Custom Query (28 matches)

RSB 4.11 fails to build rtems-tools

I checked out RSB on branch 4.11.

The execution of the following command fails

../source-builder/sb-set-builder --prefix=/mnt/sdb/rtems/4.11 4.11/rtems-powerpc --jobs 16

with the following error:

Waf: Entering directory `/mnt/sdb/rtems/src_4.10/rsb/rtems/sources/git/rtems-tools.git/build'
[ 16/258] Compiling rtemstoolkit/elftoolchain/libelf/elf.c
../rtemstoolkit/elftoolchain/libelf/elf.c:34:29: error: expected expression before ‘,’ token
  .libelf_arch  = LIBELF_ARCH,
                             ^

Build failed

How can I obtain a working 4.11 toolchain?

Failure to download mpc tar.gz during 4.11 build

Building from 4.11 branch results in following error.

download: ​http://www.multiprecision.org/mpc/download/mpc-1.0.3.tar.gz: error: HTTP Error 500: Internal Server Error error: downloading ​http://www.multiprecision.org/mpc/download/mpc-1.0.3.tar.gz: all paths have failed, giving up Build FAILED

See error report: rsb-report-arm-rtems4.11-gcc-4.9.3-newlib-2.2.0.20150423-x86_64-linux-gnu-1.txt

error: downloading ​http://www.multiprecision.org/mpc/download/mpc-1.0.3.tar.gz: all paths have failed, giving up Build Set: Time 0:04:43.144951 Build FAILED

4.11 GCC Does not build with Recent GCC

The older gcc used with 4.11.3 does not build with the GCC 7.2 included with Ubuntu 17.10. Log attached.

rtems-tools 4.11 branch is missing rtems bsp builder

$ ls rtems-tools/tester/ config covoar rt rtems rtems-test wscript

Seems like a good idea to backport it and adjust the BSPs in each tier.

Classic API Barrier Wait Section Title Has Wrong Name

The section title says obtain not wait.

source-builder failure building qemu and qemu4

The command and the failure:

time ../source-builder/sb-set-builder --mail --mail-to=build@rtems.org --mail-from=joel@rtems.org --log=l-qemu4-4.11.txt --prefix=/home/joel/rtems-cron-411/tools/4.11 devel/qemu4
RTEMS Source Builder - Set Builder, 4.11 (158ad680aed1)
Build Set: devel/qemu4
error: no build set file found: devel/qemu4.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-411/rtems-source-builder/source-builder/sb/setbuilder.py", line 735, in run
    b.build(deps, mail = mail)
  File "/home/joel/rtems-cron-411/rtems-source-builder/source-builder/sb/setbuilder.py", line 633, in build
    if sizes_valid:
UnboundLocalError: local variable 'sizes_valid' referenced before assignment

rtems-tools tester 4.11 branch hosts defaults error on Linux

The 4.11 branch of rtems-tools/tester/rtems-test fails on Ubuntu (and other Linux hosts) with an error message regarding "failed to load Linux host support"

c/src/lib/libbsp/sparc/shared/1553/gr1553b.c: 2 * memory leak ?

1.

[rtems/c/src/lib/libbsp/sparc/shared/1553/gr1553b.c:231]: (error) Memory leak: priv

if ( ambadev == NULL ) {

return DRVMGR_FAIL;

}

2.

[rtems/c/src/lib/libbsp/sparc/shared/1553/gr1553b.c:260]: (error) Memory leak: priv

return DRVMGR_OK;

}

rtems/c/src/lib/libbsp/sparc/shared/1553/b1553brm.c:992: bad if test ?

[rtems/c/src/lib/libbsp/sparc/shared/1553/b1553brm.c:992]: (style) Unsigned variable 'count' can't be negative so it is unnecessary to test it.

Source code is

if (count >= 0) {

return RTEMS_SUCCESSFUL;

} return RTEMS_UNSATISFIED;

Which means that the second return is dead code. Suggest code rework.

rtems/c/src/lib/libbsp/sparc/shared/amba/ambapp.c:258: possible bad test ?

[rtems/c/src/lib/libbsp/sparc/shared/amba/ambapp.c:258]: (error) Memory leak: apbdev

apbdev = ambapp_alloc_dev_struct(DEV_APB_SLV); if (!dev)

return -1;

Maybe better code

apbdev = ambapp_alloc_dev_struct(DEV_APB_SLV); if (!apbdev)

return -1;

Looks like a cut-n-paste error and then some inadequate testing to me.

