE_MAXIMUM_THREAD_LOCAL_STORAGE_SIZE
Description	This configuration option can be used to reserve space for the dynamic linking of modules with thread-local storage objects. See also: ​https://lists.rtems.org/pipermail/devel/2020-September/061867.html
#4076	4 years ago	fixed	arch/i386	Gedare Bloom	Gedare Bloom	4 years ago
Summary	i386: bad asm in smp mode
Description	A note for me (or someone) to make the fix outlined on the mailing list. cpukit/score/cpu/i386/include/rtems/asm.h:157   movb imps_apic_cpu_map(\REG),\REG        /* CPU ID in REG */ The assembler is right to complain. movb has to target one of the 1-byte mnemonics, so this should be %al for the LSB of %eax. One needs to check through this logic carefully, but I think the right thing to do here is: movzbl     imps_apic_cpu_map(\REG),\REG That should do the trick. Can you test it locally? Sure! Builds fine and testsuite testing reveals:
#4077	4 years ago	fixed	network/libbsd	Chris Johns	Chris Johns	4 years ago
Summary	Implement LibBSD module dependencies in the build system
Description	The build system has a partially implemented module dependency system. Complete the implementation. Control tests being built. If a test depends on a module and the module is disabled do not build the test. Add dependency checking to the freebsd-to-rtems.py command. If the module definition is not consistent generate an error.
#4078	4 years ago	fixed	arch/powerpc	Sebastian Huber	Sebastian Huber	20 months ago
Summary	GCC C++ compiler for powerpc defines _GNU_SOURCE as a built-in
Description	For example: powerpc-rtems6-g++ -E -P -v -dD empty.cc 2>&1 | grep GNU_SOURCE Yields: /opt/rtems/6/libexec/gcc/powerpc-rtems6/10.2.1/cc1plus -E -quiet -v -P -D_GNU_SOURCE -Dmpc750 empty.cc -dD #define _GNU_SOURCE 1 This is a GCC configuration bug. The reason is that the RTEMS GCC configuration includes gnu-user.h.
#4079	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber <sebastian.huber@…>	4 years ago
Summary	build: i386/pc686/appstart.o is created more than once
Description	In this build specification item ​https://git.rtems.org/rtems/tree/spec/build/bsps/i386/pc386/objsmpstart.yml is an error: * Node /home/EB/sebastian_h/git-rtems-6/build/i386/pc686/appstart.o is created more than once. The task generators are: 1. bld(target=['appstart.o'], idx=14, tg_idx_count=14, meths=['check_err_features', 'check_err_order', 'process_rule', 'process_source'], rule='${OBJCOPY} -I binary -O elf32-i386 -B i386 ${SRC} ${TGT}', _name='appstart.o', source=[], path=/home/EB/sebastian_h/git-rtems-6, posted=True, features=[]) in /home/EB/sebastian_h/git-rtems-6 2. bld(target='/bsps/i386/pc386/objsmpstart', idx=15, tmp_use_objects=False, tg_idx_count=15, meths=['check_err_features', 'check_err_order', 'process_rule', 'process_source'], tmp_use_stlib=True, _name='/bsps/i386/pc386/objsmpstart', source=[/home/EB/sebastian_h/git-rtems-6/build/i386/pc686/appstart.o], tmp_use_var='', path=/home/EB/sebastian_h/git-rtems-6, posted=True, typ='objects', features=[]) in /home/EB/sebastian_h/git-rtems-6 If you think that this is an error, set no_errcheck_out on the task instance All what we have to do is to add a specially produced appstart.o to librtemsbsp.a.
#4080	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	3 years ago
Summary	build: Add start.o dependency to the executable link step
Description	RTEMS is a real-time operating system and the applications using RTEMS end up in a single executable which contains also the full system initialization. The linker uses a start file to determine what should be linked into the executable. The start file is also built by the build system, however, it is only indirectly referenced via a linker script, for example: ​https://git.rtems.org/rtems/tree/bsps/arm/shared/start/linkcmds.base#n28 So, every executable depends on the start file (start.o usually). If start.o changes, then the executable needs to be re-linked. Also the start file needs the name expected by the linker script, for example not start.S.17.o. The start files are build with this commands: ​https://git.rtems.org/rtems/tree/wscript#n535 ​https://git.rtems.org/rtems/tree/wscript#n244 Somehow we have to add a dependency on start.o to every executable. To experiment with the build system use the following commands to set it up: cd $HOME mkdir rtems-sdk mkdir rtems-sdk/src cd rtems-sdk/src git clone git://git.rtems.org/rtems.git git clone git://git.rtems.org/rtems-source-builder.git rsb cd rsb/rtems ../source-builder/sb-set-builder --prefix=$HOME/rtems-sdk/6 6/rtems-sparc cd ../../rtems echo "[sparc/leon3]" > config.ini ./waf configure --prefix=$HOME/rtems-sdk/6 ./waf The host must be capable to build GCC and needs the Python development support (header files) installed.
#4081	4 years ago	fixed	build	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Remove Autoconf/Automake? based build system
Description	Remove Autoconf/Automake? based build system after a testing phase of the waf based build system. The planned date to remove the old build system is 1st November 2020. For a status of the new build system see: ​https://devel.rtems.org/wiki/Release/6/Waf%20BSP%20Checklist
#4084	4 years ago	fixed	tool/gcc	Joel Sherrill		4 years ago
Summary	throwing std::runtime on PC BSP fails
Description	Copied from #2830 for new release milestone. The solution is to update gcc/config/i386/t-rtems. Whether for gcc 10 or gcc master is the difference on this CR versus #2830 which had to be to gcc 7 as a patch via the RSB. Throwing a std::runtime() exception locks up. The lock up is in the exception clean up handler where the exception object is destructed. The destructor loops distructing the std::string object. The path ends up in libstdc++-v3/include/ext/atomicity.h line 48 or exchange_and_add. At a guess it would seem like the C++ atomics on i386 is broken or fragile. UPDATE: This was broken when gcc i386 eliminated -mcpu in favor of -march/-mtune. The multilibs were built with -mtune and not -march.
#4093	4 years ago	fixed	bsps	Joel Sherrill	Jan Sommer <jan.sommer@…>	4 years ago
Summary	bsps/pc386: Add missing license header
Description	bspsmp.c is missing a licence header.
#4097	4 years ago	fixed	shell	Frank Kuehndel	Sebastian Huber	4 years ago
Summary	shell: CRTL-U sets the cursor to the wrong position (cloned)
Description	Cloned from #4096: This patch fixes a tiny bug in the command line editing of the RTEMS shell. Typing CTRL-U in the shell should remove all characters left of the cursor. After pressing CTRL-U, the current implementation does wrongly place the cursor at the end of the line instead at its beginning. To reproduce the bug, start the shell and type 'abc123' (no <RETURN>): ~/src/rtems $ qemu-system-arm -net none -nographic -M realview-pbx-a9 -m 256M -kernel build/arm/realview_pbx_a9_qemu/testsuites/libtests/dl10.exe *** BEGIN OF TEST libdl (RTL) 10 *** *** TEST VERSION: 6.0.0.d9bdf166644f612dd628fe4951c12c6f8e94ba5f *** TEST STATE: USER_INPUT *** TEST BUILD: RTEMS_DEBUG RTEMS_NETWORKING RTEMS_POSIX_API RTEMS_SMP *** TEST TOOLS: 10.2.1 20200904 \ (RTEMS 6, RSB 31f936a7b74d60bda609a9960c6e1a705ba54974, Newlib a0d7982) RTL (libdl) commands: dl, rtl RTEMS Shell on /dev/foobar. Use 'help' to list commands. SHLL [/] # abc123 Then move the cursor onto the '1' by hitting three times the <ARROW-LEFT> key. Next type <CTRL>-U: SHLL [/] # 123 Note that the cursor is at the end of the line (after '3') instead of correctly at the beginning (on the '1'), now. Continuing typing 'echo ' incorrectly results in the output: SHLL [/] # 123echo 123 The patch changes this behavior so that the cursor in the second last step will be on the '1' and typing 'echo ' will then correctly reflected as: SHLL [/] # echo 123
#4098	4 years ago	fixed	tool/rsb	Joel Sherrill		2 years ago
Summary	All BSP bsets fail to build on master
Description	On the master building for RTEMS 6, all of the BSP bsets fail like this: RTEMS Source Builder - Set Builder, 6 (ed5030bc24db) Build Set: 6/bsps/xilinx_zynq_zedboard error: no build set file found: 6/bsps/xilinx_zynq_zedboard.bset Mailing report: build@rtems.org Traceback (most recent call last): File "../source-builder/sb-set-builder", line 25, in <module> sb.setbuilder.run() File "/home/joel/rtems-cron-6/rtems-source-builder/source-builder/sb/setbuilder.py", line 735, in run b.build(deps, mail = mail) File "/home/joel/rtems-cron-6/rtems-source-builder/source-builder/sb/setbuilder.py", line 633, in build if sizes_valid: UnboundLocalError: local variable 'sizes_valid' referenced before assignment
#4100	4 years ago	fixed	arch/arm	Joel Sherrill		4 years ago
Summary	arm/edb7312 Build Failure When SMP is Enabled
Description	This is from the edb7312 but may apply to others. Build is still underway and this is a common file. ''''gmake[4]: Entering directory `/home/joel/rtems-cron-6/b-edb7312/arm-rtems6/c/edb7312/cpukit' depbase=`echo score/src/percpu.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`;\ arm-rtems6-gcc --pipe -DHAVE_CONFIG_H -I. -I/home/joel/rtems-cron-6/b-edb7312/arm-rtems6/c/edb7312/include -I/home/joel/rtems-cron-6/rtems/cpukit/include -I/home/joel/rtems-cron-6/rtems/cpukit/score/cpu/arm/include -mcpu=arm7tdmi -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -MT score/src/percpu.o -MD -MP -MF $depbase.Tpo -c -o score/src/percpu.o ../../../../../rtems/c/src/../../cpukit/score/src/percpu.c &&\ mv -f $depbase.Tpo $depbase.Po {standard input}: Assembler messages: {standard input}:157: Error: selected processor does not support `sev' in ARM mode {standard input}:347: Error: selected processor does not support `wfe' in ARM mode {standard input}:525: Error: selected processor does not support `wfe' in ARM mode gmake[4]: *** [score/src/percpu.o] Error 1
#4101	4 years ago	fixed	arch/arm	Joel Sherrill		4 years ago
Summary	atsamv fails to build on master
Description	This may also fail on 5. $ ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=atsamv --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --enable-smp --disable-multiprocessing --disable-rtems-debug --disable-profiling --enable-tests --enable-cxx --enable-maintainer-mode rm-rtems6-gcc --pipe -DHAVE_CONFIG_H -I. -I/home/joel/rtems-cron-6/b-atsamv/arm-rtems6/c/atsamv/include -I/home/joel/rtems-cron-6/rtems/cpukit/include -I/home/joel/rtems-cron-6/rtems/cpukit/score/cpu/arm/include -mthumb -mcpu=cortex-m7 -mfpu=fpv5-d16 -mfloat-abi=hard -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -MT libcsupport/src/error.o -MD -MP -MF $depbase.Tpo -c -o libcsupport/src/error.o ../../../../../rtems/c/src/../../cpukit/libcsupport/src/error.c &&\ mv -f $depbase.Tpo $depbase.Po In file included from /home/joel/rtems-cron-6/rtems/cpukit/include/rtems/score/threaddispatch.h:21, from /home/joel/rtems-cron-6/rtems/cpukit/include/rtems/score/objectimpl.h:29, from /home/joel/rtems-cron-6/rtems/cpukit/include/rtems/score/threadimpl.h:31, from ../../../../../rtems/c/src/../../cpukit/libcsupport/src/error.c:16: /home/joel/rtems-cron-6/rtems/cpukit/include/rtems/score/percpu.h:439:3: error: unknown type name 'CPU_Interrupt_frame' 439 | CPU_Interrupt_frame Interrupt_frame; | ^~~~~~~~~~~~~~~~~~~
#4103	4 years ago	fixed	admin	Joel Sherrill		4 years ago
Summary	BeagleBoard? BSPs Fail to Build
Description	This may also fail on 5. ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=beagleboardorig --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --enable-smp --disable-multiprocessing --disable-rtems-debug --disable-profiling --enable-tests --enable-cxx --enable-maintainer-mode arm-rtems6-gcc -mcpu=cortex-a8 -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -B./../../lib/libbsp/arm/beagle -B/home/joel/rtems-cron-6/rtems/bsps/arm/beagle/start -specs bsp_specs -qrtems -L./../../cpukit -L/home/joel/rtems-cron-6/rtems/bsps/arm/shared/start -Wl,--wrap=printf -Wl,--wrap=puts -Wl,--wrap=putchar -Wl,--gc-sections -o base_sp.exe base_sp/base_sp-init.o base_sp/base_sp-apptask.o ./../../lib/libbsp/arm/beagle/librtemsbsp.a ./../../cpukit/librtemscpu.a ./../../cpukit/librtemstest.a /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(threadrestart.o): in function `_Thread_Dispatch_request': /home/joel/rtems-cron-6/rtems/cpukit/include/rtems/score/threaddispatch.h:245: undefined reference to `_CPU_SMP_Send_interrupt' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(smp.o): in function `_SMP_Handler_initialize': /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/smp.c:135: undefined reference to `_CPU_SMP_Initialize' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(smp.o): in function `_SMP_Start_processors': /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/smp.c:78: undefined reference to `_CPU_SMP_Start_processor' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(smp.o): in function `_SMP_Handler_initialize': /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/smp.c:151: undefined reference to `_CPU_SMP_Finalize_initialization' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/smp.c:135: undefined reference to `_CPU_SMP_Initialize' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(smp.o): in function `_SMP_Send_message': /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/smp.c:221: undefined reference to `_CPU_SMP_Send_interrupt' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: ./../../cpukit/librtemscpu.a(threadstartmultitasking.o): in function `_Thread_Start_multitasking': /home/joel/rtems-cron-6/b-beagleboardorig/arm-rtems6/c/beagleboardorig/cpukit/../../../../../rtems/c/src/../../cpukit/score/src/threadstartmultitasking.c:44: undefined reference to `_CPU_SMP_Prepare_start_multitasking' collect2: error: ld returned 1 exit status gmake[5]: *** [base_sp.exe] Error 1
#4105	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add RTEMS_PARTITION_ALIGNMENT
Description	The storage area for partitions is usually statically allocated. Add RTEMS_PARTITION_ALIGNMENT so that these areas can meet the minimum alignment required by rtems_partition_create(). This is related to #4013.
#4107	4 years ago	fixed	build	Joel Sherrill	Sebastian Huber	4 years ago
Summary	waf -j24 fails on powerpc/psim -- probable dependency issue
Description	While experimenting with -j options since waf defaults to fewer jobs than I have been using with make, I tripped across this failure on powerpc/psim which I assume is a missing dependency somewhere: 1561/4127] Linking build/powerpc/psim/testsuites/fstests/fsjffs2gc01.exe [1562/4127] Linking build/powerpc/psim/testsuites/fstests/fsclose01.exe [1563/4127] Linking build/powerpc/psim/libdebugger.a [1564/4127] Linking build/powerpc/psim/libz.a [1565/4127] Compiling testsuites/fstests/fslink/test.c /home/joel/rtems-work/tools/6/lib/gcc/powerpc-rtems6/10.2.1/../../../../powerpc-rtems6/bin/ld: cannot find -ljffs2 collect2: error: ld returned 1 exit status ../../../testsuites/fstests/fsdosfsname01/init.c: In function 'test_creating_invalid_directories': ../../../testsuites/fstests/fsdosfsname01/init.c:430:19: warning: '%s' directive output may be truncated writing up to 6424 bytes into a region of size 257 [-Wformat-truncation=] 430 | "%s/%s", | ^~ ../../../testsuites/fstests/fsdosfsname01/init.c:428:5: note: 'snprintf' output between 6 and 6430 bytes into a destination of size 262 428 | snprintf( dirname, | ^~~~~~~~~~~~~~~~~~ 429 | sizeof( dirname ), | ~~~~~~~~~~~~~~~~~~ 430 | "%s/%s", | ~~~~~~~~ 431 | MOUNT_DIR, | ~~~~~~~~~~ 432 | DIRECTORY_NAMES_INVALID[index] ); | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ../../../cpukit/mghttpd/mongoose.c: In function 'handle_request': ../../../cpukit/mghttpd/mongoose.c:1919:45: warning: '.gz' directive output may be truncated writing 3 bytes into a region of size between 1 and 255 [-Wformat-truncation=] 1919 | snprintf(gz_path, sizeof(gz_path), "%s.gz", buf); | ^~~ ../../../cpukit/mghttpd/mongoose.c:1919:7: note: 'snprintf' output between 4 and 258 bytes into a destination of size 255 1919 | snprintf(gz_path, sizeof(gz_path), "%s.gz", buf); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Waf: Leaving directory `/home/joel/rtems-work/rtems/build/powerpc/psim' Build failed -> task in 'testsuites/fstests/fsjffs2gc01.exe' failed with exit status 1 (run with -v to display more information) INI file was: [powerpc/psim] # Selects the compiler used to build the BSP (allowed values are "gcc" and # "clang"). Please note that the values of some options depend on the compiler # selection and changing the compiler may lead to unpredictable behaviour if # these options are not adjusted as well. Use the --rtems-compiler command line # option to get the default values for a particular compiler via # ./waf bsp_defaults. COMPILER = gcc # Flags passed to the library archiver ARFLAGS = crD # Warning flags passed to the C and C++ compiler WARNING_FLAGS = -Wall # Warning flags passed to the C compiler CC_WARNING_FLAGS = -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs # Warning flags passed to the C++ compiler CXX_WARNING_FLAGS = # Enable the Ada support __RTEMS_ADA__ = False # Enable the RTEMS internal debug support RTEMS_DEBUG = False # Enable the Driver Manager startup RTEMS_DRVMGR_STARTUP = False # Enable the message passing based multiprocessing support (MPCI); do # not confuse this with the SMP support RTEMS_MULTIPROCESSING = False # Enable the legacy TCP/IP network support RTEMS_NETWORKING = False # Enable the Newlib C library support RTEMS_NEWLIB = True # Enable the para-virtualization support RTEMS_PARAVIRT = False # Enable support for POSIX signals RTEMS_POSIX_API = True # Enable the SMP lock profiling support RTEMS_PROFILING = False # Enable the Symmetric Multiprocessing (SMP) support RTEMS_SMP = False # Build the Ada test programs (may be also enabled by BUILD_TESTS) BUILD_ADATESTS = False # Build the test programs BUILD_TESTS = True # Build the benchmark programs (may be also enabled by BUILD_TESTS) BUILD_BENCHMARKS = False # Build the file system test programs (may be also enabled by # BUILD_TESTS) BUILD_FSTESTS = False # Build the library test programs (may be also enabled by BUILD_TESTS) BUILD_LIBTESTS = False # Build the MPCI test programs (may be also enabled by BUILD_TESTS) BUILD_MPTESTS = False # Build the POSIX test programs (may be also enabled by BUILD_TESTS) BUILD_PSXTESTS = False # Build the POSIX timing test programs (may be also enabled by # BUILD_TESTS) BUILD_PSXTMTESTS = False # Build the Rhealstone benchmark programs (may be also enabled by # BUILD_TESTS) BUILD_RHEALSTONE = False # Build the sample programs (may be also enabled by BUILD_TESTS) BUILD_SAMPLES = True # Build the SMP test programs (may be also enabled by BUILD_TESTS) BUILD_SMPTESTS = False # Build the single-processor test programs (may be also enabled by # BUILD_TESTS) BUILD_SPTESTS = False # Build the timing test programs (may be also enabled by BUILD_TESTS) BUILD_TMTESTS = False # Build the validation test programs (may be also enabled by # BUILD_TESTS) BUILD_VALIDATIONTESTS = False # Optimization flags passed to the C and C++ compiler OPTIMIZATION_FLAGS = -O2 -g -fdata-sections -ffunction-sections # ABI flags ABI_FLAGS = -Dppc603e -meabi -mcpu=603e -msdata=sysv # This sets a mode where the time runs as fast as possible when a # clock ISR occurs while the IDLE thread is executing. This can # significantly reduce simulation times. CLOCK_DRIVER_USE_FAST_IDLE = True # If defined then the BSP may reduce the available memory size # initially. This can be useful for debugging (reduce the core size) # or dynamic loading (std gcc text offsets/jumps are < +/-32M). Note # that the policy can still be defined by the application (see sbrk.c, # BSP_sbrk_policy). By undefining CONFIGURE_MALLOC_BSP_SUPPORTS_SBRK # this feature is removed and a little memory is saved. CONFIGURE_MALLOC_BSP_SUPPORTS_SBRK = True # If defined, then the PowerPC specific code in RTEMS will use some of # the special purpose registers to slightly optimize interrupt # response time. The use of these registers can conflict with other # tools like debuggers. PPC_USE_SPRG = False # This defines the base address of the exception table. NOTE: Vectors # are actually at 0xFFF00000 but file starts at offset. PPC_VECTOR_FILE_BASE = 0xfff00100 # If defined to a non-zero value, prints the some information in case # of a fatal error. BSP_VERBOSE_FATAL_EXTENSION = 1 # Flags passed to the linker (GNU ld) LDFLAGS = -Wl,--gc-sections # Install the legacy application Makefile framework. INSTALL_LEGACY_MAKEFILES = True # If defined to a non-zero value, prints the exception context when an # unexpected exception occurs. BSP_PRINT_EXCEPTION_CONTEXT = 1 # If defined to a non-zero value, reset the board when the application # exits. BSP_RESET_BOARD_AT_EXIT = 1 # If defined to a non-zero value, print a message and wait until # pressed before resetting board when application exits. BSP_PRESS_KEY_FOR_RESET = 0
#4108	4 years ago	fixed	test	Joel Sherrill	Sebastian Huber	20 months ago
Summary	Review minimum and ensure it really is a minimum configuration
Description	Does minimum actually reflect what you think is the smallest possible configuration now? That as always the test's purpose -- to have a single that was the smallest we thought we could produce. It seems all your changes that it needs a review for tweaks to the configuration. Inspecting the size of this test should tell you the smallest executable possible on this BSP with the current configuration settings. There is a version in rtems-examples/misc which will either need to be synchronized or deleted. This needs to be checked on 5 and 6.
#4110	4 years ago	fixed	unspecified	Chris Johns	Ryan Long <ryan.long@…>	2 years ago
Summary	Ubuntu package list is broken
Description	The list found here: ​https://docs.rtems.org/branches/master/user/hosts/posix.html#xubuntu failed with libpython-dev. A Linux user needs to check this and update the list.
#4115	4 years ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	4 years ago
Summary	Cortex-A9 MPCore based BSPs should include the workaround for Errata 794072 and 845369
Description	Cloned from #4114: The workaround may be already set up by the bootloader. Not having the workaround applied may lead to system crashes, so make sure the BSP does it during system startup.
#4116	4 years ago	fixed	arch/arm	Joel Sherrill	Sebastian Huber	4 years ago
Summary	raspberrypi fails to build on waf and autoconf
Description	This may apply to 5 as well but I didn't check. The following configurations for pi didn't build: FAILED (2) autoconf build of arm raspberrypi (SMP/NOMP/DEBUG/PROFILE) FAILED (1) waf build of arm raspberrypi (SMP/NOMP/DEBUG/PROFILE) FAILED (2) autoconf build of arm raspberrypi (SMP/NOMP/DEBUG/NOPROFILE) FAILED (1) waf build of arm raspberrypi (SMP/NOMP/DEBUG/NOPROFILE) FAILED (2) autoconf build of arm raspberrypi (SMP/NOMP/NODEBUG/PROFILE) FAILED (1) waf build of arm raspberrypi (SMP/NOMP/NODEBUG/PROFILE) FAILED (2) autoconf build of arm raspberrypi (SMP/NOMP/NODEBUG/NOPROFILE) FAILED (1) waf build of arm raspberrypi (SMP/NOMP/NODEBUG/NOPROFILE) This one of the configure commands: ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=raspberrypi --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --enable-smp --disable-multiprocessing --enable-rtems-debug --enable-profiling --enable-tests --enable-cxx --enable-maintainer-mode And this is the failure: gmake[4]: Entering directory `/home/joel/rtems-cron-6/b-raspberrypi/arm-rtems6/c/raspberrypi/cpukit' depbase=`echo score/cpu/arm/cpu_asm.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`;\ arm-rtems6-gcc --pipe -DHAVE_CONFIG_H -I. -I/home/joel/rtems-cron-6/b-raspberrypi/arm-rtems6/c/raspberrypi/include -I/home/joel/rtems-cron-6/rtems/cpukit/include -I/home/joel/rtems-cron-6/rtems/cpukit/score/cpu/arm/include -DASM -mcpu=arm1176jzf-s -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -mcpu=arm1176jzf-s -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -MT score/cpu/arm/cpu_asm.o -MD -MP -MF $depbase.Tpo -c -o score/cpu/arm/cpu_asm.o ../../../../../rtems/c/src/../../cpukit/score/cpu/arm/cpu_asm.S &&\ mv -f $depbase.Tpo $depbase.Po ../../../../../rtems/c/src/../../cpukit/score/cpu/arm/cpu_asm.S: Assembler messages: ../../../../../rtems/c/src/../../cpukit/score/cpu/arm/cpu_asm.S:75: Error: selected processor does not support `dmb' in ARM mode ../../../../../rtems/c/src/../../cpukit/score/cpu/arm/cpu_asm.S:93: Error: selected processor does not support `dmb' in ARM mode gmake[4]: *** [score/cpu/arm/cpu_asm.o] Error 1
#4117	4 years ago	fixed	arch/arm	Joel Sherrill	Sebastian Huber	4 years ago
Summary	lpc1768_mbed_ahb_ram_eth sp01 and others fail to link
Description	This fails on both waf and autoconf. This is troubling because the test is sp01 which should build and link. Did not check if this also applies to 5 branch. ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=lpc1768_mbed_ahb_ram_eth --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --disable-smp --disable-multiprocessing --disable-rtems-debug --disable-profiling --enable-tests --enable-cxx --enable-maintainer-mode Fails with: arm-rtems6-gcc -mthumb -mcpu=cortex-m3 -O2 -ggdb3 -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -B./../../lib/libbsp/arm/lpc176x -B/home/joel/rtems-cron-6/rtems/bsps/arm/lpc176x/start -specs bsp_specs -qrtems -L./../../cpukit -L/home/joel/rtems-cron-6/rtems/bsps/arm/shared/start -Wl,--wrap=printf -Wl,--wrap=puts -Wl,--wrap=putchar -Wl,--gc-sections -o sp01.exe sp01/sp01-init.o sp01/sp01-task1.o ./../../lib/libbsp/arm/lpc176x/librtemsbsp.a ./../../cpukit/librtemscpu.a ./../../cpukit/librtemstest.a /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: sp01.exe section `.bss' will not fit in region `RAM_AHB1' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: region `RAM_AHB1' overflowed by 2324 bytes collect2: error: ld returned 1 exit status gmake[5]: *** [sp01.exe] Error 1
#4118	4 years ago	wontfix	arch/sparc	Joel Sherrill		4 years ago
Summary	Add clang variants for sparc BSPs
Description	There are clang variants for RISV-V BSPs but not SPARC BSPs even though clang is supported on them. If there are other architectures with clang support, I don't know.
#4120	4 years ago	fixed	arch/arm	Joel Sherrill	Sebastian Huber	4 years ago
Summary	altcycv_devkit and imx7 fail to build all tests on master
Description	This appears to not happen with waf but it needs to be resolved or we don't have parity to achieve a switchover. Happens to at least two BSPs. Not sure if more until all arm BSP configurations finish. Configuration: ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=altcycv_devkit --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --enable-smp --disable-multiprocessing --enable-rtems-debug --enable-profiling --enable-tests --enable-cxx --enable-maintainer-mode Results in: arm-rtems6-gcc -DHAVE_CONFIG_H -I. -I../../../../../../rtems/c/src/../../testsuites/smptests -I. -I/home/joel/rtems-cron-6/b-altcycv_devkit/arm-rtems6/c/altcycv_devkit/include -I/home/joel/rtems-cron-6/rtems/cpukit/include -I/home/joel/rtems-cron-6/rtems/cpukit/score/cpu/arm/include -I/home/joel/rtems-cron-6/b-altcycv_devkit/arm-rtems6/c/altcycv_devkit/lib/libbsp/arm/altera-cyclone-v/include -I/home/joel/rtems-cron-6/rtems/bsps/include -I/home/joel/rtems-cron-6/rtems/bsps/arm/include -I/home/joel/rtems-cron-6/rtems/bsps/arm/altera-cyclone-v/include -DT_FILE_NAME='"init.c"' -I../../../../../../rtems/c/src/../../testsuites/smptests/../support/include -march=armv7-a -mthumb -mfpu=neon -mfloat-abi=hard -mtune=cortex-a9 -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -MT smpmulticast01/smpmulticast01-init.o -MD -MP -MF smpmulticast01/.deps/smpmulticast01-init.Tpo -c -o smpmulticast01/smpmulticast01-init.o `test -f 'smpmulticast01/init.c' || echo '../../../../../../rtems/c/src/../../testsuites/smptests/'`smpmulticast01/init.c ../../../../../../rtems/c/src/../../testsuites/smptests/smpmulticast01/init.c:37:10: fatal error: rtems/testopts.h: No such file or directory 37 | #include <rtems/testopts.h> | ^~~~~~~~~~~~~~~~~~ compilation terminated. gmake[5]: *** [smpmulticast01/smpmulticast01-init.o] Error 1
#4121	4 years ago	fixed	arch/arm	Joel Sherrill	Sebastian Huber	4 years ago
Summary	lm3s3749 fails to build all tests on master
Description	FAILED (2) autoconf build of arm lm3s3749 (NOSMP/NOMP/DEBUG/PROFILE) FAILED (1) waf build of arm lm3s3749 (NOSMP/NOMP/DEBUG/PROFILE) FAILED (2) autoconf build of arm lm3s3749 (NOSMP/NOMP/DEBUG/NOPROFILE) FAILED (1) waf build of arm lm3s3749 (NOSMP/NOMP/DEBUG/NOPROFILE) ../rtems/configure --target=arm-rtems6 --enable-rtemsbsp=lm3s3749 OPERATION_COUNT=10 --prefix=/home/joel/rtems-cron-6/tools/6/bsp-install --disable-networking --enable-posix --disable-smp --disable-multiprocessing --enable-rtems-debug --enable-profiling --enable-tests --enable-cxx --enable-maintainer-mode arm-rtems6-gcc -mthumb -mcpu=cortex-m3 -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -B./../../lib/libbsp/arm/lm3s69xx -B/home/joel/rtems-cron-6/rtems/bsps/arm/lm3s69xx/start -specs bsp_specs -qrtems -L./../../cpukit -L/home/joel/rtems-cron-6/rtems/bsps/arm/shared/start -Wl,--wrap=printf -Wl,--wrap=puts -Wl,--wrap=putchar -Wl,--gc-sections -o math.exe math/math-init.o math/math-domath.o -lm ./../../lib/libbsp/arm/lm3s69xx/librtemsbsp.a ./../../cpukit/librtemscpu.a ./../../cpukit/librtemstest.a /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: math.exe section `.rodata' will not fit in region `ROM_INT' /home/joel/rtems-cron-6/tools/6/lib/gcc/arm-rtems6/10.2.1/../../../../arm-rtems6/bin/ld: region `ROM_INT' overflowed by 2872 bytes
#4122	4 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Change RTEMS_NO_RETURN to use C11 and C++11 standard means
Description	C11 has a _Noreturn keyword and C++11 has a noreturn? attribute. The standards are a bit less flexible than the GNU attribute with respect to the position of the attribute.
#4123	4 years ago	fixed	tool	Joel Sherrill	Sebastian Huber <sebastian.huber@…>	3 years ago
Summary	rtems-bsps needs to be switched to look at waf info for BSPs
Description	rtems-bsps depends on autoconf. It must be fixed to depend on waf BSP configuration information before we can switch from autoconf.
#4124	4 years ago	fixed	tool	Joel Sherrill	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-bsp-builder needs to be switched to waf builds
Description	rtems-bsp-tester depends on autoconf. It must be fixed to depend on waf BSP configuration information before we can switch from autoconf.
#4126	4 years ago	fixed	network/legacy	Chris Johns		2 years ago
Summary	Remove legacy networking stack from rtems.git and place in a new repo
Description	Remove the legacy networking stack from rtems.git and place it in new top level repo rtems-legacy-network.git. The initial tasks are: Move the code to a new repo Add rtems_waf support to build it Clean out BSP references to network interfaces or drivers Merge in the network demos repo
#4127	4 years ago	fixed	test	Joel Sherrill	Sebastian Huber	4 years ago
Summary	x86_64/amd64 fails to build
Description	This is with waf and autoconf on master. Short versions of failing configurations: FAILED (2) autoconf build of x86_64 amd64 (NOSMP/NOMP/DEBUG/PROFILE) FAILED (1) waf build of x86_64 amd64 (NOSMP/NOMP/DEBUG/PROFILE) FAILED (2) autoconf build of x86_64 amd64 (NOSMP/NOMP/DEBUG/NOPROFILE) FAILED (1) waf build of x86_64 amd64 (NOSMP/NOMP/DEBUG/NOPROFILE) This is the error: In file included from ../../../testsuites/libtests/termios11/init.c:2: ../../../testsuites/libtests/termios11/../termios10/termios10impl.h: In function 'POSIX_Init': ../../../testsuites/libtests/termios11/../termios10/termios10impl.h:171:1: error: no return statement in function returning non-void [-Werror=return-type] 171 | } | ^ cc1: some warnings being treated as errors In file included from ../../../testsuites/libtests/termios10/init.c:1: ../../../testsuites/libtests/termios10/termios10impl.h: In function 'POSIX_Init': ../../../testsuites/libtests/termios10/termios10impl.h:171:1: error: no return statement in function returning non-void [-Werror=return-type] 171 | } | ^ This is the waf ini file for this specific build: [x86_64/amd64] # Selects the compiler used to build the BSP (allowed values are "gcc" and # "clang"). Please note that the values of some options depend on the compiler # selection and changing the compiler may lead to unpredictable behaviour if # these options are not adjusted as well. Use the --rtems-compiler command line # option to get the default values for a particular compiler via # ./waf bsp_defaults. COMPILER = gcc # Flags passed to the library archiver ARFLAGS = crD # Warning flags passed to the C and C++ compiler WARNING_FLAGS = -Wall # Warning flags passed to the C compiler CC_WARNING_FLAGS = -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs # Warning flags passed to the C++ compiler CXX_WARNING_FLAGS = # Enable the Ada support __RTEMS_ADA__ = False # Enable the RTEMS internal debug support RTEMS_DEBUG = True # Enable the Driver Manager startup RTEMS_DRVMGR_STARTUP = False # Enable the legacy TCP/IP network support RTEMS_NETWORKING = False # Enable the Newlib C library support RTEMS_NEWLIB = True # Enable the para-virtualization support RTEMS_PARAVIRT = False # Enable support for POSIX signals RTEMS_POSIX_API = True # Enable the SMP lock profiling support RTEMS_PROFILING = False # Build the Ada test programs (may be also enabled by BUILD_TESTS) BUILD_ADATESTS = False # Build the test programs BUILD_TESTS = True # Build the benchmark programs (may be also enabled by BUILD_TESTS) BUILD_BENCHMARKS = False # Build the file system test programs (may be also enabled by # BUILD_TESTS) BUILD_FSTESTS = False # Build the library test programs (may be also enabled by BUILD_TESTS) BUILD_LIBTESTS = False # Build the MPCI test programs (may be also enabled by BUILD_TESTS) BUILD_MPTESTS = False # Build the POSIX test programs (may be also enabled by BUILD_TESTS) BUILD_PSXTESTS = False # Build the POSIX timing test programs (may be also enabled by # BUILD_TESTS) BUILD_PSXTMTESTS = False # Build the Rhealstone benchmark programs (may be also enabled by # BUILD_TESTS) BUILD_RHEALSTONE = False # Build the sample programs (may be also enabled by BUILD_TESTS) BUILD_SAMPLES = True # Build the SMP test programs (may be also enabled by BUILD_TESTS) BUILD_SMPTESTS = False # Build the single-processor test programs (may be also enabled by # BUILD_TESTS) BUILD_SPTESTS = False # Build the timing test programs (may be also enabled by BUILD_TESTS) BUILD_TMTESTS = False # Build the validation test programs (may be also enabled by # BUILD_TESTS) BUILD_VALIDATIONTESTS = False # Sets the test verbosity of the RTEMS Test Framework. Valid values # are "Verbose", "Normal", and "Quiet". For normal test runs you can # use a verbosity of "Normal". If you need a more detailed output, # then you can set it to "Verbose". For example, in case a regression # ocurred. Reducing the verbosity level can significantly reduce the # test duration especially on targets with a slow serial output # device. All verbosity levels produce enough output to notice failed # test cases. In particular, the begin and end of test information is # not affected by the verbosity level. RTEMS_TEST_VERBOSITY = Normal # Optimization flags passed to the C and C++ compiler OPTIMIZATION_FLAGS = -O2 -g -fdata-sections -ffunction-sections # ABI flags ABI_FLAGS = -mno-red-zone -mcmodel=large -Werror=return-type # If defined to a non-zero value, prints the some information in case # of a fatal error. BSP_VERBOSE_FATAL_EXTENSION = 1 # Flags passed to the linker (GNU ld) LDFLAGS = -Wl,--gc-sections # Install the legacy application Makefile framework. INSTALL_LEGACY_MAKEFILES = True # If defined to a non-zero value, prints the exception context when an # unexpected exception occurs. BSP_PRINT_EXCEPTION_CONTEXT = 1 # If defined to a non-zero value, reset the board when the application # exits. BSP_RESET_BOARD_AT_EXIT = 1 # If defined to a non-zero value, print a message and wait until # pressed before resetting board when application exits. BSP_PRESS_KEY_FOR_RESET = 0
#4128	4 years ago	fixed	bsps	Joel Sherrill	Jiri Gaisler	4 years ago
Summary	all griscv variants fail to build on master
Description	Looks to be independent of waf or autoconf or configuration settings since a code issue in the BSP: ./../../../../../../../rtems/c/src/lib/libbsp/riscv/griscv/../../../../../../bsps/riscv/griscv/clock/clockdrv.c: In function 'grlib_clock_initialize': ../../../../../../../../rtems/c/src/lib/libbsp/riscv/griscv/../../../../../../bsps/riscv/griscv/clock/clockdrv.c:174:10: warning: implicit declaration of function 'grlib_irqmp_has_timestamp' [-Wimplicit-function-declaration] 174 | if (!grlib_irqmp_has_timestamp(irqmp_ts)) { | ^~~~~~~~~~~~~~~~~~~~~~~~~ ../../../../../../../../rtems/c/src/lib/libbsp/riscv/griscv/../../../../../../bsps/riscv/griscv/clock/clockdrv.c:174:10: warning: nested extern declaration of 'grlib_irqmp_has_timestamp' [-Wnested-externs] ../../../../../../../../rtems/c/src/lib/libbsp/riscv/griscv/../../../../../../bsps/riscv/griscv/clock/clockdrv.c:175:17: error: 'GRLIB_FATAL_CLOCK_NO_IRQMP_TIMESTAMP_SUPPORT' undeclared (first use in this function); did you mean 'LEON3_FATAL_CLOCK_NO_IRQMP_TIMESTAMP_SUPPORT'? 175 | bsp_fatal(GRLIB_FATAL_CLOCK_NO_IRQMP_TIMESTAMP_SUPPORT); | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ | LEON3_FATAL_CLOCK_NO_IRQMP_TIMESTAMP_SUPPORT ../../../../../../../../rtems/c/src/lib/libbsp/riscv/griscv/../../../../../../bsps/riscv/griscv/clock/clockdrv.c:175:17: note: each undeclared identifier is reported only once for each function it appears in gmake[6]: *** [clockdrv.o] Error 1
#4129	4 years ago	fixed	tool/website	Joel Sherrill		2 years ago
Summary	x86_64 not listed as architecture in components in Trac
Description	Just needs to be added.
