Timeline
Mar 8, 2006:
- 1:17 AM Changeset [2580] by
- OS/2 doesn't do this properly.
- 1:04 AM Ticket #49 (libc: st_ctime is incorrect and st_birthtime isn't converted to UTC) closed by
- fixed: The former is a stat backend bug. Both fixed.
- 1:03 AM Changeset [2579] by
- #49: Fixed stat backend so ctime is set, and made fstat, lstat and …
- 12:48 AM Changeset [2578] by
- skip hardlink tests if not supported.
- 12:30 AM Ticket #61 (libc: symlink(,existing) returns the wrong errno) closed by
-
fixed: This broke
touch bar ; ln -sf foo bar. - 12:30 AM Changeset [2577] by
- #61: Fixed incorrect ENOENT errno on symlink when it should return EEXIST.
- 12:18 AM Ticket #61 (libc: symlink(,existing) returns the wrong errno) created by
- DosOpen fails with 110, which is converted to ENOENT. It should be EEXIST.
Mar 7, 2006:
- 9:47 PM Changeset [2576] by
- binary mode hacking on OS/2
- 9:46 PM Changeset [2575] by
- binary mode hacking on OS/2
- 9:11 PM Changeset [2574] by
- removed generated stuff.
- 8:08 PM Changeset [2573] by
- set mime to application/octet-stream because svn doesn't do eol style …
- 7:44 PM Changeset [2572] by
- svn:eol-style=LF
- 7:17 PM Changeset [2571] by
- dos \r\n
- 2:55 AM Ticket #60 (libc: filehandle inheritance doesn't work right!) closed by
- fixed: This is probably what's been responsible for some of the really weird …
- 2:54 AM Changeset [2570] by
- #60: Fixed blunder in exec/spawn filehandle inheritance code.
- 2:43 AM Ticket #60 (libc: filehandle inheritance doesn't work right!) created by
- There is a bug in the file handle inheritance code (exec/spawn) where …
- 1:23 AM Changeset [2569] by
- native.
- 1:22 AM Changeset [2568] by
- skip hardlink testcases if hardlinks aren't supported.
- 1:20 AM Changeset [2567] by
- Fixed emx perl problems.
- 1:03 AM Changeset [2566] by
- line ending fuckup.
- 12:59 AM Changeset [2565] by
- line ending fuckup.
- 12:50 AM Changeset [2564] by
- skip hardlink testcases if hardlinks aren't supported.
- 12:49 AM Changeset [2563] by
- skip hardlink testcases if hardlinks aren't supported.
- 12:48 AM Changeset [2562] by
- initial hacking.
Mar 6, 2006:
- 11:59 PM Changeset [2561] by
- rewinddir is broken in EMX, avoid it and do close/opendir instead.
- 11:35 PM Ticket #59 (libc: rewinddir) created by
- Rewinding a directory doesn't read the directory again, it simply …
- 7:02 PM Changeset [2560] by
- Initial porting.
- 1:38 AM Changeset [2559] by
- diffutils 2.8.1
- 1:21 AM Changeset [2558] by
- grep 2.5.1a
- 1:07 AM Changeset [2557] by
- grep 2.5.1a
- 12:55 AM Changeset [2556] by
- diffutils 2.8.1
- 12:38 AM Changeset [2555] by
- coreutils 5.94
- 12:25 AM Changeset [2554] by
- coretuils-5.94
Mar 5, 2006:
- 12:48 AM Ticket #55 (emxomfld doesn't delete temporary files) closed by
- fixed: Applied.
- 12:47 AM Changeset [2553] by
- #55: delete the response file when reinit the args.
- 12:20 AM Ticket #56 (emxomfld doesn't properly allocate space for map file name) closed by
- invalid: This was a fix for you own bug in #46 which wasn't committed.
- 12:16 AM Ticket #46 (ilink map file name is wrong when multiple dots are in target name) closed by
- fixed: if the filename was wihtout an extension the strcat may cause a …
- 12:13 AM Changeset [2552] by
- #46: real fix.
- 12:01 AM Changeset [2551] by
- Applies #46.
Mar 4, 2006:
- 10:45 PM Ticket #54 (libc: fchdir(open('.', O_RDONLY)) may corrupt path resolving) closed by
- fixed
- 1:47 AM Changeset [2550] by
- baselayout hacking.
- 1:35 AM Changeset [2549] by
- my good old ports stuff.