PDF documentation is formatted as multi-column

The PDF generated documents is in a 2 column format.

RSB VERSION in unicode environment fails.

This is the issue in #3312

Move expat's home site to github from SF.

See #3315

mhttpd's http etag can result in invalid caching in a browser.

The mhttp's http etag uses the mtime and file length and this can cause

subtle issues if a target has no RTC or it is incorrect and files are being copied without preserving the mtime or changes happen that do not change the length.

The cp and untar code do not update a file's time.

Add support for an etag callback so a user can manage the tag, ie MD5 or something similar.

JFFS2 returns a stat directory size as 0

A stat of a directory on JFFS2 returns a size of 0. The IMFS and RFS return a non-zero size.

Returning 0 in st_size for a directory stat call breaks scandir.

4.11: infinite loop in RSB's path when a prefix path is not writable

The code gets the dirname() of the path stepping up until there is no path however dirname('/') is / so the path never has a length of 0.

RSB and RTEMS Tools Support for python2 and python3 (cloned)

Cloned from #3537:

Upstream python does not create a python command any more and creates python2 and python3. Distributions and operating systems are starting to ship without the python command.

The RSB and RTEMS Tools python commands need to be updated and tested so they run on Python2 and Python3 and support added to use the available commands.

libc_reent set up is broken

The patch ...

​https://git.rtems.org/rtems/commit/?id=69aa33490b1cd357519ab70b15ad150e11bb752e

... removes pieces of the initialisation of newlib's re-entrant struct. Are they are needed? I think they are. This breaks code such as telnet clients is subtle and difficult to find ways.

With this change you cannot assign anything to stdint or stdout unless a suitable call to newlib is made that has the internal CHECK_INIT() macro. Any assignment is wiped out as _sdidinit is false and that results in _sinit being called.

Is the lazy init in newlib on purpose? If so should __getreent in our libcsupport be returning a struct that is not initialised?

Is the change in the linked patch standards compliant?

A simple grep of code for an assignment of libc_reent did not show how this TCB field was initialised and it is not apparrent or easy to find. Why is the code written this way? There are no comments with _Thread_Control_add_ons and Thread_Configured_control explaining what are doing and how to use them and how they get loaded into the TCB.

No output from joel scripts in telnet (cloned)

Cloned from #3859:

Running a joel script in a telnet session results in the output being sent to the global stdout. For example:

$ telnet 1.2.3.4
Trying 1.2.3.4...
Connected to 1.2.3.4.
Escape character is '^]'.

RTEMS Shell on /dev/pty0. Use 'help' to list commands.
[/] # cat j
#! joel
ls -las /
[/] # ./j
[/] #

The bug is a new shell main loop task will default to the global stdout, stdin etc and has no information about the parent's std handles. A joel script runs in it's own work task and does not know the telnet's std handles.

There are a related set of issues in the handling of the shell_env variable, POSIX key handling and the use of the external call rtems_shell_main_loop.

The telnet example in libbsd has:

static void
telnet_shell(char *name, void *arg)
{
	rtems_shell_env_t env;

	memset(&env, 0, sizeof(env));

	env.devname = name;
	env.taskname = "TLNT";
	env.login_check = NULL;
	env.forever = false;

	rtems_shell_main_loop(&env);
}

This is problematic as control of the env has been lost and this make backwards comptatable changes difficult. Control of this struct needs to be brought back under the shell code.

Currently the posix key is set in the parent task only when the run entry point is used. The run's created shell_env is then passed to the shell's main loop task as an argument from which it is cloned. This means an env is malloced in each run call and again in the main loop of the shell.

The current code leaks memory as repeated calls to a joel script in a shell will set the key over and over. The destructor is only called when the task is deleted. We have to assume the cleanup of any shell_env allocated externally to the shell code has to be handled externally.

Setting the key in the main loop task is problematic because telnet code such as the example in libbsd uses a local stack shell_env and the key has a destructor that blindly free's the key's memory when a task is released.

Changes:

1. Add parent_stdout, parent_stdin, and parent_stderr to the shell_env and set to the parent's std handles.
2. Add a managed flag to shell_env and only set when allocated by rtems_shell_init_env. Change rtems_shell_env_free to only free the shell_env if managed.
3. Remove all key sets and have only one in the shell's main loop code.
4. Change rtems_shell_init_env to get the current tasks key and clone that before cloning the global env.
5. Update rtems_shell_dup_current_env to set the parent std handles.
6. Have the main loop use the parent std handles rather than the global handles.
7. Check the magic field has been set in the shell's main loop and raise an error if not set. The only code to set this field should reside in shell.c. Code such as libbsd will need to call rtems_shell_dup_current_env.