#4131	4 years ago	fixed	shell	Frank Kuehndel	Frank Kühndel <frank.kuehndel@…>	4 years ago
Summary	"fdisk DISK_NAME mount" in bdpart-mount.c uses non-existing file system type "msdos"
Description	The function rtems_bdpart_mount() seems to be only used by the 'fdisk' shell command to mount all partitions of a disk with a single command: fdisk DISK_NAME mount mounts the file system of each partition of the disk fdisk DISK_NAME unmount unmounts the file system of each partition of the disk The whole command does not work because in file cpukit/libblock/src/bdpart-mount.c line 103 specifies the file system type of each partition to be "msdos". Yet, "msdos" does not exist. The name must be "dosfs". Beside of this fundamental problem, there are more issues with the code in bdpart-mount.c: 1) The function returns RTEMS_SUCCESSFUL despite the mount always fails. 2) The reason for errors is not written to the terminal. 3) The directory '/mnt' is created but not deleted later on (failure or not). 3) There is no documentation about this special 'fdisk' feature in the ​RTEMS Shell Guide. ('fdisk' is mentioned but its documentation is a bit short). 4) Only "msdos" formatted partitions can be mounted and all partitions are mounted read only. This is hard coded and cannot be changed by options. Moreover, there is no information about this to the user of the shell (i.e. using 'fdisk' mount requires insider knowledge). How to reproduce: 1) For testing I use the 'testsuites/samples/fileio/init.c' sample and since I do not know the 'root' password, the password must be removed from that file to login as 'root' without password: index 86b34b99dd..51507f9a53 100644 --- a/testsuites/samples/fileio/init.c +++ b/testsuites/samples/fileio/init.c @@ -563,8 +563,7 @@ static void fileio_start_shell(void) writeFile( "/etc/passwd", 0644, - "root:$6$$FuPOhnllx6lhW2qqlnmWvZQLJ8Thr/09I7ESTdb9VbnTOn5.65" - "/Vh2Mqa6FoKXwT0nHS/O7F0KfrDc6Svb/sH.:0:0:root::/:/bin/sh\n" + "root::0:0:root::/:/bin/sh\n" "rtems::1:1:RTEMS Application::/:/bin/sh\n" "test:$1$$oPu1Xt2Pw0ngIc7LyDHqu1:2:2:test account::/:/bin/sh\n" "tty:*:3:3:tty owner::/:/bin/false\n" 2) I run the fileio sample using qemu: cd rtems env QEMU_AUDIO_DRV="none" qemu-system-arm -net none -nographic -M realview-pbx-a9 -m 256M -kernel build/arm/realview_pbx_a9_qemu/testsuites/samples/fileio.exe 3) Type any key to stop the timer and enter the sample tool. Type 's' to enter the shell, login as 'root' with empty password. 4) Type the following shell commands (they create a RAM disk, partition it, regsiter it, format it and try to mount it): mkrd fdisk /dev/rda fat32 16 write mbr fdisk /dev/rda register mkdos /dev/rda1 fdisk /dev/rda mount 5) The last line above is the command which fails - without an error message. There exists a '/mnt' directory but no '/mnt/rda1' directory as it shoud be: ls -la /mnt 6) If you change line 103 of 'cpukit/libblock/src/bdpart-mount.c' from "msdos" to "dosfs", compile and build the executable and re-run the above test, '/mnt/rda1' exists (but the file system is mounted read-only).
#4134	4 years ago	fixed	tool	Sebastian Huber	Sebastian Huber	4 years ago
Summary	rtemsspec: Support user-defined ordering of directives in header files (Doxygen)
Description	Currently, the generated interfaces in header files are defined by their dependencies and a topological sorting (alphabetical order is applied to unrelated interfaces). In order to be in line with the Sphinx documentation it would be nice to apply also the user-defined ordering specified in groups.
#4140	4 years ago	fixed	build	Joel Sherrill	Sebastian Huber <sebastian.huber@…>	4 years ago
Summary	waf install of RTEMS breaks Makefile system in rtems-examples
Description	This boils down to bsp_specs being referenced in Makefile.inc and it no longer being at the install point. This can be worked around by copying an empty file to the install location for bsp_specs enough to compile the programs but I assume that the -B... list of arguments needs to be replaced Properly the BSP compile flags need to be processed into the installed target.cfg and the line with -B.. modified as needed in gcc-target-default.cfg. i386-rtems6-gcc --pipe -B/home/joel/rtems-work/tools/6/bsp-install/i386-rtems6/pc686/lib/ -specs bsp_specs -qrtems -Wall -mtune=pentiumpro -march=pentium -c -o o-optimize/test.o test.c i386-rtems6-gcc: fatal error: cannot read spec file 'bsp_specs': No such file or directory compilation terminated. make: *** [o-optimize/test.o] Error 1 [joel@localhost hello_world_c]$ cp /dev/null /home/joel/rtems-work/tools/6/bsp-install/i386-rtems6/pc686/lib/bsp_specs [joel@localhost hello_world_c]$ RTEMS_MAKEFILE_PATH=/home/joel/rtems-work/tools/6/bsp-install/i386-rtems6/pc686/ make clean all rm -f a.out core mon.out gmon.out rm -f -r rm -f -r a.out *.o *.BAK Depends-o-optimize.tmp rm -f -r o-optimize o-debug test -d o-optimize || mkdir o-optimize i386-rtems6-gcc --pipe -B/home/joel/rtems-work/tools/6/bsp-install/i386-rtems6/pc686/lib/ -specs bsp_specs -qrtems -Wall -mtune=pentiumpro -march=pentium -c -o o-optimize/test.o test.c i386-rtems6-gcc --pipe -B/home/joel/rtems-work/tools/6/bsp-install/i386-rtems6/pc686/lib/ -specs bsp_specs -qrtems -Wall -mtune=pentiumpro -march=pentium -qrtems -Wl,-Ttext,0x00100000 -mtune=pentiumpro -march=pentium -Wl,--gc-sections -L/home/joel/rtems-work/rtems/bsps/i386/shared/start -L/home/joel/rtems-work/rtems/bsps/i386/pc386/start -mtune=pentiumpro -march=pentium -o o-optimize/hello.exe o-optimize/test.o i386-rtems6-nm -g -n o-optimize/hello.exe > o-optimize/hello.num i386-rtems6-size o-optimize/hello.exe text data bss dec hex filename 151790 9628 33056 194474 2f7aa o-optimize/hello.exe
#4142	4 years ago	fixed	lib/block	Sebastian Huber	Sebastian Huber	4 years ago
Summary	Sparse disk destroys data when it grows
Description	The qsort() in sparse_disk_get_new_block() may move the appended key which invalidates the pointer.
#4143	4 years ago	fixed	test	Frank Kuehndel	joel@…	4 years ago
Summary	fileio - Stack too small for login with password
Description	When I use the 'shell' from the fileio sample with the command below: env QEMU_AUDIO_DRV="none" \ qemu-system-arm -no-reboot -net none -nographic -M realview-pbx-a9 -m 256M \ -kernel build/arm/realview_pbx_a9_qemu/testsuites/samples/fileio.exe The executable crashes with an "BLOWN STACK!!!" as soon as I try to login as 'root' with password. (The logins without password work fine.) Increasing the stack size of the affected thread a bit (to 5 times RTEMS_MINIMUM_STACK_SIZE) solves the issue. How to reproduce: Start the fileio sample in QEMU with the above command. Type a key to stop the timer. Type "s" <return> to enter the shell. Login as "root" with the password printed to the terminal. In the line with the first shell prompt, the text "BLOWN STACK!!!" appears unexpectedly followed by further crash information. [...] /dev/foobar login: root Password: RTEMS Shell on /dev/foobar. Use 'help' to list commands. SHLL [/] # BLOWN STACK!!! task control block: 0x00205278 [...] My config.ini was [arm/realview_pbx_a9_qemu] RTEMS_DEBUG = True RTEMS_NETWORKING = True RTEMS_POSIX_API = True RTEMS_SMP = True BUILD_TESTS = True RTEMS origin.master at 558a19d3adff6a3ecf678f94408b48beccbbc9fb
#4145	4 years ago	fixed	tool/rsb	Joel Sherrill	Ryan Long <ryan.long@…>	2 years ago
Summary	rtems-source-builder: Update RTEMS Kernel Recipe to Use waf for RTEMS
Description	kernel configuration builds using autoconf-based build system. Needs to be switched to waf. Hopefully contained to rtems-kernel-common.cfg.
#4148	4 years ago	fixed	fs/rfs	Chris Johns	Chris Johns	4 years ago
Summary	RFS bit map search buffer overflow
Description	The bit search create writes past the end of its buffer. See: ​https://lists.rtems.org/pipermail/devel/2020-October/062701.html
#4155	4 years ago	fixed	arch/sparc	Sebastian Huber	Daniel Hellstrom	3 years ago
Summary	Add a workaround for the LEON3FT RETT Restart errata: TN-0018
Description	See also: ​https://www.gaisler.com/doc/antn/GRLIB-TN-0018.pdf
#4157	4 years ago	fixed	posix	Sebastian Huber	Sebastian Huber	3 years ago
Summary	pthread_spin_unlock() may corrupt the thread stack if RTEMS_PROFILING and RTEMS_SMP is enabled
Description	The call to _SMP_lock_Stats_register_or_max_section_time() registers a dummy statistics block which resides on the thread stack.
#4158	4 years ago	fixed	bsps	Sebastian Huber	Sebastian Huber	4 years ago
Summary	rtems_interrupt_server_delete() does not destroy the ISR lock of the server control
Description	This leads to memory corruption if RTEMS_PROFILING and RTEMS_SMP is enabled.
#4164	3 years ago	fixed	network/libbsd	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Fix NVMe disk synchronization and media block handling
#4166	3 years ago	duplicate	tool	Joel Sherrill		17 months ago
Summary	rtems-bsp-builder: Add way to select rtems version (6 vs 7)
Description	When testing the RTEMS master, one can use the rtems6 or rtems7 tools. The rtems-bsp-builder is hard-coded to rtems6. As part of doing this, the version should be in the email subject and report so that similar reports can be distinguished.
#4167	3 years ago	fixed	tool	Joel Sherrill	Joel Sherrill	2 years ago
Summary	Coverity Scan Does Not Run on rtems6
Description	This is a ticket to note that Coverity does not appear to like the gcc version going from one to two digits. Since the Coverity Scan support for free users is minimal, I asked on the GCC mailing list if anyone had insight (​https://gcc.gnu.org/pipermail/gcc/2020-October/234073.html). This is completely outside our control and this is just a tracking ticket to note that Coverity Scan reports will be run against the 5 branch until this is fixed. This will likely be of limited use since rtems 5 is stable and changes are minimal compared to the rtems master. If it really is a two digit major issue (unconfirmed if this works for gcc 9), I could do Coverity builds using a special toolchain based on gcc 9.
#4175	3 years ago	fixed	arch/aarch64	Kinsey Moore		3 years ago
Summary	Resolve warning during AArch64 build
Description	When compiling under the ILP32 mulitilib variant, the build generates a warning related to the saved stack pointer. This resolves the warning by sizing the saved stack pointer appropriately instead of it always being 64bits long.
#4180	3 years ago	fixed	bsps	Christian Mauderer	Christian Mauderer	2 years ago
Summary	Add BSP for i.MXRT1050
Description	Add a BSP for the ​i.MXRT1050 series. The default settings for the BSP will run on the IMXRT1050-EVKB evaluation board.
#4181	3 years ago	fixed	admin	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add CONFIGURE_INIT_TASK_STORAGE_SIZE
Description	Currently, the Classic API initialization task is created with rtems_task_create(). In order to better support applications which use the new rtems_task_construct() directive add a configuration options which constructs the Classic API initialization task with rtems_task_construct(): /* Generated from spec:/acfg/if/init-task-storage-size */ /** * @brief This configuration option is an integer define. * * The value of this configuration option defines the task storage size of the * Classic API initialization task. If this configuration option is specified, * then the Classic API initialization task is constructed by * rtems_task_construct() instead of using rtems_task_create(). * * @par Default Value * The default value is 0. * * @par Value Constraints * @parblock * The value of this configuration option shall satisfy all of the following * constraints: * * * It shall be greater than or equal to #CONFIGURE_MINIMUM_TASK_STACK_SIZE. * * * It shall be defined using RTEMS_TASK_STORAGE_SIZE(). * @endparblock * * @par Notes * @parblock * A task storage area of the specified size is defined by the configuration * for the Classic API initialization task. The * * * #CONFIGURE_INIT_TASK_STACK_SIZE and * * * ``CONFIGURE_INIT_TASK_STORAGE_SIZE`` * * configuration options are mutually exclusive. * @endparblock */ #define CONFIGURE_INIT_TASK_STORAGE_SIZE
#4184	3 years ago	fixed	bsps	Jan Sommer	Jan Sommer <jan.sommer@…>	3 years ago
Summary	arm/bsps: Small MMU pages are rounded to 1 MiB
Description	The function arm_cp15_start_set_translation_table_entries rounds the end address of a section up to 1 MiB even if small 4kiB pages are used (i.e. ARM_MMU_USE_SMALL_PAGES enabled). Will submit a patch which only rounds the end address to 4kiB boundary.
#4192	3 years ago	fixed	bsps	Jan Sommer		2 years ago
Summary	bsp/xilinx_zynq: Enable support for small MMU pages
Description	Enable the xilinx_zynq based BSPs to support the configuration of the MMU to use small 4 kiB pages. By default 1 MB pages will be used. The 4 kiB pages have to be enabled with the ARM_MMU_USE_SMALL_PAGES configuration option.
#4193	3 years ago	fixed	tool/newlib	Joel Sherrill	Sebastian Huber <sebastian.huber@…>	3 years ago
Summary	devctl.h does not compile from C++ (cloned)
Description	Cloned from #4174: The closing boilerplate of the C++ extern C wrapper is out of order with another #endif and the resulting code does not compile. The attached patch addresses this but it needs to be merged to newlib and the RSB updated before this ticket closed.
#4194	3 years ago	worksforme	tool/rsb	Sebastian Huber		2 years ago
Summary	RSB failed with mail subject: Build Linux: PASSED 6/rtems-arm on x86_64-linux-gnu
Description	error: config error: gdb-common-1.cfg:99: "gdb: python: header file not found: python2.7/Python.h, please install" Build FAILED Mailing report: build@rtems.org The mail subject should indicated that the build failed.
#4195	3 years ago	fixed	tool	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Add gr740 support available in SIS 2.25 to the RTEMS Tester
#4196	3 years ago	fixed	tool/gcc	Sebastian Huber	Sebastian Huber	22 months ago
Summary	Improve gcov support provided by GCC for RTEMS
Description	If we want to use gcov in RTEMS to get code coverage in all system states, then we need some tweaks in GCC: ​https://gcc.gnu.org/pipermail/gcc/2020-November/234164.html The linker set based gcov information registration is already included in GCC: ​https://gcc.gnu.org/onlinedocs/gcc/Instrumentation-Options.html#index-fprofile-info-section There is a discussion about how the gcov information can be transferred from the embedded system to the host: ​https://gcc.gnu.org/pipermail/gcc-patches/2020-November/559342.html The following patch set was integrated in GCC 13: ​https://gcc.gnu.org/pipermail/gcc-patches/2022-April/593536.html ​https://gcc.gnu.org/onlinedocs/gcc/Freestanding-Environments.html#Freestanding-Environments
#4197	3 years ago	fixed	arch/aarch64	Kinsey Moore		3 years ago
Summary	testsuite: ts-validation-0 broken on AArch64 due to SIZE_MAX issues in the test design
Description	The ts-validation-0 test is broken on AArch64 LP64 ABI due to issues with unexpected values of SIZE_MAX. I was not familiar enough with the test architecture in use to debug it properly. This test passes under the ILP32 ABI due to the 32bit pointers and differing value of SIZE_MAX.
#4198	3 years ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	2 years ago
Summary	arm: Add multilib for Cortex-R52 processor
Description	This processor uses the ARMv8-R instruction set.
#4200	3 years ago	fixed	doc	Chris Johns		17 months ago
Summary	Build system doco link in the User Guide is broken
Description	​https://docs.rtems.org/branches/master/user/bld/index.html#commands The "RTEMS Software Engineering" guide link at the end of this section is to a .rst file. I think cross-doc links is broken no matter how you attempt to solve this, ie PDF files will never work.
#4201	3 years ago	fixed	build	Chris Johns		17 months ago
Summary	waf building out of source tree does not find all configured bsps
Description	config.ini is: [DEFAULT] BUILD_TESTS = False RTEMS_DEBUG = True RTEMS_POSIX_API = True [arm/beagleboneblack] [arm/xilinx_zynq_a9_qemu] [i386/pc686] [powerpc/mvme2307] And configure is: $ /opt/work/chris/rtems/kernel/rtems.git/waf \ --top=/opt/work/chris/rtems/kernel/rtems.git \ --out=build configure \ --rtems-config=config.ini \ --prefix=/opt/work/chris/rtems/nfs/6 Setting top to : /opt/work/chris/rtems/kernel/rtems.git Setting out to : /opt/work/chris/rtems/nfs/bsps/build Configure board support package (BSP) : powerpc/mvme2307 Checking for program 'powerpc-rtems6-gcc' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-gcc Checking for program 'powerpc-rtems6-g++' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-g++ Checking for program 'powerpc-rtems6-ar' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ar Checking for program 'powerpc-rtems6-ld' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ld Checking for program 'ar' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ar Checking for program 'g++, c++' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-g++ Checking for program 'ar' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ar Checking for program 'gas, gcc' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-gcc Checking for program 'ar' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ar Checking for program 'gcc, cc' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-gcc Checking for program 'ar' : /opt/work/chris/rtems/nfs/6/bin/powerpc-rtems6-ar Checking for asm flags '-MMD' : yes Checking for c flags '-MMD' : yes Checking for cxx flags '-MMD' : yes Checking for program 'rtems-bin2c' : /opt/work/chris/rtems/nfs/6/bin/rtems-bin2c Checking for program 'gzip' : /usr/bin/gzip Checking for program 'rtems-ld' : /opt/work/chris/rtems/nfs/6/bin/rtems-ld Checking for program 'rtems-syms' : /opt/work/chris/rtems/nfs/6/bin/rtems-syms Checking for program 'xz' : /usr/bin/xz 'configure' finished successfully (0.245s) The same config.ini in the source tree finds all configured bsps. Being able to cleanly sandbox builds is a feature of the autotools build system.
#4202	3 years ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	19 months ago
Summary	Add BSP for ARM Fixed Virtual Platform with a Cortex-R52
#4204	3 years ago	fixed	bsps	Joel Sherrill		2 years ago
Summary	Make Termination Handler which Prints Exit Information Less Scary
Description	This task is change the default termination handler to print successful exit when the cause was exit(0). It scares users to see it and requires explanation.
#4206	3 years ago	fixed	arch/aarch64	Sebastian Huber	Joel Sherrill	3 years ago
Summary	aarch64/xilinx_zynqmp_lp64 does not run with RTEMS_DEBUG enabled
Description	qemu-system-aarch64 -serial null -serial mon:stdio -machine xlnx-zcu102 -m 4096 -kernel build/aarch64/xilinx_zynqmp_lp64/testsuites/samples/ticker.exe -nographic *** FATAL *** fatal source: 7 (RTEMS_FATAL_SOURCE_ASSERT) fatal code: 1074874896 (0x40114a10) RTEMS version: 6.0.0.644cd68d8ce0044f78f095a648bfe4c1567f351b RTEMS tools: 10.2.1 20201215 (RTEMS 6, RSB 6d0bff0eb6ed35d180491614b976277d5469a302, Newlib 6cc47c4) executing thread ID: 0x08a010001 executing thread name: UI1 assertion "cpu_self->thread_dispatch_disable_level == 1" failed: file "/home/EB/sebastian_h/src/rtems/cpukit/score/src/threadhandler.c", line 120, function: _Thread_Handler
#4207	3 years ago	fixed	network/libbsd	Chris Johns	Chris Johns <chrisj@…>	3 years ago
Summary	PZERO define in RTEMS is wrong
Description	Our libbsd support in RTEMS defines PZERO to be zero. FreeBSD defines PZERO to be (PRI_MIN_KERN + 20) and PRI_MIN_KERN is 80. The __sleep() call uses bits allocated above the bits used to hold the highest priority to signal things to do. The PDROP (0x200) bit causes a lock to be dropped before sleeping and not reclaimed after. Setting PZERO to 0 breaks FreeBSD code in libbsd because calls that end up in __sleep() can incorrectly have the PDROP bit set. A grep of FreeBSD shows how this happens: (void) nfsmsleep(&lp->nfslock_lock, mutex, PZERO - 1, "nfsv4gr", NULL); error = msleep(msqkptr, &msq_mtx, (PZERO - 4) | PCATCH, msgsnd", hz); This examples shows how either side is used... selwakeuppri(&tp->tun_rsel, PZERO + 1);
#4209	3 years ago	invalid	arch/arm	rmueller	Joel Sherrill	17 months ago
Summary	STM32H743ZI Nucleo Consle support not working without workaround
Description	I think here is still some issue with the HUART3 console support, at least for the STM32H743ZI-Nucleo board. I solves the issue by calling some hardware initializaton functions and override void HAL_UART_MspInit(UART_HandleTypeDef* huart) like this: #include "hardware_init.h" #include <stm32h7xx_hal_rcc.h> #include <stdio.h> void hardware_init() { MX_USART3_UART_Init(115200); } /** * ST-LINK UART3 * CN5 pins on board */ UART_HandleTypeDef huart3; GPIO_InitTypeDef gpio_uart_init_struct; void MX_USART3_UART_Init(uint32_t baudRate) { __HAL_RCC_USART3_CONFIG(RCC_USART3CLKSOURCE_HSI); __HAL_RCC_GPIOD_CLK_ENABLE(); __HAL_RCC_USART3_CLK_ENABLE(); int result; huart3.Instance = USART3; huart3.Init.BaudRate = baudRate; huart3.Init.WordLength = UART_WORDLENGTH_8B; huart3.Init.StopBits = UART_STOPBITS_1; huart3.Init.Parity = UART_PARITY_NONE; huart3.Init.Mode = UART_MODE_TX_RX; huart3.Init.HwFlowCtl = UART_HWCONTROL_NONE; huart3.Init.OverSampling = UART_OVERSAMPLING_16; huart3.Init.OneBitSampling = UART_ONE_BIT_SAMPLE_DISABLE; huart3.Init.ClockPrescaler = UART_PRESCALER_DIV1; //huart3.Init.FIFOMode = UART_FIFOMODE_DISABLE; //huart3.Init.TXFIFOThreshold = UART_TXFIFO_THRESHOLD_1_8; //huart3.Init.RXFIFOThreshold = UART_RXFIFO_THRESHOLD_1_8; huart3.AdvancedInit.AdvFeatureInit = UART_ADVFEATURE_NO_INIT; // we can't do error handling (simple print out first) here because UART3 is our print interface result = HAL_UART_Init(&huart3); if(result == HAL_OK) { printf("\rUART3 configured successfully!\r\n"); } } // Will be called by HAL_UART_Init void HAL_UART_MspInit(UART_HandleTypeDef* huart) { /*Configure GPIO pins : PD8 PD9 */ gpio_uart_init_struct.Pin = GPIO_PIN_8|GPIO_PIN_9; gpio_uart_init_struct.Mode = GPIO_MODE_AF_PP; gpio_uart_init_struct.Pull = GPIO_NOPULL; gpio_uart_init_struct.Speed = GPIO_SPEED_FREQ_LOW; gpio_uart_init_struct.Alternate = GPIO_AF7_USART3; HAL_GPIO_Init(GPIOD, &gpio_uart_init_struct); } I then call hardware_init() from my main application and after that, printf is working properly. Maybe the problem is also related to using the Nucleo board? Kind Regards Robin
#4212	3 years ago	invalid	fs	Chris Johns	Chris Johns	17 months ago
Summary	libio leaks location clones
Description	rtems_filesystem_location_copy overwrites a location that may have been cloned via the ops handler leaking the destination node. If a file system has allocated resources in a location node when cloned it will leak. I assume there is a free for every clone? An example of a leak is to mount of file system under IMFS and then open a file under the mounted file system. The path eval restart in the IMFS will clone a node and then open will: rtems_filesystem_eval_path_extract_currentloc( &ctx, &iop->pathinfo ); rtems_filesystem_eval_path_cleanup( &ctx ); The ctx contains a cloned location. Allocating resources to a location may be required because of the limited available file system fields and the opaque nature of the fields or a file system may have data it locks or reference counts and that needs to be managed. I do not fully understand the semantics around the location object and while I could guess at a fix there are a number functions related to locations and a large number of use cases that I would not know if my fix is OK. For example should the copy function free the destination location and then clone the source location or should it just free the destination location or should all the places a copy is done be updated to manage the free and cloning based on the context? Or is the issue in the open call and the lines pasted here be swapped?
#4214	3 years ago	fixed	arch/nios2	Sebastian Huber	Sebastian Huber	23 months ago
Summary	nios2: Add thread-local storage support
#4215	3 years ago	fixed	arch/nios2	Sebastian Huber	Sebastian Huber	22 months ago
Summary	nios2: Add multilib with all custom floating-point instructions supported by GCC
Description	See Table 13. "Floating Point Custom Instruction 2 Operation Summary" in the Nios II Custom Instruction User Guide: ​https://www.intel.com/content/dam/www/programmable/us/en/pdfs/literature/ug/ug_nios2_custom_instruction.pdf
#4217	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Change status code for nested obtain of MrsP semaphores
Description	Return RTEMS_INCORRECT_STATE to indicate a nested obtain since this is a kind of deadlock. This status code is also used for other deadlocks.
#4218	3 years ago	fixed	tool/binutils	Joel Sherrill	Ryan Long <ryan.long@…>	3 years ago
Summary	aarch64 as internal error with spconfig01
Description	This is to track ​https://sourceware.org/bugzilla/show_bug.cgi?id=27217 where there are more details on the actual failure. This impacts spconfig01 and spmisc01 on all aarch64 BSPs.
#4219	3 years ago	fixed	tool/gdb	Joel Sherrill		3 years ago
Summary	gdb master does not compile on CentOS 7
Description	There is a enumerated type and method with the same name. The CentOS 7 gdb does not like this. This ticket is to track the upstream ticket ​https://sourceware.org/bugzilla/show_bug.cgi?id=27219
#4221	3 years ago	fixed	tool	Joel Sherrill	kgardas	2 years ago
Summary	dtc recipe issue -- dependency on yaml.h
Description	This shows up on FreeBSD and may just be a matter of me not knowing the right package to install. Which yaml development package should be loaded. But it does fail. CC fstree.o CC livetree.o CC treesource.o gmake: *** No rule to make target 'yaml.h', needed by 'yamltree.o'. Stop. shell cmd failed: /bin/sh -ex /usr/home/joel/rtems-cron-6/rtems-source-builder/rtems/build/dtc-1.6.0-x86_64-freebsd12.2-1/do-build error: building dtc-1.6.0-x86_64-freebsd12.2-1 Suggestions appreciated. See error report: rsb-report-dtc-1.6.0-x86_64-freebsd12.2-1.txt
#4222	3 years ago	fixed	fs	Chris Johns	Chris Johns	3 years ago
Summary	Add a per mount flag to not create a node on O_CREAT in open
Description	The libio rtems_filesystem_eval_path_context_t context uses a token left after a path evaluation to indicate the path to the leaf node is valid and the path leaf does not exist. System calls such as mknod, mkdir and open uses this to return EEXIST or in the case of open to create a S_IFREG node for a file. Some file systems do not support creating an S_IFREG node and the open handler is required to do this. [ An aside, libio should leave the EEXIST detection to the file system and its evaluation handlers. The EEXIST error is one of many that could be returned yet it is specifically singled out using a fragile interface of a path token remaining. Anyway back to the change ... ] The open call has: if ( rtems_filesystem_eval_path_has_token( &ctx ) ) { create_regular_file( &ctx, mode ); } and create_regular_file makes a node: rv = rtems_filesystem_mknod( currentloc, token, tokenlen, S_IFREG | mode, 0 ); then wipes the evaluation context of any real and valid flags and settings: rtems_filesystem_eval_path_set_flags( ctx, 0 ); rtems_filesystem_eval_path_set_path( ctx, token, tokenlen ); rtems_filesystem_eval_path_continue( ctx ); then continues the path evaluation. The continues path evaluation is suppose to find the new node. If a file system does not support making a regular file node and the mknod call is stubbed out to lie about making the node continuing the path evaluation fails with a not found error. I cannot see a way to code around this with the current libio code. I propose adding a bool flag called no_reg_make_node to the rtems_filesystem_mount_table_entry_t struct that defaults to false. If true open does not call create_regular_file.
#4223	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Missing "extern" in RTEMS_LINKER_ROSET_ITEM_ORDERED_DECLARE()
Description	The fix for #3865 contains a typo. In RTEMS_LINKER_ROSET_ITEM_ORDERED_DECLARE() there is a missing "extern".
#4225	3 years ago	fixed	bsps	Daniel Hellstrom	Daniel Hellstrom	20 months ago
Summary	Cobham Gaisler NOEL-V RISC-V 32- and 64-bit processor BSP
Description	To add a BSP configurations for Cobham Gaisler's NOEL-V processor IP. The NOEL-V processor is compliant with the open RISC-V 32/64-bit ISA. Depending on the soft-core IP configuration different RISC-V extensions may or may not be available, a matching RTEMS BSP build configuration must be selected accordingly. Similar to Cobham Gaisler's LEON3/4/5 processor designs, the NOEL-V is also a part of GRLIB providing building blocks such as I/O peripherals and L2-cache etc. that is typically used with the processor in SoC solutions. For more information about GRLIB and the current state and road-map of NOEL-V: ​https://www.gaisler.com/grlib ​https://www.gaisler.com/noel-v The NOEL-V and GRLIB are available either under open-source GPL or commercial license. Demonstration designs with pre-built bit-streams, DTS/DTB, user's manuals, quick-start guides, etc. are freely available for the FPGA development boards listed under the NOEL-V page above. For example: Arty A7: Artix-7 FPGA Development Board (​https://www.gaisler.com/NOEL-ARTYA7) PolarFire? FPGA Splash Kit (​https://www.gaisler.com/NOEL-PF) Xilinx Kintex UltraScale? FPGA KCU105 Evaluation Kit (​https://www.gaisler.com/NOEL-XCKU) All the above demonstration designs have the same base memory layout and peripherals based on GRLIB. A DTS/DTB is needed to describe the system for RTEMS. The first BSP implementation targets only a minimal system with Memory Controller, Processor, IRQ Controller and Timer (RISC-V plic, clint, mtime). Thus the GRLIB drivers based on the driver manager is not needed to run RTEMS on NOEL-V. The free GRMON eval hardware debugger can be used with the above demonstration designs to initialize, upload RTEMS images, execute RTEMS and debug the new NOEL-V BSPs.
#4226	3 years ago	invalid	admin	Sebastian Huber	Amar Takhar	17 months ago
Summary	Buildbot needs a tool chain update
Description	On 28/01/2021 06:37, buildbot@… wrote: The Buildbot has detected a failed build on builder CHECK sparc/erc32 FreeBSD 10.1-STABLE GCC 7.3.0 POSIX while building rtems. Full details are available at: ​https://buildbot.rtems.org/#builders/20/builds/440 Buildbot URL: ​https://buildbot.rtems.org/ Worker for this Build: build-1 Build Reason: <unknown> Blamelist: Sebastian Huber <sebastian.huber@…> BUILD FAILED: make (failure) Sincerely, -The Buildbot The Buildbot needs a tool chain update. It still uses GCC 7.5. For RTEMS 6, GCC 10 is required (RSB ade089253e70d75105a8311844f82f6d20cc30a8).
#4228	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems_build_name() depends on the signedness of char type
Description	The rtems_build_name() is implemented by: {{{!c #define _Objects_Build_name( _C1, _C2, _C3, _C4 ) \ ( (uint32_t)(_C1) << 24 | \ (uint32_t)(_C2) << 16 | \ (uint32_t)(_C3) << 8 | \ (uint32_t)(_C4) ) }}} If the name is build from four "char" components, then the result depends on the signedness of the char type. The macro should first cast the chars to uint8_t.
#4229	3 years ago	fixed	doc	Ryan Long	Gedare Bloom <gedare@…>	3 years ago
Summary	conf.py files need updated for sphinxcontrib-bibtex ver. 2.0.0 and greater
Description	I followed the instructions in the README, but I still ran across this error. I was able to get around it by specifying the version ofsphinxcontrib-bibtex to be 1.0.0. This fix might probably needs to be applied to the 5 release branch as well. Command line output Checking for 'sphinxcontrib.bibtex' : not found (see README.txt) The configuration failed (complete log in /home/rlong/nta-workspace/nta-docs/build/config.log) Output in ${TOP}/build/config.log err: WARNING: Since v2.0, Sphinx uses "index" as master_doc by default. Please add "master_doc = 'contents'" to your conf.py. found ---------------------------------------- Checking for 'sphinxcontrib.bibtex' [1/1] Processing ^[[01;34mbuild/.conf_check_70571fadfd25a3c4e29a4087f18cfa38/testbuild/contents.rst^[[0m /home/rlong/nta-workspace/nta-docs/sphinx/bin/sphinx-build -b html -D extensions=sphinxcontrib.bibtex -C . out out: Running Sphinx v3.4.3 making output directory... done err: Extension error: You must configure the bibtex_bibfiles setting from /home/rlong/nta-workspace/nta-docs/user: Test does not build: Traceback (most recent call last): File "/home/rlong/nta-workspace/nta-docs/.waf3-2.0.19-1f3c580272b15a03d2566843c5fe872a/waflib/Configure.py", line 333, in run_build bld.compile() File "/home/rlong/nta-workspace/nta-docs/.waf3-2.0.19-1f3c580272b15a03d2566843c5fe872a/waflib/Build.py", line 176, in compile raise Errors.BuildError(self.producer.error) waflib.Errors.BuildError: Build failed -> task in '' failed with exit status 2 (run with -v to display more information) not found (see README.txt) from /home/rlong/nta-workspace/nta-docs/user: The configuration failed
#4230	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Timeout for automatic barriers is broken
Description	A barrier wait timeout at an automatic barrier doesn't decrement the count of waiting threads.
#4231	3 years ago	fixed	bsps	Chris Johns	Chris Johns	3 years ago
Summary	MVME 2600/2700 has no console output
Description	A single character is output then nothing further. Tests output using the printk but spconsole01 fails. It seems there is a problem with the UART driver interrupts.
#4237	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	consolesimpletask.c: Two Unchecked Return Values Spotted by Coverity
Description	CID 1437625: Unchecked return value from call to rtems_task_create(). CID 1472765: Unchecked return value from call to rtems_task_start(). Snippet from Coverity output. 234 IMFS_add_node( "/dev", &cons->Node, NULL ); 235 CID 1437625 (#1 of 1): Unchecked return value (CHECKED_RETURN) 1. check_return: Calling rtems_task_create without checking return value (as is done elsewhere 22 out of 25 times). 236 rtems_task_create( 237 rtems_build_name('C', 'O', 'N', 'S'), 238 RTEMS_MAXIMUM_PRIORITY - 1, 239 RTEMS_MINIMUM_STACK_SIZE, 240 RTEMS_DEFAULT_ATTRIBUTES, 241 RTEMS_DEFAULT_MODES, 242 &cons->task 243 ); 244 CID 1472765: Unchecked return value (CHECKED_RETURN) [select issue] 245 rtems_task_start( 246 cons->task, 247 _Console_simple_task_Task, 248 (rtems_task_argument) cons 249 ); 250}
#4238	3 years ago	fixed	bsps	Chris Johns	Chris Johns	3 years ago
Summary	Change motorola_powerpc bsp to support irq-generic
Description	Update the motorola_powerpc to support irq-generic moving the IRQ management to the IRQ server. This enables libbsd support for interrupts.
#4239	3 years ago	fixed	bsps	Chris Johns	Chris Johns	3 years ago
Summary	PowerPC shared ISA IRQ support is broken
Description	The PowerPC shared ISA IRQ management in i8259s.c is broken. It fails to handle: Interrupt acknowledgements for the slave controller when the master has a lower priority pending interrupt acknowledge the slave and the master controllers The nesting support to suspend and resume an interrupt does not correctly handle the enable/disable The initialise of an interrupt does not handle any pending request Locks up when using with the IRQ server which is executing the handler in a task context All these issues have been fixed in the i386 BSP however the code between is not shared as the i386 BSP has BSP specific IRQ management code as well.
