source: rtems/c/src/lib/libbsp/powerpc/beatnik/ChangeLog @ 438983d2

4.115
Last change on this file since 438983d2 was 438983d2, checked in by Jennifer Averett <Jennifer.Averett@…>, on 06/29/10 at 13:49:06

2010-06-29 Joel Sherrill <Joel.Sherrill@…>

  • configure.ac: Added AC_CHECK_SIZEOF macro.
  • Property mode set to 100644
File size: 9.2 KB
Line 
12010-06-29      Joel Sherrill <Joel.Sherrill@OARcorp.com>
2
3        * configure.ac: Added AC_CHECK_SIZEOF macro.
4
52010-05-10      Joel Sherrill <joel.sherrilL@OARcorp.com>
6
7        * include/bsp.h: Do not include <bsp/bsp_bsdnet_attach.h> if networkign
8        disabled.
9
102010-04-14      Ralf Corsépius <ralf.corsepius@rtems.org>
11
12        * network/support/bsp_attach.c: Include <unistd.h>.
13
142010-03-27      Thomas Doerfler <Thomas.Doerfler@embedded-brains.de>
15
16        * network/if_gfe/if_gme.c, network/if_em/if_em.c: Removed warnings
17
182010-02-09  Till Straumann <strauman@slac.stanford.edu>
19
20        * network/if_mve/mv643xx_eth.c: Fixed alignment attribute
21        in descriptor declaration. Not the pointers to the descriptors
22        have to be aligned but the descriptors themselves (didn't
23        cause problems but caused unnecessary holes in 'private' struct).
24
25        FIX: Added more robustness when number of available TX descriptors
26        drops to zero. (This can e.g., happen if the link goes bad causing
27        packets to stall in the FIFO.) At the following points the transmitter
28        is explicitly (re-)started:
29          o when link comes up and number of available TXDs is zero the
30            TX is restarted.
31          o on a failed attempt to send data due to lack of TXDs the
32            TX is restarted if swiping the TX ring doesn't yield any
33            buffers (i.e., if the # of available buffers is still zero
34            after the swipe).
35
362009-12-14      Sebastian Huber <sebastian.huber@embedded-brains.de>
37
38        * Makefile.am, preinstall.am: Removed ppc_exc_bspsupp.h include file.
39
402009-12-08      Joel Sherrill <joel.sherrill@oarcorp.com>
41
42        * include/bsp.h: BSP should not define
43        CONFIGURE_NUMBER_OF_TERMIOS_PORTS.
44
452009-12-08      Joel Sherrill <joel.sherrill@oarcorp.com>
46
47        * network/if_em/Makefile, network/if_gfe/Makefile: Removed.
48
492009-12-03      Joel Sherrill <joel.sherrill@oarcorp.com>
50
51        * Makefile.am, preinstall.am: Account for coverhd.h and tm27.h. Use
52        stubs for both.
53
542009-11-06  Till Straumann <strauman@slac.stanford.edu>
55
56        * Makefile.am, irq/discovery_pic.c, irq/irq_init.c:
57        dump raw_exceptions.rel, raw_exception.h; these have
58        disappeared. <bsp/vectors.h> must now be included
59        instead.
60
61        * Makefile.am: Add 'altivec.rel'.
62
63        * make/custom/beatnik.cfg: Use -mcpu=7400; this enables
64        AltiVec!
65
662009-11-06  Till Straumann <strauman@slac.stanford.edu>
67
68        * beatnik.cfg, make/custom/beatnik.cfg: moved beatnik.cfg
69        to new make/custom subdir.
70
712009-10-20  Till Straumann <strauman@slac.stanford.edu>
72
73        * startup/bspstart.c: leave 'work-space start' and initial
74        stack alone. These are now handled by the shared framework
75        and linker script etc. Locate interrupt stack after __rtems_end
76        and obtain its size from the configuration.
77
782009-10-20  Till Straumann <strauman@slac.stanford.edu>
79
80        * network/if_mve/mv643xx_eth.c: made mutex a binary semphore;
81        simple binary semaphore doesn't support priority inheritance.
82        This was silently ignored under previous releases but is an
83        error under 4.10.