RSB Tools newlib patch checksum fails.

The sourceware git web interface has been updated to a new git version and git adds its version number to exported patches. This has broken the checksum for:

newlib-cygwin-git-f70d9ae6adc6ed7952806056349ba9f8ba3c65c8.patch

Any build using our git repo and the 4.11 branch will fail to build.

FreeBSD 12 does not build 4.11/rtems-arm gcc

Building 4.11/rtems-arm on FreeBSD 12.1 fails due to -Wreserved-user-defined-literal warnings.

4.11 branch RSB is broken with python3 on Linux

With Ubuntu 20 the 4.11 RSB fails with:

(py) johnsy@radium:~/si/rtems/rsb/rtems$ ../source-builder/sb-set-builder --prefix=$HOME/si/rtems/4.11 --log=arm.txt 4.11/rtems-arm                                                                                        
../source-builder/sb/macros.py:233: SyntaxWarning: "is" with a literal. Did you mean "=="?
  if key.startswith('%{') and key[-1] is '}':                                                  
../source-builder/sb/macros.py:269: SyntaxWarning: "is" with a literal. Did you mean "=="?
  if c is '#' and not state.startswith('value'):                    
../source-builder/sb/macros.py:272: SyntaxWarning: "is" with a literal. Did you mean "=="?     
  if not (state is 'key' and len(token) == 0) and \                       
../source-builder/sb/macros.py:275: SyntaxWarning: "is" with a literal. Did you mean "=="?
  if state is 'key':                                                   
[blah blah blah]

rtems-tools waf is broken with python 3.8.2

$ ./waf distclean configure --prefix=$HOME/si/rtems/4.11
/home/johnsy/si/rtems/rt/wscript:140: SyntaxWarning: "is not" with a literal. Did you mean "!="?
  elif host is not 'native':
Traceback (most recent call last):
  File "/home/johnsy/si/rtems/rt/.waf3-1.8.19-b1fc8f7baef51bd2db4c2971909a568d/waflib/Node.py", line 312, in ant_iter
    raise StopIteration
StopIteration

4.11/rtems-arm does not build on ubuntu 21.10

The error is:

../../gcc-4.9.3/gcc/reload1.c: In function 'void init_reload()':
../../gcc-4.9.3/gcc/reload1.c:89:24: error: use of an operand of type 'bool' in 'operator++' is forbidden in C++17
   89 |   (this_target_reload->x_spill_indirect_levels)
      |   ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~

sb-bootstrap does not work with python3

Python module importing of sb-bootstrap is broken on the 4.11 branch with python3.

The sb-bootstrap was updated on the 5 branch and rtems-bootstrap was added to rtems.git on the 5 branch. The RSB updated version has the wrappers for the python version which is a hassle to make work so copy in the rtems-bootstrap version. It is a single file and does not have the modules issue.

Instead of updating sb-bootstrap replace the one on 4.11 in the RSB with a copy of rtems-bootstrap from the 5 branch in rtems.git and add a copy to the 4.11 branch in rtems.git.

gcc m68k inline asm does not recognize sr

unknown register when using inline assembly for m68k gcc gives: error: unknown register name 'sr' in 'asm' from code below

    asm ("movew #0x2700,%%sr \n\t" );

for other registers it works, using gcc for *.S works fine

error: no build set file found: 4.11/rtems-arm.bset

I was trying to follow the instruction here: ​https://docs.rtems.org/releases/rtems-docs-4.11.2/user/installation/index.html#installation

I was trying to build a tool chain for the ARM architecture as descibed. Below is my output and error:

$ pwd /c/opt/rtems/4_11_2

$ ./source-builder/sb-set-builder --prefix=/opt/rtems/4.11 --jobs=none 4.11/rtems-arm RTEMS Source Builder - Set Builder, 4.11.not_released Build Set: 4.11/rtems-arm error: no build set file found: 4.11/rtems-arm.bset Build Set: Time 0:00:00 Build FAILED

$ uname -a CYGWIN_NT-6.1-WOW64 AZ25CY5LJH2W 1.7.16(0.262/5/3) 2012-07-20 22:55 i686 Cygwin