#4240	3 years ago	fixed	lib/debugger	Ryan Long	Chris Johns	3 years ago
Summary	rtems-debugger-target.c: Dereference before null check spotted by Coverity
Description	CID 1468682: swbreaks is assigned a value dereferenced from target before it has been verified that target is not null. 173 rtems_debugger_target* target = rtems_debugger->target; deref_ptr: Directly dereferencing pointer target. 174 rtems_debugger_target_swbreak* swbreaks = target->swbreaks.block; 175 size_t swbreak_size; 176 uint8_t* loc = (void*) addr; 177 size_t i; 178 int r; 179 CID 1468682 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking target suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 180 if (target == NULL || swbreaks == NULL || kind != target->breakpoint_size) { 181 errno = EIO; 182 return -1; 183 }
#4241	3 years ago	fixed	lib/debugger	Ryan Long	Chris Johns	3 years ago
Summary	rtems-debugger-threads.c: Three Dereference before null check errors spotted by Coverity
Description	CID 1468681: Dereference before null check in rtems_debugger_thread_continue(). CID 1468690: Dereference before null check in rtems_debugger_thread_system_resume(). CID 1468694: Dereference before null check in rtems_debugger_thread_find_index(). 430 rtems_debugger_threads* threads = rtems_debugger->threads; 431 rtems_debugger_thread* current; 432 int r = 0; deref_ptr_in_call: Dereferencing pointer threads. [show details] 433 current = rtems_debugger_thread_current(threads); CID 1468681 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 434 if (threads != NULL && current != NULL) { 347 rtems_debugger_threads* threads = rtems_debugger->threads; 348 rtems_debugger_thread* current; 349 int r = 0; deref_ptr_in_call: Dereferencing pointer threads. [show details] 350 current = rtems_debugger_thread_current(threads); CID 1468690 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 351 if (threads != NULL && current != NULL) { 150 rtems_debugger_threads* threads = rtems_debugger->threads; deref_ptr_in_call: Dereferencing pointer threads. [show details] 151 rtems_debugger_thread* current = rtems_debugger_thread_current(threads); 152 int r = -1; CID 1468694 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 153 if (threads != NULL) {
#4242	3 years ago	duplicate	lib/debugger	Ryan Long	Chris Johns	3 years ago
Summary	rtems-debugger-threads.c: Three Dereference before null check errors spotted by Coverity
Description	CID 1468681: Dereference before null check in rtems_debugger_thread_continue(). CID 1468690: Dereference before null check in rtems_debugger_thread_system_resume(). CID 1468694: Dereference before null check in rtems_debugger_thread_find_index(). 430 rtems_debugger_threads* threads = rtems_debugger->threads; 431 rtems_debugger_thread* current; 432 int r = 0; deref_ptr_in_call: Dereferencing pointer threads. [show details] 433 current = rtems_debugger_thread_current(threads); CID 1468681 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 434 if (threads != NULL && current != NULL) { 347 rtems_debugger_threads* threads = rtems_debugger->threads; 348 rtems_debugger_thread* current; 349 int r = 0; deref_ptr_in_call: Dereferencing pointer threads. [show details] 350 current = rtems_debugger_thread_current(threads); CID 1468690 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 351 if (threads != NULL && current != NULL) { 150 rtems_debugger_threads* threads = rtems_debugger->threads; deref_ptr_in_call: Dereferencing pointer threads. [show details] 151 rtems_debugger_thread* current = rtems_debugger_thread_current(threads); 152 int r = -1; CID 1468694 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking threads suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 153 if (threads != NULL) {
#4244	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Possible infinite recursion in Classic API Signal handling
Description	Due to implementation flaws, there may be an infinite recursion in the signal processing even if ASR mode is disabled during signal processing. The problem is that rtems_task_mode() is used to set up and tear down the ASR-specific task mode. This needs to be changed to set the modes under protection of the right locks and using a loop instead of recursion to schedule the next ASR processing step if needed.
#4245	3 years ago	fixed	bsps	Chris Johns	Chris Johns	3 years ago
Summary	Add PCI regions support to RTEMS nexus bus support
Description	The RTEMS Nexus bus support in LibBSD assumes a linear address map with a !:1 address mapping between the PCI bus master and CPU. Some target hardware has a different PCI bus master address space to the CPU memory address space. Provide a BSP means to enable PCI support in machine/bus.h in LibBSD Update powerpc/motorola_powerpc to support cache coherent memory Change any PowerPC internal calls that clash with the Linux IO interface used in LibBSD. The PowerPC shared BSP calls have a different call signature
#4246	3 years ago	fixed	network/libbsd	Chris Johns	Chris Johns	17 months ago
Summary	Add libbsd DC NIC support to mvme2700 board
Description	Add support for the Tulip NIC to LibBSD for the MVME 2700 (the mvme2307 bsp).
#4250	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	b1553brm.c: Dereference before null check error spotted by Coverity
Description	CID 1399829: Dereference before null check in brm_control(). rtems_libio_ioctl_args_t *ioarg = (rtems_libio_ioctl_args_t *) arg; deref_ptr: Directly dereferencing pointer ioarg. 997 unsigned int *data = ioarg->buffer; 998 struct bc_msg *cmd_list = (struct bc_msg *) ioarg->buffer; 999 brm_priv *brm; 1000 struct drvmgr_dev *dev; 1001 rtems_device_driver ret; 1002 int len, msglen; 1003 1004 FUNCDBG("brm_control[%d]: [%i,%i]\n", minor, major, minor); 1005 1006 if ( drvmgr_get_dev(&b1553brm_drv_info.general, minor, &dev) ) { 1007 return RTEMS_UNSATISFIED; 1008 } 1009 brm = (brm_priv *)dev->priv; 1010 CID 1399829 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking ioarg suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 1011 if (!ioarg) {
#4251	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	b1553rt.c: Dereference before null check error spotted by Coverity
Description	CID 1399830: Dereference before null check in rt_control(). 621 rtems_libio_ioctl_args_t *ioarg = (rtems_libio_ioctl_args_t *) arg; deref_ptr: Directly dereferencing pointer ioarg. 622 unsigned int *data = ioarg->buffer; 623 624 rt_priv *rt; 625 struct drvmgr_dev *dev; 626 627 FUNCDBG("rt_control[%d]: [%i,%i]\n", minor, major, minor); 628 629 if ( drvmgr_get_dev(&b1553rt_drv_info.general, minor, &dev) ) { 630 return RTEMS_UNSATISFIED; 631 } 632 rt = (rt_priv *)dev->priv; 633 CID 1399830 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking ioarg suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 634 if (!ioarg) {
#4252	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	grtc.c: Dereference before null check error spotted by Coverity
Description	CID 1399840: Dereference before null check in grtc_ioctl(). 1534 rtems_libio_ioctl_args_t *ioarg = (rtems_libio_ioctl_args_t *)arg; deref_ptr: Directly dereferencing pointer ioarg. 1535 unsigned int *data = ioarg->buffer; 1536 int status,frm_len,i,ret; 1537 struct grtc_ioc_buf_params *buf_arg; 1538 struct grtc_ioc_config *cfg; 1539 struct grtc_ioc_hw_status *hwregs; 1540 struct grtc_ioc_pools_setup *pocfg; 1541 struct grtc_ioc_assign_frm_pool *poassign; 1542 struct grtc_frame *frm, *frms; 1543 struct grtc_frame_pool *pool; 1544 struct grtc_list *frmlist; 1545 struct grtc_ioc_stats *stats; 1546 unsigned int mem; 1547 IRQ_LOCAL_DECLARE(oldLevel); 1548 1549 FUNCDBG(); 1550 1551 if ( drvmgr_get_dev(&grtc_drv_info.general, minor, &dev) ) { 1552 return RTEMS_INVALID_NUMBER; 1553 } 1554 pDev = (struct grtc_priv *)dev->priv; 1555 CID 1399840 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking ioarg suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 1556 if (!ioarg)
#4253	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	grspw_pkt.c: Dereference before null check error spotted by Coverity
Description	CID 1399846: Dereference before null check in grspw_addr_ctrl(). 657 struct grspw_priv *priv = d; deref_ptr: Directly dereferencing pointer priv. 658 struct grspw_regs *regs = priv->regs; 659 unsigned int ctrl, nodeaddr; 660 SPIN_IRQFLAGS(irqflags); 661 int i; 662 CID 1399846 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking priv suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 663 if (!priv || !cfg)
#4254	3 years ago	fixed	lib	Ryan Long	Ryan Long <thisisryanlong@…>	3 years ago
Summary	shell.c: Dereference before null check error spotted by Coverity
Description	CID 1467420: Dereference before null check in rtems_shell_line_editor(). 322 int in_fileno = fileno(in); deref_ptr_in_call: Dereferencing pointer out. 323 int out_fileno = fileno(out); 324 325 /* 326 * Only this task can use this file descriptor because calling 327 * fileno will block if another thread call made a call on this 328 * descriptor. 329 */ CID 1467420 (#1 of 1): Dereference before null check (REVERSE_INULL)check_after_deref: Null-checking out suggests that it may be null, but it has already been dereferenced on all paths leading to the check. 330 output = (out && isatty(in_fileno));
#4258	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	printertask.c: Unchecked return value errors spotted by Coverity
Description	CID 1399710: Unchecked return value in printer_task(). 138 switch ( buffer->action_kind ) { 139 case ACTION_WRITE: 140 write( fd, &buffer->data[ 0 ], buffer->action_data.size ); 141 printer_task_append_buffer( ctx, &ctx->free_buffers, buffer ); 4. Breaking from switch. 8. Breaking from switch. 15. Breaking from switch. 142 break; 143 case ACTION_DRAIN: CID 1399710 (#1 of 1): Unchecked return value (CHECKED_RETURN)19. check_return: Calling fsync without checking return value (as is done elsewhere 4 out of 5 times). 144 fsync(fd);
#4259	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	grspw.c: Unchecked return value error spotted by Coverity
Description	CID 1399781: Unchecked return value in grspw_device_init(). 551 if ( value ) 552 pDev->bd_dma_area = value->i; 553 13. Condition grspw_buffer_alloc(pDev), taking false branch. 554 if (grspw_buffer_alloc(pDev)) 555 return RTEMS_NO_MEMORY; 556 557 /* Create semaphores */ CID 1399781 (#2 of 2): Unchecked return value (CHECKED_RETURN)14. check_return: Calling rtems_semaphore_create without checking return value (as is done elsewhere 15 out of 17 times). 558 rtems_semaphore_create(
#4260	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	record-client.c: Unchecked return value error spotted by Coverity
Description	CID 1456677: Unchecked return value in resolve_hold_back(). 191 for ( index = begin_index; index < per_cpu->item_index; ++index ) { 192 const rtems_record_item_64 *item; 193 194 item = &per_cpu->items[ index ]; CID 1456677 (#1 of 1): Unchecked return value (CHECKED_RETURN)12. check_return: Calling visit without checking return value (as is done elsewhere 4 out of 5 times). 195 visit( ctx, item->event, item->data );
#4261	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	getgrent.c: Unchecked return value error spotted by Coverity
Description	CID 1459004: Unchecked return value in endgrent(). 102 if (ctx->fp != NULL) { 103 fclose(ctx->fp); 104 } 105 106 free(ctx); CID 1459004 (#1 of 1): Unchecked return value (CHECKED_RETURN)3. check_return: Calling pthread_setspecific without checking return value (as is done elsewhere 5 out of 6 times). 107 pthread_setspecific(grp_key, NULL);
#4262	3 years ago	fixed	lib/debugger	Ryan Long	Chris Johns	3 years ago
Summary	rtems-debugger-threads.c: Unchecked return value error spotted by Coverity
Description	CID 1468688: Unchecked return value in snapshot_thread(). 262 else { 263 rtems_status_code sc; 264 sc = rtems_task_suspend(id); 265 if (sc != RTEMS_SUCCESSFUL && sc != RTEMS_ALREADY_SUSPENDED) { 266 rtems_debugger_printf("error: rtems-db: thread: suspend: %08" PRIx32 ": %s\n", 267 id, rtems_status_text(sc)); 268 r = -1; 269 } 270 } 271 272 /* 273 * Read the target registers into the thread register array. 274 */ CID 1468688 (#1 of 1): Unchecked return value (CHECKED_RETURN)17. check_return: Calling rtems_debugger_target_read_regs without checking return value (as is done elsewhere 4 out of 5 times). 275 rtems_debugger_target_read_regs(thread);
#4264	3 years ago	fixed	network/libbsd	GabrielMoyano	Jan Sommer	3 years ago
Summary	Activate ehci_pci in rtems-libbsd (cloned)
Description	Cloned from #4263: Import ehci_pci from freebsd-org using freebsd-to-rtems.py
#4265	3 years ago	wontfix	rtems	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Add rtems_get_build_hash()
Description	Add a function to report a hash value of the build environment in the test suite results. /* Generated from spec:/rtems/config/if/get-build-hash */ /** * @ingroup RTEMSAPIConfig * * @brief Gets the RTEMS build hash. * * The build hash is calculated from all key-value pairs of the build * environment. Local file system paths in the values do not contribute to the * hash value. * * @return Returns the pointer to the RTEMS build hash. * * @par Notes * The build hash can be used to distinguish test suite results obtained from * different build environments. * * @par Constraints * @parblock * The following constraints apply to this directive: * * * The directive may be called from within any runtime context. * * * The directive will not cause the calling task to be preempted. * @endparblock */ const char *rtems_get_build_hash( void );
#4267	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	19 months ago
Summary	Add rtems_get_target_hash()
Description	Add a function to report a hash value characterizing the target system in the test suite results. /* Generated from spec:/rtems/config/if/get-target-hash */ /** * @ingroup RTEMSAPIConfig * * @brief Gets the RTEMS target hash. * * The target hash is calculated from BSP-specific values which characterize a * target system. * * @return Returns the pointer to the RTEMS target hash. * * @par Notes * @parblock * For example, the device tree, settings of the memory controller, a serial * number of a chip may be used to calculate the target hash. * * The target hash can be used to distinguish test suite results obtained from * different target systems. * @endparblock * * @par Constraints * @parblock * The following constraints apply to this directive: * * * The directive may be called from within any runtime context. * * * The directive will not cause the calling task to be preempted. * @endparblock */ const char *rtems_get_target_hash( void );
#4269	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	19 months ago
Summary	Add rtems_get_build_label()
Description	Add the following directive: /* Generated from spec:/rtems/config/if/get-build-label */ /** * @ingroup RTEMSAPIConfig * * @brief Gets the RTEMS build label. * * The build label is a user-provided string defined by the build * configuration. * * @return Returns the pointer to the RTEMS build label. * * @par Notes * The build label can be used to distinguish test suite results obtained from * different build configurations. A use case is to record test results with * performance data to track performance regressions. For this a database of * performance limits is required. The build label and the target hash * obtained from rtems_get_target_hash() can be used as a key to obtain * performance limits. * * @par Constraints * @parblock * The following constraints apply to this directive: * * * The directive may be called from within any runtime context. * * * The directive will not cause the calling task to be preempted. * @endparblock */ const char *rtems_get_build_label( void );
#4270	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	A failing task extension produces zombi objects and resource leaks
Description	In _Thread_Initialize() we have this code: _Objects_Open_u32( &information->Objects, &the_thread->Object, config->name ); /* * We assume the Allocator Mutex is locked and dispatching is * enabled when we get here. We want to be able to run the * user extensions with dispatching enabled. The Allocator * Mutex provides sufficient protection to let the user extensions * run safely. */ extension_status = _User_extensions_Thread_create( the_thread ); if ( extension_status ) return true; If an extension fails, then the thread object is not closed. Also the delete extensions are not called. If a later create extension fails, the earlier create extensions may have allocated resources which could be freed by a corresponding delete extension.
#4277	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	grtc.c: Missing break in switch errors spotted by Coverity
Description	CID 1399761: Missing break in switch in process_dma(). CID 1399765: Missing break in switch in process_dma(). CID 1399766: Missing break in switch in process_dma(). CID 1399777: Missing break in switch in process_dma(). CID 1399761 1366 pDev->frm = NULL; 1367 pDev->frame_state = FRM_STATE_HDR; 1368 CID 1399761: Missing break in switch (MISSING_BREAK) [select issue] fallthrough: The above case falls through to this one. 1369 case FRM_STATE_HDR: 1370 DBG2("FRAME_STATE_HDR\n"); CID 1399765 1450 pDev->frame_state = FRM_STATE_PAYLOAD; 1451 pDev->frm = frm; 1452 1453 case FRM_STATE_PAYLOAD: 1454 DBG2("FRAME_STATE_PAYLOAD\n"); CID 1399766 1450 pDev->frame_state = FRM_STATE_PAYLOAD; 1451 pDev->frm = frm; 1452 CID 1399766 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value FRM_STATE_PAYLOAD is not terminated by a break statement. 1453 case FRM_STATE_PAYLOAD: 1454 DBG2("FRAME_STATE_PAYLOAD\n"); CID 1399767 1354 switch( pDev->frame_state ) { CID 1399777 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value FRM_STATE_NONE is not terminated by a break statement. 1355 case FRM_STATE_NONE: 1356 DBG2("FRAME_STATE_NONE\n"); 1357 1358 /* Find Start of next frame by searching for 0x01 */ 1359 ret = grtc_hw_find_frm(pDev); 1360 if ( ret != 0 ) { 1361 /* Frame start not found */ 1362 return 0; 1363 } 1364 1365 /* Start of frame found, Try to copy header */ 1366 pDev->frm = NULL; 1367 pDev->frame_state = FRM_STATE_HDR;
#4278	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	cpuusagetop.c: Missing break in switch errors spotted by Coverity
Description	CID 1399726: Missing break in switch in task_usage(). CID 1399728: Missing break in switch in task_usage(). CID 1399742: Missing break in switch in task_usage(). CID 1399726 233 continue; CID 1399726 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value 2U is not terminated by a break statement. 234 case RTEMS_TOP_SORT_CURRENT_PRI: 235 if ( 236 _Thread_Get_priority( thread ) 237 > _Thread_Get_priority( data->tasks[j] ) 238 ) { 239 continue; 240 } fallthrough: The above case falls through to this one. CID 1399728 230 continue; CID 1399728 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value 1U is not terminated by a break statement. 231 case RTEMS_TOP_SORT_REAL_PRI: 232 if (thread->Real_priority.priority > data->tasks[j]->Real_priority.priority) 233 continue; fallthrough: The above case falls through to this one. 234 case RTEMS_TOP_SORT_CURRENT_PRI: CID 1399742 226 continue; CID 1399742 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value 3U is not terminated by a break statement. 227 case RTEMS_TOP_SORT_TOTAL: 228 if (CPU_usage_Equal_to(&usage, &data->zero) || 229 CPU_usage_Less_than(&usage, &data->usage[j])) 230 continue; fallthrough: The above case falls through to this one. 231 case RTEMS_TOP_SORT_REAL_PRI:
#4279	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	pci_cfg_read.c: Missing break in switch error spotted by Coverity
Description	CID 1399723: Missing break in switch in pci_read_addressable(). 70 switch (type) { 71 case PCI_RES_IO: 72 range0 = &bus->dev.resources[BRIDGE_RES_IO]; 73 break; CID 1399723 (#1 of 1): Missing break in switch (MISSING_BREAK)unterminated_case: The case for value 3 is not terminated by a break statement. 74 case PCI_RES_MEM: 75 range1 = &bus->dev.resources[BRIDGE_RES_MEM]; fallthrough: The above case falls through to this one. 76 default: 77 case PCI_RES_MEMIO:
#4280	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gen_uuid.c: Unchecked return value from library errors spotted by Coverity
Description	CID 1049146: Unchecked return value from library in get_clock(). CID 1049147: Unchecked return value from library in get_random_fd(). CID 1049146 427 rewind(state_f); 428 fl.l_type = F_UNLCK; CID 1049146 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)18. check_return: Calling fcntl(state_fd, 8, &fl) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 429 fcntl(state_fd, F_SETLK, &fl); 430 } 431 432 *clock_high = clock_reg >> 32; CID 1049147 167 if (i >= 0) CID 1049147 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)5. check_return: Calling fcntl(fd, 2, i | 1) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 168 fcntl(fd, F_SETFD, i | FD_CLOEXEC); 169 }
#4281	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	main_chmod.c: Unchecked return value from library error spotted by Coverity
Description	CID 1063856: Unchecked return value from library in rtems_shell_main_chmod(). 55 for (n=2 ; n < argc ; n++) CID 1063856 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)4. check_return: Calling chmod(argv[n++], mode) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 56 chmod(argv[n++], mode); 57 58 return 0; 59}
#4282	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	pwdgrp.c: Unchecked return value from library error spotted by Coverity
Description	CID 1255518: Unchecked return value from library in pwdgrp_init(). 65 /* 66 * Do the best to create this directory. 67 */ CID 1255518 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)1. check_return: Calling mkdir("/etc", 493U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 68 mkdir("/etc", S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IXOTH);
#4283	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_rasta_spw_router.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399760: Unchecked return value from library in gr_rasta_spw_router_init1(). 412 priv->prefix[14] += dev->minor_drv; CID 1399760 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 413 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 414 priv->prefix[15] = '/'; 415 priv->prefix[16] = '\0'; 416
#4284	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_rasta_io.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399763: Unchecked return value from library in gr_rasta_io_init1(). 575 strcpy(priv->prefix, "/dev/rastaio0"); 576 priv->prefix[12] += dev->minor_drv; CID 1399763 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 577 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 578 priv->prefix[13] = '/'; 579 priv->prefix[14] = '\0';
#4285	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_leon4_n2x.c: Unchecked return value error from library error spotted by Coverity
Description	CID 1399767: Unchecked return value error from library in gr_cpci_leon4_n2x_init1(). 495 strcpy(priv->prefix, "/dev/leon4n2x0"); 496 priv->prefix[13] += dev->minor_drv; CID 1399767 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 497 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 498 priv->prefix[14] = '/'; 499 priv->prefix[15] = '\0';
#4286	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_rasta_tmtc.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399771: Unchecked return value from library in gr_rasta_tmtc_init1(). 588 strcpy(priv->prefix, "/dev/rastatmtc0"); 589 priv->prefix[14] += dev->minor_drv; CID 1399771 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 590 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 591 priv->prefix[15] = '/'; 592 priv->prefix[16] = '\0';
#4287	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_rasta_adcdac.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399780: Unchecked return value from library in gr_rasta_adcdac_init1(). 396 strcpy(priv->prefix, "/dev/rastaadcdac0"); 397 priv->prefix[16] += dev->minor_drv; CID 1399780 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 398 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 399 priv->prefix[17] = '/'; 400 priv->prefix[18] = '\0';
#4288	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_701.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399783: Unchecked return value from library in gr701_init1(). 331 strcpy(priv->prefix, "/dev/gr701_0"); 332 priv->prefix[11] += dev->minor_drv; CID 1399783 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 333 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 334 priv->prefix[12] = '/'; 335 priv->prefix[13] = '\0';
#4289	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_tmtc_1553.c: Unchecked return value from library error spotted by Coverity
Description	CID 1399785: Unchecked return value from library in gr_tmtc_1553_init1(). 315 strcpy(priv->prefix, "/dev/tmtc1553_0"); 316 priv->prefix[14] += dev->minor_drv; CID 1399785 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 317 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 318 priv->prefix[15] = '/'; 319 priv->prefix[16] = '\0';
#4290	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	gr_cpci_gr740.c: Unchecked return value from library error spotted by Coverity
Description	CID 1437630: Unchecked return value from library in gr_cpci_gr740_inti1(). 472 strcpy(priv->prefix, "/dev/gr740_0"); 473 priv->prefix[11] += dev->minor_drv; CID 1437630 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 474 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 475 priv->prefix[12] = '/'; 476 priv->prefix[13] = '\0';
#4291	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	main_help.c: Unchecked return value from library error spotted by Coverity
Description	CID 1437650: Unchecked return value from library in rtems_shell_help(). 472 strcpy(priv->prefix, "/dev/gr740_0"); 473 priv->prefix[11] += dev->minor_drv; CID 1437630 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)2. check_return: Calling mkdir(priv->prefix, 511U) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 474 mkdir(priv->prefix, S_IRWXU | S_IRWXG | S_IRWXO); 475 priv->prefix[12] = '/'; 476 priv->prefix[13] = '\0';
#4294	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	satcan.c: Unsigned compared against 0 error spotted by Coverity
Description	CID 1399768: Unsigned compared against 0 in satcan_ioctl(). 303 DBG("SatCAN: ioctl: masking register %d, with value %x\n\r", 304 regmod->reg, regmod->val); unsigned_compare: This less-than-zero comparison of an unsigned value is never true. regmod->reg < 0U. 305 if (regmod->reg < 0) 306 return RTEMS_INVALID_NAME; 307 else if (regmod->reg <= SATCAN_FILTER_STOP)
#4295	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	b1553brm.c: Unsigned compared against 0 error spotted by Coverity
Description	CID 1399773: Unsigned compared against 0 in brm_write(). 983 rw_args->bytes_moved = count; 984 CID 1399773 (#1 of 1): Unsigned compared against 0 (NO_EFFECT)unsigned_compare: This greater-than-or-equal-to-zero comparison of an unsigned value is always true. count >= 0U. 985 if (count >= 0) { 986 return RTEMS_SUCCESSFUL; 987 }
#4301	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	grcan.c: Unused value error spotted by Coverity
Description	CID 1437613: Unused value in grcan_set_sfilter(). 1788 /* disable TX/RX SYNC filtering */ CID 1437613 (#1 of 1): Unused value (UNUSED_VALUE)assigned_value: Assigning value 4294967295ULL to pDev->sfilter.mask here, but that stored value is overwritten before it can be used. 1789 pDev->sfilter.mask = 0xffffffff; value_overwrite: Overwriting previous write to pDev->sfilter.mask with value 0ULL. 1790 pDev->sfilter.mask = 0;
#4302	3 years ago	fixed	arch/powerpc	Joel Sherrill	Joel Sherrill <joel@…>	2 years ago
Summary	Obsolete powerpc/haleakala board
Description	This BSP was merged in 1995 and there has been no mention of it recently. There do not appear to be any recent modifications to this BSP that are not part of a generic sweep of BSPs. Removing this will reduce the number of BSPs which need modification to account for changes being introduced to support libbsd bus interface properly on PowerPC.
#4320	3 years ago	fixed	bsps	Jan Sommer	Jan Sommer	3 years ago
Summary	Add driver for cadence-spi device for xilinx based BSPs
Description	Implement a spidev compatible driver for using the cadence-spi device of the Zynq devices in standard SPI mode.
#4321	3 years ago	fixed	bsps	Jan Sommer	Jan Sommer	3 years ago
Summary	Add spi driver for AXI SPI ip core from Xilinx
Description	Add a spidev compatible driver for the xilinx axi quad spi ip core for standard SPI mode
#4323	3 years ago	fixed	bsps	Daniel Hellstrom	Daniel Hellstrom	3 years ago
Summary	grlib,can: introduce a new common CAN baud-rate timing calculating functions
Description	There is no point in having two different implementations specific per GRLIB CAN device driver (OCCAN and GRCAN). Instead we can use the common CAN timing definitions and translate that into the device specific parameters using a new common calculation function.
#4324	3 years ago	fixed	bsps	Daniel Hellstrom	Daniel Hellstrom	3 years ago
Summary	grlib,grcanfd: extend the GRCAN driver with GRCANFD support
Description	The new GRCAN_FD IP supports CAN FD standard and is mostly backwards compatible with GRCAN SW interface. It introduces support for CAN-FD frames and a new bit-rate timing registers for new baud-rate. The GRCAN driver have been extended to support the GRCANFD IP using the same API. Device IP-core specific functions are split into different files (grcanstd.c and grcanfd.c). (this was fixed in 5 branch, see ticket #4307).
#4326	3 years ago	fixed	score	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	objectextendinformation.c: Dereference after null check error spotted by Coverity
Description	CID 26033: Dereference after null check in _Objects_Extend_information(). 172 if ( old_maximum > extend_count ) { 173 /* 174 * Copy each section of the table over. This has to be performed as 175 * separate parts as size of each block has changed. 176 */ CID 26033 (#1 of 1): Dereference after null check (FORWARD_NULL)11. var_deref_model: Passing null pointer information->object_blocks to memcpy, which dereferences it. 177 memcpy( 178 object_blocks, 179 information->object_blocks, 180 block_count * sizeof( *object_blocks ) 181 );
#4329	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	main_rtrace.c: Dereference after null check error spotted by Coverity
Description	CID 1399709: Dereference after null check in rtems_trace_buffering_shell_save(). 6. var_compare_op: Comparing buf to null implies that buf might be null. 472 if (!buf) 473 { 474 close (out); 475 printf ("error: no memory\n"); 476 } 477 CID 1399709 (#1 of 1): Dereference after null check (FORWARD_NULL)7. var_deref_model: Passing null pointer buf to memset, which dereferences it. 478 memset (buf, 0, SAVE_BUF_SIZE);
#4331	3 years ago	fixed	lib	Ryan Long	Daniel Hellstrom	14 months ago
Summary	main_pci.c: Dereference after null check error spotted by Coverity
Description	CID 1399716: Dereference after null check in shell_pci_infodev(). 10. var_compare_op: Comparing bus to null implies that bus might be null. 358 if (!bus) { CID 1399716 (#1 of 1): Dereference after null check (FORWARD_NULL)11. var_deref_op: Dereferencing null pointer bus. 359 printf(" AT BUS: 0x%x via Host Bridge\n", bus->num); 360 } else {
#4336	3 years ago	fixed	arch/sparc	Daniel Hellstrom	Daniel Hellstrom	3 years ago
Summary	Improve the workaround for the LEON3FT store-store errata: TN-0009 on master
Description	See also: ​​https://www.gaisler.com/doc/antn/GRLIB-TN-0009.pdf Some areas were already fixed by #3057. RTEMS-5 branch ticket #4154.
#4338	3 years ago	fixed	rtems	Frank Kuehndel	Sebastian Huber	21 months ago
Summary	rtems_clock_set(): Cannot set future dates later than approximately 2105
Description	Short Problem Description ​RTEMS Classic API Guide says: RTEMS can represent time points of this clock in nanoseconds ranging from 1988-01-01T00:00:00.000000000Z to 2514-05-31T01:53:03.999999999Z. Yet, years larger than roughly 2105 to 2110 cannot be set (or at least the date should be wrong but this never occured in my tests). The possible RETURN VALUES are RTEMS_SUCCESSFUL, RTEMS_INVALID_ADDRESS, RTEMS_INVALID_CLOCK Yet, rtems_clock_set() can return status RTEMS_INVALID_NUMBER, too. (Only for such dates in the far future.) How To Replicate? Call rtems_clock_set() with this time_of-day: { year = 2514, month = 5, day = 31, hour = 1, minute = 53, second = 3, ticks = 995 } The return value will be RTEMS_INVALID_NUMBER. Bugs in _TOD_To_seconds() cpukit/rtems/src/clockset.c: rtems_clock_set() calls cpukit/rtems/src/clocktodvalidate.c: _TOD_Validate() and cpukit/rtems/src/clocktodtoseconds.c: _TOD_To_seconds() and cpukit/score/src/coretodset.c: _TOD_Set() First issue: _TOD_To_seconds() converts the time_of_day structure into seconds using a variable time of type uint32_t. This simply overflows when in comes close to year 2110. Debugger output at the end of _TOD_To_seconds(): (gdb) print the_tod->year $16 = 2104 (gdb) print time $17 = 4233686400 with a higher year: (gdb) print the_tod->year $28 = 2514 (gdb) print *the_tod $31 = { year = 2514, month = 5, day = 31, hour = 1, minute = 53, second = 3, ticks = 995 } (gdb) print time $29 = 192272 Second issue: _TOD_To_seconds() can (most likely) not handle the leap year issues of the years 2200, 2300, 2400, 2500 because it has dedicated code for the 2100 case only: /* The year 2100 is not a leap year */ if ( time >= (TOD_SECONDS_AT_2100_03_01_00_00 - TOD_SECONDS_1970_THROUGH_1988)) { time -= TOD_SECONDS_PER_DAY; } _TOD_Check_time_of_day_and_run_hooks() causes STATUS_INVALID_NUMBER cpukit/score/src/coretodset.c: _TOD_Set() calls cpukit/score/src/coretodset.c: _TOD_Check_time_of_day_and_run_hooks() in this code snippet (note return status): status = _TOD_Check_time_of_day_and_run_hooks( tod ); if ( status != STATUS_SUCCESSFUL ) { _TOD_Release( lock_context ); return status; } _TOD_Check_time_of_day_and_run_hooks( tod ) can return status STATUS_INVALID_NUMBER which is not documented for rtems_clock_set(). The small time in seconds value 192272 from the integer overrun discussed above triggers the middle if clause: static Status_Control _TOD_Check_time_of_day_and_run_hooks( const struct timespec *tod ) { if ( !_Watchdog_Is_valid_timespec( tod ) ) { return STATUS_INVALID_NUMBER; } if ( tod->tv_sec < TOD_SECONDS_1970_THROUGH_1988 ) { return STATUS_INVALID_NUMBER; } if ( _Watchdog_Is_far_future_timespec( tod ) ) { return STATUS_INVALID_NUMBER; } return _TOD_Hook_Run( TOD_ACTION_SET_CLOCK, tod ); } Final Notes I found #2665 and #2548 but I do not say these are relevant here. #define WATCHDOG_MAX_SECONDS 0x3ffffffff in cpukit/include/rtems/score/watchdogimpl.h covers 17179869183 seconds which are some 544 years _TOD_Check_time_of_day_and_run_hooks() seems to be only used by _TOD_Set() but _TOD_Set() has three users. I did not check whether the invers conversation (rtems_clock_get_tod()) works for dates which are centuries in the future.
#4339	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	main_cp.c: Unused value error spotted by Coverity
Description	CID 1255344: Unused value in copy(). 315 case FTS_DC: /* Warn, continue. */ 316 warnx("%s: directory causes a cycle", curr->fts_path); CID 1255344 (#5 of 5): Unused value (UNUSED_VALUE)assigned_value: Assigning value from rval to badcp here, but that stored value is overwritten before it can be used. 317 badcp = rval = 1; 318 continue;
#4340	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	print-ls.c: Unused value error spotted by Coverity
Description	CID 1255346: Unused value in rtems_shell_ls_printacol 327 if (col >= numcols) { CID 1255346 (#1 of 1): Unused value (UNUSED_VALUE)assigned_value: Assigning value from col to chcnt here, but that stored value is overwritten before it can be used. 328 chcnt = col = 0; 329 (void)putchar('\n'); 330 } value_overwrite: Overwriting previous write to chcnt with value from printaname(globals, p, dp->s_inode, ((globals->f_humanize && 0) ? dp->s_size : dp->s_block)). 331 chcnt = printaname(globals, p, dp->s_inode,
#4341	3 years ago	fixed	lib/dl	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	fastlz.c: Unused value error spotted by Coverity
Description	CID 1399751: Unused value in fastlz2_compress(). 219 if(ip[0] == ip[-1] && FASTLZ_READU16(ip-1)==FASTLZ_READU16(ip+1)) 220 { 221 distance = 1; CID 1399751 (#1 of 1): Unused value (UNUSED_VALUE)assigned_pointer: Assigning value from ip + 3 to ip here, but that stored value is overwritten before it can be used. 222 ip += 3;
#4342	3 years ago	fixed	bsps	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	disp_hcms29xx.c: Unused value error spotted by Coverity
Description	CID 1399752: Unused value in disp_hcms29xx_update_task(). 598 if (rc == RTEMS_SUCCESSFUL) { CID 1399752 (#1 of 1): Unused value (UNUSED_VALUE)returned_value: Assigning value from rtems_semaphore_release(softc_ptr->disp_param.trns_sema_id) to rc here, but that stored value is overwritten before it can be used. 599 rc = rtems_semaphore_release(softc_ptr->disp_param.trns_sema_id); 600 }
#4345	3 years ago	fixed	network/libbsd	Till Straumann	Heinz Junkes <junkes@…>	2 years ago
Summary	Nexus Driver for Marvell mv643xx Ethernet Adapter (powerpc/beatnik BSP, MVME6100 board)
Description	This ​pull request implements a nexus driver (on top of the 5-freebsd-12 branch) for the Marvell 643xx ethernet adapter which is used on the MVME6100 board (powerpc/beatnik BSP). Note that this driver builds on top of the low-level driver which is part of the beatnik BSP and requires a patch to the rtems bsp which was submitted in a related ​ticket.
#4346	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Require RTEMS_PRIORITY for MrsP semaphores
Description	The MrsP semaphores are a generalization the priority ceiling semaphores for SMP configurations. Priority ceiling semaphores are required to use the priority task wait queue discipline. Require this discipline also for MrsP semaphores.
#4355	3 years ago	fixed	test	Ryan Long	joel@…	3 years ago
Summary	Remove leon3 test configurations which assume sis is built into gdb from rtems-test
Description	When trying to run the tests with rtems-test on leon3, every test was invalid except for minimum. rtems-test is still using sis through gdb when it should be using the dedicated sis executable.
#4356	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	2 years ago
Summary	rtems_semaphore_set_priority() uses an invalid SMP lock
Description	If the priority of a locked priority ceiling mutex is set, the used SMP lock sequence is invalid.
#4358	3 years ago	fixed	admin	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Priority discipline is broken for semaphores and message queues in SMP configurations
Description	The priority queues in clustered scheduling configurations use a per scheduler priority queue rotation to ensure FIFO fairness across schedulers. This mechanism is implemented in the thread queue surrender operation. Unfortunately some semaphore and message queue directives use wrongly the thread queue extract operation.
#4363	3 years ago	wontfix	tool/gdb	Chris Johns	Chris Johns	17 months ago
Summary	GDB does not build on Windows for MSYS2
Description	The details are in the GDB bug ... ​https://sourceware.org/bugzilla/show_bug.cgi?id=27657
#4365	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	main_cp.c: Ignore return value from stat()
Description	CID 26051: Unchecked return value from library in main_cp(). 257 if (Rflag && (Lflag || Hflag)) CID 26051 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)34. check_return: Calling stat(*argv, &tmp_stat) without checking return value. This library function may fail and return an error code. [Note: The source code implementation of the function has been overridden by a builtin model.] 258 stat(*argv, &tmp_stat);
#4368	3 years ago	fixed	tool	Chris Johns	Joel Sherrill	3 years ago
Summary	rtems--tools tool kit temp file constructor does nothing
Description	A change to add an overloaded constructor ... ​https://git.rtems.org/rtems-tools/commit/?id=420d7a13672991a1480d06ac02190f2976b9253b does not do anything. The body of the constructor is empty. Thinking about this I think the change is not right because the code depends on libiberty make temp call to find the temp directory and that is designed to crate safe temp files. I am now wondering how the code was tested? The files registered are not known to the clean up handler that runs on exit.
#4374	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	Handle symbol sets in covoar
Description	The covoar tool should be changed to handle all symbol sets for a coverage run rather than being called by coverage.py for each symbol set. Currently, coverage.py calls covoar once for each symbol set found in the symbol-sets.ini configuration file. This leads to a lot of redundancy. For each symbol set, covoar has to process the objdumps and the coverage information of all the executables even though they do not change. Changing covoar to process all symbol sets and generate reports in one invocation will result in a large speedup. This will also require a minor modification to coverage.py to generate the correct .ini file and only call covoar once.
