https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121273
Bug ID: 121273
Summary: ice for
gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c with
-O2
Product: gcc
Version: 16.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: dcb314 at hotmail dot com
Target Milestone: ---
>From the gcc testsuite on a raspberry Pi 5:
dcb@raspberrypi:~/gcc/trunk/gcc/testsuite $ ~/gcc/results/bin/gcc -c -w
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c
dcb@raspberrypi:~/gcc/trunk/gcc/testsuite $ ~/gcc/results/bin/gcc -c -w -O1
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c
Going nicely so far.
dcb@raspberrypi:~/gcc/trunk/gcc/testsuite $ ~/gcc/results/bin/gcc -c -w -O2
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c: In function
‘f__Float16_uint64_t_b_i_a_i’:
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c:22:3: error: unrecognizable
insn:
22 | }
| ^
./gcc.target/aarch64/sve/unpacked_cond_fmaxnm_1.c:25:3: note: in expansion of
macro ‘TEST_FN’
25 | TEST_FN (FN, TYPE0, TYPE1, COUNT, b_i, a_i) \
| ^~~~~~~
This bug seems to have existed since sometime before g:47a7dcd4c027bcc9
