| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
| |
This completes the workaround for bfs_spawn() hanging on FreeBSD with
ASan enabled.
Link: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=280318
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This should have been benign (just a performance regression), but
FreeBSD has a bug that causes the fork()-based fallback to lock up when
ASAN is enabled, due to fork() and dl_iterate_phdr(). This is not a
complete workaround for that issue, since the fork() fallback may be
used for other reasons.
Fixes: 29ddac2 ("config: Check for posix_spawn_file_actions_addfchdir{,_np}()")
|
| |
|
|
|
|
|
|
| |
This reverts commit 4f80c17192f2b28c96a489969d4435151d68d0ce.
Link: https://www.austingroupbugs.net/view.php?id=1133
|
|
|
|
|
|
|
|
| |
This fixes an ABA problem where sigdispatch() could think no handlers
are registered for a signal even when there are.
Link: https://unix.stackexchange.com/a/779594/56202
Fixes: 375caac ("sighook: New utilities for hooking signals")
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
It should be at least half the max char distance so that we mostly get
replacements, not inserts + deletes.
|
|
|
|
|
|
|
|
|
|
| |
Not all posix_spawn() implementations use errno to report execv()
failures from the child process, as that requires either a kernel
posix_spawn() implementation or a pipe to pass the error back.
This should fix tests/posix/exec_nonexistent on OpenBSD and HPPA.
Link: https://buildd.debian.org/status/fetch.php?pkg=bfs&arch=hppa&ver=3.3.1-1&stamp=1717489148&raw=0
|
| |
|
| |
|
|
|
|
| |
This fixes bfs under Valgrind, which reserves SIGRTMAX for its own use.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Prior to GCC 14.1, the __builtin_riscv_pause() can cause an error if the
appropriate extension is not enabled in -march:
/tmp/ccR1L1lA.s: Assembler messages:
/tmp/ccR1L1lA.s:670: Error: unrecognized opcode `pause', extension `zihintpause' required
Link: https://gcc.gnu.org/pipermail/gcc-patches/2023-August/626748.html
Link: https://gcc.gnu.org/git/?p=gcc.git;a=commit;h=c2d04dd659c499d8df19f68d0602ad4c7d7065c2
Link: https://buildd.debian.org/status/fetch.php?pkg=bfs&arch=riscv64&ver=3.3.1-1&stamp=1717488400&raw=0
|
| |
|
| |
|
| |
|
| |
|
| |
|
|\ |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
POSIX requires an error if (for example) -L encounters a symlink loop.
The GNU find change was restricted to -xtype, so add a manual ELOOP test
to eval_xtype() for compatibility.
This reverts commit 470589cbd9ca3e73d8c01ac3a96cbc065179dcc5.
Link: https://savannah.gnu.org/bugs/?19605
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The POSIX Utility Syntax Guidelines specify that flag groups like -HL
should be handled like -H -L. GNU find doesn't support grouping flags
in this way, but BSD find does.
To avoid conflicts with non-flag primaries, for now we require at least
one flag in a group to be a capital letter. That is, we support things
like -Lds but not -ds. We also do not support -fPATH (without a space)
as it would conflict with -follow, -fprint, etc. It is impossible to be
compatible with both GNU and BSD find here:
user@gnu$ find -follow
link
link/file
...
user@bsd$ find -follow
find: ollow: No such file or directory
Link: https://pubs.opengroup.org/onlinepubs/9699919799/basedefs/V1_chap12.html
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
glibc added an implementation in 2.39.
|
|/
|
|
|
|
|
|
|
|
| |
Based on the latest C23 draft (N3220):
- Argument types to generic bit functions should be unsigned
- Bit functions return unsigned int
- Byte-swapping functions (stdc_memreverse8*()) weren't added
- stdc_rotate_{left,right}() weren't added
- first_leading_*() counts from the *left*
|
|
|
|
| |
Closes: https://github.com/tavianator/bfs/issues/21
|
|
|
|
|
| |
Link: https://savannah.gnu.org/bugs/index.php?65770
Link: https://github.com/kkos/oniguruma/issues/296
|
|
|
|
|
| |
Other utilities on FreeBSD like dd, sleep, etc. that respond to SIGINFO
still print the kernel info line too.
|
| |
|
|
|
|
|
|
|
| |
GNU findutils just made the same change, so looping symlinks will be
considered broken rather than a hard error.
Link: https://savannah.gnu.org/bugs/?51926
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
| |
If we raise RLIMIT_NOFILE, we have to lower it before calling exec() for
compatibility with select(). If posix_spawn() doesn't support that, we
fall back to fork(), which is quite a bit slower.
Therefore, if we're going to exec() on most files, it's better to keep
RLIMIT_NOFILE the same to avoid the fork() cost, even though it makes
bftw() somewhat slower.
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
This has been fixed in Cosmopolitan.
Link: https://github.com/jart/cosmopolitan/issues/944
|
| |
|