aco: Fix 8 and 16-bit addition reductions on GFX6-7
On GFX6-7, the 8 and 16-bit integer add reductions use the 32-bit v_add instruction, which clobbers the VCC register. But we forgot to mark it as such, which lead to test failures.
Due to an influx of spam, we have had to impose restrictions on new accounts. Please see this wiki page for instructions on how to get full permissions. Sorry for the inconvenience.
Our infrastructure migration is complete. Please remember to update your SSH remote to point to ssh.gitlab.freedesktop.org; SSH to the old hostname will time out. You should not see any problems apart from that. Please let us know if you do have any other issues.
On GFX6-7, the 8 and 16-bit integer add reductions use the 32-bit v_add instruction, which clobbers the VCC register. But we forgot to mark it as such, which lead to test failures.