2221
2221
- Fix stack_size assertion in GC_pthread_create.
2222
2222
- Fix assertion in GC_steal_mark_stack.
2225
- Fix CPU count detection for Irix and FreeBSD. (Thanks to Dan Bonachea.)
2226
- Integrate Dan Bonachea's patch for the IBM XLC compiler on Darwin.
2227
- Integrated Andreas Tobler's FreeBSD/PowerPC patch.
2228
- Don't access the GC thread structure from the restart handler. It's
2229
unsafe, since the handler may run too late. (Thanks to Ben Maurer for
2230
tracking this down.)
2231
- Applied Christian Thalinger's patch to change comment syntax in
2233
- Added test for GC_no_dls in GC_dyld_image_add for DARWIN. (Thanks to
2234
Juan Jose Garcia Ripoli).
2235
- Use LINUX_STACKBOTTOM for Linux/SH and LINUX/ARM. (Thanks to Sugioka
2236
Toshinobu and Christian Thalinger.)
2237
- Rewrote GC_parse_map_entry. This assumed a fixed column layout of
2238
/proc/self/maps on Linux. This ceased to be true about 2 years ago.
2239
The old code is probably quite problemetic with -DREDIRECT_MALLOC. It
2240
is also used by default for IA64, though I haven't seen actual failures
2242
- More consistently define HBLKSIZE to 4096 on 64 bit architectures with
2243
4K pages. (Thanks to Andrew Haley.)
2244
- With win32 threads, GC_stop_world needs to acquire GC_write_cs. (Thanks
2245
to Ben Hutchings for the observation and patch.)
2246
- Move up struct callinfo declaration to make gcc 4.0.2. happy.
2249
- Add "int" to Solaris "end" and "etext" declaration in gc.h. Declared
2250
the symbols with underscores and as arrays, since that's what's actually
2251
used. Perhaps this could all just be removed? (Thanks to John Bowman.)
2252
- Fixed ARM GC_test_and_set code. (Thanks to Kazu Hirata and Paul Brook.)
2253
- Added casts for assignments to hb_last_reclaimed, which truncate the
2254
value. Added a cast to GC_adj_words_allocd. Use GetModuleHandleA
2255
when retrieving a handle to kernel32.dll under win32. (Thanks to the
2256
Visual Prolog developers.)
2257
- Added Tandem S-Series support. (Thanks to Craig McDaniel. A modified
2258
version of his patch was applied, and hence breakage is probably not
2260
- Remove spurious gc:: qualifier for operator delete[] in gc_cpp.h.
2261
(Thanks to Hanno Boeck.)
2262
- Changed a test for LINUX in config_macros.h to one for __linux__.
2263
- Fix ppc 64 test_and_set code by removing it. (Thanks to Christian
2265
- Add prototypes for GC_finalizer_notifier and GC_thr_init. (Thanks to
2267
- Use ld instead of nonexistent ldz instruction in Darwin FindTopOfStack.
2268
(Thanks to Andreas Tobler.)
2269
- Add support for Darwin/X86. (Thanks to Geoff Norton and the Mono
2271
- Merge in some recent gcc fixes. Add ppc64 asm code. (Thanks to Bryce
2272
McKinley and other gcj developers.)
2273
- Scan MEM_PRIVATE sections under Windows ME and predecessors.
2274
- Interior pointers with some largish offsets into large objects could
2275
be ignored, if GC_all_interior_pointers was set. (Oddly this worked
2276
correctly for stack references if it was not set. Otherwise it failed
2277
for both stack and heap references.) Thanks to Andrew McKinlay for the
2279
- Integrated Tatsuya Bizenn's NETBSD threads support, with some
2280
minimally tested changes.
2281
- Added GC_strdup and friends to make leak detection work correctly
2282
for strdup clients. (Thanks to Jon Moore.) Fixed the existing strdup
2283
with malloc redirection to handle a null malloc return correctly.
2284
- Fix Makefile.am, so it handles exe extensions under Cygwin correctly
2225
2288
- The USE_MUNMAP code should really use a separate data structure
2226
2289
indexed by physical page to keep track of time since last use of
added
removed
src/gc/doc/README.changes
