aboutsummaryrefslogtreecommitdiffhomepage
path: root/changelog.md
Commit message (Collapse)AuthorAge
* Update changelog and increment package versionGravatar Herbert Valerio Riedel2017-05-01
|
* Prepare for 2.7.2.1 releaseGravatar Herbert Valerio Riedel2016-11-12
|
* Don't use readdir_r if deprecatedGravatar Erik de Castro Lopo2016-09-07
| | | | | | | | | | | | | | | GNU glibc 2.23 and later deprecate `readdir_r` in favour of plain old `readdir` which in some upcoming POSIX standard is going to required to be re-entrant. Eventually we want to drop `readder_r` all together, but want to be compatible with older unixen which may not have a re-entrant `readdir`. Solution is to make systems with *known* re-entrant `readir` use that and use `readdir_r` whereever we have it and don't *know* that `readdir` is re-entrant. Closes: https://github.com/haskell/unix/issues/70
* Update changelog for 2.7.2.0 releaseGravatar Herbert Valerio Riedel2016-04-19
|
* Bump to 2.7.2.0Gravatar Herbert Valerio Riedel2016-01-31
| | | | | | There's been some subtle API changes that may be useful to detect via `MIN_VERSION_unix(2,7,2)`, consequently perform a minor version bump relative to the last release.
* Change `drainOutput`'s `tcdrain(3)` into a `safe` FFI callGravatar Herbert Valerio Riedel2016-01-30
| | | | | Since the primary purpose of `tcdrain(3)` is to block it makes much more sense to use a `safe` FFI import.
* Don't assume `tcdrain` and `ctermid` exist alwaysGravatar Herbert Valerio Riedel2016-01-30
| | | | | | This follows the scheme suggested in #24 This fixes #55
* Improve detection of `fdatasync(2)`Gravatar Herbert Valerio Riedel2015-12-16
| | | | | | | | This attempts a simpler and hopefully more robust test for `fdatasync(2)` See 94d8824bae10c9d91f56c1aee9c45a90136a1770 / #42 for the previous attempt. This hopefully addresses #52 in a better way than #53
* Kill `-fwarn-unused-imports`-induced sillinessGravatar Herbert Valerio Riedel2015-11-16
| | | | | | | | | This all started in e968172cb0c9f4fc653c775faf3ecb661f5b1948 which tried to get rid of redundant import warnings. But we should rather err on having to tolerate a few warnings on exotic configurations rather than breaking a build... This hopefully fixes #36
* Don't assume existence of termios constants beyond `B38400`Gravatar Herbert Valerio Riedel2015-11-16
| | | | | | | | | | This patch does not modify the `BaudRate` structure but rather causes some functions taking a `BaudRate` to throw exceptions on systems which don't provide `B57600` and/or `B115200`. I'm not totally happy with this patch yet, but this unbreaks compilation on platforms which don't define at least of these constants. This fixes #51
* Don't assume non-POSIX `WCOREDUMP(x)` macro existsGravatar Herbert Valerio Riedel2015-11-16
| | | | This fixes #50
* Prepare for 2.7.1.0 releaseGravatar Herbert Valerio Riedel2014-12-19
|
* Tweak markdown markup again [skip ci]Gravatar Herbert Valerio Riedel2014-12-19
|
* Fix markdown markup [skip ci]Gravatar Herbert Valerio Riedel2014-12-19
|
* Fix SIGINFO and SIGWINCH.Gravatar Erik de Castro Lopo2014-12-19
| | | | | | | | | | | | It seems these two signals have not been working since at least 2009. Detection of these signals seems to have never been added to the configure.ac script and the code guarded by #ifdef then bit-rotted (the idiom used to handle these signals seems to have been abandoned for something simpler/better in 2009). This fix simply handles these signals the same way the other signals are handled in System/Posix/Signals.hsc. Closes #30 and #31
* Wrap posix_fadvise(2) and posix_fallocate(2)Gravatar Herbert Valerio Riedel2014-12-15
| | | | | | | | | | | | This adds two new functions in `System.Posix.Unistd` - `fileAdvise` (aka `posix_fadvise(2)`), and - `fileAllocate` (aka `posix_fallocate(2)`) This is based in part on #7 and has been heavily refactored from its original patch submission by Ricardo Catalinas Jiménez. Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
* Wrap fsync(2) and fdatasync(2)Gravatar Herbert Valerio Riedel2014-12-07
| | | | | | | | | | | | | | This adds two new functions in `System.Posix.Unistd` - `fileSynchronise` (aka `fsync(2)`), and - `fileSynchroniseDataOnly` (aka `fdatasync(2)`) This is based on part of #7 and has been heavily refactored from its original patch submission by Ricardo Catalinas Jiménez. This also bumps version to 2.7.1.0 as a minor version bump is now needed. Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
* Retry process execution in case of ENOTDIRGravatar Iku Iwasa2014-12-06
| | | | | | | If `PATH` environment variable contains non directory component, `__hsunix_execvpe()` failed by `ENOTDIR`. This fixes #11 for all platforms.
