]> git.ipfire.org Git - thirdparty/gcc.git/commit
testsuite: Create a variant of uninit-pred-7_a.c [PR123377]
authorAndrew Pinski <andrew.pinski@oss.qualcomm.com>
Sat, 3 Jan 2026 05:25:12 +0000 (21:25 -0800)
committerAndrew Pinski <andrew.pinski@oss.qualcomm.com>
Sun, 4 Jan 2026 18:18:57 +0000 (10:18 -0800)
commitbcde241ac56de6e168fc16f778be60ff13bbcbca
tree7eae4113f0c914eb01715e2a1a6033226b389921
parent1a2b5eeb5d24a44db9cb757906bd06912af15379
testsuite: Create a variant of uninit-pred-7_a.c [PR123377]

So it turns out the xfail uninit-pred-7_a.c didn't always happen
depending on the setting of logical-op-non-short-circuit.
So this creates a second copy of the testcase for the case
of `logical-op-non-short-circuit=0` without the xfail and then sets
`logical-op-non-short-circuit=1` for uninit-pred-7_a.c with still
the xfail.

Tested on x86_64-linux-gnu to make sure both pass and uninit-pred-7_a.c
xfails like it should. Also tested manually on powerpc64-linux-gnu to
see the bogus warning happen with logical-op-non-short-circuit=1.

PR testsuite/123377
gcc/testsuite/ChangeLog:

* gcc.dg/uninit-pred-7_a.c: Add
`--param logical-op-non-short-circuit=1` to the options.
* gcc.dg/uninit-pred-7_a_a.c: New test.

Signed-off-by: Andrew Pinski <andrew.pinski@oss.qualcomm.com>
gcc/testsuite/gcc.dg/uninit-pred-7_a.c
gcc/testsuite/gcc.dg/uninit-pred-7_a_a.c [new file with mode: 0644]