-
Notifications
You must be signed in to change notification settings - Fork 15.5k
Closed
Labels
backend:X86backend:X86 Scheduler ModelsAccuracy of X86 scheduler modelsAccuracy of X86 scheduler modelsbugzillaIssues migrated from bugzillaIssues migrated from bugzilla
Description
| Bugzilla Link | 36931 |
| Resolution | FIXED |
| Resolved on | Sep 24, 2018 08:57 |
| Version | trunk |
| OS | Windows NT |
| Depends On | #35432 |
| Blocks | #31672 #35898 #36479 #36256 |
| CC | @alexey-bataev,@adibiagio,@legrosbuffle,@topperc,@rotateright |
| Fixed by commit(s) | rL331770, rL342892 |
Extended Description
MUL64 definitely needs its own scheduler class, possibly the others as well. Multiply by constant (IMUL32rmi etc.) may need splitting off as well.
Note: Integer multiplies seem to be a common issue in vectorization regressions (incorrectly comparing scalar/vector integer multiply costs), so all need to be accurate for [Bug #36550] to work well.
Metadata
Metadata
Assignees
Labels
backend:X86backend:X86 Scheduler ModelsAccuracy of X86 scheduler modelsAccuracy of X86 scheduler modelsbugzillaIssues migrated from bugzillaIssues migrated from bugzilla