Error building 4.11.2 rtems-tools

So I am struggling to build the 4.11.2 rtems-tools. I am building as follows:

$pwd /c/opt/rtems/4_11_2/rtems

$ ../source-builder/sb-set-builder --prefix=/opt/rtems/4.11 --jobs=none --no-download 4.11/rtems-tools

I am getting the following error:

x rtems-tools-4.11.2/wscript x tools/ x tools/4.11/ x tools/4.11/gdb/ x tools/4.11/gdb/sparc/ x tools/4.11/gdb/sparc/7.9/ x tools/4.11/gdb/sparc/7.9/0001-sim-erc32-Disassembly-in-stand-alone-mode-did-not-wo.patch x tools/4.11/gdb/sparc/7.9/0002-sim-erc32-Corrected-wrong-CPU-implementation-and-ver.patch x tools/4.11/gdb/sparc/7.9/0003-sim-erc32-Perform-pseudo-init-if-binary-linked-to-no.patch x tools/4.11/gdb/sparc/7.9/0005-sim-erc32-Remove-unused-defines-in-Makefile-and-swit.patch x tools/4.11/gdb/sparc/7.9/0006-sim-erc32-Fix-incorrect-simulator-performance-report.patch x tools/4.11/gdb/sparc/7.9/0007-sim-erc32-File-loading-via-command-line-did-not-work.patch x tools/4.11/gdb/sparc/7.9/0008-sim-erc32-Added-v-command-line-switch-for-verbose-ou.patch x tools/4.11/gdb/sparc/7.9/0009-sim-erc32-Removed-type-mismatch-compiler-warnings.patch x tools/4.11/gdb/sparc/7.9/0010-sim-erc32-Switched-emulated-memory-to-host-endian-or.patch x tools/4.11/gdb/sparc/7.9/0011-sim-erc32-use-SIM_AC_OPTION_HOSTENDIAN-to-probe-for-.patch x tools/4.11/gdb/sparc/7.9/0012-sim-erc32-Use-memory_iread-function-for-instruction-.patch x tools/4.11/gdb/sparc/7.9/0014-sim-erc32-Use-gdb-callback-for-UART-I-O-when-linked-.patch x tools/4.11/gdb/sparc/7.9/0015-sim-erc32-Access-memory-subsystem-through-struct-mem.patch x tools/4.11/gdb/sparc/7.9/0016-sim-erc32-Use-readline.h-for-readline-types-and-func.patch x tools/4.11/gdb/sparc/7.9/0017-sim-erc32-Move-local-extern-declarations-into-sis.h.patch x tools/4.11/gdb/sparc/7.9/0018-sim-erc32-Add-support-for-LEON3-processor-emulation.patch x tools/4.11/gdb/sparc/7.9/0019-sim-erc32-Add-support-for-LEON2-processor-emulation.patch x tools/4.11/gdb/sparc/7.9/0022-Add-watchpoint-support-to-gdb-simulator-interface.patch x tools/4.11/gdb/sparc/7.9/0023-sim-erc32-ELF-loading-could-fail-on-unaligned-sectio.patch + cd rtems-tools-4.11.2 + chmod -R a+rX,g-w,o-w . + cd /c/opt/rtems/4_11_2/rtems/build/rtems-tools-4.11.2-1 + SB_CXC=no ==> clean %{buildroot}: /c/opt/rtems/4_11_2/rtems/build/tmp/rtems-tools-4.11.2-1-root-fred.decaro + echo '==> clean %{buildroot}: /c/opt/rtems/4_11_2/rtems/build/tmp/rtems-tools-4.11.2-1-root-fred.decaro' + rm -rf /c/opt/rtems/4_11_2/rtems/build/tmp/rtems-tools-4.11.2-1-root-fred.decaro + /bin/mkdir -p /c/opt/rtems/4_11_2/rtems/build/tmp/rtems-tools-4.11.2-1-root-fred.decaro + echo '==> %build:' ==> %build: ++ pwd + build_top=/c/opt/rtems/4_11_2/rtems/build/rtems-tools-4.11.2-1 + test i686-pc-cygwin '!=' i686-pc-cygwin + RT_HOST= + cd rtems-tools-4.11.2 + ./waf configure --prefix=/opt/rtems/4.11 Error: run waf-light from a folder containing waflib shell cmd failed: sh -ex /c/opt/rtems/4_11_2/rtems/build/rtems-tools-4.11.2-1/doit error: building rtems-tools-4.11.2-1

Any ideas?