On 8/8/23 02:55, Robin Dapp via Gcc-patches wrote:
Looks reasonable to me - I couldn't read from above whether you did
testing on riscv and thus verified the runtime correctness of the fallback?
If not may I suggest to force matching the pattern on a target you can
test for this purpose?
I tested on riscv (manually and verified the run test) but didn't bootstrap.
The vector test suite (or autovec) is not yet enabled by default anyway but
that's going to change soon.
Presumably this is an alternative to the approach Juzhe posted a week or
two ago and ultimately dropped?
Jeff