* Tighten Safe Haskell bounds, fixes new warning in GHC 7.10.Gravatar David Terei2014-12-06
| | | | Closes #27
* Kill spurious line from changelogGravatar Herbert Valerio Riedel2014-12-06
| | | [skip ci]
* Use CAPI FFI imports for `truncate`Gravatar Herbert Valerio Riedel2014-10-18
| | | | | This makes sure we pick up the LFS version of `truncate` in case `off_t` is affected by CPP defines such as `_FILE_OFFSET_BITS`.
* Use correct POSIX offset-type for tell/seekdirGravatar Herbert Valerio Riedel2014-10-18
| | | | | | | | | | | | This fixes the FFI imports to use the proper `CLong` type over the previous incorrect `COff` type, as using the wrong argument type can cause problems when the `long` and `off_t` types have different size. Historic note from the manual page: In glibc up to version 2.1.1, the return type of telldir() was off_t. POSIX.1-2001 specifies long, and this is the type used since glibc 2.1.2 (released in 1999).
* Merge branch 'sol-fix-getgrgid_r-v3' of https://github.com/kgardas/unix into ↵Gravatar Herbert Valerio Riedel2014-09-10
|\ | | | | | | | | | | | | kgardas-sol-fix-getgrgid_r-v3 Conflicts: changelog.md
| * fix getGroupEntryForID/Name on SolarisGravatar Karel Gardas2014-09-10
| | | | | | | | | | | | | | | | | | This patch fixes getGroupEntryForID and getGroupEntryForName on Solaris The issue on Solaris is that it defines both required getgrgid_r and getgrnam_r functions as CPP macros which depending on configuration are mapped to real function implementations with different names. The issue is solved by using C API calling convention instead of platform C ABI calling convention.
* | Bump `base` constraint for AMPGravatar Herbert Valerio Riedel2014-09-09
|/
* Update changelog and prepare for 2.7.0.1 releaseGravatar Herbert Valerio Riedel2014-03-22
| | | | Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org>
* Handle EROFS/ETXTBSY as permission denied in `fileAccess` (re #8741)Gravatar Alain O'Dea2014-02-28
| | | | | | | | | | | | This extends `System.Posix.Files.`access` to map EROFS & ETXTBSY to mean permission denied just like EACCESS. Based on a patch by Alain O'Dea and comments by Duncan Coutts Authored-by: Alain O'Dea <alain.odea@verafin.com> Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org> (cherry picked from commit ecc92abad017cf12d8eb83509d4d57ae14ad47f9)
* Convert `changelog` to markdown formatGravatar Herbert Valerio Riedel2014-02-28
Signed-off-by: Herbert Valerio Riedel <hvr@gnu.org> (cherry picked from commit 4a08984afe5390d91f10f9b1caf7365e6a93595b)
generated by cgit v1.2.3 (git 2.39.1) at 2024-05-05 14:45:07 +0000