Mar 2, 2006:
- 2:30 AM Ticket #58 (libc: closed standard handles cause uninitialized stdout/stdin/stderr ...) closed by
- fixed: The stdin/stdout/stderr streams should apparently always be …
- 2:26 AM Changeset [2548] by
- #58: The standard file handles are always initialized as open.
- 2:24 AM Changeset [2547] by
- fclose / flush testcase.
- 1:20 AM Ticket #58 (libc: closed standard handles cause uninitialized stdout/stdin/stderr ...) created by
- smoketests/fclose-1.c exhibit a problem where the parent closes one or …
- 1:16 AM Changeset [2546] by
- fclose / flush testcase.
Mar 1, 2006:
Feb 21, 2006:
- 9:53 PM Ticket #57 (Dynamic libraries cannot be freed when sockets (at least) are used) closed by
- invalid: This is not an error, and it has nothing to do with GCC. This problem …
- 7:13 PM Ticket #57 (Dynamic libraries cannot be freed when sockets (at least) are used) created by
- When a dll uses a socket function (but not only in this case), the dll …
Feb 13, 2006:
- 1:51 PM Ticket #56 (emxomfld doesn't properly allocate space for map file name) created by
- When the output file name doesn't include the extension (this is not …
- 1:49 PM Ticket #55 (emxomfld doesn't delete temporary files) created by
- when arg_init() is called, the temporary files is not deleted, so at …
Feb 7, 2006:
- 6:08 AM Changeset [2545] by
- #54: Fixed unitialized directory handle member fInUnixTree.
- 6:02 AM Ticket #54 (libc: fchdir(open('.', O_RDONLY)) may corrupt path resolving) created by
- dirOpen isn't initializing the fInUnixTree member, and a fchdir …
- 5:28 AM Ticket #53 (libc: freopen("missing",,) causes panic) closed by
- fixed: Hmm.. incomplete fix here, _openstream got a lock bool argument which …
- 5:28 AM Changeset [2544] by
- #53: Fixed LIBC panic when _openstream locks the stream vector in any …
- 5:17 AM Ticket #53 (libc: freopen("missing",,) causes panic) created by
- in _fopen there are at least two fialure paths which are locking and …
- 4:12 AM Ticket #52 (libc: path resolver fails resolving a directory symlink for lstat) closed by
- fixed: This was actually two symlink bugs, the first one hiding the 2nd. Both …
- 4:10 AM Changeset [2543] by
- #52: Fixed two 'symlink/' problems in the path resolver.
- 3:36 AM Ticket #52 (libc: path resolver fails resolving a directory symlink for lstat) created by
- […]
- 3:25 AM Ticket #51 (libc: unlink fails on readonly file) closed by
- fixed: We also ended up with the wrong rc. Corrected the code such that it …
- 3:24 AM Changeset [2542] by
- #51: Made unlink able to delete readonly files.
- 2:57 AM Ticket #51 (libc: unlink fails on readonly file) created by
- rm from coreutil fails to delete read-only files. Will change the code …
- 1:55 AM Changeset [2541] by
- Rewrote the freopen(NULL) code to use fcntl(,F_SETFL,) for updating …
- 12:32 AM Changeset [2540] by
- #50: Inherit umask.
- 12:30 AM Changeset [2539] by
- #50: Inherit umask.
Feb 6, 2006:
- 11:13 PM Ticket #50 (libc: umask isn't inherited) closed by
- fixed: Implemented umask inhertiance. The testcase works.
- 11:12 PM Changeset [2538] by
- #50: Inherit umask.
- 11:10 PM Changeset [2537] by
- #30: Fixed toupper / tolower EOF regression. (forgot to commit it)
- 10:05 PM Ticket #50 (libc: umask isn't inherited) created by
- The last test in the 'equal' chmod testcase (coreutils) is apparently …
- 6:35 AM Ticket #49 (libc: st_ctime is incorrect and st_birthtime isn't converted to UTC) created by
- Tittle says it all. coreutils stat shows unix epoch for ctime (change …
- 6:27 AM Ticket #48 (libc: freopen(NULL, "rb", stdio) panics (doesn't work as expected)) closed by
- fixed: Did a quick implenentation of this NULL feature.