#4375	3 years ago	fixed	fs	Joel Sherrill		2 years ago
Summary	Add futimens() and utimensat()
Description	RTEMS currently only supports utimes() which is marked as Legacy in the POSIX standard. Support needs to be added for futimens() and utimensat() per the following specification: ​https://pubs.opengroup.org/onlinepubs/9699919799/functions/futimens.html utimes() only has second granularity on its times. The newer methods greatly improve on that even if many of the RTEMS file systems do not support nanosecond times on the file access, create, etc timestamps. The following changes have been identified: The utimes file handler entry will need to be replaced with a handler geared to the futimens() API. File systems will have to be updated to account for the new handler. Existing methods like utimes() will have to be adjusted to be implemented in terms of the futimens() style method. File system test cases will have to be added to the fstests.
#4376	3 years ago	fixed	tool	Alex White	Ryan Long <ryan.long@…>	2 years ago
Summary	covoar/ReportsBase Coverage::GenerateReports?() uses raw pointer to std::list
Description	The Coverage::GenerateReports() function uses a variable of type std::list<ReportsBase *> to store the ReportsBase-derived classes used to generate the reports. The function should use an std::vector<std::unique_ptr<ReportsBase>> instead. An std::vector is the preferred container type in this case because the constant-time insertion and removal properties of std::list are not needed. The smart pointer type std::unique_ptr<ReportsBase> is preferred over ReportsBase * because it eliminates the need for manual pointer deletion at the end of the function.
#4377	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	covoar/Explanations Explanations::writeNotFound() NULL check of wrong variable
Description	The Explanations::writeNotFound() method contains two NULL checks for the fileName variable. The second check should be checking if notFoundFile is NULL because it relates to the error being reported in the true case. It appears that this was a simple mistake that was never caught until now.
#4378	3 years ago	fixed	tool	Alex White		3 years ago
Summary	covoar/symbol-sets add libuuid to symbol-sets.ini
Description	libuuid should be added to the list of symbol sets in the symbol-sets.ini configuration file.
#4380	3 years ago	fixed	doc	Joel Sherrill	Joel Sherrill	3 years ago
Summary	Update RTEMS POSIX Compliance Spreadsheet
Description	There have been additions to RTEMS and the tracking spreadsheet now includes FACE Technical Standard, Edition 3.1
#4382	3 years ago	fixed	doc	Joel Sherrill	Joel Sherrill	17 months ago
Summary	mingw - add flex to packages needed
Description	New install of msys2 and pacman instructions in manual missed flex.
#4383	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	covoar keeps DWARF info in memory too long
Description	To get covoar working again, the ELF and DWARF info cleanup was moved from the ExecutableInfo constructor to the ExecutableInfo destructor. Without this change, covoar would report the locations of symbols as unknown:-1. This is because the DWARF info is used later in the ExecutableInfo::getSourceAndLine() method with a call to debug.get_source(). This caused a new issue: the ELF and DWARF info is now held in memory for the life of the program. This causes very high memory usage when running covoar, more than 10 GB for some BSPs. A quick and dirty solution is to find what is being referenced in debug.get_source() and preserve it after debug.end() is called. It appears that refraining from calling cus.clear() in rld::dwarf::file::end() would be a good place to start. Additionally, the rld::dwarf::sources class will likely need to be changed to copy out the DWARF source file info.
#4384	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	Add SMTP login to rtems-tools mailer
Description	The ability to provide a username and password to perform an SMTP login should be added so that rtems-test results can be mailed via an external SMTP server that requires authentication. The following additional command line options should be enough to make it work: --smtp-port --smtp-user --smtp-password The mail-related command line options should be filtered from the mailed logs so that credentials do not get leaked.
#4386	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	covoar missing conditional branch instructions on ARM
Description	Two ARM instructions, cbz and cbnz, are not being treated as conditional branch instructions by covoar currently. They should be added to the list of conditional branch instructions in Target_arm.
#4387	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	covoar marks taken/not taken incorrectly for AArch64
Description	The AArch64 annotated assembly produced by covoar reveals that branches that are never taken are marked "ALWAYS TAKEN" and branches that are always taken are marked "NEVER TAKEN". It appears that this is caused by incorrect return values in TargetBase::qemuTakenBit() and TargetBase::qemuNotTakenBit(). These methods should be overridden in Target_aarch64 so that the return values are switched.
#4388	3 years ago	fixed	build	Alex White	Alex White	3 years ago
Summary	Copy improved mailer.py from rtems-tools
Description	The mailer.py script from rtems-tools has had various improvements made to it that have been left out of the mailer.py script in rtems-source-builder. The mailer.py in rtems-source-builder should be made to mirror the mailer.py script in rtems-tools so that it incorporates the same improvements.
#4389	3 years ago	fixed	lib	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Undefined behaviour if the area size calculation in calloc() and rtems_calloc() overflows
#4390	3 years ago	fixed	admin	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Make zero size allocation result consistent across directives
Description	There is a special case in malloc() which results in a malloc( 0 ) == NULL invariant. This special case should move to rtems_heap_allocate_aligned_with_boundary() to have a consistent behaviour across the API level memory allocation directives.
#4391	3 years ago	fixed	posix	Joel Sherrill	Ryan Long <ryan.long@…>	3 years ago
Summary	SPARC GCC no longer needs sysconf(515) (_SC_STACK_PROT)
Description	In reviewing Ryan's rework of psx13, Gedare asked why there was a test for sysconf(515) conditionally compiled for SPARC. The answer was GCC needed it. But investigated showed that GCC has not had code to call this for over 20 years. It was removed per this commit. commit ac4904440b4733f461bcd2c45915f502b0345f80 Author: Alexandre Oliva <oliva@lsd.ic.unicamp.br> Date: Mon Jan 24 00:47:38 2000 +0000 sol2-sld-64.h (TRANSFER_FROM_TRAMPOLINE): Moved... * config/sparc/sol2-sld-64.h (TRANSFER_FROM_TRAMPOLINE): Moved... * config/sparc/sol2.h: ... here. From-SVN: r31578 This ticket is to remove the code from posix/src/sysconf.c and update psx13 accordingly.
#4392	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems: Constify timer fire when directives
Description	The time of day pointer can be made constant in the timer fire when directives.
#4393	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems: Change rtems_task_get_affinity() status
Description	In case the processor set is not large enough to contain the processor affinity set of the task return RTEMS_INVALID_SIZE instead of RTEMS_INVALID_NUMBER. This is more in line with other directives since the issue is related to the size of an object.
#4395	3 years ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Workspace initialization is broken for arm/imx7 and arm/raspberrypi
Description	Cloned from #4394: The ARMV7_CP15_START_WORKSPACE_ENTRY_INDEX has a wrong value. The bug was introduced by the fix for #4185.
#4396	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Add futimens() and utimensat()
Description	Implement the futimens and utimensat APIs
#4397	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Make utime() use utimensat()
Description	Make utime use the utimensat API
#4398	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Make utimes() use utimensat()
Description	Make utimes use the utimensat API
#4399	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Add tests for futimens() and utimensat()
Description	Create tests to test the futimens() and utimensat() APIs
#4400	3 years ago	fixed	lib	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Change filesystem utime_h to utimens_h to support nanosecond granularity
Description	Change utime_h to utimens_h for the new nanosecond granularity APIs
#4401	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems: Change rtems_scheduler_get_processor_set() status
Description	In case the processor set is not large enough to contain the processor set of the scheduler return RTEMS_INVALID_SIZE instead of RTEMS_INVALID_NUMBER. This is more in line with other directives since the issue is related to the size of an object. A similar change was done in #4393.
#4402	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	rtemstoolkit/mailer.py option added twice
Description	In the add_arguments function, the intent is that the first two options are added to the arguments before the rest of them are iterated and added. The iteration of the remaining options, however, starts with the first option when it should start with the second. Additionally, the order of keys in a dictionary cannot be relied on for all versions of Python we support. The dictionary should be changed to an OrderedDict to preserve key order to facilitate iteration.
#4403	3 years ago	fixed	rtems	Frank Kuehndel	Sebastian Huber	3 years ago
Summary	rtems_timer_fire_when() returns wrong status code when wall_time argument is NULL
Description	The RTEMS Classic API Guide (​https://docs.rtems.org/branches/master/c-user/timer/directives.html#rtems-timer-fire-when) elaborates for rtems_timer_fire_when(): RTEMS_INVALID_ADDRESS The wall_time parameter was NULL. RTEMS_INVALID_CLOCK The time of day was invalid. Yet, calling rtems_timer_fire_when() with wall_time = NULL and all other parameters valid results in RTEMS_INVALID_CLOCK. Have a look into cpukit/rtems/src/timercreate.c function _Timer_Fire_when() to see the cause for this behavior. I attach two patches: sptimer_err01.patch -- extends an existing test to reproduce this issue timercreate.patch -- fixes the issue There are at least two ways to handle this bug and I do not want to decide which way to choose: Change the documentation Change the code For the records: RTEMS_VERSION = 6.0.0 remote origin = git://git.rtems.org/rtems.git commit HEAD = 2c5199bb049efe8e29cd12461dc57bd6e30388e8 (HEAD -> master, origin/master, origin/HEAD) config.ini = [arm/realview_pbx_a9_qemu] RTEMS_DEBUG = True RTEMS_NETWORKING = True RTEMS_POSIX_API = True RTEMS_SMP = True BUILD_TESTS = True NEWLIB_VERSION = 3.2.0 SOURCE_BUILDER = Set Builder, 6 (5e449fb5c2cb) GCC_VERSION = arm-rtems6-gcc (GCC) 10.2.1 20210309 (RTEMS 6, RSB 5e449fb5c2cb6812a238f9f9764fd339cbbf05c2, Newlib d10d0d9) BINUTILS_VERSION = GNU ld (GNU Binutils) 2.36.1.20210309
#4406	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems: Constify rtems_task_wake_when()
Description	The time of day pointer can be made constant in this directive.
#4410	3 years ago	fixed	admin	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems_task_start() does not check that the entry point is not equal to NULL (cloned)
Description	Cloned from #4409: Bug was introduced by commit 33829ce155069462ba410d396da431386369ed08 related to #2555.
#4411	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems_task_restart() should set the real priority to the initial priority
Description	The rtems_task_restart() directive should set the real priority to the initial priority. If another task is restarted this is only under certain conditions. The RTEMS 4.5.0 behaviour was to unconditionally set the real priority to the initial priority during a task restart.
#4412	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Unexpected rtems_task_restart() behaviour if called from within interrrupt context
Description	In rtems_task_restart() there is a check if the executing thread is restarted. However, in interrupt context, this simple check results in an internal error INTERNAL_ERROR_BAD_THREAD_DISPATCH_DISABLE_LEVEL if the executing thread is restarted. Check also if an ISR is in progress to make sure the thread restart works within interrupt context.
#4413	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Allow pthread_cancel() from within interrupt context
Description	The pthread_cancel() does not immediately delete resources if invoked for threads other than the executing thread. With minor changes it may be called also from within interrupt context.
#4414	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Return RTEMS_CALLED_FROM_ISR in rtems_task_delete()
Description	Calling rtems_task_delete() from within interrupt context has an unpredictable behaviour. It depends on which thread is interrupted. Return an error status instead.
#4415	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	ReportsBase?.cc: Division or modulo by float zero error spotted by Coverity
Description	CID 1503709: Division or modulo by float zero in WriteSummaryReport?(). 485 percentage = (double) notExecuted; CID 1503709 (#1 of 1): Division or modulo by float zero (DIVIDE_BY_ZERO)7. divide_by_zero: In expression percentage /= (double)totalBytes, division by expression totalBytes which may be zero has undefined behavior. 486 percentage /= (double) totalBytes; 487 percentage *= 100.0;
#4416	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-exeinfo.cpp: Division or modulo by zero errors spotted by Coverity
Description	CID 1503008: Division or modulo by zero in output_inlined(). CID 1503008 (#1 of 1): Division or modulo by zero (DIVIDE_BY_ZERO)9. divide_by_zero: In expression funcs_inlined.size() * 100UL / total, division by expression total which may be zero has undefined behavior. 801 std::cout << "inlined funcs : " << funcs_inlined.size () << std::endl 802 << " total funcs : " << total << std::endl 803 << " % inline funcs : " << (funcs_inlined.size () * 100) / total << '%' 804 << std::endl CID 1503015: Division or modulo by zero in output_inlined(). CID 1503015 (#1-2 of 2): Division or modulo by zero (DIVIDE_BY_ZERO)18. divide_by_zero: In expression inlined_size * 100UL / total_size, division by expression total_size which may be zero has undefined behavior. 805 << " total size : " << total_size << std::endl 806 << " inline size : " << inlined_size << std::endl 807 << " % inline size : " << (inlined_size * 100) / total_size << '%' 808 << std::endl;
#4417	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	Explanations.cc: Resource leak error spotted by Coverity
Description	CID 1399608: Resource leak in load(). CID 1399608 (#1 of 1): Resource leak (RESOURCE_LEAK)9. leaked_storage: Variable explain going out of scope leaks the storage it points to.
#4418	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovData?.cc: Resource leak error spotted by Coverity
Description	CID 1399610: Resource leak in readFrame(). 250 newFunction = new GcovFunctionData; 5. noescape: Resource newFunction is not freed or pointed-to in readFunctionFrame. [show details] 6. Condition !this->readFunctionFrame(header, gcovFile, newFunction), taking true branch. 251 if ( !readFunctionFrame(header, gcovFile, newFunction) ){ 252 fprintf( stderr, "Error while reading FUNCTION from gcov file...\n" ); CID 1399610 (#1 of 1): Resource leak (RESOURCE_LEAK)7. leaked_storage: Variable newFunction going out of scope leaks the storage it points to. 253 return false; 254 }
#4419	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	TraceReaderLogQEMU.cc: Resource leak error spotted by Coverity
Description	CID 1399615: Resource leak in processFile(). 113 // 114 // Discard Header section 115 // 6. noescape: Resource logFile is not freed or pointed-to in ReadUntilFound. [show details] 7. Condition !ReadUntilFound(logFile, "----------------"), taking true branch. 116 if (! ReadUntilFound( logFile, QEMU_LOG_SECTION_END ) ) { 117 fprintf( stderr, "Unable to locate end of log file header\n" ); CID 1399615 (#2 of 2): Resource leak (RESOURCE_LEAK)8. leaked_storage: Variable logFile going out of scope leaks the storage it points to. 118 return false;
#4420	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	TraceWriterQEMU.cc: Resource leak error spotted by Coverity
Description	CID 1399621: Resource leak in writeFile(). 5. noescape: Resource traceFile is not freed or pointed-to in fwrite. 116 status = ::fwrite( &header, sizeof(trace_header), 1, traceFile ); 6. Condition status != 1, taking true branch. 117 if (status != 1) { 118 std::cerr << "Unable to write header to " << file << std::endl; CID 1399621 (#1 of 1): Resource leak (RESOURCE_LEAK)7. leaked_storage: Variable traceFile going out of scope leaks the storage it points to. 119 return false;
#4421	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	DesiredSymbols?.cc: Resource leak error spotted by Coverity
Description	CID 1503018: Resource leak in load(). 13. alloc_fn: Storage is returned from allocation function operator new. 14. noescape: Resource new Coverage::SymbolInformation is not freed or pointed-to in operator =. [show details] CID 1503018 (#2 of 2): Resource leak (RESOURCE_LEAK)15. leaked_storage: Failing to save or free storage allocated by new Coverage::SymbolInformation leaks it. 104 set[sym.name()] = *(new SymbolInformation);
#4422	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	ReportsBase?.cc: Resource leak error spotted by Coverity
Description	CID 1503711: Resource leak in WriteSummaryReport?(). 548 fprintf( 549 report, 550 "Percentage branch paths covered : %4.4g\n", 551 100.0 - percentageBranches 552 ); 553 } CID 1503711 (#1 of 1): Resource leak (RESOURCE_LEAK)34. leaked_storage: Variable report going out of scope leaks the storage it points to. 554} 555
#4423	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	record-client.c: Unchecked return value error spotted by Coverity
Description	CID 1503011: Unchecked return value in resolve_hold_back(). 191 for ( index = begin_index; index < per_cpu->item_index; ++index ) { 192 const rtems_record_item_64 *item; 193 194 item = &per_cpu->items[ index ]; CID 1503011 (#1 of 1): Unchecked return value (CHECKED_RETURN)12. check_return: Calling visit without checking return value (as is done elsewhere 4 out of 5 times). 195 visit( ctx, item->event, item->data );
#4424	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-bin2c.c: Copy into fixed size buffer spotted by Coverity
Description	CID 1063892: Copy into fixed size buffer in process(). 141 char *ifbasename; 142 ifbasename = basename(ifbasename_to_free); 143 CID 1063892 (#1 of 1): Copy into fixed size buffer (STRING_OVERFLOW)13. fixed_size_dest: You might overrun the 1025-character fixed-size string buf by copying ifbasename without checking the length. 144 strcpy(buf, ifbasename);
#4425	3 years ago	fixed	tool	Ryan Long		2 years ago
Summary	GcovFunctionData?.cc: Copy into fixed size buffer errors spotted by Coverity
Description	CID 1063892: Copy into fixed size buffer in addBlock(). CID 1399613 (#1 of 1): Copy into fixed size buffer (STRING_OVERFLOW)1. fixed_size_dest: You might overrun the 256-character fixed-size string block.sourceFileName by copying sourceFileName without checking the length. 2. parameter_as_source: Note: This defect has an elevated risk because the source argument is a parameter of the current function. 184 strcpy (block.sourceFileName, sourceFileName); CID 1399616: Copy into fixed size buffer in setBlockFileName(). CID 1399616 (#1 of 1): Copy into fixed size buffer (STRING_OVERFLOW)1. fixed_size_dest: You might overrun the 256-character fixed-size string block->sourceFileName by copying fileName without checking the length. 2. parameter_as_source: Note: This defect has an elevated risk because the source argument is a parameter of the current function. 298 strcpy(block->sourceFileName, fileName);
#4426	3 years ago	fixed	posix	Sebastian Huber	Sebastian Huber	3 years ago
Summary	clock_nanosleep() may use wrong clock for relative times
Description	For relative times, the clock identifier is not used to select the clock and instead always the CLOCK_MONOTONIC is used. A side-effect is that sleep() and nanosleep() use the wrong clock (CLOCK_MONOTONIC instead of CLOCK_REALTIME).
#4427	3 years ago	fixed	tool	Ryan Long		3 years ago
Summary	GcovData?.cc: Dereference after null check error spotted by Coverity
Description	CID 1399597: Dereference after null check in readGcnoFile(). 53 strcpy( gcnoFileName, fileName ); 54 strcpy( gcdaFileName, fileName ); 55 strcpy( textFileName, fileName ); 56 strcpy( cFileName, fileName ); 57 tempString = strstr( gcdaFileName,".gcno" ); 58 tempString2 = strstr( textFileName,".gcno" ); 59 tempString3 = strstr( cFileName,".gcno" ); 60 2. Condition tempString == NULL, taking true branch. 3. var_compare_op: Comparing tempString to null implies that tempString might be null. 4. Condition tempString2 == NULL, taking false branch. 61 if ( (tempString == NULL) && (tempString2 == NULL) ){ 62 fprintf(stderr, "ERROR: incorrect name of *.gcno file\n"); 63 } 64 else 65 { CID 1399597 (#1 of 1): Dereference after null check (FORWARD_NULL)5. var_deref_model: Passing null pointer tempString to strcpy, which dereferences it. 66 strcpy( tempString, ".gcda"); // construct gcda file name 67 strcpy( tempString2, ".txt"); // construct report file name 68 strcpy( tempString3, ".c"); // construct source file name 69 }
#4428	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	ReportsBase?.cc: Dereference after null check error spotted by Coverity
Description	CID 1503710: Dereference after null check in WriteAnnotatedReport?(). 12. var_compare_op: Comparing info.uncoveredRanges to null implies that info.uncoveredRanges might be null. 13. Condition info.uncoveredBranches == NULL, taking false branch. 219 if ((info.uncoveredRanges == NULL) && 220 (info.uncoveredBranches == NULL)) 5. Continuing loop. 9. Continuing loop. 221 continue; 222 223 // If uncoveredRanges and uncoveredBranches are empty, then everything 224 // must have been covered for this symbol. Just skip it. CID 1503710 (#1 of 1): Dereference after null check (FORWARD_NULL)14. var_deref_model: Passing null pointer info.uncoveredRanges->set to empty, which dereferences it. [show details] 225 if ((info.uncoveredRanges->set.empty()) && 226 (info.uncoveredBranches->set.empty())) 227 continue;
#4430	3 years ago	fixed	tool	Ryan Long		3 years ago
Summary	DesiredSymbols?.cc: Uninitialized pointer read error spotted by Coverity
Description	CID 1503005: Uninitialized pointer read in load(). 111 if (sym.type() == sym.st_func) { 20. alloc_fn: Calling operator new which returns uninitialized memory. CID 1503005 (#2 of 2): Uninitialized pointer read (UNINIT)21. uninit_use_in_call: Using uninitialized value (new Coverage::SymbolInformation).sourceFile when calling operator =. [show details] 112 set[sym.name()] = *(new SymbolInformation);
#4432	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovFunctionData?.cc: Uninitialized pointer field error spotted by coverity
Description	CID 1399640: Uninitialized pointer field in GcovFunctionData?(). 20 GcovFunctionData::GcovFunctionData() 21 { 22 numberOfArcs = 0; 23 numberOfBlocks = 0; 24 coverageMap = NULL; 2. uninit_member: Non-static class member id is not initialized in this constructor nor in any functions that it calls. 4. uninit_member: Non-static class member checksum is not initialized in this constructor nor in any functions that it calls. 6. uninit_member: Non-static class member firstLineNumber is not initialized in this constructor nor in any functions that it calls. CID 1399640 (#1 of 1): Uninitialized pointer field (UNINIT_CTOR)8. uninit_member: Non-static class member symbolInfo is not initialized in this constructor nor in any functions that it calls. 25 }
#4433	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	DesiredSymbols?.h: Uninitialized pointer field error spotted by coverity
Description	CID 1399637: Uninitialized pointer field in SymbolInformation?(). 176 SymbolInformation() : 177 baseAddress( 0 ), 178 uncoveredBranches( NULL ), 179 uncoveredRanges( NULL ), 180 unifiedCoverageMap( NULL ) 181 { CID 1399637 (#1 of 1): Uninitialized pointer field (UNINIT_CTOR)2. uninit_member: Non-static class member sourceFile is not initialized in this constructor nor in any functions that it calls. 182 }
#4434	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	record-filter-zlib.cc: Uninitialized pointer field error spotted by Coverity
Description	CID 1503020: Uninitialized pointer field in ZlibFilter?(). 36ZlibFilter::ZlibFilter() : buffer_(65536) 37{ CID 1503013: Unchecked return value from library (CHECKED_RETURN) [select issue] 38 inflateInit(&stream_); 2. uninit_member: Non-static class member stream_.next_in is not initialized in this constructor nor in any functions that it calls. 4. uninit_member: Non-static class member stream_.avail_in is not initialized in this constructor nor in any functions that it calls. 6. uninit_member: Non-static class member stream_.total_in is not initialized in this constructor nor in any functions that it calls. 8. uninit_member: Non-static class member stream_.next_out is not initialized in this constructor nor in any functions that it calls. 10. uninit_member: Non-static class member stream_.avail_out is not initialized in this constructor nor in any functions that it calls. 12. uninit_member: Non-static class member stream_.total_out is not initialized in this constructor nor in any functions that it calls. 14. uninit_member: Non-static class member stream_.msg is not initialized in this constructor nor in any functions that it calls. 16. uninit_member: Non-static class member stream_.zalloc is not initialized in this constructor nor in any functions that it calls. 18. uninit_member: Non-static class member stream_.zfree is not initialized in this constructor nor in any functions that it calls. 20. uninit_member: Non-static class member stream_.opaque is not initialized in this constructor nor in any functions that it calls. 22. uninit_member: Non-static class member stream_.data_type is not initialized in this constructor nor in any functions that it calls. 24. uninit_member: Non-static class member stream_.adler is not initialized in this constructor nor in any functions that it calls. CID 1503020 (#1 of 1): Uninitialized pointer field (UNINIT_CTOR)26. uninit_member: Non-static class member stream_.reserved is not initialized in this constructor nor in any functions that it calls. 39}
#4435	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Thread cancellation may produce ready threads with an active thread timer
Description	The _Thread_Cancel() (in contrast to _Thread_Restart() which uses a similar code block) may produce ready threads with an active timer in case the thread to cancel has its thread life protection enabled. The problem is this code block: Priority_Control priority; _Thread_Add_life_change_request( the_thread ); if ( _Thread_Is_life_change_allowed( previous ) ) { _Thread_State_release( the_thread, &lock_context ); _Thread_queue_Extract_with_proxy( the_thread ); _Thread_Timer_remove( the_thread ); } else { _Thread_Clear_state_locked( the_thread, STATES_SUSPENDED ); _Thread_State_release( the_thread, &lock_context ); } priority = _Thread_Get_priority( executing ); _Thread_Raise_real_priority( the_thread, priority ); _Thread_Remove_life_change_request( the_thread ); The life change request should only be added/removed if a life change is allowed (see _Thread_Restart()).
#4437	3 years ago	fixed	bsps	Joel Sherrill		2 years ago
Summary	powerpc/shared/sbrk.c: Do not reference errno
Description	Using errno implicitly references the newlib reentrancy structure. Changing this eliminates one reason the reentrancy code may be pulled into minimum.exe.
#4438	3 years ago	fixed	bsps	Joel Sherrill	Joel Sherrill <joel@…>	3 years ago
Summary	powerpc/shared/irq/ppc-irq-legacy.c: Use rtems_malloc() not malloc()
Description	malloc() is required to set errno which implicitly references the newlib reentrancy structure. Changing this eliminates one reason the reentrancy code may be pulled into minimum.exe.
#4439	3 years ago	fixed	rtems	Joel Sherrill	Joel Sherrill <joel@…>	3 years ago
Summary	Open of /dev/console should only be in SysInit? List if Console Enabled
Description	The current code implicitly requires that the code to open /dev/console be included in the system initialization set anytime a dependency exists on the newlib reentrancy structure. This means that referencing errno pulls in at least open(), libio code, close(), atexit(), and unmount support. This was shown to result in about 20K of text an 5K of data being included in powerpc/psim minimum.exe. A similar impact would happen on all BSPs.
#4442	3 years ago	fixed	lib	Harrison Gerber	Gedare Bloom	3 years ago
Summary	CID 1399721: Copy into fixed size buffer
Description	In pci_cfg_print_device: A source buffer of statically unknown size is copied into a fixed-size destination buffer (CWE-120)
#4444	3 years ago	fixed	rtems	Harrison Gerber	Gedare Bloom	3 years ago
Summary	CID 1399727: Copy into fixed size buffer
Description	In rtems_monitor_line_editor: A source buffer of statically unknown size is copied into a fixed-size destination buffer (CWE-120) NOTE: Issue is more regarding source/dest overlapping with use of strcpy: needs change to use memory copy for overlapping source/dest
#4445	3 years ago	fixed	lib/debugger	Harrison Gerber	Gedare Bloom	3 years ago
Summary	CID 1468684: Uninitialized scalar variable
Description	In tcp_remote_connect: Use of an uninitialized variable (CWE-457)
#4448	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	covoar reports uncovered ranges of size 1
Description	The uncovered range reports produced by covoar for the ARM architecture contain ranges of size 1. When clicked, they link to instructions that appears to be covered. This should be fixed to remove these erroneous ranges.
#4451	3 years ago	wontfix	tool	Ryan Long		3 years ago
Summary	DesiredSymbols?: Implement move assignment operator
Description	CID 1505133: Missing move assignment operator for SymbolInformation? class. 128 */ CID 1505133 (#1 of 1): Missing move assignment operator (MISSING_MOVE_ASSIGNMENT)missing_move_assignment: Class Coverage::SymbolInformation may benefit from adding a move assignment operator. See other events which show the copy assignment operator being applied to rvalues, where a move assignment may be faster. 129 class SymbolInformation {
#4453	3 years ago	fixed	doc	Joel Sherrill	Joel Sherrill	17 months ago
Summary	Global vs Local Attribute Unclear for SMP/Distributed MP
Description	The description for global vs local attributes on objects do not make it clear whether they apply to SMP or distributed multiprocessing configuration. This applies to at least the rtems_task_create description but likely needs to be addressed in the description in every create with global/local and in the SMP section.
#4454	3 years ago	duplicate	bsps	Jan Sommer		2 years ago
Summary	bsps/i386: TSC calibration inaccurate
Description	The current implementation of the TSC calibration during startup yields inaccurate results. There are 2 issues: 1. Rounding error: The calibrate_tsc function calibrates for 1s by waiting for rtems_clock_get_ticks_per_second() ticks via the PIT. The PIT has a frequency of 1193182 Hz. For a rtems tick of 1ms this yields 1193. Waiting for 1s means then waiting for 1193000 PIT ticks (instead of the full 1193182). For a 1 GHz TSC clock this would yield already ~150 kHz less then the correct TSC frequency. 2. Waiting for the first loop at begin of timer period for (i = rtems_clock_get_ticks_per_second() * pc386_isrs_per_tick; i != 0; --i ) { /* We know we've just completed a tick when timer goes from low to high */ then_lsb = then_msb = 0xff; do { <== First loop does not start at begin of clock period READ_8254(now_lsb, now_msb); if ((then_msb < now_msb) || ((then_msb == now_msb) && (then_lsb < now_lsb))) break; then_lsb = now_lsb; then_msb = now_msb; } while (1); } This can yield deviations of several MHz for the TSC frequency. For example, for a tested CPU with a TSC frequency of 19167.1 MHz yields a calibration result of 1912.3 MHz. Proposed solution Just wait 1193182 PIT ticks for 1 s instead of a number of rtems ticks Use the raw timer value instead of waiting for clock ticks.
#4455	3 years ago	fixed	admin	Jan Sommer	Jan Sommer	3 years ago
Summary	bsps/i386: TSC calibration inaccurate
Description	The current implementation of the TSC calibration during startup yields inaccurate results. There are 2 issues: 1. Rounding error: The calibrate_tsc function calibrates for 1s by waiting for rtems_clock_get_ticks_per_second() ticks via the PIT. The PIT has a frequency of 1193182 Hz. For a rtems tick of 1ms this yields 1193. Waiting for 1s means then waiting for 1193000 PIT ticks (instead of the full 1193182). For a 1 GHz TSC clock this would yield already ~150 kHz less then the correct TSC frequency. 2. Waiting for the first loop at begin of timer period for (i = rtems_clock_get_ticks_per_second() * pc386_isrs_per_tick; i != 0; --i ) { /* We know we've just completed a tick when timer goes from low to high */ then_lsb = then_msb = 0xff; do { <== First loop does not start at begin of clock period READ_8254(now_lsb, now_msb); if ((then_msb < now_msb) || ((then_msb == now_msb) && (then_lsb < now_lsb))) break; then_lsb = now_lsb; then_msb = now_msb; } while (1); } This can yield deviations of several MHz for the TSC frequency. For example, for a tested CPU with a TSC frequency of 19167.1 MHz yields a calibration result of 1912.3 MHz. Proposed solution Just wait 1193182 PIT ticks for 1 s instead of a number of rtems ticks Use the raw timer value instead of waiting for clock ticks.
#4458	3 years ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	10 months ago
Summary	Simplify trap table initialization
Description	Currently, the trap table is statically initialized (see start.S). Unexpected traps end up in system error mode. Later during startup most traps are changed to jump to _ISR_Handler(). This is a side-effect of bsp_spurious_initialize(). The used functions for this dynamic initialization are a bit complex and need cache manager directives which involve all online processors. This should be simplified to cover only basic requirements. The _ISR_Vector_table should be initialized by _CPU_Initialize_vectors() to an new default handler _SPARC_ISR_handler_default() which just invokes _Terminate(). The traps associated with external interrupts (0x11 up to including 0x1f) should be statically initialized in start.S to jump to _ISR_Handler(). Standard synchronous exceptions such as 0x01 instruction access exception 0x02 illegal instruction 0x03 privileged instruction 0x04 fp disabled 0x07 memory address not aligned 0x08 fp exception 0x09 data access exception 0x0A tag overflow 0x24 cp_disabled 0x28 cp_exception should also be statically initialize to jump to _ISR_Handler() to invoke the fatal error handler. If software needs to change the default initialization, then it can use set_vector() or alternatives on demand.
#4460	3 years ago	fixed	admin	Harrison Gerber	Gedare Bloom	18 months ago
Summary	CID 1437647: Explicit null dereferenced
Description	A null pointer dereference will occur. In rtems_fdt_unload: Dereference of an explicit null value (CWE-476)
#4461	3 years ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	grlib: Register system console as "/dev/console"
Description	The driver manager console driver did not register the system console as "/dev/console" this resulted in some test failures (for example spconsole01.exe).
#4462	3 years ago	fixed	admin	Harrison Gerber	Gedare Bloom	18 months ago
Summary	CID 1437638: Logically dead code
Description	In rtems-fdt.c: The indicated dead code may have performed some action; that action will never occur. In rtems_fdt_init_index: Code can never be reached because of a logical contradiction (CWE-561) 159 entries = calloc(num_entries, sizeof(rtems_fdt_index_entry)); cond_notnull: Condition entries, taking true branch. Now the value of entries is not NULL. 160 if (!entries) 161 { 162 return -RTEMS_FDT_ERR_NO_MEMORY; 163 } 164 165 names = calloc(1, total_name_memory); notnull: At condition entries, the value of entries cannot be NULL. dead_error_condition: The condition !entries cannot be true. 166 if (!entries) 167 { CID 1437638 (#1 of 1): Logically dead code (DEADCODE) dead_error_begin: Execution cannot reach this statement: free(entries);. 168 free(entries); 169 return -RTEMS_FDT_ERR_NO_MEMORY; 170 }
#4463	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems_condition_variable_signal() is implemented as a broadcast
Description	rtems_condition_variable_signal() should use _Condition_Signal().
#4468	3 years ago	fixed	build	Chris Johns	Chris Johns <chrisj@…>	3 years ago
Summary	waf build system does not generate the same bspopts.h
Description	Building with autoconf and the command line of: $ ../rtems.git/configure --target=powerpc-rtems6 --enable-rtemsbsp=mvme2307 gives: $ grep BSP_CONSOLE_BAUD `find . -name bspopts.h` #define BSP_CONSOLE_BAUD 9600 Building with waf and a config of ... [DEFAULT] BUILD_TESTS = True RTEMS_POSIX_API = True [powerpc/mvme2307] gives: $ grep BSP_CONSOLE_BAUD `find . -name bspopts.h` #define BSP_CONSOLE_BAUD 115200 The default should be 9600 and the motorola_powerpc BSP is listed in the options under 9600. Something is broken in the waf BSP opts handling.
#4469	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-exeinfo.cpp: Not restoring ostream format errors spotted by Coverity
Description	CID 1503006: Not restoring ostream format in output_compilation_unit(). 185. format_changed: left changes the format state of std::cout for category adjustfield. 478 std::cout << " | " 479 << std::setw (source_max + 1) << std::left 480 << rld::path::basename (s.source); CID 1503007: Not restoring ostream format in output_tls(). 11. format_changed: right changes the format state of std::cout for category adjustfield. 12. format_changed: hex changes the format state of std::cout for category basefield. 13. format_changed: setfill changes the format state of std::cout for category fill. 14. format_restored: dec changes the format state of std::cout for category basefield. 15. format_restored: setfill changes the format state of std::cout for category fill. 704 std::cout << " data size : " << tls_data_size->value () << std::endl 705 << " bss size : " << tls_bss_size->value () << std::endl 706 << " alignment : " << tls_alignment->value () << std::endl 707 << std::right << std::hex << std::setfill ('0') 708 << " data addr : 0x" << std::setw (8) << tls_data_begin->value () 709 << std::endl 710 << std::dec << std::setfill (' ') 711 << std::endl;
#4470	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	CoverageMapBase?.cc: Not restoring ostream format error spotted by Coverity
Description	CID 1503022: Not restoring ostream format in dump(). 1. format_changed: hex changes the format state of out for category basefield. 2. format_changed: setfill changes the format state of out for category fill. 78 out << std::hex << std::setfill('0') 79 << "Address range: low = " << std::setw(8) << lowAddress 80 << " high = " << std::setw(8) << highAddress 81 << std::endl;
#4471	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	ReportsHtml?.cc: Not restoring ostream format error spotted by Coverity
Description	CID 1505939: Not restoring ostream format in PutSymbolSummaryLine?(). 5. format_changed: fixed changes the format state of report for category floatfield. 6. format_changed: setprecision changes the format state of report for category precision. 782 report << "<td class=\"covoar-td\" align=\"center\">" 783 << std::fixed << std::setprecision( 2 ) 784 << ( symbolInfo.stats.uncoveredInstructions * 100.0 ) / 785 symbolInfo.stats.sizeInInstructions 786 << "</td>" << std::endl;
#4472	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	ReportsText?.cc: Not restoring ostream format error spotted by Coverity
Description	CID 1505940: Not restoring ostream format in PutSymbolSummaryLine?(). 5. format_changed: fixed changes the format state of report for category floatfield. 6. format_changed: setprecision changes the format state of report for category precision. 231 report << "============================================" << std::endl 232 << "Symbol : " 233 << symbolName << std::endl 234 << "Total Size in Bytes : " 235 << symbolInfo.stats.sizeInBytes << std::endl 236 << "Total Size in Instructions : " 237 << symbolInfo.stats.sizeInInstructions << std::endl 238 << "Total number Branches : " 239 << symbolInfo.stats.branchesNotExecuted + 240 symbolInfo.stats.branchesExecuted 241 << std::endl 242 << "Total Always Taken : " 243 << symbolInfo.stats.branchesAlwaysTaken << std::endl 244 << "Total Never Taken : " 245 << symbolInfo.stats.branchesNeverTaken << std::endl 246 << std::fixed << std::setprecision( 2 ) 247 << "Percentage Uncovered Instructions : " 248 << uncoveredInstructions << std::endl 249 << "Percentage Uncovered Bytes : " 250 << uncoveredBytes << std::endl;
#4473	3 years ago	fixed	tool	Alex White	Alex White	3 years ago
Summary	AddressToLineMapper?.h: Pointer to local outside scope error spotted by Coverity
Description	CID 1505281: Pointer to local outside scope SourceLine() 28 : address(0), CID 1505281 (#1 of 1): Pointer to local outside scope (RETURN_LOCAL) 1. escape_local_addr: Returning, through this->path_, the address of stack variable <temporary>. 29 path_("unknown"), 30 line_num(-1), 31 is_end_sequence(true) 32 { 2. return: Returning here. 33 }
#4475	3 years ago	fixed	network/libbsd	Chris Johns	Chris Johns	23 months ago
Summary	Add NFSv4 client support to libbsd
Description	Add NFSv4 client support to libbsd. The client support is part of kernel and provides support for a range of the NFS protocol version including version 4. The task is made of the following: Update the locks and locking to support the lock manager. VFS requires this be supported Add support for the FreeBSD Virtual File System (VFS) Add the kernel RPC implementation Add NFS client support VFS VFS is built around the FB's file structure and file descriptor structure. These need to be supported. The file struct and file descriptor form the basis for all fd based interactions with the FB kernel code such as files, devices and sockets. LibIO and FB fd The FB fd and file support clashes with the existing libio implementations used for things like socket, kqueue and select. The work to port the VFS requires: socket, kqueue, select support FB fd support. This is closer to the kernel. No direct libio access Provide an interface to libio mapping the iop descriptors to FB descriptors. This approach nests the FB file descriptors inside the libio support. This is not optimal in terms of some activities but the hot paths in the code should not see a significant change in performance. Some paths like open are not optimal because the RTEMS file system op handlers imply a specific implementation that does not match up well with the FB namei interface. The long term approach is to look at libio and the handlers and consider allowing closer integration of FB's fd support. BIO and Paging The bio layer is required for VFS. No paging support other than the calls needed to have VFS link are provided. No VM paging is need and none of that code need be ported. System Calls Collect all the syscall functionality exported to RTEMS into a single place. Currently the calls are spread around the code and this: Does not centralize the code Add a lot of code to FB being ported Proc0 FreeBSD's initialization creates the initial process called proc0. Add support to create proc0 and let all thread' reference it. This allow the creds to be managed, kernel trace to work and it removes a number of changes that do not need to be made. File Descripter Reuse The libio functionality of cycling across the fd range rather than reusing lower fd numbers is not compatible with the select call. Not reusing fd numbers once closed until the range as been used provides some protection against bugs where a user does not cleanly handle the fd they own however this is not compatible with FB and makes supporting select harder. The select interface should not be used on RTEMS and users should be encourages to use threads for concurrent IO or kqueue. The means select may exist for legacy code porting reasons however the cycling fd count reduces how well the code ported will actually work. For example port a piece of select code, open a socket and use, it may work. Then open and close a file 64 times then run the same code. It will fail. FB fd's are reused. RPC The legacy NFSv2 client the the NFSv4 cannot be built at the same time. The kernel RPC and the user land RPC cannot exist in the same build. I have not looked extensively into this but I was seeing clashes with headers for the same types but different contents. The legacy NFSv2 client uses the user land RPC code while the kernel NFSv4 uses the kernel's RPC. Module support is already in libbsd to handle the separate building of each piece. Security NFSv4 provides a number of extra security features. The code will be ported is possible but little or no security tested is planned at this point in time. Testing The NFSv4 stack will be tested against Linux and FreeBSD servers.
