https://blogs.gentoo.org/ago/2020/07/04/gentoo-tinderbox/ Issue: dev-libs/isl-0.24 does not respect CFLAGS. Discovered on: amd64 (internal ref: tinderbox) NOTE: To reproduce this issue you may want to add "-frecord-gcc-switches" to CFLAGS/CXXFLAGS/FFLAGS/FCFLAGS. If this is not something about c/c++ context see this bug as an hint to hide the QA warning (with QA_FLAGS_IGNORED) where is not possible to respect CFLAGS.
Created attachment 742422 [details] build.log build log and emerge --info
I have no idea how you hit this, but CFLAGS is respected. You can try, for example: CFLAGS=-O1 ebuild isl-0.24.ebuild clean install or CFLAGS=-O0 and look in build.log, build-abi_x86_32.x86.log, or build-abi_x86_64.amd64.log. I don't know how the QA script is finding this, but it is a false positive. Maybe it has something to do with the fact that isl's build system uses CFLAGS_FOR_BUILD to pick up CFLAGS for the build, but the ebuild handles that correctly. (Incidentally the build system introduces this extra env var for cross compiling, so it does make sense.) Anyhow, I'm closing this as invalid. If you feel I'm wrong about this, please reopen, but I'm going to need better quality information on how this is being hit, because I can't reproduce it.
(In reply to Anthony Basile from comment #2) > I have no idea how you hit this, but CFLAGS is respected. You can try, for > example: > > CFLAGS=-O1 ebuild isl-0.24.ebuild clean install > > or CFLAGS=-O0 and look in build.log, build-abi_x86_32.x86.log, or > build-abi_x86_64.amd64.log. > > I don't know how the QA script is finding this, but it is a false positive. > Maybe it has something to do with the fact that isl's build system uses > CFLAGS_FOR_BUILD to pick up CFLAGS for the build, but the ebuild handles > that correctly. (Incidentally the build system introduces this extra env > var for cross compiling, so it does make sense.) > > Anyhow, I'm closing this as invalid. If you feel I'm wrong about this, > please reopen, but I'm going to need better quality information on how this > is being hit, because I can't reproduce it. I just looked at build log and I don't find my cflags in several lines, one for example: x86_64-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I/var/tmp/portage/dev-libs/isl-0.24/work/isl-0.24 -I/var/tmp/portage/dev-libs/isl-0.24/work/isl-0.24/include -Iinclude/ -O3 -fomit-frame-pointer -malign-double -fstrict-aliasing -ffast-math -c -o isl_test.o /var/tmp/portage/dev-libs/isl-0.24/work/isl-0.24/isl_test.c my CFLAGS are: CFLAGS="-O2 -pipe -march=x86-64 -frecord-gcc-switches -fno-diagnostics-color"
I definitely see this here.
The bug has been referenced in the following commit(s): https://gitweb.gentoo.org/repo/gentoo.git/commit/?id=d19a7f9d9ba8362f9e12a28805957bf17c0b6cb2 commit d19a7f9d9ba8362f9e12a28805957bf17c0b6cb2 Author: Sam James <sam@gentoo.org> AuthorDate: 2021-10-10 22:57:25 +0000 Commit: Sam James <sam@gentoo.org> CommitDate: 2021-10-10 22:57:25 +0000 dev-libs/isl: respect CFLAGS Drop the fun (and aggressive) AX_CC_MAXOPT macro call which was causing us to not respect the user's CFLAGS. Note that 0.24 seemed to be triggering maintainer mode; have chucked eautoreconf into the ebuild for 0.24-r1 given it seemingly wasn't doing any harm to be doing it informally in 0.24 anyway, but will keep an eye on it. Revbump because of the aforementioned churn in a critical dependency. Bug: https://bugs.gentoo.org/815700 Signed-off-by: Sam James <sam@gentoo.org> dev-libs/isl/files/isl-0.24-gdb-autoload-dir.patch | 35 ++++++++++ dev-libs/isl/files/isl-0.24-respect-flags.patch | 14 ++++ dev-libs/isl/isl-0.24-r1.ebuild | 80 ++++++++++++++++++++++ dev-libs/isl/isl-0.24.ebuild | 6 +- 4 files changed, 133 insertions(+), 2 deletions(-)
Let's give it a bit of time first.
sparc done
amd64 stable
x86 stable
arm done
arm64 done
ppc done
ppc64 done all arches done