Debian Patches
Status for keyutils/1.6.3-6
| Patch | Description | Author | Forwarded | Bugs | Origin | Last update |
|---|---|---|---|---|---|---|
| Debian-specific-kernel-feature-tests.patch | Debian-specific kernel feature tests Upstream assumes features based on the kernel version, but some of these features aren't available in the Debian kernel, or were only enabled at a later point. Also, the release number comparison used by upstream breaks with sid's non-numeric release number of "unstable". This adjust the feature tests to reflect Debian's kernels: * Drop $OSRELEASE tests * Drop ancient RedHat-specific test * Debian's kernel generally lacks CONFIG_BIG_KEYS * Debian's kernel has CONFIG_KEY_DH_OPERATIONS since 4.19~exp1 |
Christian Kastner <ckk@debian.org> | not-needed | 2019-02-20 | ||
| Make-build-reproducible.patch | Make build reproducible Make the build reproducible by making it possible to set the build date, instead of embedding the current date. |
Christian Kastner <ckk@kvr.at> | not-needed | 2015-10-07 | ||
| pkg-config-install-tweaks.patch | pkg-config install tweaks | Christian Kastner <ckk@debian.org> | invalid | 2019-02-09 | ||
| man-page-fixes.patch | man page fixes | Christian Kastner <ckk@debian.org> | invalid | 2019-02-09 | ||
| Tests-for-KEYCTL_MOVE-require-kernel-5.3-or-above.patch | Tests for KEYCTL_MOVE require kernel 5.3 or above | Christian Kastner <ckk@kvr.at> | no | 2020-02-10 | ||
| Mark-test-requiring-root-as-such.patch | Mark test requiring root as such | Christian Kastner <ckk@kvr.at> | no | 2022-06-19 | ||
| Ensure-maxsquota-is-set-correctly.patch | test: Ensure maxsquota is set correctly Based on a fix provided by Aurelien Jarno, we check for the quota of the invoking user, rather than root. And we set it to 0 if quota lookup fails for some reason. Discovered when running tests in containers. |
Christian Kastner <ckk@kvr.at> | no | 2025-04-04 | ||
| man-fix-a-formatting-on-the-keyctl.1-man-page.patch | man: fix a formatting on the keyctl.1 man page Fix a formatting and a command output on the keyctl.1 man page in a couple of places. |
Vladis Dronov <vdronov@redhat.com> | no | https://web.git.kernel.org/pub/scm/linux/kernel/git/dhowells/keyutils.git/commit/?id=81a8c963db50681c7470dcfe8fe6c79a70d7aadd | 2022-05-10 | |
| cflags.patch | Avoid setting Intel Architecture specific CFLAGS | Nobuhiro Iwamatsu <iwamatsu@debian.org> | not-needed | debian | 2014-05-18 | |
| hardening.patch | Use build flags from environment (dpkg-buildflags) Necessary for hardening flags. Also strip -pie/-fPIE from CFLAGS/LDFLAGS when used to link libraries. It conflicts with -fPIC. |
Simon Ruderich <simon@ruderich.org> | not-needed | debian | 2014-05-18 | |
| Fix-format-specifier-for-pointer-subtraction.patch | Fix format specifier for pointer subtraction Resolves the following compiler error: keyctl_watch.c: In function 'consumer': keyctl_watch.c:177:61: error: format '%zx' expects argument of type 'size_t', but argument 3 has type 'int' [-Werror=format=] 177 | fprintf(stderr, "NOTIFY[%03zx]: ty=%06x sy=%02x i=%08x\n", | ~~~~^ | | | long unsigned int | %03x 178 | p - buffer, n.n.type, n.n.subtype, n.n.info); | ~~~~~~~~~~ | | | int |
Alyssa Ross <hi@alyssa.is> | no | https://web.git.kernel.org/pub/scm/linux/kernel/git/dhowells/keyutils.git/commit/?id=ac2d0f3340a819362da460a35dc8e0cd81c0e004 | 2023-03-01 | |
| test-Fix-test-expectation-based-on-kernel-config.patch | test: Fix test expectation based on kernel config Some test results are dependent on the kernel configuration option CONFIG_SYSTEM_BLACKLIST_AUTH_UPDATE. Check the kernel configuration file for its presence and expect appropriate test results. Function has_kernel_config is based on its xfstsests counterpart. |
Pavel Reichl <preichl@redhat.com> | no | https://web.git.kernel.org/pub/scm/linux/kernel/git/dhowells/keyutils.git/commit/?id=c076dff259e99d84d3822b4d2ad7f3f66532f411 | 2022-12-20 | |
| test-Don-t-rely-on-PAGE_SIZE-for-maxdesc-length.patch | test: Don't rely on PAGE_SIZE for maxdesc length The maxdesc length was fixed at 4096 in kernel 3.18 rather than PAGE_SIZE, and toolbox.inc.sh accounts for that when setting $maxdesc. Rather than introducing yet another variable, calculate the length of $maxdesc at runtime in the few spots where this is needed. This was discovered on a POWER9 host with 64K page size when running tests as non-root. Because 64K page size < 20000 quota on this host, the else branch was entered, where the expected EDQUOT did not occur because the hard 4096 $maxdesc introduced in 3.18 was of course within the 20000 quota. |
Christian Kastner <ckk@kvr.at> | no | 2025-04-13 |