#4482	3 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Document kernel character I/O support in Classic API Guide
Description	Directives like printk() are available for a long time. Document them in the Classic API Guide.
#4483	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovData?.cc: Not restoring ostream format error spotted by Coverity
Description	CID 1506202: Not restoring ostream format 3. format_changed: hex changes the format state of std::cerr for category basefield. 407 std::cerr << "File is not a valid *.gcno output (magic: 0x" 408 << std::hex << std::setw( 4 ) << preamble->magic 409 << ")" << std::endl; CID 1506202 (#1 of 1): Not restoring ostream format (STREAM_FORMAT_STATE)4. end_of_path: Changing format state of stream std::cerr for category basefield without later restoring it.
#4484	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovFunctionData?.cc: Not restoring ostream format error spotted by Coverity
Description	CID 1506205: Not restoring ostream format 1. format_changed: hex changes the format state of textFile for category basefield. 360 textFile << " > BLOCK " << std::setw( 3 ) << block->id 361 << " from " << block->sourceFileName << std::endl 362 << " -counter: " << std::setw( 5 ) << block->counter << std::endl 363 << " -flags: 0x" << std::hex << block->flags << std::endl 364 << " -lines: ";
#4485	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovData?.cc: Out-of-bounds acces errors spotted by Coverity
Description	CID 1506208: Out-of-bounds access CID 1506208 (#1 of 1): Out-of-bounds access (OVERRUN)2. overrun-buffer-arg: Overrunning struct type _ZN4Gcov13gcov_preambleE of 12 bytes by passing it to a function which accesses it at byte offset 47 using argument 48L. 132 gcdaFile.write( (char *) &preamble , 4 * sizeof( preamble ) ); 133 if ( gcdaFile.fail() ) { 134 std::cerr << "Error while writing gcda preamble to a file " 135 << gcdaFileName << std::endl; 136 } CID 1506209: Out-of-bounds access CID 1506209 (#1 of 1): Out-of-bounds access (OVERRUN)1. overrun-buffer-arg: Overrunning buffer pointed to by (char *)&preamble of 8 bytes by passing it to a function which accesses it at byte offset 47 using argument 48L. 400 gcovFile.read( (char *) &preamble, 4 * sizeof( gcov_preamble ) ); 401 if ( gcovFile.gcount() != 4 * sizeof( gcov_preamble ) ) { 402 std::cerr << "Error while reading file preamble" << std::endl; 403 return -1; 404 }
#4486	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	TraceWriterQEMU.cc: String not null terminated issue spotted by Coverity
Description	CID 1506203: String not null terminated CID 1506203 (#1 of 1): String not null terminated (STRING_NULL)6. string_null: Passing unterminated string header.magic to operator <<, which expects a null-terminated string. [Note: The source code implementation of the function has been overridden by a builtin model.] 124 std::cerr << "magic = " << header.magic << std::endl 125 << "version = " << header.version << std::endl 126 << "kind = " << header.kind << std::endl 127 << "sizeof_target_pc = " << header.sizeof_target_pc << std::endl 128 << "big_endian = " << header.big_endian << std::endl 129 << std::hex << std::setfill('0') 130 << "machine = " << std::setw(2) << header.machine[0] 131 << ':' << header.machine[1] 132 << std::dec << std::setfill(' ') 133 << std::endl;
#4487	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	GcovData?.cc: Uninitialized scalar field error spotted by Coverity
Description	CID 1506206: Uninitialized scalar field 22 GcovData::GcovData( Coverage::DesiredSymbols& symbolsToAnalyze ): 23 numberOfFunctions( 0 ), 24 symbolsToAnalyze_m( symbolsToAnalyze ) 25 { 2. uninit_member: Non-static class member gcnoPreamble.magic is not initialized in this constructor nor in any functions that it calls. 4. uninit_member: Non-static class member gcnoPreamble.version is not initialized in this constructor nor in any functions that it calls. CID 1506206 (#1 of 1): Uninitialized scalar field (UNINIT_CTOR)6. uninit_member: Non-static class member gcnoPreamble.timestamp is not initialized in this constructor nor in any functions that it calls. 26 }
#4488	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	TraceWriterQEMU.cc: Uninitialized scalar variable issue spotted by Coverity
Description	CID 1506204: Uninitialized scalar variable 115 header.machine[1] = 0; // XXX ?? CID 1506204 (#1 of 1): Uninitialized scalar variable (UNINIT)4. uninit_use_in_call: Using uninitialized value header. Field header._pad is uninitialized when calling fwrite. 116 status = ::fwrite( &header, sizeof(trace_header), 1, traceFile );
#4489	3 years ago	wontfix	build	Chris Johns		17 months ago
Summary	waf configure's --rtems-config and build create an error
Description	Waf allows distclean, configure, build and install on a single command line however adding --rtems-config to the configure stage generates this error: The --rtems-config command line option is not allowed in the build command This is error is incorrect.
#4490	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	rtems_partition_return_buffer() wrongly accepts buffers which are exactly at the buffer area end
Description	The range checking in rtems_partition_return_buffer() is wrong.
#4491	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	TraceWriterQEMU.cc: Buffer not null terminated issue spotted by Coverity
Description	CID 1506207: Buffer not null terminated CID 1506207 (#1 of 1): Buffer not null terminated (BUFFER_SIZE)3. buffer_size: Calling strncpy with a source string whose length (12 chars) is greater than or equal to the size argument (12) will fail to null-terminate header.magic. 109 strncpy( header.magic, QEMU_TRACE_MAGIC, sizeof(header.magic) );
#4492	3 years ago	fixed	tool	Ryan Long		3 years ago
Summary	ObjdumpProcessor?.cc: Uninitialized pointer field
Description	CID 1506210: Uninitialized pointer field 125 ObjdumpProcessor::ObjdumpProcessor( 126 DesiredSymbols& symbolsToAnalyze, 127 std::shared_ptr<Target::TargetBase>& targetInfo 128 ): symbolsToAnalyze_m( symbolsToAnalyze ), 129 targetInfo_m( targetInfo ) 130 { CID 1506210 (#1 of 1): Uninitialized pointer field (UNINIT_CTOR)2. uninit_member: Non-static class member inputBuffer_m is not initialized in this constructor nor in any functions that it calls. 131 }
#4499	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-exeinfo.cpp: Uninitialized scalar field error spotted by Coverity
Description	CID 1471637: Uninitialized scalar field 211 section::section (const section& orig) 212 : sec (orig.sec), 213 data (orig.data) 214 { CID 1471637 (#1 of 1): Uninitialized scalar field (UNINIT_CTOR)2. uninit_member: Non-static class member byteorder is not initialized in this constructor nor in any functions that it calls. 215 }
#4500	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rld-dwarf.cpp: Uninitialized scalar field
Description	CID 1503019: Uninitialized scalar field 677 function::function (file& debug, debug_info_entry& die) 678 : debug (debug), 679 machine_code_ (false), 680 external_ (false), 681 declaration_ (false), 682 inline_ (DW_INL_not_inlined), 683 entry_pc_ (0), 684 has_entry_pc_ (false), 685 pc_low_ (0), 686 pc_high_ (0), 687 ranges_ (debug), 688 call_line_ (0) CID 1503019 (#1 of 1): Uninitialized scalar field (UNINIT_CTOR)14. uninit_member: Non-static class member prototyped_ is not initialized in this constructor nor in any functions that it calls.
#4501	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	2 years ago
Summary	TraceConverter?.cc: Uncaught exception issue spotted by Coverity
Description	CID 1471639: Uncaught exception CID 1471639 (#1-9 of 9): Uncaught exception (UNCAUGHT_EXCEPT)root_function: In function main(int, char **) an exception of type rld::error is thrown and never caught. . . . fun_call_w_exception: Called function throws an exception of type rld::error. [show details] 144 executableInfo = new Coverage::ExecutableInfo( 145 executable, 146 "", 147 false, 148 symbolsToAnalyze 149 );
#4502	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rld-rap.cpp: Uncaught exception issue spotted by Coverity
Description	CID 1471643: Uncaught exception CID 1471643 (#1 of 1): Uncaught exception (UNCAUGHT_EXCEPT)exn_spec_violation: An exception of type rld::error is thrown but the throw list noexcept doesn't allow it to be thrown. This will cause a call to unexpected() which usually calls terminate(). 724 section_merge::~section_merge () 725 { 726 if (rld::verbose () >= RLD_VERBOSE_FULL_DEBUG) fun_call_w_exception: Called function throws an exception of type rld::error. [show details] 727 std::cout << "rap:section-merge: " << sec.name 728 << " size=" << sec.size () 729 << " offset=" << sec.offset 730 << " " << obj.obj.name ().full () << std::endl; 731 }
#4506	3 years ago	fixed	posix	Joel Sherrill	Ryan Long <ryan.long@…>	3 years ago
Summary	posix_devctl() should return the errno directly not -1 and set errno
Description	The RTEMS implementation of posix_devctl() from POSIX 1003.26 does not return error codes as required by the standard. "Upon successful completion, posix_devctl() shall return zero; otherwise an error number shall be returned to indicate the error. The value returned via the dev_info_ptr argument is driver dependent." It should return the errno value and not -1 and set the errno variable. The current behavior is wrong but understandable because the implementation is a wrapper for close() and ioctl() which do set errno and return -1.
#4507	3 years ago	wontfix	fs	Chris Johns	Chris Johns	3 years ago
Summary	mount call retruns error number and it should return -1
Description	The mount call should return -1 for an any error with the errno set. Some paths in the mount do this and others so not. The implementation needs to be changed to only return -1 when there is an error.
#4508	3 years ago	fixed	doc	Frank Kuehndel	Sebastian Huber	3 years ago
Summary	rtems_message_queue_receive(): flush() does not release waiting tasks
Description	The RTEMS Classic API Guide ​https://docs.rtems.org/branches/master/c-user/index.html in Message Manager → Directives → section 14.4.8 rtems_message_queue_receive() says: RTEMS_UNSATISFIED: The queue was flushed while the calling task was waiting to receive a message. Yet, when I tested it, this does not happen. A waiting task is not released from the message queue when calling rtems_message_queue_flush(). Moreover, I did not find any sign in the source code of rtems_message_queue_flush() which look liked it would dequeue waiting threads. (In contrast rtems_message_queue_delete() does!) Furthermore, there is no other hint in the documentation of either function that flush() would kick out waiting receivers. I may misinterpret the documentation above or oversee something in the code. Yet, I believe the text I cited above is a mistake in the documentation and should simply be removed. How to reproduce/How I tested: # I created a message queue and a second task with a higher priority than the first one. # I made the second task wait on the message empty queue by calling rtems_message_queue_receive(). # Mean while the first task called rtems_message_queue_flush() on the empty queue. # Result: The second task does still wait on the queue after rtems_message_queue_flush() returned. It does not return with RTEMS_UNSATISFIED. The debug output of my test looks like (the worker task is not released by rtems_message_queue_flush() but afterwards by a timeout of 3 ticks): P:574:0:UI1/Valid/Valid/Valid/Flushed/Local/Fifo/Timeout/Empty/Nop:tc-message-re ceive.c:1316 L:#### rtems_message_queue_construct() with L:#### .maximum_pending_messages = 3 L:#### .maximum_message_size = 5 L:#### .storage_size = 72 L:#### .storage_free = 0 L:#### .attributes = 0 (LOCAL=0, FIFO=0, PRIORITY=4) L:#### rtems_task_restart BEFORE L:#### rtems_message_queue_receive(option_set=0, timeout_param=3 ) BEGIN L:#### rtems_task_restart AFTER L:#### rtems_message_queue_flush() BEFORE L:#### rtems_message_queue_flush() AFTER L:#### WaitForWorker wait ticks BEFORE L:#### rtems_message_queue_receive() END status = 6 (0=RTEMS_SUCCESSFUL, 13=RTEMS_UNSATISFIED, 6=RTEMS_TIMEOUT ) L:#### WorkerTask send event L:#### WaitForWorker wait ticks AFTER F:586:0:UI1/Valid/Valid/Valid/Flushed/Local/Fifo/Timeout/Empty/Nop:tc-message-receive.c:1004:RTEMS_TIMEOUT == RTEMS_UNSATISFIED F:587:0:UI1/Valid/Valid/Valid/Flushed/Local/Fifo/Timeout/Empty/Nop:tc-message-receive.c:1043:3 == 0 L:#### rtems_message_queue_delete() For the records: RTEMS_VERSION = 6.0.0 remote origin = ​git://git.rtems.org/sebh/rtems.git /* The HASH is also in ​git://git.rtems.org/rtems.git */ commit HEAD = bb77a82f619752ba147c317d982ac47f56f4afec config.ini = [arm/realview_pbx_a9_qemu] RTEMS_DEBUG = True RTEMS_NETWORKING = True RTEMS_POSIX_API = True RTEMS_SMP = True BUILD_TESTS = True NEWLIB_VERSION = 3.2.0 SOURCE_BUILDER = Set Builder, 6 (f12dee02d52d) GCC_VERSION = arm-rtems6-gcc (GCC) 10.3.1 20210409 (RTEMS 6, RSB c938bd7cbe16bd03d3b382f2d7cfee6a86aa9424, Newlib 0c0f3df) BINUTILS_VERSION = GNU ld (GNU Binutils) 2.36.1.20210409
#4509	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Message queue priority dicipline is broken in SMP configurations
Description	The message queue handler uses the thread queue extract operation to dequeue a receiver thread. This is wrong in SMP configurations since the extract thread queue operation does not ensure FIFO fairness across schedulers. The thread queue surrender operation should be used instead.
#4511	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Count of postponed jobs is not set to zero for a newly created rate-monotonic period object
Description	rtems_rate_monotonic_get_status() returns an arbitrary number for the count of postponed jobs if it is called for a newly created period object. The count of postponed jobs should be cleared to zero at object creation.
#4513	3 years ago	fixed	doc	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Document parts of the Cache Manager in the Classic API Guide
Description	The Cache Manager directives are available via <rtems.h>. Document most of them in the Classic API Guide.
#4517	3 years ago	wontfix	tool	Chris Johns		17 months ago
Summary	Build failures for MacOS
Description	This ticket is a collection of the following archs that fail to build: bfin lm32 mips or1k powerpc v850 bfin ../../../sourceware-mirror-binutils-gdb-7ab567f/sim/bfin/dv-bfin_dma.c:124:35: error: implicitly declaring library function 'abs' with type 'int (int)' [-Werror,-Wimplicit-function-declaration] if (dma->ele_size != (unsigned) abs (dma->x_modify)) lm32 ../../../sourceware-mirror-binutils-gdb-7ab567f/sim/lm32/dv-lm32cpu.c:151:8: error: implicit declaration of function 'lm32bf_h_csr_get' is invalid in C99 [-Werror,-Wimplicit-function-declaration] im = lm32bf_h_csr_get (cpu, LM32_CSR_IM); mip ../../../sourceware-mirror-binutils-gdb-7ab567f/sim/mips/cp1.c:495:7: error: implicitly declaring library function 'abort' with type 'void (void) __attribute or1k ../../../sourceware-mirror-binutils-gdb-7ab567f/sim/or1k/../common/cgen-trace.c:213:6: error: implicitly declaring library function 'abort' with type 'void (void) __attribute__((noreturn))' [-Werror,-Wimplicit-function-declaration] abort (); powerpc semantics.c:42:7: error: implicit declaration of function 'tree_find_property' is invalid in C99 [-Werror,-Wimplicit-function-declaration] if (tree_find_property(root, "/options/mpc860c0")) v850 ../../../sourceware-mirror-binutils-gdb-7ab567f/sim/v850/simops.c:1651:6: error: implicit declaration of function 'free' is invalid in C99 [-Werror,-Wimplicit-function-declaration] free (path);
#4518	3 years ago	fixed	network/libbsd	Chris Johns	Chris Johns	3 years ago
Summary	RTEMS and LibBSD both provide competing pipe() interfaces
Description	RTEMS provides a pipe() call in cpukit/libfs/src/pipe/pipe.c and LibBSD also provides a call in rtemsbsd/rtems/rtems-bsd-syscall-api.c. We should not have competing interfaces. The selection of the version you get depends on the link order and the linker version. The RTEMS 6 linker seems to have a different preference to RTEMS 5. Apart from this fragility LibBSD should not be providing this interface. A LibBSD pipe() is problematic for a number of reasons and so I am going to remove the call from it. If LibBSD support is needed it needs to be via a file system op handles like all other file system interfaces for descriptor based access. I will not be adding this support. The removal of pipe() from LibBSD removes the kqueue support for pipes. See the test selectpollkqueue01. If this support is needed please contact the RTEMS developers.
#4519	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	record-main-lttng.cc: Out-of-bounds access
Description	CID 1503017: Out-of-bounds access in CopyThreadName?(). 1. Condition api_index < 3, taking false branch. 298 if (api_index < THREAD_API_COUNT) { 299 name = thread_names_[api_index][GetObjIndexOfID(item.data)]; 300 } else { 2. alias: Assigning: name = kEmptyThreadName. name now points to byte 0 of kEmptyThreadName (which consists of 3 bytes). 301 name = kEmptyThreadName; 302 } 303 CID 1503017 (#1 of 1): Out-of-bounds access (OVERRUN)3. overrun-buffer-arg: Overrunning buffer pointed to by name of 3 bytes by passing it to a function which accesses it at byte offset 15 using argument 16UL. 304 std::memcpy(dst, name, THREAD_NAME_SIZE);
#4522	3 years ago	fixed	tool/newlib	Sebastian Huber	Sebastian Huber	3 years ago
Summary	Optimize red-black tree insert/extract
Description	Code coverage analysis of the red-black tree insert/extract operations defined in <sys/tree.h> showed that the macros contain dead code. To fix this simplify some expressions and add specialized rotations: ​https://sourceware.org/pipermail/newlib/2021/018599.html
#4524	3 years ago	fixed	rtems	Joel Sherrill	Joel Sherrill <joel@…>	12 months ago
Summary	Re-add lost capability for custom stack allocator to allocate IDLE thread stacks
Description	The move to statically allocate the stacks for the IDLE threads resulted in the loss of the capability for a custom stack allocator to be able to allocate the idle threads' stacks. This occurred close to the 5 branching point and is addressed for the 5 branch by #4520. This ticket is for the master (6).
#4525	3 years ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	3 years ago
Summary	rtems-tld.cpp: Logically dead code error spotted by Coverity
Description	CID 1399595: Logically dead code in generate_traces(). 1263 if (sig.has_ret () && !generator_.ret_trace.empty ()) 1264 { 1265 if (ds_added) 1266 ds += " + "; 1267 else 1268 ds_added = true; const: At condition drs_added, the value of drs_added must be equal to 0. dead_error_condition: The condition drs_added cannot be true. 1269 if (drs_added) CID 1399595 (#1 of 1): Logically dead code (DEADCODE)dead_error_line: Execution cannot reach this statement: drs += " + ";. 1270 drs += " + "; 1271 else 1272 drs_added = true; 1273 ds += "sizeof(" + sig.ret + ')'; 1274 drs += "sizeof(" + sig.ret + ')'; 1275 }
#4527	3 years ago	fixed	rtems	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Add new directives to get all time values provided by the timehands
Description	Add directives for all time values provided by the timehands to the Clock Manager.
#4528	3 years ago	fixed	rtems	Frank Kuehndel	Sebastian Huber	2 years ago
Summary	rate monotonic: reset of CPU usage time not always detected
Description	rtems_rate_monotonic_get_statistics() and rtems_rate_monotonic_get_status() must detect whether the used CPU time was resetted during a period (rtems_cpu_usage_reset()). This is achieved by storing the CPU usage time counter value when the period is started. To detect whether there was a reset, the CPU usage time value from the period start and the current value are compared. If the current value is lower than the one from the start there was a reset. The bug is that when the current CPU usage time value is greater than the value at the start, a reset may still have occurred. For example, when the period starts the used CPU time is at 1000 and 1000 is stored. Then the CPU time counter is resetted - so it is 0. Assume the period thread does some work and the CPU usage time is at 100 when rtems_rate_monotonic_get_status() is called. In this case the current code detects the reset and returns correctly RTEMS_NOT_DEFINED. Yet, when more CPU usage time passes without rtems_rate_monotonic_period() being called - lets say the used CPU time is at 1200 - and rtems_rate_monotonic_get_status() is called, the stored value (1000) is smaller than the current one (1200) so the reset is not detected. This bug has been found with a test which manipulates the time directly. I doubt this bug is very relevant in practice. It requires that there are periods with long intervals and much CPU load and that either the system is shortly started or rtems_cpu_usage_reset() is used frequently. The background: The Classic API guide ​https://docs.rtems.org/branches/master/c-user/rate-monotonic/directives.html#rtems-rate-monotonic-get-status says for rtems_rate_monotonic_get_status(): RTEMS_NOT_DEFINED: There was no status available due to a reset of the processor time usage of the owner task of the period. The directive rtems_cpu_usage_reset() sets the_thread->cpu_time_used 0 in line 43 of cpukit/libmisc/cpuuse/cpuusagereset.c. The detection whether there was an reset or not is done in file cpukit/rtems/src/ratemonperiod.c line 56 in function _Rate_monotonic_Get_status(): [...] /* * Determine cpu usage since period initiated. */ _Thread_Get_CPU_time_used( owning_thread, &used ); /* * The cpu usage info was reset while executing. Can't * determine a status. */ if ( _Timestamp_Less_than( &used, &the_period->cpu_usage_period_initiated ) ) return false; [...] The directive rtems_rate_monotonic_get_statistics() is affected indirectly because collecting its data is skipped when there was a reset. cpukit/rtems/src/ratemonperiod.c line 149 in function _Rate_monotonic_Update_statistics() (which in turn calls the above code). For the records: RTEMS_VERSION = 6.0.0 remote origin = git://git.rtems.org/sebh/rtems.git commit HEAD = 1df2666ca1d73807a338e0d209ed1f42f67f00a9 (HEAD, origin/qual-43) config.ini = [arm/realview_pbx_a9_qemu] RTEMS_DEBUG = True RTEMS_NETWORKING = True RTEMS_POSIX_API = True RTEMS_SMP = True BUILD_SAMPLES = False BUILD_VALIDATIONTESTS = True BUILD_UNITTESTS = True NEWLIB_VERSION = 3.2.0 SOURCE_BUILDER = Set Builder, 6 (f12dee02d52d) GCC_VERSION = arm-rtems6-gcc (GCC) 10.3.1 20210409 (RTEMS 6, RSB c938bd7cbe16bd03d3b382f2d7cfee6a86aa9424, Newlib 0c0f3df) BINUTILS_VERSION = GNU ld (GNU Binutils) 2.36.1.20210409
#4531	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Data corruption in SMP schedulers
Description	Certain operations involving sticky thread queues, thread to processor affinity changes, priority updates, or thread pinning lead to data corruption in SMP schedulers, in particular the default SMP scheduler (SMP EDF).
#4532	3 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Priority inversion issues with MrsP locking protocol implementation
Description	While a thread is scheduled on a helping scheduler, while it does not own a MrsP semaphore, if it obtains a MrsP semaphore, then no scheduler node using an idle thread and the ceiling priority of the semaphore is unblocked for the home scheduler. This could lead to priority inversion issues and is not in line with the MrsP protocol.
#4533	2 years ago	fixed	lib/debugger	Chris Johns	Chris Johns	2 years ago
Summary	libdebugger should only build for archs with a backend
Description	Libdebugger has limited backend support however the generic parts are bring built for all architectures and libdebugger.a is being install however it will never link because of the missing backend support. LibBSD has been fixed to correctly detect libdebugger.a and enable the test. This now fails to build.
#4534	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	SMP EDF scheduler violates priority group ordering
Description	The SMP EDF scheduler supports one-to-one and one-to-all thread to processor affinities. The one-to-one thread to processor affinity introduces a constraint on the ordering of threads. The implementation uses one ready queue for threads which have a one-to-all affinity and one for each one-to-one affinity group. To order threads across the ready queues, a generation number is used. However, the approach to update the generation number each time a thread is inserted into a ready queue is wrong. The generation number need to be updated only in the enqueue and enqueue scheduled operations where an insert priority is available. The scheduled chain needs to take the generation number into account. An example scenario which shows the bug is this. Let T be a high priority task affine to processor X. Let A be a lower priority task affine to processor X. Let B be a lower priority task with no affinity to a particular processor which executes on processor Y. Let B be in the same priority group than A and after A. Let T set the affinity to all processors. Now A (higher priority relative to B) should execute on X and T (high priority) should execute on Y.
#4543	2 years ago	worksforme	arch/aarch64	mohamedosama94		2 years ago
Summary	cannot build multiprocessor application on eclipse
Description	Hi, hope my mail finds you well, i have an issue as i am using rtems6 for zinq a53_lp64 i downloaded the source and installed the toolchain aarch64-rtems6 and i built the bsp successfully by using waf and i used the qemu to run the test samples exe files successfully , so i installed the eclipse and added rtems plugin to it and i built the sample rtems example successfully but when i try to build an application task has test_begin() or test_end() the make is uncomplete and error (.text.Init+0x50): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `rtems_test_begin' always appear i tried open declaration and i found the reference prototype int rtems_test_begin(const char* name, const RTEMS_TEST_STATE state); on test-info.h so eclipse refuse to build as cannot see the source code of the function and this also for rtems_test_end so what i have to do to make the eclipse build the rtems-application, kindly check the logs /home/mohamed/quick-starts/rtems/6/lib/gcc/aarch64-rtems6/10.3.1/../../../../aarch64-rtems6/bin/ld: o-optimize/init.o: in function `Init': /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:64: undefined reference to `rtems_test_begin' /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:64:(.text.Init+0x50): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `rtems_test_begin' /home/mohamed/quick-starts/rtems/6/lib/gcc/aarch64-rtems6/10.3.1/../../../../aarch64-rtems6/bin/ld: /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:66: undefined reference to `locked_print_initialize' /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:66:(.text.Init+0x54): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `locked_print_initialize' /home/mohamed/quick-starts/rtems/6/lib/gcc/aarch64-rtems6/10.3.1/../../../../aarch64-rtems6/bin/ld: o-optimize/init.o: in function `success': /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:31: undefined reference to `rtems_test_end' /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:31:(.text.Init+0x5c): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `rtems_test_end' /home/mohamed/quick-starts/rtems/6/lib/gcc/aarch64-rtems6/10.3.1/../../../../aarch64-rtems6/bin/ld: /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:32: undefined reference to `rtems_test_exit' /home/mohamed/eclipse-ground/rtems/rtems-appinit.c:32:(.text.Init+0x64): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `rtems_test_exit' /home/mohamed/quick-starts/rtems/6/lib/gcc/aarch64-rtems6/10.3.1/../../../../aarch64-rtems6/bin/ld: o-optimize/init.o:(.rodata._User_extensions_Initial_extensions+0x80): undefined reference to `rtems_test_fatal_extension' collect2: error: ld returned 1 exit status best regards, Mohamed
#4544	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	The last processor must not be removed if it is owned by a helping scheduler
Description	The following error condition is currently not checked by rtems_scheduler_remove_processor(): While an attempt is made to remove a processor from a scheduler, while the processor is the only processor owned by the scheduler, if a thread exists which uses the scheduler as a helping scheduler, then shall not be removed. The reason is that ask for help requests and withdraw node requests are processed asynchronously in any order. An ask for help request carried out on a scheduler without a processor is undefined behaviour.
#4545	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	The SMP EDF scheduler can only support more restricted affinity sets of a thread
Description	The SMP EDF scheduler supports a one-to-one and one-to-all thread to processor affinity. Currently it accepts affinity sets which are a proper subset of the online processors containing at least two processors owned by the scheduler. In this case it sets a one-to-one thread to processor affinity. This leads to undefined behaviour if a processor is removed since the higher level check in rtems_scheduler_remove_processor() does not account for this implementation detail.
#4546	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	A thread restart does not update the priority of related threads
Description	Consider the following scenario. Let a high priority task H wait on a priority inheritance mutex M owned by a low priority task L. Restart task H. This should remove the priority inherited by H from L. This is currently not the case.
#4548	2 years ago	fixed	build	Chris Johns	Sebastian Huber <sebastian.huber@…>	2 years ago
Summary	Add RTEMS version option to waf configure
Description	We need a way to override the default release for testing. This is a developer option and is not for production. A waf option is considered the best way to manage this.
#4550	2 years ago	duplicate	tool	Joel Sherrill	Chris Johns	2 years ago
Summary	Add --rtems-version to rtems-test
Description	The rtems-test does not have a way to override the RTEMS version. This results in some tools defaulting to rtems6 when you want to test with the rtems7 version.
#4551	2 years ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Workaround for GRLIB-TN-0011 required for sparc/leon3 BSPs in SMP configuration
Description	The check script tn-0011.tcl reported the following affected code block in cpu_asm.S: sparc-rtems6-objdump -d build/sparc/gr712rc/testsuites/validation/ts-validation-0.exe | tn-0011.tcl INFO: Scanning utility for INFO: GRLIB-TN-0011 - LEON3/FT AHB Lock Release during Atomic Operation INFO: Script revision: 3 (20180412) INFO: INFO: Scanning objdump -d output on standard input... 0x400ad5f8 : seqa INFO: 400ad5f8: c2 7a 60 58 swap [ %o1 + 0x58 ], %g1 INFO: 400ad5fc: 80 a0 60 00 cmp %g1, 0 INFO: 400ad600: 12 80 00 14 bne 400ad650 <done_flushing+0x7c> INFO: INFO: Error locations found: 1
#4556	2 years ago	fixed	shell	chenjin_zhong	Chris Johns	16 months ago
Summary	rtems_shell_main_mmove problem
Description	Hi, I think mmove command in shell means that the memory can be overlapped when copy happens. but I find that when implementing this function, the memcpy is used in RTEMS4.13/RTEMS5.1.Should the memcpy be replaced with memmove ? The src code is listed as follows. static int rtems_shell_main_mmove( int argc, char *argv[] ) { unsigned long tmp; void *src; void *dst; size_t length; if ( argc < 4 ) { fprintf(stderr,"%s: too few arguments\n", argv[0]); return -1; } /* Convert arguments into numbers */ if ( rtems_string_to_pointer(argv[1], &dst, NULL) ) { printf( "Destination argument (%s) is not a number\n", argv[1] ); return -1; } if ( rtems_string_to_pointer(argv[2], &src, NULL) ) { printf( "Source argument (%s) is not a number\n", argv[2] ); return -1; } if ( rtems_string_to_unsigned_long(argv[3], &tmp, NULL, 0) ) { printf( "Length argument (%s) is not a number\n", argv[3] ); return -1; } length = (size_t) tmp; /* Now copy the memory. */ memcpy(dst, src, length); return 0; }
#4557	2 years ago	fixed	shell	chenjin_zhong	Chris Johns	14 months ago
Summary	medit problem in RTEMS6 (was 4.11/5.1)
Description	Hi, I find when open medit editor in shell console, and then input CTRL+E/CTRL+X instructions in medit.the selected text cannot be cut. I review the code in main_medit file. First, when I input CTRL+E instruction, the function select_toggle is executed and selects the text. Second, when I input CTRL+X, the function cut_selection copies the selected text to clipboard. but after copy, the selected text is non-selected. Therefore, the erase_selection in cut_selection cannot take effect. the src code is listed as follows. I suggest that remove select_toggle in copy_selection function. and add after copy_selection. the code frament is marked in bold. static void select_toggle(struct editor *ed) { ed->selecting = ed->selecting ? 0 : 1; update_selection(ed, ed->selecting); adjust(ed); } static void cut_selection(struct editor *ed) { copy_selection(ed); erase_selection(ed); select_toggle(ed); } static void copy_selection(struct editor *ed) { int selstart, selend; if (!get_selection(ed, &selstart, &selend)) return; ed->env->clipsize = selend - selstart; ed->env->clipboard = (unsigned char *) realloc(ed->env->clipboard, ed->env->clipsize); if (!ed->env->clipboard) return; copy(ed, ed->env->clipboard, selstart, ed->env->clipsize); select_toggle(ed); } case ctrl('c'): copy_selection(ed); select_toggle; break;
#4560	2 years ago	fixed	tool/newlib	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Use thread-local storage for Newlib reentrancy objects
Description	Problem The state of the art architectures supported by RTEMS have all efficient support for thread-local storage (MIPS has issues with thread-local storage, however, is MIPS state of the art?). Newlib currently uses a huge object of type struct _reent to store thread-specific data. This object is returned by __getreent(). It is related to the __DYNAMIC_REENT__ Newlib configuration option which is always defined for RTEMS. The reentrancy structure contains errno and also the standard input, output, and error file streams. This means that if an application only uses errno it has a dependency on the file stream support event if it does not use it. This is an issue for lower end targets and the pre-qualification of RTEMS. Solution One approach to disentangle the dependencies introduced by struct _reent is to get rid of this structure and replace the individual members of the structure with thread-local objects. For example, instead of struct _reent { int _errno; __FILE *_stdin; __FILE *_stdout; __FILE *_stderr; }; use _Thread_local int _errno; _Thread_local __FILE *_stdin; _Thread_local __FILE *_stdout; _Thread_local __FILE *_stderr; Newlib already has access macros for the struct _reent members, for example: #define _REENT_SIGNGAM(ptr) ((ptr)->_new._reent._gamma_signgam) #define _REENT_RAND_NEXT(ptr) ((ptr)->_new._reent._rand_next) #define _REENT_RAND48_SEED(ptr) ((ptr)->_new._reent._r48._seed) #define _REENT_RAND48_MULT(ptr) ((ptr)->_new._reent._r48._mult) #define _REENT_RAND48_ADD(ptr) ((ptr)->_new._reent._r48._add) How-to Implement The member access macros are incomplete. The first step is to use the Newlib configuration for RTEMS as is and rename all struct _reent members, for example add an TEMPORARY prefix to all member names, _errno to TEMPORARY_errno. Then add member access macros until Newlib builds again. Install this Newlib and check that RTEMS and libbsd compiles. Run the RTEMS and libbsd test suites to check for regressions. In a second step to this for the _REENT_SMALL configuration of Newlib. The third step is to add a new Newlib configuration option, for example _REENT_THREAD_LOCAL which turns the struct _reent members into thread-local objects with corresponding "member" access macros. Define _REENT to NULL. Skills C and assembly Difficulty This is a large (350-hour) project of hard difficulty.
#4566	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Atomic store does not use the order parameter for C++
Description	This could result in invalid inline code for the SMP ticket locks if used from C++.
#4569	2 years ago	fixed	arch/aarch64	Sebastian Huber	Joel Sherrill	2 years ago
Summary	aarch64 cache support does not provide rtems_cache_disable_data()
Description	The aarch64 cache support defines: bsps/aarch64/shared/cache/cache.c:#define CPU_CACHE_SUPPORT_PROVIDES_DISABLE_DATA However, it does not provide rtems_cache_disable_data().
#4572	2 years ago	fixed	doc	Joel Sherrill	Sebastian Huber	2 years ago
Summary	Improve Clarity for New Clock Manager Directives
Description	Looking through the Clock Manager at the new APIs, I noticed that there is nothing about what bintime and sbintime are in the "Time and Date Data Structures" section. Explaining what these time representations are is needed. struct timeval and timespec are also not described in "Time and Data Data Structures" but this may be OK since they are from C and/or POSIX. There is no discussion of what coarse is versus the non-coarse and what this means in practice. There is no background on what the various clock sources are. There may be other similar typos but rtems_clock_get_monotonic() describes its sole argument as being of the wrong type. Without some background on the various time formats, sources, and terminology, the distinction between these APIs and their intended uses is impossible to figure out.
#4581	2 years ago	fixed	admin	Kinsey Moore	Kinsey Moore <kinsey.moore@…>	2 years ago
Summary	Add AArch64 to Tier 1 BSPs
Description	I finally got around to producing an emailed test run on the Avnet ZU3EG development board. ​The test results are available here
#4583	2 years ago	fixed	tool	Joel Sherrill	Joel Sherrill	2 years ago
Summary	rtems-bsp-builder: Do not vary on network
Description	With the network stack removed from the rtems.git repository, there is no need to build BSPs with and without networking. This should be removed. It should reduce the number of BSP build configurations in a run.
