[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

This revision was not accepted when it landed; it landed in state "Needs Revision". This revision was automatically updated to reflect the committed changes. Closed by commit rGfd39b1bb20ce: Revert Revert As a follow-up to my initial mail to llvm-dev heres a first… (authored by echristo). Herald

Re: [PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

Aha, found it in the revert. Bunch of tests use -O1 and probably have something specific in O1 they wanted. -eric On Mon, Nov 25, 2019 at 8:55 PM Eric Christopher wrote: > > Usually nice to give a link if there is one :) > > On Mon, Nov 25, 2019, 8:42 PM Muhammad Omair Javaid via Phabricator >

Re: [PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

Usually nice to give a link if there is one :) On Mon, Nov 25, 2019, 8:42 PM Muhammad Omair Javaid via Phabricator < revi...@reviews.llvm.org> wrote: > omjavaid reopened this revision. > omjavaid added a comment. > This revision is now accepted and ready to land. > > Re-opening this because I

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

omjavaid reopened this revision. omjavaid added a comment. This revision is now accepted and ready to land. Re-opening this because I have reverted the commit due to failures seen on LLDB AArch64 buildbot with this commit. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

This revision was automatically updated to reflect the committed changes. Closed by commit rG8ff85ed905a7: As a follow-up to my initial mail to llvm-dev heres a first pass at the O1… (authored by echristo). Changed prior to commit: https://reviews.llvm.org/D65410?vs=228339=230987#toc

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

chandlerc accepted this revision. chandlerc added a comment. This revision is now accepted and ready to land. Minor nits around redundant predicates for SROA. With thouse fixed, LGTM. I'd really love to find a way to make TCO debuggable so that we don't lose that. I'm particularly worried about

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

mehdi_amini added a comment. LGTM to land this and iterate, but you likely want someone else to confirm :) Comment at: llvm/test/Other/new-pm-defaults.ll:253 +; CHECK-Os-NEXT: Finished llvm::Function pass manager run. +; CHECK-Oz-NEXT: Finished llvm::Function pass manager run.

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

echristo added a comment. Ping :) CHANGES SINCE LAST ACTION https://reviews.llvm.org/D65410/new/ https://reviews.llvm.org/D65410 ___ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

echristo updated this revision to Diff 228339. echristo added a comment. Update to remove comments around SROA addition. CHANGES SINCE LAST ACTION https://reviews.llvm.org/D65410/new/ https://reviews.llvm.org/D65410 Files: clang/test/CodeGen/2008-07-30-implicit-initialization.c

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

echristo updated this revision to Diff 228338. echristo marked an inline comment as done. echristo added a comment. Herald added subscribers: aheejin, sbc100, nhaehnle, jvesely. I've gone ahead and enabled SROA here. In the testing I've done so far it's helped execute time quite a bit and

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

echristo added a comment. In D65410#1613555 , @hfinkel wrote: > Thanks for starting on this. Can you go ahead and replace the sroa calls with > mem2reg calls for `O1` and then see what that does to the performance? That > strikes me as a major change,

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

chandlerc added a comment. One high level point that is at least worth clarifying, and maybe others will want to suggest a different approach: The overall approach here is to have as small of a difference between the O1 and O2

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

hfinkel added a comment. Thanks for starting on this. Can you go ahead and replace the sroa calls with mem2reg calls for `O1` and then see what that does to the performance? That strikes me as a major change, but certainly one that potentially makes sense, so I'd rather we go ahead and test it

[PATCH] D65410: [PassManager] First Pass implementation at -O1 pass pipeline

echristo created this revision. echristo added reviewers: chandlerc, hfinkel. Herald added subscribers: cfe-commits, jfb, dexonsmith, steven_wu, hiraditya, javed.absar, mcrosier, mehdi_amini. Herald added projects: clang, LLVM. As a follow-up to my initial mail to llvm-dev here's a first pass at