Re: [Linaro-TCWG-CI] gcc-17-2325-g4302e16581b4: 10 regressions on arm

On Wed, Jul 15, 2026 at 10:10=E2=80=AFAM <[email protected]> wrote:
>
> Dear contributor,
>
> Our automatic CI has detected problems related to your patch(es). Please =
find some details below.
>
> In  armv8l-unknown-linux-gnueabihf native, after:
>   | commit gcc-17-2325-g4302e16581b4
>   | Author: Andrew Pinski <[email protected]>
>   | Date:   Fri Jul 10 16:55:19 2026 -0700
>   |
>   |     vect: Add back testcases for trapping math with comparisons [PR36=
190]
>   |
>   |     These were originally renamed to no-trapping-math-vect-*.c in r0-=
87196-g04b03edb56633f
>   |     but were fixed in r0-107313-gf9613c9a919462 but never added back =
to the
>   |     trapping versions.
>   | ... 17 lines of the commit log omitted.
>
> Produces 10 regressions:
>   |
>   | regressions.sum:
>   | Running gcc:gcc.dg/vect/vect.exp ...
>   | FAIL: gcc.dg/vect/vect-111.c (test for excess errors)
>   | FAIL: gcc.dg/vect/vect-111.c -flto -ffat-lto-objects (test for excess=
 errors)
>   | FAIL: gcc.dg/vect/vect-ifcvt-12.c (test for excess errors)
>   | FAIL: gcc.dg/vect/vect-ifcvt-12.c -flto -ffat-lto-objects (test for e=
xcess errors)
>   | ... and 6 more

FAIL: gcc.dg/vect/vect-111.c (test for excess errors)
Excess errors:
cc1: warning: '-fassociative-math' disabled; other options take precedence

This is caused by an interaction with -ftrapping-math added by the new
testcases and this from check_vect_support_and_set_flags:
```
    } elseif [is-effective-target arm_neon_ok] {
        eval lappend DEFAULT_VECTCFLAGS [add_options_for_arm_neon ""]
        # NEON does not support denormals, so is not used for vectorization=
 by
        # default to avoid loss of precision.  We must pass -ffast-math to =
test
        # vectorization of float operations.
        lappend DEFAULT_VECTCFLAGS "-ffast-math"
        if [is-effective-target arm_neon_hw] {
            set dg-do-what-default run
        } else {
            set dg-do-what-default compile
        }
```
I am not sure what the best way of solving this. Do we just not test
these testcases on arm neon?
or something else?


>
> Used configuration :
>  *CI config* tcwg_gcc_check armv8l-unknown-linux-gnueabihf
>  *configure and test flags:* none, autodetected on armv8l-unknown-linux-g=
nueabihf--disable-multilib --with-float=3Dhard --with-mode=3Dthumb --with-t=
une=3Dcortex-a9 --with-arch=3Darmv7-a --with-fpu=3Dneon-fp-armv8
>
> We track this bug report under https://linaro.atlassian.net/browse/GNU-19=
69. (Require Linaro login) Please let us know if you have a fix.
>
> If you have any questions regarding this report, please ask on linaro-too=
[email protected] mailing list.
>
> -----------------8<--------------------------8<--------------------------=
8<--------------------------
>
> The information below contains the details of the failures, and the ways =
to reproduce a debug environment:
>
> You can find the failure logs in *.log.1.xz files in
>  * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds=
/10423/archive/artifacts/00-sumfiles/
> The full lists of regressions and improvements as well as configure and m=
ake commands are in
>  * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds=
/10423/archive/artifacts/notify/
> The list of [ignored] baseline and flaky failures are in
>  * http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-arm-build/builds=
/10423/archive/artifacts/sumfiles/xfails.xfail
>
> Current build   : http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-a=
rm-build/builds/10423/archive/artifacts/
> Reference build : http://54.172.246.49:9090/jobs/tcwg_gcc_check--master-a=
rm-build/builds/10422/archive/artifacts/
>
> Instruction to reproduce the build : https://gitlab.com/LinaroLtd/tcwg/ci=
/interesting-commits/-/raw/master/gcc/sha1/4302e16581b490a196a11de3a54a424f=
52e27340/tcwg_gcc_check/master-arm/reproduction_instructions.txt
>
> Full commit : See in git+ssh://[email protected]/git/gcc.git