84
852009-10-20  Till Straumann <strauman@slac.stanford.edu>
86
87        * Makefile.am, bsp_specs, preinstall.am, flash/flashcfg.c,
88        include/bspopts.h.in, irq/discovery_pic.c, irq/irq_init.c,
89        marvell/gt_timer.c, marvell/gt_timer.h, marvel/gti2c.c,
90        network/if_gfe/if_gfe.c, network/if_gfe/if_gfe_rtems.c,
91        network/if_mve/mv643xx_eth.c, network/support/bsp_attach.c,
92        pci/gt_pci_init.c, pci/motload_fixup.c, startup/bspstart.c,
93        startup/i2c_init.c:
94        Ported to rtems HEAD (to become rtems-4.10). This consisted
95        mainly of fixing compiler warnings (mostly: adding prototypes
96        to function declarations and moving extern declarations to
97        global scope).
98
99        A pecularity: if_gfe.c had to remove 'queue.h' inclusion.
100        we have two versions of queue.h: one in newlib another one in
101        rtems - don't know how this is supposed to work...     
102
1032009-10-17  Till Straumann <strauman@slac.stanford.edu>
104
105        * Makefile.am, network/if_mve_mv643xx_eth.c,
106        network/if_mve_pub.h: Enhanced low-level API allowing
107        the user to selectively enable/disable/acknowledge
108        interrupts and to install their own ISR (rather than having
109        the driver posting an event to a single task).
110
1112009-10-03  Till Straumann <strauman@slac.stanford.edu>
112        * network/if_mve/mv643xx_eth.c:
113        BUGFIX: mbuf leak; consume_rx_mbuf() must release mbuf
114        if 'len'<=0.
115
116        BUGFIX: Must initialize 'media-word' argument before
117        calling BSP_mve_media_ioctl() (defines PHY instance).
118
1192009-06-05  Till Straumann <strauman@slac.stanford.edu>
120        * network/if_mve/mv643xx_eth.c, network/if_mve/if_mve_pub.h,
121        Makefile.am: Added MC address reference count and
122        BSP_mve_mcast_filter_accept_del() to remove a single
123        entry from the filter.
124
1252009-06-05  Till Straumann <strauman@slac.stanford.edu>
126        * network/if_mve/mv643xx_eth.c, network/if_mve/if_mve_pub.h,
127        Makefile.am: first stab at adding multicast support.
128
1292009-06-05  Till Straumann <strauman@slac.stanford.edu>
130        * network/if_gfe/if_gfe.c:
131        o propagate PROMISC flag to hardware (SIOCSIFFLAGS)
132        o handle case where IFF_PROMISC is set (and wasn't before)
133          in gfe_hash_fill() routine.
134
1352009-06-02  Till Straumann <strauman@slac.stanford.edu>
136        * network/if_gfe/if_gfe.c: activated and fixed multicast
137        support.
138
1392009-06-01  Till Straumann <strauman@slac.stanford.edu>
140        * network/if_em/if_em.c: activated multicast support.
141
1422008-10-30      Till Straumann <strauman@slac.stanford.edu>
143
144    MERGED from rtems-4-7-branch:
145        * Makefile.am, network/if_mve/mv643xx.c, network/if_mve/if_mve_pub.h:
146        o Exported new low-level driver entry points:
147            - BSP_mve_ack_link_chg() so that changes in PHY link status can be
148            propagated to the serial port when handling link-change interrupts.
149            - BSP_mve_dump_stats() for printing statistics.
150        o FIXED reading of statistics counters.
151        o Count interrupts (statistics) in ISR rather than network daemon
152          (which is only used by the BSD driver).
153
1542008-10-04      Till Straumann <strauman@slac.stanford.edu>
155
156        * beatnik.cfg: updated to 4.9; removed make-exe
157        make-cxx-exe commands. Replaced CPU cflags to use
158        -mpowerpc -D__ppc_generic.
159
1602008-10-04      Till Straumann <strauman@slac.stanford.edu>
161
162        * startup/linkcmds: increased size of CODE memory
163        area to 32M.
