search for: fedorized

Does git-commit-after-all print correctly after all the passes? Maybe I messed it up and it skip some passes, therefore having less to do? Either that, or piping has a higher cost than writing to file. Looks like it surprisingly spends much less time in system more when going through file. Maybe that's because the file is consistently around the same size and is mmapped into memory

On 03/15/2018 01:32 PM, Fedor Sergeev via llvm-dev wrote: > For this to be really usable in this setup we need additionally to: >   - extend -print-module-scope to cover basic block passes >   - introduce a clear way to separate module IRs as those are being > printed by -print-after-all > > But yes, it should work, and a wrapper that pipes to git fast-import > seems to be

Hello Fedor. Thank you for the information. I made a simple patch that exposes PassInstrumentationCallback so llvmGetPassPluginInfo can use it: https://reviews.llvm.org/D71086 . Would this change make sense? Thanks, Juneyoung Lee On Thu, Dec 12, 2019 at 12:44 AM Fedor Sergeev <fedor.sergeev at azul.com> wrote: > > > On 12/3/19 8:01 PM, Juneyoung Lee via llvm-dev wrote: > >

On 03/15/2018 06:09 PM, Alexandre Isoard wrote: > Does git-commit-after-all print correctly after all the passes? Maybe > I messed it up and it skip some passes, therefore having less to do? I did verify that total amount of lines committed to git is reasonably high: ] git rev-list master | while read cmt; do git show $cmt:some-ir.ll; done | wc -l 1587532 corresponding number for

Hmm... I tried Alexandre's fix from D44244 and surprisingly it appears that just using -print-module-scope w/o any additional git actions is waaaay slower on my testcase than -git-commit-module-all. Hell, even a plan -print-after-all is slower:  ] time R/bin/opt -O3 some-ir.ll -disable-output -git-commit-after-all 2>/dev/null real    0m8.041s user    0m7.133s sys     0m0.936s ] time

Is there a reason for why `-asan` and `-asan-module` can be mixed but Function passes and Module passes with the new PM can't be mixed? - Leo On Thu, Sep 27, 2018 at 3:21 AM Fedor Sergeev <fedor.sergeev at azul.com> wrote: > > On 09/27/2018 12:25 PM, Philip Pfaffe wrote: >> >> `opt < %s -passed='asan' -asan-module -S` > > asan-module is another

The print-module-after-all type of option exists in upstream: -print-module-scope - When printing IR for print-[before|after]{-all} always print a module IR commit 7d160f714357f6784ead669ce516e94991c12e5a Author: Fedor Sergeev <fedor.sergeev at azul.com<mailto:fedor.sergeev at azul.com>> Date: Fri Dec 1 17:42:46 2017 +0000 IR

For this to be really usable in this setup we need additionally to:   - extend -print-module-scope to cover basic block passes   - introduce a clear way to separate module IRs as those are being printed by -print-after-all But yes, it should work, and a wrapper that pipes to git fast-import seems to be the best way to handle it. regards,   Fedor. On 03/15/2018 12:31 AM, Daniel Neilson via

Frontends _are_ using PassBuilder, but they need to hook into the default pipeline creation to insert the sanitizer passes. On Tue, Sep 25, 2018 at 12:15 PM Fedor Sergeev <fedor.sergeev at azul.com> wrote: > Hmm... frontends should be using PassBuilder anyway. > And if they are using PassBuilder then they are using PassRegistry.def as > well - all the >

But they're deeply connected. I debug codegen problems all the time. That opt-bisect doesn't work with codegen is really unfortunate. If opt-bisect should work with codegen then we need to think about how codegen will work with the new PM. I agree that whether or not the new PM becomes default is somewhat orthogonal but eventually it will and at that point I hope we have a functioning

On 3/14/19 2:04 AM, Hiroshi Yamauchi wrote: > > > On Wed, Mar 13, 2019 at 2:37 PM Fedor Sergeev <fedor.sergeev at azul.com > <mailto:fedor.sergeev at azul.com>> wrote: > >> >> - Add a new proxy ModuleAnalysisManagerLoopProxy for a loop pass >> to be able to get to the ModuleAnalysisManager in one step and >> PSI through it. >

Hello! I've discovered that `CallSiteSplitting` optimization doesn't support musttail calls. The easiest fix as it stands is disabling it for such call sites: https://reviews.llvm.org/D43729 . However, I'm not happy with such contribution. My more sophisticated attempt has failed due to my poor understanding of llvm internals. Here is the attempted patch:

Update: I was able to make progress on it today ( See https://reviews.llvm.org/D43729 ). Apparently my problems were: * Iterating through the instruction/block list after erasing block/instruction * Trying to split block after removing one predecessor Regarding the latter, it appears that semantics of `DuplicateInstructionsInSplitBetween` change significantly in such case, and it starts to loop

Fedor, Yes that is what happens in my case that the loop is fully unrolled and hence ‘removed’. My objection though is that there is still IR that could be dumped (i.e. the function containing the loop that was removed or the entire module) and that is what I want to have dumped after each pass when I specify -print-after-all. Of course there may be certain implementation details that could make

Overall seems fine to me. On 3/11/19 8:12 PM, Hiroshi Yamauchi wrote: > Here's a revised approach based on the discussion: > > - Cache PSI right after the profile summary in the IR is written in > the pass pipeline. This would avoid the need to insert > RequireAnalysisPass for PSI before each non-module pass that needs it. > PSI can be technically invalidated but unlikely

Hi Leonard, Fedor, while it's true that RegisterPass is not applicable for new-pm passes, PassRegistry.def is not the whole story. Passes in PassRegistry are available for the opt tool. The sanitizers are passes that usually get added to the pipeline by the frontend. There, you need to use PassBuilder's callbacks mechanism to hook the sanitizer into the optimizer. Assuming you're

git-commit-after-all solution has one serious issue - it has a hardcoded git handling which makes it look problematic from many angles (picking a proper git, selecting exact way of storing information, creating repository, replacing the file etc etc). Just dumping information in a way that allows easy subsequent machine processing seems to be a more flexible, less cluttered and overall clean

On 8/7/19 6:20 PM, Hiroshi Yamauchi wrote: > I basically run "clang > -fexperimental-new-pass-manager -print-after-all ..." > > It's conceivable that something is different in our setup or in clang > (from opt)... I'll see if I can reproduce it outside our setup. Does it depend on machine architecture? I generally use x86... regards,   Fedor. > > Thanks.

Oh, well... as usually the answer appears to be pretty obvious. 99% of the time is spent inside the plain write. -print-after-all prints into llvm::errs(), which is an *unbuffered* raw_fd_stream. And -git-commit-after-all opens a *buffered* raw_fd_stream. As soon as I hacked -print-after-all to use a buffered stream to stderr performance went up to the normal expected values: ] time bin/opt

On 8/6/19 7:31 PM, Fedor Sergeev via llvm-dev wrote: > > > On 8/6/19 3:02 AM, Hiroshi Yamauchi via llvm-dev wrote: >> I had a chance to try -print-after-all with NPM. >> >> It seems like there's still no output for the passes >> before objc-arc-contract (which is basically what I saw before.) Does >> anyone else see this? >> >> Are we