- 6:22 AM Changeset [2536] by
- #48: Quick implementation of the freopen(NULL) operation to make cat …
- 4:48 AM Ticket #48 (libc: freopen(NULL, "rb", stdio) panics (doesn't work as expected)) created by
- If filename is a null pointer, the freopen() function shall attempt to …
Feb 5, 2006:
- 4:26 PM Ticket #47 (Create .sym files directly from emxomfld) created by
- This patch allows emxomfld to produce a .sym file directly when -Zsym …
- 3:41 PM Ticket #46 (ilink map file name is wrong when multiple dots are in target name) created by
- When the target name has multiple dots (e.g. my.code.dll), ilink will …
- 1:21 PM Ticket #45 (libc: speed up path resolving) created by
- There are several tricks that should be used in the unix path resolver …
- 10:20 AM Ticket #44 (libc: statvfs reports 14 chars max filename length everywhere) created by
- I think this is a left over from the DOS days. We should return 255 on …
- 7:55 AM Changeset [2535] by
- memalign and valloc shouldn't be in USE_GNU in malloc.h.
- 7:28 AM Changeset [2534] by
- 0.6.1 notes.
- 7:02 AM Ticket #43 (libc: Runtime configuration of where to use UnixEAs) created by
- This is a followup to #21. Should add some kind of env.var. for …
- 6:47 AM Changeset [2533] by
- #30: Fixed toupper / tolower EOF regression. (test_ctype.c found it).
- 5:53 AM Ticket #42 (libc: Add logging option for not creating the log file until it's ...) created by
-
Even with
LIBC_LOGGING=-allwe get log files with the default … - 5:42 AM Changeset [2532] by
- fixed the test to check for type too and ignore the readonly problem …
- 5:33 AM Changeset [2531] by
- #41: typo breaking fchmod().
- 5:13 AM Changeset [2530] by
- Fixed incorrect assertion in libc_back_fsInfoObjRelease.
- 5:07 AM Changeset [2529] by
- Fixed incorrect assertion in libc_back_fsInfoObjRelease.
- 4:57 AM Changeset [2528] by
- #41: And added special file mode mask getter and extra validations to …
- 4:48 AM Changeset [2527] by
- added symlink-1.
- 4:40 AM Changeset [2526] by
- #41: And added special file mode mask getter and extra validations to …
- 3:54 AM Ticket #41 (libc: incorrect __libc_back_fsUnixAttribsGet() usage caused st_dev ...) closed by
- fixed: That's not quite right. the fstat backend is delaying the dev/ino calc …
- 3:54 AM Changeset [2525] by
- #41: Calc st_ino and st_dev at the end of the unix attrib getter if …
- 3:13 AM Ticket #41 (libc: incorrect __libc_back_fsUnixAttribsGet() usage caused st_dev ...) created by
- we're not checking if we actually read any UnixEAs in …
- 3:11 AM Ticket #40 (libc: open("/", RDONLY) fails.) closed by
- fixed: That's it. fixed.
- 3:11 AM Changeset [2524] by
- #40: Fixed double slash preventing root directories from being opened.
- 3:00 AM Ticket #40 (libc: open("/", RDONLY) fails.) created by
- We fail to open the root directory, probably because of the trailing '/'.
- 2:56 AM Changeset [2523] by
- 2:54 AM Ticket #21 (libc: limit Unix EAs to specific file systems) closed by
- fixed: We're now only reading and writing UnixEAs on HPFS, JFS, FAT and RAMFS …
- 2:53 AM Changeset [2522] by
- #21: Restrict reading and writing of UnixEAs to HPFS, JFS, FAT and RAMFS.
- 12:05 AM Ticket #36 (emxbind: export truncation bug and allow export & imports up to 255 chars) closed by
- fixed: The export.c patch is quite obviously wrong. I've changed it to emit a …
- 12:05 AM Changeset [2521] by
- o #38: Fixed truncation bug writing to the LX nametable. Imports and …
Feb 4, 2006:
- 11:09 PM Ticket #35 (ld crashes when map files are enabled) closed by
- duplicate: Duplicate of #27.
- 8:14 PM Ticket #39 (dllar.cmd enhancements) created by
- This patch allows dllar.cmd to produce also omf import libraries and …
- 7:10 PM Changeset [2520] by
- #38: Added DosSysCtl prototype and use it on panic to check for debugger.
- 7:09 PM Ticket #38 (libc: when debugged default to breakpoint on panic) closed by
- fixed: Implemented.
- 7:08 PM Changeset [2519] by
- #38: Added DosSysCtl prototype and use it on panic to check for debugger.
