llvm-project

Author SHA1 Message Date

Author	SHA1	Message	Date
Nemanja Ivanovic	69289cc10f	[PowerPC] Fix broken kill flag after MI peephole The test case in https://bugs.llvm.org/show_bug.cgi?id=47373 exposed two bugs in the PPC back end. The first one was fixed in commit `2771407584` but the test case had to be added without -verify-machineinstrs due to the second bug. This commit fixes the use-after-kill that is left behind by the PPC MI peephole optimization.	2020-09-02 17:07:49 -05:00
Nemanja Ivanovic	2771407584	[PowerPC] Do not legalize vector FDIV without VSX Quite a while ago, we legalized these nodes as we added custom handling for reciprocal estimates in the back end. We have since moved to target-independent combines but neglected to turn off legalization. As a result, we can now get selection failures on non-VSX subtargets as evidenced in the listed PR. Fixes: https://bugs.llvm.org/show_bug.cgi?id=47373	2020-09-02 16:03:36 -05:00

Nemanja Ivanovic

69289cc10f

[PowerPC] Fix broken kill flag after MI peephole

The test case in https://bugs.llvm.org/show_bug.cgi?id=47373 exposed
two bugs in the PPC back end. The first one was fixed in commit
2771407584 but the test case had to
be added without -verify-machineinstrs due to the second bug.
This commit fixes the use-after-kill that is left behind by the
PPC MI peephole optimization.

2020-09-02 17:07:49 -05:00

Nemanja Ivanovic

2771407584

[PowerPC] Do not legalize vector FDIV without VSX

Quite a while ago, we legalized these nodes as we added custom
handling for reciprocal estimates in the back end. We have since
moved to target-independent combines but neglected to turn off
legalization. As a result, we can now get selection failures on
non-VSX subtargets as evidenced in the listed PR.

Fixes: https://bugs.llvm.org/show_bug.cgi?id=47373

2020-09-02 16:03:36 -05:00

2 Commits