When an ignored file is removed (e.g. a package patch is no longer
needed after a version bump), the corresponding entry in the ignore list
is no longer needed.
However, we currently only validate that an ignored *test* still fails,
not that a ignore files is now missing.
Add a new test to check-package that does that check, and add a
test-case for that check.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Cc: Ricardo Martincoski <ricardo.martincoski@gmail.com>
(cherry picked from commit 5eac4f81eb)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The toolchain-builder project reported an issue with Qemu 7.2.0 for
x86-core2--glibc--bleeding-edge toolchain [1]:
Run /sbin/init as init process
random: fast init done
EXT4-fs (vda): warning: mounting unchecked fs, running e2fsck is recommended
EXT4-fs (vda): re-mounted. Opts: (null). Quota mode: disabled.
Starting syslogd: OK
traps: syslogd[52] general protection fault ip:b7e21465 sp:bfe59e6c error:0 in libc.so.6[b7d9b000+123000]
Starting klogd: OK
traps: klogd[56] general protection fault ip:b7e94465 sp:bf8f069c error:0 in libc.so.6[b7e0e000+123000]
Running sysctl: traps: logger[62] general protection fault ip:b7e48b6c sp:bfd7d194 error:0 in libc.so.6[b7e05000+123000]
Segmentation fault
traps: logger[64] general protection fault ip:b7dd3b6c sp:bf9b8604 error:0 in libc.so.6[b7d90000+123000]
Segmentation fault
(Followed by a kernel panic.)
Testing with the pevious Qemu release (7.1.0) allows to boot the system without any problem.
Building qemu sources between 7.1.0 and 7.2.0 allows to identify the first "bad" commit [2] and
report to the Qemu project [3].
Thanks to Qemu maintainers review, several issues was noticed:
"The default i386 busybox build config does not respect glibc's requirements around stack alignment
(see [4] for previous discussions and a workaround)."
Disabling CONFIG_STACK_OPTIMIZATION_386 option (as suggested in the Gentoo bug report) fixed the issue!
This option has been added and enabled by default in buxybox 1_29_0, so it was used since then the for
Buildroot's qemu defconfig.
Note: The x86-i686--glibc--bleeding-edge (generic x86) doesn't trigger the issue with
CONFIG_STACK_OPTIMIZATION_386 enabled.
Fixes:
https://gitlab.com/buildroot.org/toolchains-builder/-/jobs/3731683337
[1] https://gitlab.com/buildroot.org/toolchains-builder/-/jobs/3731683337
[2] 958e1dd130
[3] https://gitlab.com/qemu-project/qemu/-/issues/1478
[4] https://bugs.gentoo.org/725674
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit f3fa74a8f7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
According to release notes from Microchip, this fixes a memory
leak in AP modes, fixes a failure with an undisclosed but specific
AP, and re-certified with WFA for WILC1000.
Signed-off-by: Kris Bahnsen <kris@embeddedTS.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 2003877ceb)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit 2dff6e93ca (package/readline: add upstream patch to fix crash
with invalid locale specification) fixed a regression in readline 8.2
[0], that could have been caught with a runtime test. readline is a
library, so we need an executable that exercises readline.
Since readline and bash are developped in tandem [1], it is only logical
to use bash to test readline.
Add a new runtime test for bash, that checks that we can indeed run an
interactive shell, and that an non-existing locale does not cause the
dreaded segfault. We do not use the default configuration, because it
uses a uclibc toolchain, and we want to reproduce against a glibc one.
[0] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1021109
[1] https://tiswww.case.edu/php/chet/readline/rltop.html#Bugs
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 511f28ae5c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The current mupdf version fails to build when zlib-ng is used as the
zlib implementation. Add a patch to fix it.
This fixes the following build failure raised since the bump to
version 1.21.0 of the mupdf package in commit
3c2033cd2a:
source/fitz/encode-basic.c: In function 'deflate_write':
source/fitz/encode-basic.c:343:27: error: 'UINT_MAX' undeclared (first use in this function)
343 | newbufsize = n >= UINT_MAX ? UINT_MAX : deflateBound(&state->z, n);
| ^~~~~~~~
source/fitz/encode-basic.c:26:1: note: 'UINT_MAX' is defined in header '<limits.h>'; did you forget to '#include <limits.h>'?