164
1652008-05-10      Till Straumann <strauman@slac.stanford.edu>
166
167        * pci/gt_pci_init.c, pci/pci_io_remap.c: fixed 32-bit
168        types. RTEMS' pci_config access functions now use uint32_t,
169        earlier versions used unsigned. Both are incompatible,
170        unfortunately (gcc regards unsigned and unsigned long different
171        beasts leading to warnings and alias-issues :-()
172
1732008-05-10      Till Straumann <strauman@slac.stanford.edu>
174
175        * network/porting/rtemscompat1.h, network/porting/rtemscompat.h,
176        network/porting/if_xxx_rtems.c, network/if_gfe/if_gfe_rtems.c:
177        Fixed 32-bit types (pci config access, byteorder macros differ
178        depending on RTEMS version :-(). We now check for version and
179        use appropriate types (unsigned vs. uint32_t).
180        Silenced more warnings (ifndef DEBUG_MODULAR the METHODSPTR
181        is always non-zero; hence I ifdef'ed the affected code snippet).
182       
1832008-03-20      Till Straumann <strauman@slac.stanford.edu>
184
185        * include/bsp.h, startup/bspstart.c: confdefs.h now wants
186        us to use BSP_INTERRUPT_STACK_SIZE instead of
187        CONFIGURE_INTERRUPT_STACK_MEMORY.
188
1892008-03-19      Till Straumann <strauman@slac.stanford.edu>
190
191        * irq/discovery_pic.c: must spare GPP7_0 etc. summary
192        interrupts in BSP_enable_irq_at_pic() etc.
193        New 'new-exceptions/bspsupport' code scans all IRQS
194        and enables or disables depending on the initial config
195        having a handler connected. This initial disable operation
196        switched-off the summaries and I had no GPP interrupts...
197
1982008-01-04      Till Straumann <strauman@slac.stanford.edu>
199
200        * startup/bspstart.c: changed Kate's copyright note
201        again as requested by her email 1/04/2008.
202
2032008-01-04      Till Straumann <strauman@slac.stanford.edu>
204
205        * startup/bspstart.c: changed Kate's copyright note
206        as requested by her email 8/16/2007.
207
2082007-12-11      Till Straumann <strauman@slac.stanford.edu>
209
210        * irq/discovery_pic.c: don't print warnings if an
211        invalid irq number is passed to BSP_disable_irq_at_pic(),
212        BSP_enable_irq_at_pic(). irq_supp.h says we must
213        silently ignore.
214
2152007-12-11      Till Straumann <strauman@slac.stanford.edu>
216
217        * Makefile.am: use new irq_bspsupport.rel which was
218        split out of exc_bspsupport.rel to provide finer-grained
219        control over what BSPs want to use.
220
2212007-12-10      Till Straumann <strauman@slac.stanford.edu>
222
223        * Makefile.am, startup/vpd.c, startup/vpd.h:
224        moved VPD support to ../shared/motorola.
225
2262007-12-08      Till Straumann <strauman@slac.stanford.edu>
227
228        * Makefile.am: merged shared/vme/vme_universe.c and
229        shared/vme/vme_universe_dma.c into one file.
230        Use vme_universe.c, vmeconfig.c from shared area.
231
2322007-11-30      Till Straumann <strauman@slac.stanford.edu>
233
234        * startup/bspstart: removed _Cpu_table.exceptions_in_RAM.
235
2362007/11/27 (TS):
237 - Generalized flash support and moved to shared area (libchip would probably
238   more appropriate).
2392007/10/22 (TS):
240 - DECREMENTER interrupt is now handled the same way external interrupts are.
241   It can also be assigned a priority and the handler is executed in priority
242   order, i.e., it can be preempted by higher-priority interrupts and
243   is protected from being preempted by lower-priority irqs.
2442007/10/08 (TS):
245 - ChangeLog added
246 - (Makefile.am) MUST NOT use -msoft-float because this also prevents CR7
247   to be set/cleared when calling vararg routines (which may then save/restore
248   FP args on the stack or do other bad things) :-(
249   Still don't know how to deal with implicit usage of the FPU by GCC
250   (problem in ISRs and integer-only tasks).
Note: See TracBrowser for help on using the repository browser.