- 7:04 PM Ticket #38 (libc: when debugged default to breakpoint on panic) created by
- There is a DosSysCtl function for detecting whether we're being …
- 7:00 PM Changeset [2518] by
- #33: Fixed a runpath in libc_back_signalWait which didn't reacquire …
- 6:57 PM Ticket #33 (libc: panic when using timers) closed by
- fixed: There was a return path in libc_back_signalWait which didn't …
- 6:38 PM Ticket #37 (export for symbols longer than 255 chars) created by
- emxbind cannot parse .def files with symbols longer than 255 chars. …
- 6:32 PM Ticket #36 (emxbind: export truncation bug and allow export & imports up to 255 chars) created by
- Symbols longer than 127 bytes are truncated, code removed appear bad …
- 6:26 PM Ticket #35 (ld crashes when map files are enabled) created by
- ld crashes when map file name differs from module name, because of …
- 5:26 PM Ticket #34 (weakld: generate duplicate weak aliases) closed by
- fixed: The hashing changes in emxomf broke some silly attempt at keeping …
- 5:25 PM Changeset [2517] by
- #34: Removed all the silliness trying to deadl with truncated symbols.
- 4:35 PM Ticket #34 (weakld: generate duplicate weak aliases) created by
- weakld sometimes (wonder if symbol hashing is required) outputs more …
- 1:26 PM Changeset [2516] by
- #25: Ensure correct address length returns from recvmsg and recvfrom.
- 12:45 PM Changeset [2515] by
- #32: Fixed incorrect readdir_r return code when out of files.
- 12:43 PM Ticket #32 (libc: readdir_r returns -1 instead of errno) closed by
- fixed: There was already a todo comment on the problem. I've just changed it …
Feb 3, 2006:
- 10:52 PM Ticket #20 (libc: random panic SIGSEGV when doing parallel builds) closed by
- fixed: Rewritten code using mktemp. The problem should be gone now.
- 10:51 PM Changeset [2514] by
- Bug 20, fixed: use mkstemp + close instead of mktemp for the response file.
- 10:48 PM Changeset [2513] by
- Bug 20, fixed: use libiberty make_temp_file instead of big, clumsy and …
- 10:30 PM Changeset [2512] by
- Bug 20, fixed: use libiberty make_temp_file instead of big, clumsy and …
Feb 1, 2006:
- 7:46 PM Ticket #33 (libc: panic when using timers) created by
- With IPRT tstTimer I get a LIBC PANIC! on the release of a mutex …
- 7:37 PM Ticket #32 (libc: readdir_r returns -1 instead of errno) created by
- Title says it all.
Jan 22, 2006:
Jan 19, 2006:
- 3:47 AM Ticket #30 (setlocale / ctype is broken.) closed by
- fixed
- 3:46 AM Ticket #31 (libc: is*() testcase) created by
- Need a proper isxxxx() testcase so locale bustage can be found earlier.
- 3:44 AM Changeset [2511] by
- Fixes #30: Fixed two locale bugs. First, setlocale called a locale …
- 2:11 AM Ticket #30 (setlocale / ctype is broken.) created by
-
We're using
___wctype()which is checking the globals instead …
Jan 16, 2006:
Jan 15, 2006:
- 5:40 AM Ticket #28 (emxbind doesn't read library name from .def file) closed by
- fixed
- 5:40 AM Changeset [2510] by
- Use DLL name from the .DEF file when present. Fixes #28.
- 5:32 AM Ticket #27 (ld bug when map files are requested) closed by
- fixed
- 5:32 AM Changeset [2509] by
- Applied fix from Yuri. Fixes #27.
Jan 14, 2006:
- 11:13 AM Ticket #28 (emxbind doesn't read library name from .def file) created by
- The module name used by emxbind when building dynamic libraries is not …
Jan 11, 2006:
- 5:27 PM Ticket #27 (ld bug when map files are requested) created by
- In write_output(), map file name overwrites memory because of wrong …
Jan 7, 2006:
Jan 2, 2006:
- 1:13 PM Ticket #26 (using gcc -h* gives fatal error) created by
- Hi, using commandlines like gcc -h* or gcc --help=topic …
Dec 19, 2005:
- 2:50 PM Ticket #25 (libc: recvmsg/recvfrom doesn't set correct address length) created by
- The OS/2 tcpip stack is very picky about address sizes (see mozilla …
Dec 18, 2005:
- 9:10 PM Ticket #19 (gcc: gcc limits.h overrides limits.h from libc) closed by
- fixed: The PATH wasn't the clue. The problem was in t-emx, first we used a PC …
- 9:10 PM Changeset [2508] by
- Ref #19 - testcase.