#4585	2 years ago	fixed	tool	Joel Sherrill	Joel Sherrill <joel@…>	2 years ago
Summary	Multiple Prototype Issues/Mismatches? in rtems-examples from trace linker
Description	This is some of the build failures in rtems-examples if the BSP has POSIX turned on: /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: hello-test.c:185:199: error: expected declaration specifiers or '...' before 'Thread_CPU_budget_algorithm_callout' /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: 185 | bool __wrap__Thread_Initialize(Objects_Information* a1, Thread_Control* a2, const Scheduler_Control* a3, void* a4, size_t a5, bool a6, Priority_Control a7, bool a8, Thread_CPU_budget_algorithms a9, Thread_CPU_budget_algorithm_callout a10, uint32_t a11, Objects_Name a12) /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: | ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ error: compiling wrapper: Compiler error /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: hello-deep.c:1296:6: error: conflicting types for '_Thread_Start' /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: 1296 | bool _Thread_Start(Thread_Control* a1, const Thread_Entry_information* a2, ISR_lock_Context* a3); /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: | ^~~~~~~~~~~~~ /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: In file included from hello-deep.c:30: /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: /home/joel/rtems-class/bsp-install/aarch64-rtems6/xilinx_zynqmp_lp64_qemu/lib/include/rtems/score/threadimpl.h:279:16: note: previous declaration of '_Thread_Start' was here /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: 279 | Status_Control _Thread_Start( /home/joel/rtems-class/tools/bin/aarch64-rtems6-gcc: | ^~~~~~~~~~~~~ error: compiling wrapper: Compiler error
#4588	2 years ago	fixed	admin	Kinsey Moore	Kinsey Moore <kinsey.moore@…>	2 years ago
Summary	RTEMS Stack Checker Report Bad Access When Not Enabled
Description	The RTEMS stack checker reporting function can be called without the stack checker infrastructure being enabled. This causes a bad access to 0x0 because the static interrupt stack checker data remains null instead of being initialized by rtems_stack_checker_begin_extension. The actual failure occurs in Stack_check_Find_high_water_mark while searching for U32_PATTERN. This was found when the trailing D was left off of the CONFIGURE_STACK_CHECKER_ENABLED define in a test program.
#4607	2 years ago	wontfix	bsps	ostyche		21 months ago
Summary	support for SATA（AHCI） drives
Description	Added support for SATA（AHCI） drives.The relevant file path is \cpukit\sata\ The files involved in the question were uploaded by git user ostyche who created a branch at 10:55 am on February 11, 2022
#4613	2 years ago	fixed	arch/sparc64	Gedare Bloom	Gedare Bloom	15 months ago
Summary	Deprecate sparc64 port in rtems6 and remove in rtems7
Description	This port is no longer maintained. No one has been testing it and it does not appear to have any users. Deprecate for RTEMS 6.x Remove in RTEMS 7.x
#4614	2 years ago	fixed	arch/sh	Gedare Bloom	Gedare Bloom	15 months ago
Summary	Deprecate sh port in rtems6 and remove in rtems7
Description	This port is no longer maintained. No one has been testing it and it does not appear to have any users. Deprecate for RTEMS 6.x Remove in RTEMS 7.x
#4615	2 years ago	fixed	arch/v850	Gedare Bloom	Gedare Bloom	15 months ago
Summary	Deprecate v850 port in rtems6 and remove in rtems7
Description	This port is no longer maintained. No one has been testing it and it does not appear to have any users. Deprecate for RTEMS 6.x Remove in RTEMS 7.x
#4617	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Potential non-monotonic CLOCK_MONOTONIC
Description	The RTEMS timecounter implementation could lead to a non-monotonic CLOCK_MONOTONIC if the clock interrupt is delayed for more than one second.
#4619	2 years ago	worksforme	bsps	Ryan Long		17 months ago
Summary	riscv/rv32iac: minimum sample test fails to build
Description	This issue was found while investigating #4618. The specifics of the error are listed there.
#4627	2 years ago	fixed	score	Sebastian Huber	Sebastian Huber	2 years ago
Summary	Multitasking start is broken on SMP targets which do not restore the interrupt state during context switching
Description	The current SMP multitasking start assumed that the initial heir thread of a processor starts execution in _Thread_Handler() which sets the interrupt state explicitly by _ISR_Set_level(). Under certain timing conditions processors may perform an initial context switch to a thread which already executes its thread body. In this case, interrupts are disabled after the context switch on targets which do not save/restore the interrupt state during a context switch (for example arm and riscv).
#4629	2 years ago	fixed	shell	Sebastian Huber	Sebastian Huber	2 years ago
Summary	shell: Potential stack corruption in rtems_shell_init()
Description	The rtems_shell_init() passes the address of a stack variable (exit_code) to rtems_shell_run(). If wait == false, then the stack variable goes out of scope but may be accessed by the created shell thread.
#4631	2 years ago	fixed	tool/gdb	Frank Kuehndel	Chris Johns	2 years ago
Summary	RSB fails to build gdb with Python 3.10 (and newer)
Description	The rtems-source-builder fails to build the tools in a Fedora 35 container. I am building ARM for RTEMS 6 and get this error message: config: tools/rtems-gdb-11.2.cfg error: shell macro failed: /home/minna/src/rtems-source-builder/source-builder/sb/rtems-build-dep -c gcc -l : 2: error: no library (-l) provided Build Set: Time 0:00:33.396165 The same error appears for other architectures like AARCH64, POWERPC, RISCV on Fedora 35. The source for this error is in file source-builder/config/gdb-common-1.cfg in line 118: 104 %ifnos darwin 105 %if %{host_ldflags} == %{nil} 106 %define gdb-host-libs %{nil} 107 %else 108 %define gdb-host-libs -L '%{host_ldflags}' 109 %endif 110 %if %{gdb-python-config} != %{nil} 111 %define gdb-python-lib-filter awk 'BEGIN{FS=" "}/python/{for(i=1;i<NF;++i)if(match($i,".*lpython.*")) print "lib"substr($i,3)"*";}' 112 %if %{gdb-python-ver-mm} < 3.8 113 %define gdb-python-config-lib-check-flags --ldflags 114 %else 115 %define gdb-python-config-lib-check-flags --ldflags --embed 116 %endif 117 %define gdb-python-config-libs %(%{gdb-python-config} %{gdb-python-config-lib-check-flags} | %{gdb-python-lib-filter}) 118 %define gdb-python-lib-check %(%{_sbdir}/sb/rtems-build-dep -c %{__cc} %{gdb-host-libs} -l %{gdb-python-config-libs}) 119 %else 120 %define gdb-python-lib-check %(%{_sbdir}/sb/rtems-build-dep -c %{__cc} %{gdb-host-libs} -l %{gdb-python-ver-lib}) 121 %endif 122 %if %{gdb-python-lib-check} == not-found && !%{_rsb_getting_source} 123 %error "gdb: python: library file not found: %{gdb-python-ver-lib}, please install" 124 %endif 125 %endif Some values of involved variables: gdb-python-ver-mm = 3.10 gdb-python-config = /usr/bin/python3-config gdb-python-config-lib-check-flags = --ldflags gdb-python-lib-filter = awk 'BEGIN{FS=" "}/python/{for(i=1;i<NF;++i)if(match($i,".*lpython.*")) print "lib"substr($i,3)"*";}' gdb-python-config-libs = '' The trouble is that gdb-python-config-libs is empty when line 118 gets called. It is empty because there is no python-library in the output of this call on Fedora 35: $ /usr/bin/python3-config --ldflags -L/usr/lib64 -lcrypt -ldl -lm -lm Just to compare, on OpenSUSE 15.3 the same call returns (note the -lpython3.6m ): $ /usr/bin/python3-config --ldflags -L/usr/lib64 -lpython3.6m -lpthread -ldl -lutil -lm -Xlinker -export-dynamic Commenting out line 118 makes the build go through smoothly but I would prefer a permanent fix. Would be skipping line 118 when gdb-python-config-libs is empty be an appropriate solution? Is the "real" error in line 112? (%if 3.10 < 3.8 evaluates unexpectedly to true) Would the "error" in line 123 print out the correct library name? Does the whole check make sense when python3-config is obviously installed (both the command and the library are "usually" in the same pythonX-devel package)? Notes: On Fedora 35 /usr/lib64/libpython3.10.so exists. On Fedora 35 adding --embed provides the library: $ /usr/bin/python3-config --ldflags --embed -L/usr/lib64 -lpython3.10 -lcrypt -ldl -lm -lm For the records: Command: ../source-builder/sb-set-builder --rtems-version 6 --prefix /opt/rtems/6 --with_cxx --with_fortran --with_objc --jobs=12 --log ~/sb-set-builder.log 6/rtems-arm RTEMS rtems-source-builder git: 6fe98f91d94bbf965bc0e78015585ff8823d17bd RTEMS: 6 Target architecture: ARM Host architecture: x86_64 Hoast Kernel: 5.3.18-150300.59.49-default OS: Fedora Linux 35 (Container Image) Container image base: docker.io/library/fedora:latest Python version in container: 3.10.2 Test date: 2022-03-18T11:02:10+01:00 rtems-source-builder/source-builder/sb-check log: RTEMS Source Builder - Check, 6 (6fe98f91d94b modified) warning: exe: absolute exe found in path: (__chown) /usr/sbin/chown warning: exe: absolute exe found in path: (__install_info) /usr/bin/install-info Environment is ok
#4632	2 years ago	fixed	lib	Sebastian Huber	Chris Johns	2 years ago
Summary	Error in rtems/thread.hpp with GCC 12
Description	There is a compile error with GCC 12: In file included from /home/EB/sebastian_h/src/rtems/cpukit/librtemscxx/thread.cpp:35: /home/EB/sebastian_h/src/rtems/cpukit/include/rtems/thread.hpp:449:14: error: redefinition of default argument for 'class<template-parameter-1-4>' 449 | class = thread::enable_if_attributes<A>> | ^~~~~ /home/EB/sebastian_h/src/rtems/cpukit/include/rtems/thread.hpp:312:17: note: original definition appeared here 312 | class = enable_if_attributes<A>> | ^~~~~
#4634	2 years ago	fixed	tool/rsb	Ryan Long		15 months ago
Summary	Fix hash for glib-2.48
Description	The hash needs updated in order for glib to be built. This is currently preventing
#4635	2 years ago	fixed	fs/jffs2	Chris Johns	Chris Johns <chrisj@…>	2 years ago
Summary	untar on JFFS slow down
Description	Using untar on JFFS is slower on rtems 6 compared to 4.11. The application writes files to the flash disk from an in memory tarfile and the difference is easy to see for 10M or so of files. I reviewed the changes to JFFS in rtems 6 and they look OK so it is not a regression the libfs. I also timed the flash driver and it is the same on 6 as 4.11. I noticed on 6 there is a lot more blocks being written at the flash driver level compared to 4.11. Looking at untar I see the block size written is 512 bytes and with compression this results in a lot of small blocks. Changing the untar block write size to 8K speeds returns the performance. I am not sure why there are more blocks on 6. A guess is the compression is better, which is is nice. I will not look into this any more than changing untar to use 8K blocks.
#4639	2 years ago	fixed	fs	Joel Sherrill	Christian Mauderer	2 years ago
Summary	Error in imfs memfile when device is full
Description	Cloned from #4638: The problem from #2353 still exists in 5 and 6: Original description: From the reporter (Круглов Сергей <skruglov@…>): First bug I detect with attempt write file to full device (zero empty blocks). And after I send too many files, disk overflow and system not responding (receive and write to disk have high priority). A test, and found block=0, after -- = too big value. In imfs_memfile.c function IMFS_memfile_extend block declared: unsigned int block, and after block— work incorrect (if block = 0). for ( ; block>=old_blocks ; block-- ) { IMFS_memfile_remove_block( memfile, block ); } Must be declare int block;
#4641	2 years ago	fixed	arch/x86_64	Ryan Long	Ryan Long <ryan.long@…>	2 years ago
Summary	x86_64 elf_machdep header file missing body
Description	The elf_machdep header file is missing the body under cpukit/score/cpu/x86_64/include/machine. Joel found the code that needed to be put in at ​http://cvsweb.netbsd.org/bsdweb.cgi/src/sys/arch/amd64/include/elf_machdep.h.
#4644	2 years ago	fixed	score	Matthew J Fletcher	Sebastian Huber	2 years ago
Summary	wkspace "failed allocations" increase with heap protection enabled (cloned)
Description	Cloned from #4643: Investigating the "Total number of failed allocations:" from the wkspace cmd i see that; /* Statistics */ ++stats->failed_allocs; Are incremented when _Heap_Allocate_aligned_with_boundary fails, however it seems this will routinely fail due to; _Heap_Protection_free_all_delayed_blocks seemingly on purpose creating an impossibly large allocation. uintptr_t large = 0 (uintptr_t) HEAP_BLOCK_HEADER_SIZE (uintptr_t) HEAP_ALLOC_BONUS (uintptr_t) 1; void *p = _Heap_Allocate( heap, large ); This makes it difficult to place a breakpoint on failed_allocs to investigate true low memory situations with heap protection enabled.
#4646	2 years ago	fixed	tool/rsb	Chris Johns	Chris Johns	2 years ago
Summary	Control buildsets using the --with-* command line option
Description	The following thread has raise an important issue on how we can test newer versions of tools. This ticket is about providing a solution: ​https://lists.rtems.org/pipermail/devel/2022-May/071458.html The solution is allow users to specific a tool set on the command line for a build of tools. The standard option is --with-.*. For example a command line could be: ../source-builder/sb-set-builder \ --prefix=/opt/rtems/6 \ --log=arm.txt \ --with-rtems-gcc=tools/rtems-gcc-head-newlib-head \ 6/rtems-arm The buildset logic is to be changed to expand macros for buildset file names and to add a new buildset specific keyword %defineifnot. The tools default file becomes: # # Default tools configuration. # %include 6/rtems-base.bset # # Default RTEMS build. Override on the command line to the # the RSB. For example to test the latest versino of gcc use: # # --with-rtems-gcc=tools/rtems-gcc-head-newlib-head # %defineifnot with_rtems_dtc devel/dtc-1.6.1-1 %defineifnot with_rtems_expat devel/expat-2.4.8-1 %defineifnot with_rtems_gmp devel/gmp-6.2.1 %defineifnot with_rtems_gdb tools/rtems-gdb-11.2 %defineifnot with_rtems_binutils tools/rtems-binutils-2.38 %defineifnot with_rtems_gcc tools/rtems-gcc-10-newlib-head %defineifnot with_rtems_tools tools/rtems-tools-6 # # Build gdb first to raise the Python install error as early as # possible. GDB needs expat so it needs to be built before. # %{with_rtems_dtc} %{with_rtems_expat} %{with_rtems_gmp} %{with_rtems_gdb} %{with_rtems_binutils} %{with_rtems_gcc} %{with_rtems_tools} This approach lets cron based CI flows configure a build with the latest gcc, binutils etc so they can always be tested as a workflow. The new keyword lets a define in a nested buildset file override a command line value. That may be needed.
#4650	2 years ago	fixed	network/libbsd	Sebastian Huber	Sebastian Huber	2 years ago
Summary	tcpdump: Fix dumping to file and reading from file (cloned)
Description	Cloned from #4649: When dumping to file, the file needs to be closed at program exit. When reading from file, we do not need the loop monitor. Do not use signals, since they are not cleaned up at program exit.
#4652	2 years ago	fixed	network/libbsd	Sebastian Huber	Sebastian Huber	23 months ago
Summary	if_atsam: Fix checksum offload, add multicast and VLAN support (cloned)
Description	Cloned from #4651: The if_atsam network interface driver has several issues. Do not disable the FCS if the checksum offload is disabled. Make sure the interface capabilities are enabled. Add multicast and VLAN support. Use the interface transmit method instead of the interface start approach. Correct a potential receive starvation error. Fix the interface up/down.
#4654	2 years ago	fixed	network/libbsd	Sebastian Huber	Sebastian Huber	2 years ago
Summary	pfctl: Fix global state initialization (cloned)
Description	Cloned from #4653: The last update introduced a static variable in get_query_socket() which was not initialized.
#4656	2 years ago	fixed	fs	Sebastian Huber	Sebastian Huber	2 years ago
Summary	sync() whould synchronize all file descriptors (cloned)
Description	Cloned from #4655: According to POSIX ​https://pubs.opengroup.org/onlinepubs/9699919799/functions/sync.html we have The sync() function shall cause all information in memory that updates file systems to be scheduled for writing out to all file systems. Currently, the RTEMS sync() implementation synchronizes only the file descriptors associated with a FILE object. This should be changed to call fsync() and fdatasync() for all file descriptors.
#4658	2 years ago	fixed	arch/riscv	Ryan Long		17 months ago
Summary	Fix build issue with riscv sample test
Description	When trying to build the minimum sample test for any RISC-V BSP, the following error occurs. start.o: in function .L0: /home/tester/rtems-cron-6/rtems/build/riscv/rv32i/../../../bsps/riscv/shared/start/start.S:84:(.bsp_start_text+0x2c): relocation truncated to fit: R_RISCV_GPREL_I against symbol 'bsp_section_bss_size' defined in *ABS* section in /home/tester/rtems-cron-6/rtems/build/riscv/rv32i/testsuites/samples/minimum.exe collect2: error: ld returned 1 exit status
#4661	23 months ago	fixed	tool/gcc	Chris Johns	Chris Johns <chrisj@…>	21 months ago
Summary	GCC 12 generates wrong code
Description	Calling exit() in a C++ application results in std::terminate() being called. The std::ios_base::Init::~Init() destructor flushes std::cout, std::cerr and std::clog. The flush of std::cerr results in terminate being called. The ios_base class uses a sentry to check the state. The code in it's destructor is: if (bool(_M_os.flags() & ios_base::unitbuf) && !uncaught_exception()) The std::cerr object has ios_base::unitbuf set so uncaught_exception() is called. This call gets the cxa globals: __cxa_eh_globals *globals = __cxa_get_globals (); The __cxa_get_globals() checks if the ec_globals init object: static __eh_globals_init init; to see if it is still valid and the state is: (gdb) p /x init $2 = { _M_key = 0x13010001, _M_init = 0x1 } However the static destructor has run and the POSIX key has been deleted. This results in std::terminate() being called. The code for the destructor on ARM is: 006562e0 <__eh_globals_init::~__eh_globals_init()>: 6562e0: 7903 ldrb r3, [r0, #4] 6562e2: b510 push {r4, lr} 6562e4: 4604 mov r4, r0 6562e6: b90b cbnz r3, 6562ec <__eh_globals_init::~__eh_globals_init()+0xc> 6562e8: 4620 mov r0, r4 6562ea: bd10 pop {r4, pc} 6562ec: 6800 ldr r0, [r0, #0] 6562ee: f04d fe6b bl 6a3fc8 <pthread_key_delete> 6562f2: 4620 mov r0, r4 6562f4: bd10 pop {r4, pc} and for aarch64 it is: 0000000010222c30 <__eh_globals_init::~__eh_globals_init()>: 10222c30: 39401001 ldrb w1, [x0, #4] 10222c34: 35000041 cbnz w1, 10222c3c <__eh_globals_init::~__eh_globals_init()+0xc> 10222c38: d65f03c0 ret 10222c3c: a9bf7bfd stp x29, x30, [sp, #-16]! 10222c40: 910003fd mov x29, sp return pthread_key_delete (__key); 10222c44: b9400000 ldr w0, [x0] 10222c48: 94008806 bl 10244c60 <pthread_key_delete> 10222c4c: a8c17bfd ldp x29, x30, [sp], #16 10222c50: d65f03c0 ret The _M_init boolean is not set to false.
#4662	23 months ago	fixed	tool/gcc	Chris Johns		6 months ago
Summary	Fix GCC 12 warnings
Description	The move to GCC 12 has resulted in an increase in the number of warning. This ticket covers fixing these warning.
#4666	22 months ago	fixed	fs	Sebastian Huber	Sebastian Huber	3 months ago
Summary	TFTP: Implement block and window size options
Description	Support the following TFTP options: RFC 2347 TFTP Option Extension RFC 2348 TFTP Blocksize Option RFC 7440 TFTP Windowsize Option
#4667	22 months ago	fixed	tool/newlib	Alexandre Oliva	Sebastian Huber	6 months ago
Summary	include/sys/_bitset.h defines struct bitset, pollutes namespace
Description	sys/_bitset.h gets indirectly (*) included by sys/types.h, and should thus define only names that C and C++ standards reserve for the implementation. However, the line: BITSET_DEFINE(bitset, 1); expands to a definition of struct bitset in the global namespace. Some of GCC's libstdc++ tests were failing on aarch64-rtems6.0 because of the naming ambiguity that arises. ​https://gcc.gnu.org/pipermail/gcc-patches/2022-June/596918.html It would be best if user programs didn't have to resort to such trickery. In C++, it's not so hard; in C, it can be trickier. I have not checked whether this struct bitset is actually used for any purpose on aarch64-rtems, or whether it's present on other ports. Maybe it could just be removed or renamed? Maybe upstream (FreeBSD?) would consider dropping it? (*) sys/types.h includes sys/_pthreadtypes.h includes sys/cpuset.h includes sys/_cpuset.h includes sys/_bitset.h
#4668	22 months ago	fixed	tool/gcc	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Use thread-local storage in libstdc++
Description	Fix the libstdc++ configuration so that TLS is used on targets with TLS support. Using TLS is more robust against shortages of system resources which lead to unexpected behaviour.
#4669	22 months ago	fixed	score	Sebastian Huber	Sebastian Huber	21 months ago
Summary	clock_nanosleep() uses the wrong clock to determine the start time point
Description	See gcc-patches mailing list: On 22/06/2022 08:22, Sebastian Huber wrote: > On 22/06/2022 08:01, Alexandre Oliva via Gcc-patches wrote: >> >> On rtems under qemu, the frequently-interrupted nanosleep ends up >> sleeping shorter than expected, by a margin of less than 0,3%. >> >> I figured failing the library test over a system (emulator?) bug is >> undesirable, so I put in some tolerance for the drift. >> >> Regstrapped on x86_64-linux-gnu, also tested with a cross to >> aarch64-rtems6. Ok to install? >> >> PS: I see nothing wrong with the implementation of clock_nanosleep (used >> by nanosleep) on rtems6 that could cause it to wake up too early. I >> suspect some artifact of the emulation environment. >> >> >> for libstdc++-v3/ChangeLog >> >> * testsuite/30_threads/this_thread/60421.cc: Tolerate a >> slightly early wakeup. >> --- >> .../testsuite/30_threads/this_thread/60421.cc | 3 ++- >> 1 file changed, 2 insertions(+), 1 deletion(-) >> >> diff --git a/libstdc++-v3/testsuite/30_threads/this_thread/60421.cc b/libstdc++-v3/testsuite/30_threads/this_thread/60421.cc >> index 12dbeba1cc492..f3a5af453c4ad 100644 >> --- a/libstdc++-v3/testsuite/30_threads/this_thread/60421.cc >> +++ b/libstdc++-v3/testsuite/30_threads/this_thread/60421.cc >> @@ -51,9 +51,10 @@ test02() >> std::thread t([&result, &sleeping] { >> auto start = std::chrono::system_clock::now(); >> auto time = std::chrono::seconds(3); >> + auto tolerance = std::chrono::milliseconds(10); >> sleeping = true; >> std::this_thread::sleep_for(time); >> - result = std::chrono::system_clock::now() >= (start + time); >> + result = std::chrono::system_clock::now() + tolerance >= (start + time); >> sleeping = false; >> }); >> while (!sleeping) > > This looks like a bug in RTEMS or the BSP for the test platform. I would first investigate this and then change the test which looks all right to me. This is a problem in RTEMS. RTEMS uses the FreeBSD timecounters to maintain CLOCK_REALTIME and provides two methods to get the time in a coarse and fine resolution. The std::chrono::system_clock::now() uses the fine resolution (higher overhead). The clock_nanosleep() uses the coarse resolution which may give a time before now().
#4670	22 months ago	fixed	test	Sebastian Huber	Sebastian Huber	20 months ago
Summary	Add option to build RTEMS with coverage instrumentation
Description	Add build options to build RTEMS with coverage instrumentation. Dump the gcov information after each test.
#4672	22 months ago	fixed	score	Sebastian Huber	Sebastian Huber	6 months ago
Summary	Thread-local storage should be usable once an idle thread exists
Description	At some point during system initialization the idle threads are created. Afterwards the boot processor basically executes within the context of an idle thread with thread dispatching disabled. On some architectures, the thread-local storage state must be applied to processor registers. Add a new CPU port function to do this: _CPU_Use_thread_local_storage( Context_Control *executing ).
#4673	22 months ago	fixed	lib/dl	Ryan Long	Ryan Long <ryan.long@…>	21 months ago
Summary	Runtime loader exported symbols address size not consistent
Description	In rtl-sym.c, where the exported symbols table is read in rtems_rtl_symbol_global_add, the addresses are expected to be of size sizeof(unsigned long). This works fine for most architectures because most of them evaluate to 4. For Aarch64 however, this evaluates to 8.
#4677	22 months ago	fixed	score	Adrian Varlan	Sebastian Huber	6 months ago
Summary	incorrect handling of "inactive_per_block" from "Objects_Information" structure (cloned)
Description	Cloned from #4676: The inactive_per_block for block 1 is not handled correctly. For the first object in the second block, the inactive_per_block value is not decremented. As a consequence, the block might be deleted when the second to last object from block 1 is deleted. This is only valid when the application is configured with "CONFIGURE_UNLIMITED_OBJECTS". The reason is the following: in function "_Objects_Activate_unlimited" from "objectimpl.h" block = _Objects_Get_index( the_object->id ) - OBJECTS_INDEX_MINIMUM; if ( block > objects_per_block ) { block /= objects_per_block; information->inactive_per_block[ block ]--; information->inactive--; } in a block with 8 objects per block, when creating the 9th object, _Objects_Get_index( the_object->id ) returns 9. Therefore, "block" will be 8 (9 - 1). The next "if" will not be taken (8 > 8 ? No) and the inactive_per_block will not be decremented to 7 and remains at 8. This means that block 1 is a candidate for "_Objects_Shrink_information" function although object id 9 is still valid. Steps to reproduce: -assuming 8 objects per allocation block -using semaphores as objects create 17 "RTEMS_SIMPLE_BINARY_SEMAPHORE" semaphore objects id's 1 -> 17. delete semaphores 11, 12, 13, 14, 15, 16, 17 (order is important) lock semaphore 9 delete semaphore 10 -> at this point, block 1 is freed although semaphore 9 is still active unlock semaphore 9 -> this can lead to errors. In my case due to the free the semaphore object was altered and changed from a "SEMAPHORE_VARIANT_SIMPLE_BINARY" to "SEMAPHORE_VARIANT_MUTEX_INHERIT_PRIORITY" resulting in a return value of "RTEMS_NOT_OWNER_OF_RESOURCE". Simple fix: the condition for the "if" should be ">=" instead of ">".
#4678	21 months ago	fixed	score	Sebastian Huber	Sebastian Huber	6 months ago
Summary	System initialization time increased compared to RTEMS 4.11
Description	Since RTEMS 5, the objects based on Objects_Control are statically allocated in the .bss sections. The zero-initialization is unnecessary since each object is fully initialized after allocation by an object-specific initialization routine. The memory for objects is statically allocated or allocated from the workspace if the unlimited objects configuration option is used. Workspace memory is uninitialized by default. Optimize the static allocation to place the object controls, local tables, and thread queue heads into a .noinit section. The longer system initialization time compared to RTEMS 4.11 may lead to hardware watchdog problems.
#4679	21 months ago	fixed	score	Sebastian Huber	Sebastian Huber	12 months ago
Summary	Use priority inheritance for thread join
Description	Threads may join the thread termination of another thread using the pthread_join() or rtems_task_delete() directives. Currently, the thread cancel operation uses a special case priority boosting mechanism: static void _Thread_Raise_real_priority( Thread_Control *the_thread, Priority_Control priority ) { Thread_queue_Context queue_context; _Thread_queue_Context_initialize( &queue_context ); _Thread_queue_Context_clear_priority_updates( &queue_context ); _Thread_Wait_acquire( the_thread, &queue_context ); if ( priority < the_thread->Real_priority.priority ) { _Thread_Priority_change( the_thread, &the_thread->Real_priority, priority, PRIORITY_GROUP_LAST, &queue_context ); } _Thread_Wait_release( the_thread, &queue_context ); _Thread_Priority_update( &queue_context ); } The problem is that this approach is not transitive, it does not account for priority adjustments of the calling task while waiting for the join, clustered scheduling is not supported, and deadlocks are not detected. All these problems are fixed by using a priority inheritance thread queue for the join operation.
#4680	21 months ago	fixed	posix	Sebastian Huber	Sebastian Huber	12 months ago
Summary	POSIX tasks cancelled through rtems_task_delete() should have an exit value of PTHREAD_CANCELED
Description	The rtems_task_delete() directive is basically just a combined pthread_cancel() and pthread_join(). In addition, it removes the PTHREAD_DETACHED state. The exit value returned by pthread_join() of threads cancelled by rtems_task_delete() should reflect this by getting a PTHREAD_CANCELED value and not NULL.
#4682	21 months ago	fixed	lib/dl	Ryan Long	Ryan Long <ryan.long@…>	21 months ago
Summary	Add AArch64 support to libdl
Description	Add AArch64 support to libdl using NetBSD code as necessary.
#4684	21 months ago	fixed	bsps	Chris Johns	Chris Johns	21 months ago
Summary	Support Versal DDRMC0 regions 0 and 1
Description	The Versal's DDRMC0 supports region 0 and 1. Region 0's base address defaults to 0x00000000000 and can be up to 2G in length. If the hardware has more than 2G of DDR memory it is placed in region 1 and it's default base address is 0x00800000000 placing it in the A64 address space. Add support so the user defined memory length BSP_XILINX_VERSAL_RAM_LENGTH spans region 0 and 1. Extend the heap with the region 1 memory.
#4686	21 months ago	fixed	lib/dl	Ryan Long	Ryan Long <ryan.long@…>	21 months ago
Summary	Refactor shared code in Arm and AArch64
Description	rtl-mdreloc-arm.c and rtl-mdreloc-aarch64.c share code. The shared code should be pulled out and included into each file.
#4689	21 months ago	fixed	arch/mips	Joel Sherrill	Joel Sherrill <joel@…>	18 months ago
Summary	mips/jmr3904 Massive Number of Test Failures
Description	jmr3904 last had a good test run in January 2022 (now July). Since then, there has not been one in which more than 1 test pass and I suspect that's because the one doesn't expect output. ​https://lists.rtems.org/pipermail/build/2022-January/030809.html This has been tracked down to a gdb simulator bug which has been reported as: ​https://sourceware.org/bugzilla/show_bug.cgi?id=29439 Many details follow in the comments.
#4690	21 months ago	fixed	score	Chris Johns	Chris Johns <chrisj@…>	21 months ago
Summary	CLOCK_REALTIME thread queue not updated as part of clock_settime call
Description	Setting the CLOCK_REALTIME clock using clock_settime(CLOCK_REALTIME, , ) clock does not update thread relative period entries on the watchdog timer queue. A system's stability is effected if a user calls clock_settime on the CLOCK_REALTIME clock. The watchdog's REALTIME clock queue is references against the realtime clock and movement of the time means The effected parts are: User C code that calls nanosleep, usleep, sleep or clock_nanosleep(CLOCK_REALTIME, , ). User C++ code that uses std::this_thread::sleep_for() as GCC uses the nanosleep() call. LibBSD uses CLOCK_REALTIME. The ​nanosleep call is explained in more detail in the ​clock_nanosleep rational. The nanosleep clock must be CLOCK_REALTIME. The ​clock_settime standard says: Setting the value of the CLOCK_REALTIME clock via clock_settime() shall have no effect on threads that are blocked waiting for a relative time service based upon this clock, including the nanosleep() function; nor on the expiration of relative timers based upon this clock. Consequently, these time services shall expire when the requested relative interval elapses, independently of the new or old value of the clock.
#4691	21 months ago	fixed	posix	Joel Sherrill	Joel Sherrill	6 months ago
Summary	Add support for FACE Technical Standard Restriction on POSIX Absolute Timers
Description	Overview Most of the time, the FACE Technical Standard (​https://opengroup.org/face) defines a subset (e.g. profile) of a standard. It currently includes four profiles of the POSIX 1003.1-2008 standard with a few optional features like multiple processes and IPV6. However, in one case, it defines an error case for something POSIX defines as required behavior. This ticket is intended to track adding support for an application being able to configure whether it wants the POSIX behavior or that defined by the FACE Technical Standard. FACE Technical Standard Requirements POSIX allows for setting a timer on CLOCK_REALTIME[1] while the FACE Technical Standard defines this as an error. This behavior and deviation from POSIX has been in the FACE Technical Standard since the beginning and was present in Edition 1.0[2]. This is the language from Edition 1.0 of the FACE Technical Standard which was written in the tone of a white paper rather than as a set of requirements: 3.9.4 Time Management Partial support for the POSIX CLOCK_REALTIME, which permits a calendar-based clock time to be set that becomes available to all partitions. Due to safety and security-related considerations in the POSIX definitions of this clock’s behavior, the configuration data will specify which partitions (preferably one partition) have permissions to set (using the clock_settime API) this clock. All partitions may have access to read this clock, but an error will be returned for any partition that attempts to set an alarm based on this clock that is based on absolute time (setting of an alarm based on a relative time is supported). Partitions authorized to set the calendar time will also be authorized to set the time zone value that is visible to all other partitions. Subsequent editions of the FACE Technical Standard have fine-tuned the language and made requirements clearer with simpler direct "shall" statements. Also, although the term OSS UoC was defined in Edition 1.0, the requirements now use this acronym for Operating System Segment Unit of Conformance to refer to the operating system and foundational services. Other types of segments are defined for applications, communications, and device management. In Edition 3.1[3], the language has the same intent as Edition 1.0 but the requirements are stated in a more direct and clearer manner. 4.1.1.3 OSS POSIX Clock Requirements An OSS UoC provides support for POSIX clocks. An OSS UoC shall support the POSIX Monotonic Clock option (CLOCK_MONOTONIC), including setting absolute and relative timers on this clock. An OSS UoC shall support clock read-access based POSIX CLOCK_REALTIME clock (calendar-based clock time). An OSS UoC shall support setting a relative timer-based POSIX CLOCK_REALTIME clock (calendar-based clock time). An OSS UoC shall support clock time write access for POSIX CLOCK_REALTIME clock value under the constraint that configuration data specifies the partition is authorized to set the clock time (using the clock_settime() API). Note: The time value set by one partition becomes visible to all other partitions. The timezone value set by one POSIX process using the tzset() or setenv() APIs is local to that POSIX process. An OSS UoC shall return an error when a software component attempts to set an absolute timer on the POSIX CLOCK_REALTIME clock. Note: The behavior is necessary (e.g., to prevent backwards and forwards shifts in sequential timers) due to the setting of POSIX_CLOCK_REALTIME clock impacting all partitions. The use of the word "partition" is intentional because the FACE Technical Standard includes ARINC 653 and POSIX standards. ARINC 653 defines a time and space partitioned real-time environment. POSIX or ARINC 653 applications can be hosted in a partition. In three of the four profiles, ARINC 653 is required while in the fourth it is optional. RTEMS can be hosted in an ARINC 653 partition using a paravirtualized configuration with a BSP that adapts it to the environment. The requirement being addressed by this ticket is (5) since RTEMS currently meets the other requirements. (5) contradicts the POSIX standard and modifying RTEMS to not support absolute POSIX timers is not an acceptable option. Proposed Solution The proposed solution is to add a confdefs.h configure option that enables this behavior when the OSS UoC (e.g. operating system) must provide FACE conformant behavior. When not defined, the POSIX behavior will be in effect. This will require additions to the confdefs.h family of headers, adding a test, and updating documentation. References [1] POSIX timer_create - ​https://pubs.opengroup.org/onlinepubs/9699919799/functions/timer_create.html [2] FACE Technical Standard, Edition 1.0. January 2012. ​https://publications.opengroup.org/c122. [3] FACE Technical Standard, Edition 3.1. July 2020. ​https://publications.opengroup.org/c207.
#4693	21 months ago	fixed	arch/aarch64	Chris Johns	Chris Johns	20 months ago
Summary	Remove ILP32 aarch64/xilinx_versal_* BSPs
Description	The BSPs: aarch64/xilinx_versal_ilp32_qemu aarch64/xilinx_versal_ilp32_vck190 support ILP32 addressing and the Versal has default memory regions above the 4G address boundary. There is value in supporting ILP32 if there is not practical use for it. If a use case is identified they can be added back.
#4694	21 months ago	invalid	test	Lucian-Raul Silistru	joel@…	21 months ago
Summary	sptests/sp37 fails to compile if RTEMS_SMP is not defined
Description	Issue seems to be caused by lines 191-193 and 207 in testsuites/sptests/sp37/init.c name is undefined _ISR_lock_Initialize is "(void)name" In file included from ../../../../../../RTEMS/cpukit/include/rtems/rtems/intr.h:70, from ../../../../../../RTEMS/cpukit/include/rtems.h:69, from ../../../../../../RTEMS/bsps/include/bsp/default-initial-extension.h:40, from ../../../../../../RTEMS/bsps/myriad2/ma2x8x/include/bsp.h:27, from ../../../../../../RTEMS/testsuites/support/include/tmacros.h:41, from ../../../../../../RTEMS/testsuites/sptests/sp37/system.h:35, from ../../../../../../RTEMS/testsuites/sptests/sp37/init.c:41: ../../../../../../RTEMS/testsuites/sptests/sp37/init.c: In function 'test_isr_locks': ../../../../../../RTEMS/testsuites/sptests/sp37/init.c:207:42: error: 'name' undeclared (first use in this function) 207 | _ISR_lock_Initialize( &container.lock, name ); | ~ ../../../../../../RTEMS/cpukit/include/rtems/score/isrlock.h:202:12: note: in definition of macro '_ISR_lock_Initialize' 202 | (void) _name | ../../../../../../RTEMS/testsuites/sptests/sp37/init.c:207:42: note: each undeclared identifier is reported only once for each function it appears in 207 | _ISR_lock_Initialize( &container.lock, name ); | ~ ../../../../../../RTEMS/cpukit/include/rtems/score/isrlock.h:202:12: note: in definition of macro '_ISR_lock_Initialize' 202 | (void) _name |
#4697	21 months ago	invalid	network/libbsd	Joel Sherrill		17 months ago
Summary	Documentation for libbsd Configure options
Description	libbsd appears to have some options to configure it similar to confdefs.h. We need user facing documentation on those options.
