https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123414
Jeffrey A. Law <law at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Ever confirmed|0 |1
Status|UNCONFIRMED |NEW
Last reconfirmed| |2026-01-06
--- Comment #2 from Jeffrey A. Law <law at gcc dot gnu.org> ---
The good news is we can throttle back the zvl to 256 and still trip the failure
which allows the test to be run on the BPIs natively. It's also not sensitive
to LTO, so one less variable on this one.
./xgcc -B./ -march=rv64gcv_zvl256b -mrvv-max-lmul=m8 -O3 -fsigned-char
-fno-strict-aliasing -fwrapv j.c -c && riscv64-elf-gcc j.o && ./a.out
I'll play with it under a debugger to see if anything stands out. I wouldn't
be at all surprised if it's related to wrapping.
