summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* wordesc: Don't allow braces in bare wordsTavian Barnes2023-10-181-1/+1
| | | | Things like {a,b} should be quoted to avoid brace expansion.
* parse: Set BFTW_WHITEOUTS when parsing -type wTavian Barnes2023-10-173-0/+105
|
* bftw: New flag to control whiteout visibilityTavian Barnes2023-10-173-2/+19
|
* dir: New flag to control whiteout visibilityTavian Barnes2023-10-172-3/+13
|
* dir: Fix FreeBSD union mountsTavian Barnes2023-10-175-8/+74
|
* dir: Add a flags parameter to bfs_opendir()Tavian Barnes2023-10-176-21/+45
|
* tests: Move closed_std* tests out of the POSIX groupTavian Barnes2023-10-146-2/+8
| | | | | | | | | | | | | POSIX actually says > If the utility would be executed with file descriptor 0, 1, or 2 > closed, implementations may execute the utility with the file > descriptor open to an unspecified file. So we're not guaranteed to be able to detect the situation in the first place. Add a best-effort check for these platforms and skip the test. Link: https://pubs.opengroup.org/onlinepubs/9699919799/utilities/V3_chap02.html#tag_18_09_01_01
* dir: Fix Android buildTavian Barnes2023-10-141-1/+1
| | | | Android doesn't provide a getdents() wrapper, so use the syscall.
* docs: Add a list of related utilitiesTavian Barnes2023-10-131-0/+42
|
* tests: Always create the scratch directoryTavian Barnes2023-10-131-0/+2
| | | | | | | | | | | | | | | | Otherwise, some tests that try to check for flag support like invoke_bfs scratch -quit -xattr || skip can be accidentally skipped if ./scratch doesn't exist yet: $ ./tests/tests.sh bsd/xattr tests skipped: 1 After this patch, we get the expected $ ./tests/tests.sh bsd/xattr tests passed: 1
* Merge tag '3.0.4'Tavian Barnes2023-10-125-4/+16
|\
| * Release 3.0.43.0.4Tavian Barnes2023-10-123-2/+14
| |
| * tests: Don't crash if unbuffer isn't installedTavian Barnes2023-10-121-1/+1
| |
| * bfstd: Actually use the copied localeTavian Barnes2023-10-121-1/+1
| | | | | | | | This fixes a segfault in xstrerror() when using musl.
* | build: Re-enable liburingTavian Barnes2023-10-121-1/+1
|/ | | | This reverts commit 735a2979a37251dc447274aa888ad77ded126457.
* Release 3.0.33.0.3Tavian Barnes2023-10-123-13/+5
|
* build: Disable liburing for nowTavian Barnes2023-10-121-1/+1
| | | | liburing will be enabled for the next feature release.
* docs: Start writing the 3.1 changelogTavian Barnes2023-10-121-0/+33
|
* bench: Look for ./bfs if ./bin/bfs doesn't existTavian Barnes2023-10-121-1/+5
|
* bench: Check for bfs -j support before using itTavian Barnes2023-10-121-3/+13
|
* bftw: Make sure we don't close a directory while we unwrap itTavian Barnes2023-10-121-2/+6
| | | | | | | bftw_cache_reserve() can lead to bftw_cache_pop(), which could close the directory we're trying to unwrap! If that happened, we would try dup_cloexec(-1), which would fail with EBADF, so there was no observable bug. But it's better to avoid the whole situation.
* bftw: Fix to_close list corruption with !BFS_USE_UNWRAPDIRTavian Barnes2023-10-121-6/+13
| | | | | | | | | | It's possible for pincount to drop to zero, then get incremented and drop back to zero again. If this happens, we shouldn't add it to the to_close list twice. This should fix the intermittent hang on the macOS CI. Fixes: 815798e1eea7fc8dacd5acab40202ec4d251d517
* list: Assert that we're not inserting already-attached nodesTavian Barnes2023-10-122-21/+46
|
* diag: Move enum debug_flags out of ctx.hTavian Barnes2023-10-125-53/+55
|
* bftw: Don't force buffering for parallel dfsTavian Barnes2023-10-121-5/+30
|
* bftw: Fix unbuffered depth-first searchesTavian Barnes2023-10-121-15/+41
| | | | | | | | | | | | | | | bftw() implements depth-first search by appending files to a batch, then prepending the batch to the queue. When we switched to separate file/ directory queues, this was only implemented for the file queue. Unbuffered searches don't use the file queue, so they were all breadth- first in practice. This meant that iterative deepening (-S ids) was actually "iterative deepening *breadth*-first search," a horrible strategy with no advantage over regular breadth-first search. Now it performs iterative deepening *depth*-first search, which at least limits its memory consumption. Fixes: c1b16b49988ecff17ae30978ea14798d95b80018
* bftw: Let iterative deepening work depth-first when sortingTavian Barnes2023-10-121-1/+1
|
* bench: Default to --strategies=rustTavian Barnes2023-10-101-1/+1
|
* bench: New --jobs benchmarkTavian Barnes2023-10-101-4/+64
|
* bench: Group by strategy in comparison modeTavian Barnes2023-10-101-4/+13
|
* bench: Don't print paths in the --strategies benchmarkTavian Barnes2023-10-101-1/+1
|
* bench: Allow format specifiers in (sub)group headersTavian Barnes2023-10-101-10/+10
|
* bar: Replace volatile sig_atomic_t with actual atomicsTavian Barnes2023-10-061-8/+12
|
* alloc: Test allocation size overflowsTavian Barnes2023-10-062-4/+12
|
* tests/bfs/status: Try to test SIGWINCHTavian Barnes2023-10-061-1/+1
|
* tests: Check for expect_unbuffer tooTavian Barnes2023-10-061-2/+8
| | | | That's where FreeBSD installs it.
* tests: New bfs_pty wrapper for unbufferTavian Barnes2023-10-064-7/+17
|
* ci: Install expect for unbufferTavian Barnes2023-10-052-0/+6
|
* tests/bfs: Add a test that runs -statusTavian Barnes2023-10-051-0/+3
|
* tests/bfs: Add tests for color autodetectionTavian Barnes2023-10-054-0/+65
|
* tests: Respect $NO_COLORTavian Barnes2023-10-051-9/+15
|
* README: Realign badgesTavian Barnes2023-10-051-4/+3
| | | | A recent GitHub change broke the nice alignment of the badges to the top heading, so just put the badges underneath it for now.
* bfstd: Add a thread-safe wrapper for strerror()Tavian Barnes2023-10-0511-13/+69
|
* thread: Define thread_localTavian Barnes2023-10-052-0/+16
|
* config: Fold !__FreeBSD__ into BFS_USE_SYS_CAPABILITY_HTavian Barnes2023-10-052-2/+2
|
* dstring: Limit the special dchar typedef to lint buildsTavian Barnes2023-10-041-7/+2
| | | | | Clang still thinks that alignof(dstr[1]) == 2, so out of an abundance of caution, don't mess with dchar alignment in normal builds.
* build: New lint flagTavian Barnes2023-10-041-1/+8
|
* thread: s/call_once/invoke_once/Tavian Barnes2023-10-033-3/+3
| | | | call_once() is a reserved identifier from C11.
* ci/freebsd: Use tailscale sshTavian Barnes2023-10-031-10/+3
|
* ci/macos: Don't install coreutilsTavian Barnes2023-10-031-4/+0
| | | | | Since we build our own touch(1) implementation, we no longer need to work around the macOS bug.