#4700	21 months ago	fixed	tool	Ryan Long	Ryan Long <ryan.long@…>	21 months ago
Summary	Fix instances of "is" operator used with literals
Description	Python 3.8 introduced new warnings when using the "is" operator when comparing literals.
#4701	21 months ago	fixed	bsps	Sebastian Huber	Sebastian Huber	21 months ago
Summary	Relocatable write data is placed in read-only region on some BSPs
Description	The .data.rel.ro* linker input section pattern accidentally matches with writeable data those symbol name starts with "ro".
#4703	20 months ago	fixed	lib	Chris Johns	Chris Johns	20 months ago
Summary	NetSNMP 5.7.2.1 Patch
Description	This ticket is a place holder for the NetSNMP patch moving it out of the rtems-tools repo.
#4705	20 months ago	fixed	build	Chris Johns	Chris Johns	17 months ago
Summary	Installed header clean up
Description	A number of headers are installed when they should not or are not installed.
#4706	20 months ago	fixed	rtems	Chris Johns	Chris Johns	6 months ago
Summary	Installed header break C++
Description	Some headers generate C++ errors and so cannot be included.
#4710	20 months ago	fixed	tool/rsb	Chris Johns	Chris Johns	20 months ago
Summary	RSB does not expand dir type macros correctly
Description	The expansion of the dir type macros is not done correctly. Macro replacement only appends the file part to the last element of a path list. For example: %{mydir}/abc.cfg with a path of: %define mydir x:y:z results in: x:y:z/abc.cfg The config load code handles this but it is wrong. The expansion of dir type macro variables should handle path separators (:) so you get: x/abc.cfg:y/abc.cfg:z.abc.cfg This has not been a problem in the RSB because the _configdir has only had two elements and the last one only ever had subdirectories.
#4712	20 months ago	fixed	arch/powerpc	Sebastian Huber	Sebastian Huber	20 months ago
Summary	powerpc: Add AltiVec? VRSAVE support
Description	The VRSAVE feature of the Altivec unit can be used to reduce the amount of Altivec registers which need to be saved/restored during interrupt processing and context switches.
#4713	20 months ago	fixed	posix	Joel Sherrill	Joel Sherrill	20 months ago
Summary	Make pthread_atfork follow FACE Technical Standard Behavior for Single Process Environments
Description	The POSIX specification for pthread_atfork() does not address the behavior when in a single process environment. The RTEMS implementation was an arbitrary decision on our part with no basis from the wider POSIX community. The FACE Technical Standard includes profiles without multiple process support and has defined the behavior for this method as follows: When multiple POSIX processes are supported, pthread_atfork(), the functionality is per the POSIX standard. When multiple POSIX processes are not supported, pthread_atfork() should return 0. The provision to support only a single process is not addressed in the POSIX standard. The RTEMS implementation currently returns ENOSYS instead of 0. This ticket is to track the change to RTEMS and the documentation.
#4714	20 months ago	fixed	network/lwip	Chris Johns		17 months ago
Summary	Use waf for LwIP walk sources
Description	The lwip.py support has a call to walk the source tree for source files. Waf supports glob'ing of sources so why not use that? This ticket is to remove the walk.
#4718	19 months ago	fixed	fs	Frank Kuehndel	Sebastian Huber	17 months ago
Summary	Coverity issues in cpukit/libfs/src/ftpfs/tftpDriver.c
Description	Coverity indicates the following issues in cpukit/libfs/src/ftpfs/tftpDriver.c: CID 1506523 (#1 of 1): Unchecked return value from library (CHECKED_RETURN) check_return: Calling sendto(tp->socket, send_buf, plen, 0, (struct sockaddr *)&tp->farAddress, 16U) without checking return value. This library function may fail and return an error code. 727 sendto ( 728 tp->socket, 729 send_buf, 730 plen, 731 0, 732 (struct sockaddr *) &tp->farAddress, 733 sizeof (tp->farAddress) 734 ); CID 1506522 (#1 of 1): Unchecked return value from library (CHECKED_RETURN) check_return: Calling sendto(tp->socket, (char *)&msg, len, 0, (struct sockaddr *)to, 16U) without checking return value. This library function may fail and return an error code. 529 sendto (tp->socket, (char *)&msg, len, 0, (struct sockaddr *)to, sizeof *to); CID 1437618 (#1 of 3): Unchecked return value from library (CHECKED_RETURN) check_return: Calling setsockopt(tp->socket, 65535, 4102, &tv, 16U) without checking return value. This library function may fail and return an error code. 594 setsockopt (tp->socket, SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof tv); CID 1437618 (#2 of 3): Unchecked return value from library (CHECKED_RETURN) check_return: Calling setsockopt(tp->socket, 65535, 4102, &tv, 16U) without checking return value. This library function may fail and return an error code. 555 setsockopt (tp->socket, SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof tv); CID 1437618 (#3 of 3): Unchecked return value from library (CHECKED_RETURN) check_return: Calling setsockopt(tp->socket, 65535, 4102, &tv, 16U) without checking return value. This library function may fail and return an error code. 559 setsockopt (tp->socket, SOL_SOCKET, SO_RCVTIMEO, &tv, sizeof tv);
#4724	19 months ago	fixed	doc	Chris Johns	Joel Sherrill	16 months ago
Summary	Git commit message format instructions
Description	The Eng docs ​Creating A Patch references the Wiki. The content should be in the manual. Note lots of git related bits are in the Eng manual and still on the Wiki so which is correct?
#4726	19 months ago	fixed	admin	Chris Johns	Chris Johns	19 months ago
Summary	RSB decode exception stops build
Description	Building in a Rocky VM on FB 12 with 5 I got: Traceback (most recent call last): File "/usr/lib64/python3.9/threading.py", line 973, in _bootstrap_inner self.run() File "/usr/lib64/python3.9/threading.py", line 910, in run self._target(*self._args, **self._kwargs) File "/opt/work/chris/rtems/rsb/rtems-source-builder.git/source-builder/sb/execute.py", line 204, in _readthread data = data.decode(sys.stdout.encoding) UnicodeDecodeError: 'utf-8' codec can't decode byte 0xe2 in position 4095: unexpected end of data If the data is corrupted or broken things stop. Fix to attempt to keep going. This issue also effect rtems-tools.
#4729	19 months ago	fixed	lib	Chris Johns	Chris Johns	17 months ago
Summary	Add 64bit address support to RTEMS FDT
Description	RTEMS's FDT wrapper does not handle the reg property correctly. Update to handle 64bit register addresses.
#4730	19 months ago	fixed	tool/rsb	Christian Mauderer	Chris Johns	19 months ago
Summary	rtems-source-builder doesn't generate tar archives for all packages any more
Description	Like discussed in the following mailing list thread, the rtems-source-builder doesn't generate tar archives for all packages anymore: ​https://lists.rtems.org/pipermail/devel/2022-September/073369.html All RTEMS tools build sets 6/rtems-* work fine except for 6/rtems-microblaze. devel/qemu and devel/dtc do not generate a tar archive. If the patch ​https://git.rtems.org/rtems-source-builder/commit/?id=6205068c5a429e9ee1b471f4c8a3a119bb6757b2 is reverted, the tar files are generated just like before.
#4732	19 months ago	fixed	tool/rsb	Chris Johns	Chris Johns	19 months ago
Summary	Set top in RSB version.py
Description	Setting the top lets the code be used in deployment to get the version of the RSB being used. An RSB version lets a user track the exact tools being built and version numbers reported by the tools can match the packaged build, eg an rpm name and version info.
#4741	18 months ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	18 months ago
Summary	Fix GCC PR 107248
Description	​https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107248
#4744	18 months ago	fixed	lib/debugger	Chris Johns	Chris Johns	18 months ago
Summary	libdebugger arm backend does not single step
Description	Single stepping with the ARM backend does not cleanly step instructions.
#4746	18 months ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	18 months ago
Summary	Armv7-M SysTick? clock driver uses off by one reload value
#4750	18 months ago	invalid	rtems	dsa93		18 months ago
Summary	Linker discards RTEMS_SYSINIT_ITEM object
Description	Hi I am developing device driver in custom BSP in RTEMS. I applied my BSP on top of master branch. My driver code looks like this: /* driver logic in "static" functions (omitted) */ void custom_driver_init(void); RTEMS_SYSINIT_ITEM( custom_driver_init, RTEMS_SYSINIT_DEVICE_DRIVERS, RTEMS_SYSINIT_ORDER_SEVENTH ); It is discarded by linker. custom_driver_init is never called. I discovered hack how to make linker not discard this: if I put at least one extern linkage function and call it in other file then linker dont discard "custom_driver_init" /* driver logic in "static" functions (omitted) */ /* Call this function for example from main function */ int workaraound(void) { return 5; } void custom_driver_init(void); RTEMS_SYSINIT_ITEM( custom_driver_init, RTEMS_SYSINIT_DEVICE_DRIVERS, RTEMS_SYSINIT_ORDER_SEVENTH ); Driver file compilation arguments: -MMD, -Wall, -Wmissing-prototypes, -Wimplicit-function-declaration, -Wstrict-prototypes, -Wnested-externs, -O2, -g, -fdata-sections, -ffunction-sections, -mthumb, -mcpu=cortex-m7, -mfpu=fpv5-d16, -mfloat-abi=hard, -save-temps, -Icpukit/include, -I../../../cpukit/include, -Icpukit/score/cpu/arm/include, -I../../../cpukit/score/cpu/arm/include, -Ibsps/include, -I../../../bsps/include, -Ibsps/arm/include, -I../../../bsps/arm/include, ../../../bsps/arm/someplatform/console/console.c, -c, -DHAVE_CONFIG_H=1, -DBOARD_MAINCK_FREQ_HZ=10000000 Application linkage files /home/dsa/rtems/rsb-6/bin/arm-rtems6-gcc -g -mthumb -mcpu=cortex-m7 -mfpu=fpv5-d16 -mfloat-abi=hard -DBOARD_MAINCK_FREQ_HZ=10000000 -I/home/dsa/arm-rtems6/someplatform/lib/include -B/home/dsa/arm-rtems6/someplatform/lib -qrtems -Wl,--gc-sections -Wl,-Map=ecu-app.map,--print-memory-usage,--cref CMakeFiles/ecu-app.dir/src/main.c.obj -o ecu-app.elf -lrtemsbsp -lrtemscpu I tried to research this topic and I am not sure if there is a way to prevent this optimisation, so I decided to post it here. I can provide you some more details or perform some tests
#4751	18 months ago	fixed	shell	chenjin_zhong	Chris Johns	16 months ago
Summary	chmod problem in shell of RTEMS 6
Description	Cloned from #4558: Hi, I find one problem in rtems_shell_main_chmod function. the usage of this instruction is descripted as "chmod 0777 n1 n2... #change filemode", .the src code is listed as follows. the bold part should be replaced with chmod(argv[n], mode)? static int rtems_shell_main_chmod( int argc, char *argv[] ) { int n; mode_t mode; unsigned long tmp; if (argc < 2) { fprintf(stderr,"%s: too few arguments\n", argv[0]); return -1; } /* * Convert arguments into numbers */ if ( rtems_string_to_unsigned_long(argv[1], &tmp, NULL, 0) ) { printf( "Mode argument (%s) is not a number\n", argv[1] ); return -1; } mode = (mode_t) (tmp & 0777); /* * Now change the files modes */ for (n=2 ; n < argc ; n++) chmod(argv[n++], mode); /* <<<< here */ return 0; }
#4756	18 months ago	fixed	doc	Chris Johns	Chris Johns <chrisj@…>	17 months ago
Summary	Docs build system does not build singlehtml (cloned)
Description	Cloned from #4755: The configure logic does not handle the inliner detection correctly.
#4759	17 months ago	fixed	arch/arm	jkillelea	Sebastian Huber <sebastian.huber@…>	16 months ago
Summary	TLS support is broken for ARMv6
Description	I have found two possible memory exceptions under rtems_shell_init when running on the raspberry pi 1 BSP. First, a possible null pointer dereference from getpwuid_r(0, &pwd, NULL, 0, &pwd_res); in rtems_shell_init_once. The NULL pointer is never checked before it is used as a backing buffer for pwd. Second, usage of stdin, stdout, and stderr cause memory exceptions in rtems_shell_run when used, for example shell_env->parent_stdin = stdin;.
#4762	17 months ago	fixed	network/libbsd	Aaron N	Chris Johns	17 months ago
Summary	Add Versal SDHCI Arasan driver to libbsd
Description	Added support for SD card slots in libbsd for the Versal. Tested on VCK190.
#4763	17 months ago	fixed	shell	Chris Johns	Chris Johns	17 months ago
Summary	Detect terminal size
Description	Add support to set the rows and columns in the shell. RTEMS does not have a full tty system and does not carry terminfo so implement a shell based approach to let some commands manage the terminal size. Update the edit command to support the terminal size. Fix the help lines.
#4766	17 months ago	fixed	doc	Joel Sherrill	Joel Sherrill	17 months ago
Summary	Classic API User Guide Index has Functions without Parentheses
Description	Some index entries in the Classic API Users Guide has some method names without parentheses like rtems_clock_get instead of rtems_clock_get() which is used most of the time. Fix to improve consistency.
#4769	17 months ago	fixed	config	Sebastian Huber	Sebastian Huber	16 months ago
Summary	config: Add CONFIGURE_RECORD_INTERRUPTS_ENABLED
Description	/** * @brief This configuration option is a boolean feature define. * * In case * * * this configuration option is defined * * * and #CONFIGURE_RECORD_PER_PROCESSOR_ITEMS is properly defined, * * then the interrupt event recording is enabled. * * @par Default Configuration * If this configuration option is undefined, then the described feature is not * enabled. * * @par Notes * The interrupt event recording generates interrupt entry and exit events when * interrupt entries are dispatched. */ #define CONFIGURE_RECORD_INTERRUPTS_ENABLED
#4771	17 months ago	fixed	arch/aarch64	Chris Johns	Chris Johns	11 months ago
Summary	Versal UART issues
Description	The Versal polled mode it broken when building. The disable all interrupt call is not built in polled mode. The UART interrupt mode behaves differently with a small stand alone app than a full app with libbsd. A full application with libbsd seems to behave while the small app exhibit the "not tx interrupt on FIFO load" issue.
#4772	17 months ago	fixed	doc	Chris Johns	Kinsey Moore	9 months ago
Summary	Remove use of interval from rtems_task_wake_after() documentation
Description	The Classic API documentation for rtems_task_wake_after() refers to an interval while the call accepts ticks. The exact interval is not known because the period of time to next tick is not known when the call is made. Clarify the wording Recommend new applications use the POSIX based calls nanosleep() and clock_nanosleep() for applications that have intervals in a time base rather than system ticks. The ​Gemini, Verify the fidelity of RTEMS System Tick issue provides some back ground to this change.
#4775	16 months ago	fixed	arch/riscv	Hesham Almatary	Hesham Almatary <hesham.almatary@…>	14 months ago
Summary	RISC-V: Default RV64 BSPs to medany
Description	Currently generic RISC-V BSPs (riscv/riscv) that start with rv64 and not rv64_medany will start at 0x70000000. This adds high maintenance overhead and deviates from almost all other RISC-V-based OSes and baremetal program that start at 0x80000000. Further, testing now has to account for an extra parameter (medany or not) that doubles the number of BSPs need to be tested. It would be ideal to reduce maintenance efforts by defaulting all RV64 BSPs to medany and make them start at 0x80000000 to run on both QEMU and Spike. Those that need other cmodels or start addresses could be customised (e.g., rv64imac would default to medany but rv64imac_medlow won't).
#4778	16 months ago	fixed	tool/rsb	Chris Johns		10 months ago
Summary	Deployed RSB does not find release version config
Description	Cloned from #4777: The RSB does not find the VERSION file when released.
#4782	16 months ago	fixed	doc	Joel Sherrill		15 months ago
Summary	Correct Issues in Rate Monotonic Manager Background Section
Description	Fix these issues: 11.2.5.1 last bullet is inverted logically. Non-periodic tasks must NOT displace periodic ones and should run in background at a logically lower priority. 11.2.5.4 - emphasis box needed on the second paragraph (after lead-in that ends with a colon)
#4803	15 months ago	fixed	dev	Gedare Bloom	Gedare Bloom	9 months ago
Summary	Revert or Mark CAN API as experimental
Description	The CAN API done in GSoC 2022 is not stable. There are some key deficiencies that need to be fixed, and they may cause the API to change. The API needs to be marked as experimental or reverted before it appears in a release.
#4818	15 months ago	fixed	score	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	15 months ago
Summary	Incomplete Statement in “cpu->heir”
Description	Description: The assignment statement "cpu->heir = " is incomplete and missing a value and semicolon at the end. It is recommended to revise this statement to ensure the code is not broken by the missing semicolon. The attached excel lists in a systematic way the sources where this issue applies. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-058. File list (please check the attached excel for more information): cpukit\score\src\threadcreateidle.c
#4820	15 months ago	fixed	test	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	15 months ago
Summary	Hazardous cast
Description	Description: Converting a function pointer to a different type of pointer can lead to unpredictable and undefined behavior. It is important to note that attempting to convert a pointer to a function to a pointer of a different type of function is particularly dangerous and should be avoided. The attached excel lists in a systematic way the sources where this issue applies. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-079. File list (please check the attached excel for more information): bsps\shared\irq\irq-handler-iterate.c bsps\sparc\leon3\include\tm27.h
#4821	15 months ago	fixed	doc	Joel Sherrill	Joel Sherrill <joel@…>	15 months ago
Summary	Ubuntu Host Instructions Wrong
Description	The current instructions were reported to be as being broken in email. Filing a ticket to track. So here is the quick start command line which is missing the work ‘install’ and some of the packages listed don’t exist: ​https://docs.rtems.org/branches/master/user/hosts/posix.html#xubuntu Here is the correct command line: $ sudo apt install build-essential g++ gdb unzip pax bison flex texinfo python3-dev python-is-python3 libpython2-dev libncurses5-dev zlib1g-dev ninja-build pkg-config
#4822	15 months ago	fixed	score	Daniel Páscoa	Sebastian Huber	15 months ago
Summary	Function returning unchanged function input
Description	Description: These functions appear to be returning their input without making any changes. To improve their functionality, it is recommended to re-write them to return void. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-057. File: cpukit\score\src\smp.c Function: _SMP_Process_message Line: 284
#4828	15 months ago	fixed	bsps	Daniel Páscoa	Sebastian Huber	9 months ago
Summary	Macro defined but magic number used instead
Description	Description: Each register has a dedicated SHIFT macro, however, it is not utilized in subsequent macro definitions where the shift operation is performed. To improve consistency and maintainability, consider incorporating the defined shift macro in the subsequent macro definitions instead of using hardcoded shift values. Additional Notes: These findings are related to the fact that a bunch of macros like AHBSTAT_AHBS_HMASTER_SHIFT are defined and then are not used right in the next few lines. #define AHBSTAT_AHBS_HMASTER_SHIFT 3 #define AHBSTAT_AHBS_HMASTER_MASK 0x78U #define AHBSTAT_AHBS_HMASTER_GET( _reg ) \ ( ( ( _reg ) >> 3 ) & 0xfU ) #define AHBSTAT_AHBS_HMASTER( _val ) ( ( _val ) << 3 ) As stated, in every one of these files, “This file is part of the RTEMS quality process and was automatically generated". There are YAML files where these registers are defined and the Waf build system generates the header files automatically with the current configuration. This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-013. File list: bsps\include\grlib\ahbstat-regs.h bsps\include\grlib\ahbtrace-regs.h bsps\include\grlib\apbuart-regs.h bsps\include\grlib\dsu4-regs.h bsps\include\grlib\ftmctrl-regs.h bsps\include\grlib\gptimer-regs.h bsps\include\grlib\gr1553b-regs.h bsps\include\grlib\gr740thsens-regs.h bsps\include\grlib\grcan-regs.h bsps\include\grlib\grclkgate-regs.h bsps\include\grlib\grethgbit-regs.h bsps\include\grlib\grgpio-regs.h bsps\include\grlib\grgprbank-regs.h bsps\include\grlib\grgpreg-regs.h bsps\include\grlib\griommu-regs.h bsps\include\grlib\grpci2-regs.h bsps\include\grlib\grspw2-regs.h bsps\include\grlib\grspwrouter-regs.h bsps\include\grlib\irqamp-regs.h bsps\include\grlib\l4stat-regs.h bsps\include\grlib\memscrub-regs.h bsps\include\grlib\l2cache-regs.h bsps\include\grlib\mmctrl-regs.h bsps\include\grlib\spictrl-regs.h bsps\include\grlib\spwpnp-regs.h bsps\include\grlib\spwrmap-regs.h
#4832	15 months ago	wontfix	score	Daniel Páscoa		15 months ago
Summary	CLANG flagged error
Description	Description: SPAMR/CLANG point out to a possible deference to a null pointer. However, this issue has already been checked by the RTEMS team and flagged as false positive (see rtems-6-sparc-gr740-smp-3/doc/paf/spamr/spamr.pdf page 29). Yet, the justification sentence ends with "Why CLANG complains about a possible null pointer at this place in code and not everywhere or nowhere remains a mystery" which sounds awkward and lead us to think that a complete analysis as not been done on the error. Looking at it now, "_Chain_Initialize_one" is declared as static inline in a .h. That makes it public and so not all calls are controlled. Within the function, there is already an assert operation for one of the inputs. Now, one needs to add it also to the other one. Note that this assert is only active when compiler options RTEMS_DEBUG or RTEMS_STATIC_ANALYSIS are used which shall remove the CLANG error and not influence the flight executable. (Missions might do an early compilation with RTEMS_DEGUG to spot any problem on this). Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-062. File list: cpukit\include\rtems\score\chainimpl.h (line 525)
#4834	15 months ago	fixed	shell	chenjin_zhong	Chris Johns	15 months ago
Summary	close_editor problem of RTEMS4.13/5.1 (cloned)
Description	Cloned from #4564: Hi, I find when close editor in shell console.some errors will occur. I check and analyse the source code of medit.c. the code fragment is listed as follows. #if defined(rtems) case ctrl('w'): ed = ed->env->current; close_editor(ed); break; #else case ctrl('w'): close_editor(ed); ed = ed->env->current; break; #endif static void close_editor(struct editor *ed) { struct env *env = ed->env; if (ed->dirty) { display_message(ed, "Close %s without saving changes (y/n)? ", ed->filename); if (!ask()) { ed->refresh = 1; return; } } delete_editor(ed); ed = env->current; if (!ed) { ed = create_editor(env); new_file(ed, ""); } ed->refresh = 1; } static void delete_editor(struct editor *ed) { if (ed->next == ed) { ed->env->current = NULL; } else { ed->env->current = ed->prev; } ed->next->prev = ed->prev; ed->prev->next = ed->next; if (ed->start) free(ed->start); clear_undo(ed); free(ed); } as seen above, if the macro rtems is defined. the delete_editor function will free ed pointer. Therefore, after the next loop, the ed is an invalid pointer.I have checked the code of ​https://github.com/ringgaard/sanos/blob/master/src/utils/edit/edit.c. the code is as follows. case ctrl('w'): close_editor(ed); ed = ed->env->current; break;
#4835	15 months ago	fixed	shell	chenjin_zhong	Chris Johns	15 months ago
Summary	medit malloc problem of RTEMS (cloned)
Description	Cloned from #4565: I find malloc function is called by move_gap function in medit.c. The returned value does not check. At least 32KB of memory is allocated at each time, maybe more than. The returned value "start" should be check to avoid malloc failure. The move_gap function should return immediatelty when malloc failure. the code frament is listed as follows. static void move_gap(struct editor *ed, int pos, int minsize) { int gapsize = ed->rest - ed->gap; unsigned char *p = text_ptr(ed, pos); if (minsize < 0) minsize = 0; if (minsize <= gapsize) { if (p != ed->rest) { if (p < ed->gap) { memmove(p + gapsize, p, ed->gap - p); } else { memmove(ed->gap, ed->rest, p - ed->rest); } ed->gap = ed->start + pos; ed->rest = ed->gap + gapsize; } } else { int newsize; unsigned char *start; unsigned char *gap; unsigned char *rest; unsigned char *end; if (gapsize + MINEXTEND > minsize) minsize = gapsize + MINEXTEND; newsize = (ed->end - ed->start) - gapsize + minsize; start = (unsigned char *) malloc(newsize); TODO check for out of memory gap = start + pos; rest = gap + minsize; end = start + newsize; if (p < ed->gap) { memcpy(start, ed->start, pos); memcpy(rest, p, ed->gap - p); memcpy(end - (ed->end - ed->rest), ed->rest, ed->end - ed->rest); } else { memcpy(start, ed->start, ed->gap - ed->start); memcpy(start + (ed->gap - ed->start), ed->rest, p - ed->rest); memcpy(rest, p, ed->end - p); } free(ed->start); ed->start = start; ed->gap = gap; ed->rest = rest; ed->end = end; } #ifdef DEBUG memset(ed->gap, 0, ed->rest - ed->gap); #endif }
#4838	15 months ago	duplicate	arch/aarch64	Aaron N	Chris Johns	15 months ago
Summary	Fix for UART interrupt on aarch64 Versal
Description	Fixed UART interrupts for Versal. Tested on VCK190.
#4842	15 months ago	fixed	arch/sparc	Daniel Páscoa	Sebastian Huber	9 months ago
Summary	SPARC/grlib - Registers definitions wrongly defined when there are reserved bits
Description	With respect to RTEMS 6 (ESA-promoted RTEMS SMP Qualification Data Pack release 3), the following issues are observed: File Lines Discrepancy bsps\include\grlib\ftmctrl-regs.h 84; 86; 88; 91; 96;98; 101; 106; 109; 115; 120; 123 According to table 365 of page 294 of the GR740 User Manual (v.2.50), the register addresses are incorrectly defined. bsps\include\grlib\grgpio-regs.h 225 According to table 406 of page 319 of the GR740 User Manual (v.2,50), the register address is incorrect; bsps\include\grlib\dsu4-regs.h 435 According to table 538 of page 403 of the GR740 User Manual (v.2.5), the register address is incorrect according to the table but it is correct according to the table description. bsps\include\grlib\gr1553b-regs.h 609 According to table 327 of page 259 of the GR740 User Manual (v.2.50), the register address is incorrect according to the table but it is correct according to the table description. bsps\include\grlib\spwpnp-regs.h 151 According to table 213 of page 200 of the GR740 User Manual (v.2.5), the register address is incorrect according to the table but it is correct according to the table description. bsps\include\grlib\spwrmap-regs.h 575; 591; 715; 787; 809 According to tables 191, 192, 199, 201, 202, respectively (to each identified line) starting on page 192 of the GR740 User Manual (v.2.5), the register addresses are incorrect according to the table but are correct according to the table description. It seems that the generation of these registers definitions leads to incorrect bit definitions whenever there are reserved bits specified in the Gaisler User Manual. Can you please double-check these cases? Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-CODE-VER-014.
#4843	15 months ago	fixed	bsps	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	14 months ago
Summary	Output value not assigned
Description	The Software Document Design (starting on page 181) specifies that there should be two outputs, but the RTEMS_UNSATISFIED is not returned by the function. The only failure detected may be due to the assert function, but the rtems_status_code isn't the returned value. Ensure all output values specified in the SDD are returned. Additional Notes: From code analysis and from objdumping the librtemsbsp.a library it can be observed that the asserts() currently do nothing: they are translated to #define _Assert( _e ) ( ( void ) 0 ) unless RTEMS_DEBUG or RTEMS_STATIC_ANALYSIS are defined, which in a release environment target are not defined. Since in the documentation it says that it can return RTEMS_UNSATISFIED or RTEMS_SUCCESSFUL either the documentation is wrong, or the code is not correctly implemented. This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-CODE-VER-032. File list: bsps\sparc\leon3\start\eirq.c bsp_interrupt_is_pending (line 106) bsp_interrupt_raise (line 127) bsp_interrupt_clear (line 185) bsp_interrupt_vector_enable (line 270) bsp_interrupt_vector_disable (line 299)
#4845	15 months ago	fixed	arch/sparc	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	14 months ago
Summary	Global variable declared/defined in the wrong file
Description	This source file contains both declaration and definition of the variable. It is recommended to declare global variables in a header file and define them in separate source files. Please consider this best practice and move the declarations to a header file and corresponding definitions to respective source files. Additional Notes: LEON3_Boot_Cpu is a global variable and only used by start.S file. And as such cannot be declared as static. Because is used in an assembly file there was no need to declare it as external. But the global variable is in fact declared and defined in same file. This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-043. File list: bsps\sparc\leon3\start\bspstart.c (line 54)
#4847	15 months ago	fixed	score	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	14 months ago
Summary	Goto statements
Description	The use of “goto” statements is strongly discouraged as it results in convoluted program logic, making it challenging to analyse and verify program correctness, particularly in loops. Instead, the “break” and “continue” statements provide alternatives to “goto” and should be utilized to simplify the code. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-074. File list: cpukit/score/src/threaddispatch.c (line 290)
#4850	14 months ago	fixed	arch/sparc	Daniel Páscoa	Sebastian Huber	9 months ago
Summary	LEON3 - Variable not initialized in any file / missing #ifdefs
Description	Within bsps\sparc\leon3\include\bsp\leon3.h (lines 207, 309) The *LEON3_IrqCtrl_Adev and *LEON3_Timer_Adev are declared but not initialized in any file of the QDP.: Please consider having these declared within the “else” blocks of the “#if defined(LEON3_IRQAMP_BASE)” where *LEON3_IrqCtrl_Regs and *LEON3_Timer_Regs are declared. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). Original ISVV reference for this issue is RTEMS-SMP-CODE-VER-041.
#4855	14 months ago	wontfix	unspecified	Daniel Páscoa		13 months ago
Summary	Input validity unchecked
Description	In a generic way, function inputs shall be checked for validity before being used. When successive calls are done to different functions, one might argue that inputs are to be checked only once and then considered valid in other calls. Yet this has some pitfalls like: function inputs are not checked in a consistent way throughout all functions: Each one has to constantly double-check in comments and/or callers if inputs are instead checked elsewhere in the call tree. Even if for the moment no call tree leads to invalid input issues, future maintenance operations may add new calls without validity checks. One can assume some exceptions can occur (e.g., when there is a considerable performance impact), but as long such cases are justified within the function’s comments one can easily sort out those exceptions. Bellow, there is a list of cases where input validity could be added: bsps\include\bsp\irq-generic.h: bsp_interrupt_entry_load_first function is called by bsp_interrupt_handler_dispatch_unchecked where the argument was already checked. bsps\include\bsp\irq-generic.h: bsp_interrupt_handler_index function is called by bsp_interrupt_entry_load_first, bsp_interrupt_entry_remove, bsp_interrupt_spurious, bsp_interrupt_entry_find, bsp_interrupt_entry_install, rtems_interrupt_handler_iterate functions. It was possible to trace in all of them a validation of the argument used in bsp_interrupt_handler_index before being called. bsps\shared\irq\irq-default-handler.c: bsp_interrupt_handler_default is only called by bsp_interrupt_spurious and bsp_interrupt_handler_dispatch_unchecked functions and It was possible to trace in all ot them a validation of the argument used in bsp_interrupt_handler_index before being called. bsps\shared\irq\irq-entry-remove.c: the *entry argument of bsp_interrupt_entry_do_remove is actually tested and validated inside of that function. That function is also the only place where bsp_interrupt_entry_remove is called. The arguments vector, *entry and previous_next are all validated by a call to the bsp_interrupt_entry_find and subsequent code that is run prior to call bsp_interrupt_entry_remove function. bsps\shared\irq\irq-generic.c: the bsp_interrupt_spurious function argument vector is tested before being called in bsp_interrupt_handler_dispatch_unchecked function. bsps\shared\irq\irq-generic.c: bsp_interrupt_entry_install_first function is called once by bsp_interrupt_entry_install function. The validation of the argument vector can be traceback to rtems_interrupt_entry_install function. The argument *entry is also passed by bsp_interrupt_entry_install function and explained in the previous paragraph. bsps\shared\dev\clock\clockimpl.h: in both functions (void Clock_isr and rtems_isr Clock_isr) declaration/definition (they are the same function, just declared differently depending of #if def's), both arguments are not used. Due to the evaluation of the internal #if, this functions are reduced to Clock_driver_ticks += 1; _Timecounter_Tick(); bsps\sparc\leon3\start\bspsmp.c: the bsp_start_on_secondary_processor function calls _SMP_Start_multitasking_on_secondary_processor with the same argument as it was called. This argument in that sub-function is used to get a value cpu_index_self, which itself is validated and checked if it is a proper value or not. Indirect validation. bsps\sparc\leon3\start\bspsmp.c: the _CPU_SMP_Send_interrupt function is called in 3 different places: in one place (_SMP_Request_shutdown function in smp.c file) it is inside a for loop that is limited by _SMP_Processor_configured_maximum. This is OK. For the other 2 places ( from _SMP_Send_message function in smp.c file and _Thread_Dispatch_request function in threaddispatch.h file it is not clear that the argument of _CPU_SMP_Send_interrupt cannot be limited (when it should be limited at least to _SMP_Processor_configured_maximum). bsps\sparc\leon3\start\eirq.c: bsp_interrupt_get_attributes function actually tests vector argument. But also there is only one call to this function which is from rtems_interrupt_get_attributes function in irq-enable-disable.c file where both arguments are tested prior call to bsp_interrupt_get_attributes function. bsps\sparc\shared\start\bspgetworkarea.c: _SPARC_Memory_initialize function receives its *end_of_usable_ram argument from the main function in start.S file. No validation is performed there. The argument is used to initialize a struct, which later on may be used for dynamically allocated memory. The referred pointer argument is never dereferenced, it is used as the end limit of the allocated memory position. It may make sense to test in this initialisation function if it is value is above &_Memory_Areas[ 0 ]. cpukit\libc\string\memset.c: memset function is part of space qualification profile context, and the void *<[dst]> pointer is not tested against NULL. cpukit\libc\string\memcpy.c: memcpy function is part of space qualification profile context, and the void *dst and void *src pointers is not tested against NULL. cpukit\posix\src\clocknanosleep.c: in function clock_nanosleep the *rmtp argument can be NULL if the user does not want to get the remain time, so there is no need to validated it. For the *rqtp (request) argument this should be validated if different than NULL and the tv_nsec field must be validated if between 0 and 999999999. cpukit\rtems\src\clocktodtoseconds.c: the _TOD_To_seconds function is called by _Timer_Fire_when function from timercreate.c file, by rtems_task_wake_when function from taskwakewhen.c file and by rtems_clock_set function from clockset.c file. In all three cases the argument with which the _TOD_To_seconds function is called is validated with a call to _TOD_Validate function prior to the call. cpukit\rtems\src\eventsend.c: the _Event_Seize function is called by two similar functions the rtems_event_receive from eventreceive.c file and rtems_event_system_receive from systemeventreceive.c file. The *event_out argument is validated prior the call; all other arguments are declared and defined in the caller functions prior to call the Event_Seize function. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-VAL-029.
#4857	14 months ago	duplicate	tool/rsb	Joel Sherrill	Chris Johns	10 months ago
Summary	RSB Recipe Missing %hash Does Not Fail
Description	If a recipe is missing a %hash command, it is reported as a warning. It would be preferable to have an error. warning: gdb-readline8.2.patch: no hash found From #4760 and copied as part of merging these tickets: The time has come to raise a fatal error if a hash is missing rather than a warning. This will stop missing checksums being present when releasing. It also makes it simpler to test releases for no missing checksums.
#4858	14 months ago	fixed	tool/gdb	Joel Sherrill	Joel Sherrill	14 months ago
Summary	GDB fails to build on any host with readline 8.2 or newer
Description	This showed up with gdb 11.2 building on Cygwin. The fix is already in the gdb git repository on the master but not gdb 12.x. This ticket tracks adding that to the RSB.
#4859	14 months ago	fixed	tool/gdb	Joel Sherrill	Joel Sherrill <joel@…>	14 months ago
Summary	Update gdb to 12.1
Description	While investigating the build issue tracked as #4858, I tried bumping the gdb to 12.1 hoping it included the fix. But it did not. This bumps the default gdb to 12.1 with the same patch.
#4860	14 months ago	fixed	tool	Joel Sherrill		14 months ago
Summary	Waf Failure on CentOS 8 Stream
Description	Jeff Mayes reported (​https://lists.rtems.org/pipermail/devel/2023-February/074424.html) that builds on the CentOS 8 Stream test host are failing with this: Using rtems for RTEMS source Testing: sparc erc32-sis BSP to Build: erc32 Traceback (most recent call last): File "./waf", line 167, in <module> from waflib import Scripting File "/home/tester/rtems-cron-6/rtems/.waf-2.0.20-36f5354d605298f6a89c09e0c7ef6c1d/waflib/Scripting.py", line 7, in <module> from waflib import Utils,Configure,Logs,Options,ConfigSet,Context,Errors,Build,Node File "/home/tester/rtems-cron-6/rtems/.waf-2.0.20-36f5354d605298f6a89c09e0c7ef6c1d/waflib/Configure.py", line 6, in <module> from waflib import ConfigSet,Utils,Options,Logs,Context,Build,Errors File "/home/tester/rtems-cron-6/rtems/.waf-2.0.20-36f5354d605298f6a89c09e0c7ef6c1d/waflib/Options.py", line 6, in <module> from waflib import Logs,Utils,Context,Errors File "/home/tester/rtems-cron-6/rtems/.waf-2.0.20-36f5354d605298f6a89c09e0c7ef6c1d/waflib/Context.py", line 183 raise Errors.WafError('Execution failure: %s'%str(e),ex=e),None,sys.exc_info()[2],None,sys.exc_info()[2] ^ SyntaxError: invalid syntax FAILED: Unable to waf distclean rtems In follow up emails, Chris Johns suggested updating waf which Jeff reported fixed the issue. The action for this ticket is to update waf in the following repositories: rtems rtems-examples rtems-docs rtems-libbsd rtems-lwip rtems-net-legacy (OK -- already on latest waf) rtems-tools
#4862	14 months ago	fixed	unspecified	Daniel Páscoa	Sebastian Huber	4 weeks ago
Summary	score/ & bsps/: Unused input parameters
Description	The following functions list input parameters on their signature that seem not to be used: File Function Line Resolution bsps\shared\cache\cacheimpl.h smp_cache_inst_inv_all 293 Unused due to API constraints bsps\shared\dev\clock\clockimpl.h void Clock_isr 130 Fixed: parameter can be used bsps\shared\dev\clock\clockimpl.h rtems_isr Clock_isr 134 Fixed: parameter can be used bsps\shared\irq\irq-generic.c bsp_interrupt_allocate_handler_index 62 Fixed: parameter removed bsps\sparc\leon3\start\bspclean.c bsp_fatal_extension 70 API backward compatibility cpukit\score\src\threadchangepriority.c static void _Thread_Priority_action_change( ); 92 Unused due to API constraints cpukit\score\src\threadq.c _Thread_queue_MP_callout_do_nothing 157 Unused due to API constraints cpukit\score\src\threadqenqueue.c _Thread_queue_Deadlock_fatal 382 Unused due to API constraints cpukit\score\src\threadqops.c _Thread_queue_FIFO_do_initialize 137 Unused due to API constraints cpukit\score\src\threadqtimeout.c _Thread_queue_Add_timeout_ticks 28 Unused due to API constraints cpukit\score\src\threadrestart.c _Thread_Join_flush_filter 84 Unused due to API constraints cpukit\score\src\userextiterate.c _User_extensions_Thread_begin_visitor 102 Unused due to API constraints Could one please double-check if in each case the unused parameters can be safely removed from the function’s signature? PS: The first functions are also implemented within .h files and perhaps could be moved to .c files. Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-CODE-VER-028.