Fixes: http://autobuild.buildroot.net/results/87128e9967cff20078c65ab847877aa44cc409d9
Signed-off-by: Raphaël Mélotte <raphael.melotte@mind.be>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit b20eb56440)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
If both, hicolor-icon-theme (selected by libgtk3-demo) and
adwaita-icon-theme is not selected, then the /usr/share/icons
folder is empty, causing find to fail.
Signed-off-by: Thomas Devoogdt <thomas.devoogdt@barco.com>
Tested-by: Adrian Perez de Castro <aperez@igalia.com>
Reviewed-by: Adrian Perez de Castro <aperez@igalia.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit d2f45dea4c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The most recent release of RPM is v4.18.1 [1].
Adding an upstream patch to prevent build failures for certain
toolchains (0001-fix-var-declare-in-label-error.patch). Fix should be
available next release.
[1]: https://rpm.org/wiki/Releases/4.18.1
Signed-off-by: James Knight <james.d.knight@live.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 233be1c216)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Adding a series of feature-disable configurations of options introduced
over the recent versions, but where never added.
Signed-off-by: James Knight <james.d.knight@live.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 2a068781f1)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In a previous commit [1], the use of `KCONFIG_ENABLE_OPT` was tweaked in
a way which no longer generates a default output unless the quiet flag
is unset (`Q=`). The use of `KCONFIG_SET_OPT` and `KCONFIG_DISABLE_OPT`
does not apply a quiet operation. In order to make all these kernel
configuration defines operate in a consistent way, adjusting both
`KCONFIG_SET_OPT` and `KCONFIG_DISABLE_OPT` to be default-quiet as well.
[1]: 66d2ff25ba
Signed-off-by: James Knight <james.d.knight@live.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit c110bb17ad)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Package binaries can be successfully built for and then executed on
RISC-V platforms including RV32 and RV64 variants. Tested in QEMU.
Signed-off-by: Sergey Matyukevich <geomatsi@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit f7d0ed4a78)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Buxybox's init segfault when built with x86 musl toolchain due to
recent changes about sha1 and sha156 hwaccel support in Busybox 1.36.0.
From Alpinelinux [1]:
"main/busybox: disable sha hwaccel
the flag check for this seems incorrect- it triggers sha_ni use when
the instructions aren't available, but avx512vl/bw is - Elly on irc
causes failures in github actions, which have avx512 but no sha_no"
The flag check was supposed to be fixed upstream by commit [2].
But init still crash with the latest 1_36_stable (with backported patch)
and master branch...
The same workaround has been applied to openembedded-core [3], which
states that they also have the issue with glibc.
Disable hardware sha acceleration on x86 (not x86_64), to avoid the
issue.
Checked that qemu_x86_64_defconfig with musl for x86_64 for core2 boot
as expected with sha hwaccel enabled.
Fixes:
https://gitlab.com/buildroot.org/toolchains-builder/-/jobs/4202276721
[1] https://git.alpinelinux.org/aports/commit/main/busybox?id=ae2cfdf6f6da3dc46ee09d8ceafa26921f6e058e
[2] https://git.busybox.net/busybox/commit/?id=bd76b75f72f717150b909e8c64edfda725cabe11
[3] 22f639d611
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit a57c02a4ea)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
After booting Linux the userspace is unusable since every command issued,
that is a symlink to busybox, fails because of no free ram. So let's
disable IIO in linux.config to give some more ram to userspace.
Signed-off-by: Giulio Benetti <giulio.benetti@benettiengineering.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit 65ff3e10ed)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
LDFLAGS need to be passed to the micropython build system via the
LDFLAGS_EXTRA variable, which was so far missing. This patch adds it.
Signed-off-by: Jonathan Neuschäfer <j.neuschaefer@gmx.net>
Acked-by: Chris Packham <judge.packham@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
(cherry picked from commit f89b988d33)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