- 8:53 PM Changeset [2507] by
- Reorganization continues.
- 8:49 PM Changeset [2506] by
- Reorganization continues.
- 8:47 PM Changeset [2505] by
- Reorganization continues.
- 8:43 PM Changeset [2504] by
- Reorganization continues.
- 8:36 PM Changeset [2503] by
- Reorganization continues.
- 8:33 PM Changeset [2502] by
- And we *need* STMP_FIXINC because of limits.h depends on syslimits.h …
- 6:03 PM Changeset [2501] by
- NATIVE_SYSTEM_HEADER_DIR must be all forward slashes!
- 2:31 PM Changeset [2500] by
- limits.h fix attempt (gcc forward slashes first).
- 1:41 PM Changeset [2499] by
- #23: Corrected EA name in template from FLAGS to SYMLINK.
- 1:40 PM Ticket #23 (libc: symlink is broken) closed by
- fixed: Wrong symbol name in the template structure ("FLAGS" instead of …
- 1:40 PM Changeset [2498] by
- #23: Corrected EA name in template from FLAGS to SYMLINK.
- 12:31 PM Ticket #24 (libc: fts crashes because it's still using chdir()) closed by
- fixed: Fixed in changeset [2497].
- 12:25 PM Changeset [2497] by
- Fixes #24 - chdir() crash in fts.c. - dirfd() problem in fts.c which …
- 10:17 AM Ticket #24 (libc: fts crashes because it's still using chdir()) created by
- The various fts APIs when enumerating directories because HAVE_FCHDIR …
- 9:57 AM Ticket #22 (libc: path resolver fails on open file) closed by
- fixed
- 9:56 AM Changeset [2496] by
- #22: Added DosFindFirst() as fallback when DosQueryPathInfo() fails to …
- 9:09 AM Changeset [2495] by
- Overlooked one place where the minor version was needed.
- 8:43 AM Changeset [2494] by
- o Don't display usage() on failure, just the error message.
- 8:17 AM Changeset [2493] by
- 0.6.1
- 7:50 AM Changeset [2492] by
- Created the libc06 forwarder dll.
Dec 17, 2005:
- 6:30 AM Ticket #23 (libc: symlink is broken) created by
- When porting BSD ln for using kBuild I find that symlink isn't …
- 5:35 AM BuildLibc edited by
- (diff)
Dec 15, 2005:
Dec 14, 2005:
- 9:02 PM Ticket #22 (libc: path resolver fails on open file) created by
- pwd_mkdb (baselayout) has trouble doing stat() on the temporary …
- 7:09 PM Ticket #21 (libc: limit Unix EAs to specific file systems) created by
- Fat32 does not implement EAs efficiently, and some other IFSes doesn't …
- 7:04 PM Ticket #20 (libc: random panic SIGSEGV when doing parallel builds) created by
- Two or more parallel builds of mozilla is reported causing random …
- 6:42 AM Changeset [2491] by
- Corrected directory deps.
Dec 13, 2005:
Dec 12, 2005:
- 5:30 AM Changeset [2490] by
- We cannot rename open files.
- 5:28 AM Changeset [2489] by
- FreeBSD 5.4. (src/lib/libc/gen/pw_scan.h)
- 5:26 AM Changeset [2488] by
- FreeBSD 5.4.
- 5:23 AM Changeset [2487] by
- FreeBSD 5.4.
- 5:22 AM Changeset [2486] by
- adding baselayout
- 5:22 AM Changeset [2485] by
- adding baselayout
Dec 11, 2005:
- 8:10 AM Changeset [2484] by
- kBuild development.
- 3:49 AM Changeset [2483] by
- #19: Added missing [U]LONG_LONG_MAX/MINs since gcc defines these.
- 3:47 AM Changeset [2482] by
- #19: Added missing [U]LONG_LONG_MAX/MINs since gcc defines these.
- 3:34 AM Ticket #19 (gcc: gcc limits.h overrides limits.h from libc) created by
- GCC has it's own limits.h which hides the limits.h in usr/include …
Note:
See TracTimeline
for information about the timeline view.