#4863	14 months ago	fixed	score	Daniel Páscoa	Sebastian Huber <sebastian.huber@…>	12 months ago
Summary	Operations evaluation order.
Description	Within cpukit\include\rtems\score\objectimpl.h line 545 one has: if ( !the_api the_api > OBJECTS_APIS_LAST ) This shall be equivalent to “if ( (!the_api) (the_api > OBJECTS_APIS_LAST ))”, but at first glance, this is not always obvious to everyone and one might need to read further documentation to be fully sure of it. We would suggest the parenthesis be added so that the code logic is straightforward. In this case, one would not call it an ‘excess’ as stated within the RTEMS SW Eng. Manual (“Avoid excess parentheses. Learn the operator precedence. rules.”). Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-CODE-VER-076.
#4867	14 months ago	fixed	network/lwip	Kinsey Moore	Kinsey Moore	10 months ago
Summary	Clean up rtems-lwip uLan directory
Description	The uLan directory in rtems-lwip was originally sourced from another project as documented in its ORIGIN.uLan file. This directory contains both some BSP-specific drivers as well as some of the core RTEMS/lwIP integration code in the form of sys_arch.c/h. Permission has been given by the authors of this upstream repository to relicense the code under the terms of the rtemslwip directory so as to remove references to "uLan" since this term is relatively unrelated to either project and just so happened to be the location of the source. ​https://lists.rtems.org/pipermail/devel/2023-March/074525.html Core integration components should be rehomed under rtemslwip/common and BSP-specific code should be rehomed under rtemslwip in a BSP-specific directory or in a new root directory with appropriate LICENSE and ORIGIN files depending on its provenance.
#4869	14 months ago	fixed	dev/spi	Aaron N	chrisj@…	10 months ago
Summary	Add QSPI Flash Device API
Description	Add an API for qspi flash to create device node.
#4872	14 months ago	fixed	tool/gdb	Chris Johns	Chris Johns	14 months ago
Summary	GDB 13.1 does not build on FreeBSD 13.1
Description	If the binutils package is installed GDB does not build. See ​https://sourceware.org/bugzilla/show_bug.cgi?id=30214
#4874	14 months ago	wontfix	rtems	Daniel Páscoa		14 months ago
Summary	Undefined behaviour
Description	If one searches within SRSs or ICDs for 'undefined behaviour' one will find specifications/functions on which descriptions mention conditions that will lead to 'undefined behaviour', e.g.: ... spec:/rtems/clock/if/get-boot-time ... boot_time is the pointer to a struct timespec object. The time elapsed since the Unix epoch at some time point during system initialization call will be stored in this object. Calling the directive with a pointer equal to NULL is undefined behaviour.". One created specific tests to further access the eventual impacts, for instance in the clock-related functions below: spec:/rtems/clock/if/get-boot-time spec:/rtems/clock/if/get-boot-time-bintime spec:/rtems/clock/if/get-boot-time-timeval spec:/rtems/clock/if/get-monotonic spec:/rtems/clock/if/get-monotonic-bintime spec:/rtems/clock/if/get-monotonic-coarse spec:/rtems/clock/if/get-monotonic-coarse-bintime spec:/rtems/clock/if/get-monotonic-coarse-timeval spec:/rtems/clock/if/get-monotonic-timeval spec:/rtems/clock/if/get-realtime spec:/rtems/clock/if/get-realtime-bintime spec:/rtems/clock/if/get-realtime-coarse spec:/rtems/clock/if/get-realtime-coarse-bintime spec:/rtems/clock/if/get-realtime-coarse-timeval spec:/rtems/clock/if/get-realtime-timeval The outcome was: There is no internal check for a NULL pointer in these functions, so when called with a null pointer argument there is unwanted behaviour for both the GR712rc and GR740; an attempt to access memory location 0. For the GR712rc board this memory position is not mapped and causes a trap to be raised (this trap handler leads to unpredictable results, the workflow of the program is not recoverable). For the GR740 board the actual memory position is either read or written (the previous value being overridden with unpredictable results). So, even if somehow this shall also be a user responsibility, shouldn’t one adopt a safe approach to this and add input checking for these functions? Additional Notes: This ticket was raised as an outcome of the Independent SW Verification and Validation (ISVV) for ESA-promoted RTEMS SMP Qualification Data Packs (​https://rtems-qual.io.esa.int). The original ISVV reference for this issue is RTEMS-SMP-REQ-VER-008 and RTEMS-SMP-VAL-005.
#4878	13 months ago	fixed	test	Lucian-Raul Silistru	Sebastian Huber	13 months ago
Summary	sptests/spsysinit01 sem_open call only has 3 parameters with O_CREAT flag (cloned)
Description	Cloned from #4877: When calling sem_open with the O_CREAT flag the sem_open function requires 4 parameters. It will: value = va_arg(args, unsigned int); so the "value" argument will be randomly chosen. sem_open can fail if the random value is not a valid semaphore value, failing the test.
#4881	13 months ago	fixed	tool/binutils	Chris Johns	Chris Johns	13 months ago
Summary	powerpc gas and ld warnings
Description	The upgraded binutils generates new warnings when linking executables: /opt/work/6/lib/gcc/powerpc-rtems6/12.2.1/../../../../powerpc-rtems6/bin/ld: warning: /opt/work/rtems/6-bsp/powerpc-rtems6/mvme2307/lib/vectors_entry.o: missing .note.GNU-stack section implies executable stack /opt/work/6/lib/gcc/powerpc-rtems6/12.2.1/../../../../powerpc-rtems6/bin/ld: NOTE: This behaviour is deprecated and will be removed in a future version of the linker /opt/work/6/lib/gcc/powerpc-rtems6/12.2.1/../../../../powerpc-rtems6/bin/ld: warning: /opt/work/chris/rtems/libbsd/rtems-libbsd.git/build/powerpc-rtems6-mvme2307-default/termios05.exe has a LOAD segment with RWX permissions /opt/work/6/lib/gcc/powerpc-rtems6/12.2.1/../../../../powerpc-rtems6/bin/ld: warning: /opt/work/chris/rtems/libbsd/rtems-libbsd.git/build/powerpc-rtems6-mvme2307-default/termios05.exe has a LOAD segment with RWX permissions
#4882	13 months ago	fixed	network/legacy	Chris Johns		10 months ago
Summary	mvme2703 legacy network build failure
Description	Building powerpc/mvme2703 BSP with the legacy networking fails with: ../../netinet/in.c: In function 'in_canforward': ../../netinet/in.c:117:13: warning: implicit declaration of function 'IN_CLASSA'; did you mean 'IN_BADCLASS'? [-Wimplicit-function-declaration] 117 | if (IN_CLASSA(i)) { | ^~~~~~~~~ | IN_BADCLASS ../../netinet/in.c:118:27: error: 'IN_CLASSA_NET' undeclared (first use in this function) 118 | net = i & IN_CLASSA_NET; | ^~~~~~~~~~~~~ ../../netinet/in.c:118:27: note: each undeclared identifier is reported only once for each function it appears in ../../netinet/in.c:119:41: error: 'IN_LOOPBACKNET' undeclared (first use in this function); did you mean 'IN_LOOPBACK'? 119 | if (net == 0 || net == (IN_LOOPBACKNET << IN_CLASSA_NSHIFT)) | ^~~~~~~~~~~~~~ | IN_LOOPBACK ../../netinet/in.c:119:59: error: 'IN_CLASSA_NSHIFT' undeclared (first use in this function) 119 | if (net == 0 || net == (IN_LOOPBACKNET << IN_CLASSA_NSHIFT)) | ^~~~~~~~~~~~~~~~ ../../netinet/in.c: In function 'in_ifinit': ../../netinet/in.c:495:34: error: 'IN_CLASSA_NET' undeclared (first use in this function) 495 | ia->ia_netmask = IN_CLASSA_NET; | ^~~~~~~~~~~~~ ../../netinet/in.c:496:18: warning: implicit declaration of function 'IN_CLASSB'; did you mean 'IN_BADCLASS'? [-Wimplicit-function-declaration] 496 | else if (IN_CLASSB(i)) | ^~~~~~~~~ | IN_BADCLASS ../../netinet/in.c:497:34: error: 'IN_CLASSB_NET' undeclared (first use in this function) 497 | ia->ia_netmask = IN_CLASSB_NET; | ^~~~~~~~~~~~~ ../../netinet/in.c:499:34: error: 'IN_CLASSC_NET' undeclared (first use in this function) 499 | ia->ia_netmask = IN_CLASSC_NET; | ^~~~~~~~~~~~~
#4884	13 months ago	invalid	dev/i2c	Chris Johns	Chris Johns	13 months ago
Summary	I2C default ioctl handler does not handle set clock
Description	The base struct for an I2C driver can set set_clock handler so a user can change the I2C clock speed. The default ioctl handler does not check for the I2C_BUS_SET_CLOCK command and if the set_clock handler is set calling it.
#4887	13 months ago	fixed	network/legacy	Chris Johns	Chris Johns	10 months ago
Summary	RTEMS Net Legacy does not build as a package in the RSB
Description	The waf support does not handle staged builds or the options --rtems-tools and --rtems at the same time. The build fails when reading the rtemsbsp static library because it will not be in the PREFIX. Also the build script scans for .c and .h files. This is fragile. If someone copies in a C file for some reason it is picked up. The sources and headers need to be explicit in build script.
#4889	13 months ago	fixed	network/lwip	Kinsey Moore	Kinsey Moore <kinsey.moore@…>	13 months ago
Summary	Remove file scanning from rtems-lwip
Description	rtems-lwip currently scans a subset of directories for files. Since it already supports includes in its definition format, it should instead explicitly list all files to be compiled and support for directory scanning should be removed.
#4892	13 months ago	fixed	tool/rsb	Chris Johns	Chris Johns	12 months ago
Summary	MacOS tools build failures on Ventura and M series processors
Description	Building the tools with a recent MacOS such as Ventura failed because Apple has removed support for embedded python in programs. This means GDB does not build. The solution is to teach the RSB GDB support to use a build of Python from ​https://www.python.org/. The RSB failed to build gcc on the M series processors on Ventura using Xcode 14.2. The build gcc crosscompiler (xgcc) crashes with sig faults. GCC does not bootstrap on ARM systems: *** Configuration aarch64-apple-darwin22.3.0 not supported Look at supporting Homebrew built gcc.
#4896	12 months ago	fixed	build	Chris Johns	Chris Johns	12 months ago
Summary	Makefile.inc exported BSP include path not a system include path
Description	The BSP include path exported in the Makefile.inc set of files uses -I and that is not a system include. The RTEMS 5 and earlier autoconf build system provided a spec file using the -B option and that either added a system path or the spec file included a system path. Either way the BSP include path was added as a system include path and so the compiler placed it at the end of the include list. The CPU_DEFINES macro is part of CPPFLAGS exported by Makefile.inc and we cannot control how applications uses that variable so we need to make sure the BSP include path is exported as a system include path. The issue was found with EPICS. EPICS installs a header file called link.h under its include path. RTEMS now provides a header file called link.h. The CPPFLAGS are before the EPICS include path so our link.h was seen before the EPICS one and that broken the build.
#4899	12 months ago	fixed	bsps	Chris Johns		10 months ago
Summary	BSP powerpc/qoriq_e6500_64 and powerpc/qoriq_e6500_32 fail to build
Description	[ 87/1471] Compiling bsps/powerpc/shared/vme/vmeTsi148.c /tmp//cccfu9l8.s: Assembler messages: /tmp//cccfu9l8.s:353: Error: unrecognized opcode: `eieio' /tmp//cccfu9l8.s:386: Error: unrecognized opcode: `eieio' /tmp//cccfu9l8.s:419: Error: unrecognized opcode: `eieio' /tmp//cccfu9l8.s:444: Error: unrecognized opcode: `eieio' /tmp//cccfu9l8.s:477: Error: unrecognized opcode: `eieio' /tmp//cccfu9l8.s:503: Error: unrecognized opcode: `eieio' ...
#4900	12 months ago	fixed	score	Chris Johns	Sebastian Huber <sebastian.huber@…>	6 months ago
Summary	PowerPC score warnings in score headers
Description	Building a few of the PowerPC BSPs with RTEMS_DEBUG = True generates these warnings: [ 319/1483] Compiling cpukit/score/cpu/powerpc/cpu.c In file included from ../../../cpukit/score/cpu/powerpc/include/rtems/score/cpu.h:59, from ../../../cpukit/score/cpu/powerpc/include/rtems/score/cpuimpl.h:42, from ../../../cpukit/score/cpu/powerpc/cpu.c:43: ../../../cpukit/score/cpu/powerpc/cpu.c: In function '_CPU_Fatal_halt': ../../../cpukit/include/rtems/score/basedefs.h:436:33: warning: implicit declaration of function '_Assert' [-Wimplicit-function-declaration] 436 | #define _Assert_Unreachable() _Assert( 0 ) | ^~~~~~~ ../../../cpukit/include/rtems/score/basedefs.h:865:7: note: in expansion of macro '_Assert_Unreachable' 865 | _Assert_Unreachable(); \ | ^~~~~~~~~~~~~~~~~~~ ../../../cpukit/score/cpu/powerpc/cpu.c:361:3: note: in expansion of macro 'RTEMS_UNREACHABLE' 361 | RTEMS_UNREACHABLE(); | ^~~~~~~~~~~~~~~~~ ../../../cpukit/include/rtems/score/basedefs.h:436:33: warning: nested extern declaration of '_Assert' [-Wnested-externs] 436 | #define _Assert_Unreachable() _Assert( 0 ) | ^~~~~~~ ../../../cpukit/include/rtems/score/basedefs.h:865:7: note: in expansion of macro '_Assert_Unreachable' 865 | _Assert_Unreachable(); \ | ^~~~~~~~~~~~~~~~~~~ ../../../cpukit/score/cpu/powerpc/cpu.c:361:3: note: in expansion of macro 'RTEMS_UNREACHABLE' 361 | RTEMS_UNREACHABLE(); | ^~~~~~~~~~~~~~~~~ The score headers basedefs.h header depends on assert.h and assert.h includes basedefs.h at it's top. I think this ticket should be set to blocker for the 6.1 release. We should resolve this and the warnings before we make a release.
#4901	12 months ago	fixed	doc	Joel Sherrill	Sebastian Huber	11 months ago
Summary	rtems_task_construct Documentation Missing Some Details
Description	The rtems_task_construct directive description has a couple of deficiencies: does not mention or link to RTEMS_TASK_STORAGE_SIZE does not provide rtems_task_config and discuss the fields It would be nice to discuss how each of the parameters to rtems_task_create is set/computed when using rtems_task_config. I did not look but rtems_message_queue_construct may have similar issues.
#4902	12 months ago	fixed	lib	Sebastian Huber	Sebastian Huber	12 months ago
Summary	Update zlib to version 1.2.13
Description	The last zlib update was more than ten years ago. Update it to the latest version. This includes security fixes.
#4905	11 months ago	fixed	score	Joel Sherrill	Joel Sherrill	11 months ago
Summary	SuperCore? _TOD_Adjust No Longer Referenced and Should be Removed
Description	Apparently with the rework of time handling, this method has become unused and should be removed.
#4906	11 months ago	invalid	network/libbsd	Ning Yang		10 months ago
Summary	libbsd not working on nucleo-h743ZI
Description	libbsd not working on nucleo-h743ZI. I run the libbsd testsuit selectpollkqueue01,get a fatal. The same fatal happened in my project. * FATAL * fatal source: 0 (INTERNAL_ERROR_CORE) fatal code: 2 (INTERNAL_ERROR_TOO_LITTLE_WORKSPACE) RTEMS version: 6.0.0.020163e5b7bb0b4363b57f37729b99becae9c629 RTEMS tools: 12.2.1 20230224 (RTEMS 6, RSB f0e34eab8bf33b833a7d9d0b2bddd3b89f6d83cb, Newlib 17ac400) executing i used a sh to build libbsd cd /home/yangn0/devel/rtems-arm/rtems-libbsd ./waf distclean ./waf configure --prefix=$HOME/devel/rtems-arm/tools/rtems/6 \ --rtems-bsps=arm/nucleo-h743zi \ --buildset=buildset/default.ini ./waf ./waf install
#4909	11 months ago	fixed	score	Joel Sherrill	Joel Sherrill	6 months ago
Summary	Multiple Protected Heap Functions are Ununused
Description	The following functions are unused and should be removed: _Protected_heap_Get_block_size _Protected_heap_Iterate _Protected_heap_Resize_block
#4912	11 months ago	fixed	lib	Joel Sherrill	Joel Sherrill	11 months ago
Summary	Collect Compression Library Source in Single Place
Description	Currently zlib is in cpukit/ while xz is in cpukit/libmisc. The purpose of this ticket is to create cpukit/compression and move both to it. This ticket does not add, remove, or upgrade the existing source. It is intended to just document improving the organization.
#4913	11 months ago	fixed	network/libbsd	Christian Mauderer	Christian Mauderer	11 months ago
Summary	libbsd/ipsec-tools: Problems with more than 64 open files
Description	IPSec with racoon doesn't work if there are more than 64 open file descriptors in the system.
#4915	11 months ago	wontfix	tool	Andrew Butterfield	Utkarsh Verma <utkarsh@…>	5 months ago
Summary	Building sphinx 7.0.1 fails
Description	I am trying to install sphinx on an Apple Silicon OSX machine. I've already done this successfully in the past on a similar machine. This time, after following steps in the rtems-docs README.txt, I do ./waf , it does some stuff and then: [ 7/10] Compiling common/_static/favicon.ico [ 8/10] Compiling common/_static/my-styles.css [ 9/10] Compiling common/_static/logo.png [10/10] Compiling common/_static/style.css Theme error: An error happened in rendering the page bld/index. Reason: UndefinedError("'style' is undefined") I'm using pip 22.2.1, Python 3.10.6 and pip3 installs sphinx 7.0.1 (sphinx) ab-rtems-docs% pip3 install sphinx Requirement already satisfied: sphinx in ./sphinx/lib/python3.10/site-packages (7.0.1)
#4924	10 months ago	fixed	lib	Joel Sherrill	Joel Sherrill	8 months ago
Summary	Add Regulator Helper to RTEMS
Description	This is to add the Regulator Helper. This will include the regulator itself, a test with full coverage, an example for rtems-examples, and a chapter in the Classic API Guide. The regulator is designed to sit logically between two entities -- a source and a destination, where it limits the traffic sent to the destination to prevent it from being flooded with messages from the source. This can be used to accommodate bursty input from a source and meter it out to a destination.
#4934	9 months ago	fixed	arch/arm	kgardas	kgardas	9 months ago
Summary	Update STM32 H7 HAL/LL drivers code
Description	Hello, patch series for updating STM32 H7 HAL/LL drivers code is ready for review already here: ​https://github.com/karelfv/rtems/tree/stm32h7-hal-update-2023-07. The code was tested on both M7 and M4 cores on stm32h747i-disco board and shows that update does not cause any new failures/regressions. The patch series was already submitted for review on devel@ here: ​https://lists.rtems.org/pipermail/devel/2023-July/075880.html
#4935	9 months ago	fixed	tool/rsb	Chris Johns	joel	5 months ago
Summary	7/rtems-mips get source fails
Description	Getting sources for 7/rtems-mips fails with: error: 7/rtems-mips.bset:6: cannot find file: tools/rtems-mipstx39-gdb-head Release candidates for 6 get the 7 sources because the 6-rc? are made from master.
#4940	8 months ago	fixed	arch/arm	Sebastian Huber	Sebastian Huber	8 months ago
Summary	Fix cache support for ARM926EJ-S processor
Description	Fix the CP15 cache support so that it works on ARMv5T processors such as the ARM926EJ-S. This processor has different cache maintenance operations compared to ARMv6 processors such as the ARM1176JZF-S. The current support leads to undefined instruction exceptions.
#4942	8 months ago	fixed	lib	Chris Johns	Joel Sherrill	8 months ago
Summary	Regulator warnings
Description	The recently added regulator has warnings on aarch64/xilinx_zynqmp_lp64_qemu: ../../../cpukit/libmisc/regulator/regulator.c: In function 'rtems_regulator_get_statistics': ../../../cpukit/libmisc/regulator/regulator.c:656:3: warning: implicit declaration of function 'memset' [-Wimplicit-function-declaration] 656 | memset(statistics, 0, sizeof(rtems_regulator_statistics)); | ^~~~~~ ../../../cpukit/libmisc/regulator/regulator.c:40:1: note: include '<string.h>' or provide a declaration of 'memset' 39 | #include <rtems/regulatorimpl.h> +++ |+#include <string.h> 40 | ../../../cpukit/libmisc/regulator/regulator.c:656:3: warning: incompatible implicit declaration of built-in function 'memset' [-Wbuiltin-declaration-mismatch] 656 | memset(statistics, 0, sizeof(rtems_regulator_statistics)); | ^~~~~~ ../../../cpukit/libmisc/regulator/regulator.c:656:3: note: include '<string.h>' or provide a declaration of 'memset'
#4943	8 months ago	fixed	lib/dl	Chris Johns	Chris Johns	8 months ago
Summary	ARM unwind register is not being hooked
Description	The ARM unwind support is not working and dl05 is failing on aarch64. The support uses __gnu_Unwind_Find_exidx and it is not being called. I have not tested arm but the support is shared between arm and aarch64.
#4944	8 months ago	fixed	lib/dl	Chris Johns	Chris Johns <chrisj@…>	8 months ago
Summary	libdl test dl09 fails on arm and aarch64
Description	The test crashes on arm and fails on aarch64 with: -------------------------------------------------- Run: 3 Test source (link in strstr): testsuites/libtests/dl09/dl-load.c Allocation size: 33554432 load: /dl09-o1.o handel: 0x4811d650: unresolved externals handle: 0x4811d650 loaded space alloc: /dl09-o1.o: 33554432: 0x40118940 load: /dl09-o2.o handel: 0x421191f0: unresolved externals handle: 0x421191f0 loaded space alloc: /dl09-o2.o: 33554432: 0x4c1225b0 load: /dl09-o3.o kobj_reloc: Relocation 0x40045820 too far from 0x4e122f48 (base+0x18) for 26bit word dlopen failed: (null) /opt/work/chris/rtems/kernel/rtems.git/testsuites/libtests/dl09/dl-load.c: 168 o->handle != NULL [ RTEMS shutdown ] RTEMS version: 6.0.0.10da3025186cadf2d78e34cf81cce02121069d44 RTEMS tools: 12.3.1 20230626 (RTEMS 6, RSB 8e568b2ca3489d6bfa48e1d29618ea9b48a5b408, Newlib 4c7d0df) executing thread ID: 0x0a010001 executing thread name: UI1
#4945	8 months ago	fixed	test	Chris Johns	Chris Johns	8 months ago
Summary	rtems-test SIS support is broken
Description	The sis BSP is using gdb and I am not sure gdb supports the SIS simulator any more? And sis-run is broken with the stand alone sis tool we now use. The -a option is not supported.
#4947	8 months ago	fixed	build	Chris Johns	Sebastian Huber <sebastian.huber@…>	7 months ago
Summary	Testsuite libtest tests not built with optimise or debug flags
Description	The tests are being built with the optimsize or debug options for: aarch64/xilinx_zynqmp_lp64_qemu arm/xilinx_zynq_a9_qemu The build is: 09:44:14 runner ' /opt/work/rtems/6/bin/arm-rtems6-gcc -MMD -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs -march=armv7-a -mthumb -mfpu=neon -mfloat-abi=hard -mtune=cortex-a9 -DHAVE_CONFIG_H=1 -Icpukit/include -I../../../cpukit/include -Icpukit/score/cpu/arm/include -I../../../cpukit/score/cpu/arm/include -Ibsps/include -I../../../bsps/include -Ibsps/arm/include -I../../../bsps/arm/include -Ibsps/arm/xilinx-zynq/include -I../../../bsps/arm/xilinx-zynq/include -Itestsuites/support/include -I../../../testsuites/support/include -Itestsuites/libtests/dl09 -I../../../testsuites/libtests/dl09 -c /opt/work/chris/rtems/kernel/rtems.git/testsuites/libtests/dl09/dl-load.c -o testsuites/libtests/dl09/dl-load.o ' while a cpukit file is: 09:45:00 runner ['/opt/work/rtems/6/bin/arm-rtems6-gcc', '-MMD', '-Wall', '-Wmissing-prototypes', '-Wimplicit-function-declaration', '-Wstrict-prototypes', '-Wnested-externs', '-march=armv7-a', '-mthumb', '-mfpu=neon', '-mfloat-abi=hard', '-mtune=cortex-a9', '-O2', '-g', '-fdata-sections', '-ffunction-sections', '-Icpukit/include', '-I../../../cpukit/include', '-Icpukit/score/cpu/arm/include', '-I../../../cpukit/score/cpu/arm/include', '../../../cpukit/libdl/rtl-elf.c', '-c', '-o/opt/work/chris/rtems/kernel/rtems.git/build/arm/xilinx_zynq_a9_qemu/cpukit/libdl/rtl-elf.c.59.o', '-DHAVE_CONFIG_H=1']
#4950	8 months ago	fixed	lib/dl	Chris Johns	Chris Johns	8 months ago
Summary	libdl test dl07 fails on PowerPC
Description	The libdl test dl07 fails on PowerPC (psim) with: rtl: alloc: del: READ_WRITE addr=0x91ca8 *** FATAL *** fatal source: 13 (RTEMS_FATAL_SOURCE_HEAP) heap error: heap=0x7e6d0 block=0x91cdc reason=HEAP_ERROR_BROKEN_PROTECTOR(0)=[0x91ce0,0,0,0] RTEMS version: 6.0.0.b9f11607b1731bc5f2391653cd8f4ebe48ba278e RTEMS tools: 12.3.1 20230626 (RTEMS 6, RSB 8e568b2ca3489d6bfa48e1d29618ea9b48a5b408, Newlib 4c7d0df) executing thread ID: 0x0a010001 executing thread name: UI1 The issue is the .bss section size is too small for the sections present. The bug has been triggered by the separate section support now being used in the testsuite.
#4951	8 months ago	fixed	tool	Chris Johns	Chris Johns <chrisj@…>	8 months ago
Summary	rtems-tools fail on FreeBSD to get number of cores
Description	$ ./waf test Waf: Entering directory `/opt/work/chris/rtems/tools/rtems-tools.git/build' Test python rtemstoolkit.check : fail Test python3 rtemstoolkit.check : fail Test python rtemstoolkit.config : fail Test python3 rtemstoolkit.config : fail Test python rtemstoolkit.configuration : pass Test python3 rtemstoolkit.configuration : pass Test python rtemstoolkit.error : pass Test python3 rtemstoolkit.error : pass Test python rtemstoolkit.execute : fail Test python3 rtemstoolkit.execute : fail Test python rtemstoolkit.git : fail Test python3 rtemstoolkit.git : fail Test python rtemstoolkit.host : fail Test python3 rtemstoolkit.host : fail Test python rtemstoolkit.log : pass Test python3 rtemstoolkit.log : pass Test python rtemstoolkit.macros : pass Test python3 rtemstoolkit.macros : pass Test python rtemstoolkit.mailer : fail Test python3 rtemstoolkit.mailer : fail Test python rtemstoolkit.options : fail Test python3 rtemstoolkit.options : fail Test python rtemstoolkit.path : pass Test python3 rtemstoolkit.path : pass Test python rtemstoolkit.reraise : pass Test python3 rtemstoolkit.reraise : pass Test python rtemstoolkit.rtems : pass Test python3 rtemstoolkit.rtems : pass Test python rtemstoolkit.stacktraces : pass Test python3 rtemstoolkit.stacktraces : pass Test python rtemstoolkit.textbox : pass Test python3 rtemstoolkit.textbox : pass Test python rtemstoolkit.version : fail Test python3 rtemstoolkit.version : fail Test failures
#4953	7 months ago	fixed	score	Sebastian Huber	Sebastian Huber	5 months ago
Summary	Use of symbol addresses for arbitrary values may not work for some code models
Description	The _ISR_Stack_size is defined by the application configuration like this: /** * @ingroup RTEMSAPIBaseDefs * * @brief Defines a global symbol with the name and value. * * @param _name is the user defined name of the symbol. The name shall be a * valid designator. On the name a macro expansion is performed and * afterwards it is stringified. * * @param _value is the value of the symbol. On the value a macro expansion is * performed and afterwards it is stringified. It shall expand to an integer * expression understood by the assembler. * * This macro shall be placed at file scope. */ #if defined(__USER_LABEL_PREFIX__) #define RTEMS_DEFINE_GLOBAL_SYMBOL( _name, _value ) \ __asm__( \ "\t.globl " RTEMS_XSTRING( RTEMS_SYMBOL_NAME( _name ) ) \ "\n\t.set " RTEMS_XSTRING( RTEMS_SYMBOL_NAME( _name ) ) \ ", " RTEMS_STRING( _value ) "\n" \ ) #else #define RTEMS_DEFINE_GLOBAL_SYMBOL( _name, _value ) #endif RTEMS_DEFINE_GLOBAL_SYMBOL( _ISR_Stack_size, CONFIGURE_INTERRUPT_STACK_SIZE ); Firstly, the RTEMS_DEFINE_GLOBAL_SYMBOL() is broken. It should depend on GNUC and not USER_LABEL_PREFIX. Secondly, setting a symbol to an arbitrary absolute address may not work with all code models, see: ​https://sourceware.org/pipermail/binutils/2023-September/129448.html The _ISR_Stack_size should be changed to point to the end of the ISR stack of processor 0: RTEMS_DEFINE_GLOBAL_SYMBOL( _ISR_Stack_0_end, RTEMS_SYMBOL_NAME( _ISR_Stack_area_begin ) + CONFIGURE_INTERRUPT_STACK_SIZE ); The tests for RTEMS_DEFINE_GLOBAL_SYMBOL() should be changed to use values relative to another symbol.
#4954	7 months ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	6 months ago
Summary	sparc: Simplify the clock driver and CPU counter support
Description	Unfortunately, most SPARC SoCs? supported by RTEMS lack a proper free-running counter. The only exception are the GR712RC and the GR740. For the pre-qualified GR712RC BSP, the current implementation is too complicated. Make the fall back to a free-running GPTIMER mandatory for the leon3 BSP family to simplify things. For the GR712RC use the DSU time tag counter.
#4955	7 months ago	fixed	arch/sparc	Sebastian Huber	Sebastian Huber	11 days ago
Summary	sparc: Potential stack corruption in uniprocessor configurations during start multitasking
Description	The system initialization uses the interrupt stack. A first level interrupt shall never interrupt a context which uses the interrupt stack. Such a use would lead to stack corruption and undefined system behaviour. Unfortunately, in uniprocessor configurations this is the case right now. Multiprocessing is started using _CPU_Context_restore(). The caller of this function (_Thread_Start_multitasking()) uses the interrupt stack. Later we have in cpukit/score/cpu/sparc/cpu_asm.S: mov %g1, %psr ! restore status register and ! **** ENABLE TRAPS **** ld [%o1 + G5_OFFSET], %g5 ! restore the global registers ld [%o1 + G7_OFFSET], %g7 ! Load thread specific ISR dispatch prevention flag ld [%o1 + ISR_DISPATCH_DISABLE_STACK_OFFSET], %o2 ! Store it to memory later to use the cycles ldd [%o1 + L0_OFFSET], %l0 ! restore the local registers ldd [%o1 + L2_OFFSET], %l2 ldd [%o1 + L4_OFFSET], %l4 ldd [%o1 + L6_OFFSET], %l6 ! Now restore thread specific ISR dispatch prevention flag st %o2, [%g6 + PER_CPU_ISR_DISPATCH_DISABLE] ldd [%o1 + I0_OFFSET], %i0 ! restore the input registers ldd [%o1 + I2_OFFSET], %i2 ldd [%o1 + I4_OFFSET], %i4 ldd [%o1 + I6_FP_OFFSET], %i6 ldd [%o1 + O6_SP_OFFSET], %o6 ! restore the output registers Between the ENABLE TRAPS and the restore of the output registers, we still use the stack of the caller and interrupts may be enabled. If an interrupt happens in this code block, the interrupt stack is concurrently used which may lead to a crash.
#4959	7 months ago	fixed	test	Chris Johns	Chris Johns	4 months ago
Summary	rtems-test tftp does not a session timeout
Description	The TFTP server has a timeout for session that are running however there is no timeout on the listen state. A target could fail to create a session. Add a session timeout.
#4967	5 months ago	fixed	tool/newlib	Chris Johns		2 months ago
Summary	strtof leaks memory
Description	The libc call strtof leaks memory when a thread exits if strtof has been made within the context of the thread. The code uses Balloc and Bfree to obtain pieces of memory it uses. The memory is recycled using newlib reent TLS variables. The _reclaim_reent() does not clean up _REENT_MP_P5S.
#4969	5 months ago	fixed	tool	Chris Johns	Chris Johns	7 weeks ago
Summary	Update libibery sources in rtems-tools
Description	MacOS is removing vfork. Update to gcc latest sources and use POSIX spawn support. Update all files.
#4970	5 months ago	fixed	tool	Chris Johns	Chris Johns	5 months ago
Summary	Fix C++ 17 deprecated functionality in rtems-tools
Description	The C++ code uses functionality depreciated in c++17. Update and resolve.
#4971	5 months ago	fixed	test	Sebastian Huber	Sebastian Huber	5 months ago
Summary	Report that nested validation test cases did run
Description	Some validation test cases use parameterized nested test cases. For example, the rtems_semaphore_obtain() test uses a parameterized test case to seize a mutex. The nested mutex seize test case uses a parameterized thread queue enqueue test case. Currently, there is no indication that these nested test cases did actually run. Some of these nested test cases run in a tight loop, so printing a hint on every run could lead to a massive amount of test output. Add test case remarks to the RTEMS Test Framework which may be used to give such an indication.
#4973	5 months ago	fixed	tool/rsb	Chris Johns	Chris Johns	5 months ago
Summary	Add RSB options to control a specific python
Description	Older hosts can have varying installations of pythons. Provide a means to set a version found in your path with: --with-python2 --with-python3 --with-python-version
#4974	5 months ago	fixed	tool	zhengxiaojun	zhengxiaojun <jameszxj@…>	7 weeks ago
Summary	rtems-tools buid failed with '--host=x86_64-w64-mingw32'
Description	rtems-tools build failed with "--host=x86_64-w64-mingw32" after the commit ​https://git.rtems.org/rtems-tools/commit/?id=a4b312ee4f3d8c93ab0d45d568879069e7ac7d7a 'writeargv' is undefined. add libierty/argv.c will solve this problem.
#4975	5 months ago	fixed	lib	Chris Johns	Chris Johns	4 months ago
Summary	Terminal row/col probe timeout too fast
Description	The timeout is too fast if the terminal and the target hardware are on different continents.
#4977	4 months ago	fixed	lib	Chris Johns	Chris Johns	4 months ago
Summary	tmux bug swaps lines and columns
Description	A tmux bug which is fixed but not released swaps the columns and rows fields: ​https://github.com/tmux/tmux/issues/3457 Given it will be a while before this enters downstream packages provide a work around. also improve response time to the query. At the moment a 50 msec delay is imposed and it does not have to be there.
#4980	4 months ago	fixed	doc	Chris Johns	Chris Johns	4 months ago
Summary	Build documentation using sphinx 7.2
Description	Set no max version Fix the layout to support versions later than 5.9 I have no idea when the breakage happened and if other versions are effected.
#4986	3 months ago	fixed	config	Sebastian Huber	Sebastian Huber	3 days ago
Summary	Improve CONFIGURE_TICKS_PER_TIMESLICE configuration and documentation
Description	The documentation allows a value of zero for CONFIGURE_TICKS_PER_TIMESLICE. This should be changed to make zero an invalid value and it should be rejected by an #error configuration error.
#4994	2 months ago	fixed	doc	Chris Johns	Chris Johns	2 months ago
Summary	Documentation Sphinx theme broken for HTML
Description	The theme in rtems-docs.git breaks with newer versions of sphinx-build.
#4995	8 weeks ago	fixed	lib/dl	Joel Sherrill	Sebastian Huber	7 weeks ago
Summary	powerpc/qoriq_e6500_64 rtems-syms unknown machine type (cloned)
Description	Cloned from #3627: rtems-syms does not know 64-bit powerpc. gmake[5]: Entering directory `/data/home/joel/rtems-work/rtems-testing/rtems/build-powerpc-qoriq_e6500_64-rtems/powerpc-rtems5/c/qoriq_e6500_64/testsuites/libtests' rtems-syms -e -c "-mcpu=e6500 -m64 -fno-common -O2 -g -ffunction-sections -fdata-sections -Wall -Wmissing-prototypes -Wimplicit-function-declaration -Wstrict-prototypes -Wnested-externs" -o dl07-sym.o dl07.pre error: machine-type: unknown machine type: 21
#4999	5 weeks ago	fixed	arch/powerpc	Heinz Junkes		5 weeks ago
Summary	missing bsp's in rtems 6.1 rc2
Description	Only a few bsps are available as configurations (bsets). Not all that are listed in the documentation. E.g. 'beatnik' is missing. this is the output of ../source-builder/sb-set-builder --list-bsets ... bsps/atsamv.bset bsps/beagleboneblack.bset bsps/erc32.bset bsps/gr712rc.bset bsps/gr740.bset bsps/imx7.bset bsps/pc.bset bsps/qoriq_e500.bset bsps/qoriq_e6500_32.bset bsps/qoriq_e6500_64.bset bsps/raspberrypi2.bset bsps/xilinx_zynq_zc702.bset bsps/xilinx_zynq_zc706.bset bsps/xilinx_zynq_zedboard.bset ...