search for: performloadpre

...a8a..a3387e3 100644 --- a/lib/Transforms/Scalar/GVN.cpp +++ b/lib/Transforms/Scalar/GVN.cpp @@ -1767,7 +1767,7 @@ bool GVN::processNonLocalLoad(LoadInst *LI) { } // Step 4: Eliminate partial redundancy. - if (!EnablePRE || !EnableLoadPRE) + if (!EnableLoadPRE) return false; return PerformLoadPRE(LI, ValuesPerBlock, UnavailableBlocks); This will disable Scalar PRE without disabling load PRE. (note, again, however, that load PRE can create exactly the same GEP situation you are referring to)

For doing PRE on the load, it looks like there’s only two things stopping GVN PRE from doing it: * GVN::PerformLoadPRE doesn’t hoist loads that are conditional. Probably this can be overcome with some kind of heuristic that allows it to happen in loops when the blocks where a load would have to be inserted are outside the loop. * IsFullyAvailableInBlock goes around the loop and double-counts the entry-edg...

...GVN.cpp > +++ b/lib/Transforms/Scalar/GVN.cpp > @@ -1767,7 +1767,7 @@ bool GVN::processNonLocalLoad(LoadInst *LI) { > } > > // Step 4: Eliminate partial redundancy. > - if (!EnablePRE || !EnableLoadPRE) > + if (!EnableLoadPRE) > return false; > > return PerformLoadPRE(LI, ValuesPerBlock, UnavailableBlocks); > > > > > This will disable Scalar PRE without disabling load PRE. > > > (note, again, however, that load PRE can create exactly the same GEP situation you are referring to) >

...ough so I'm not sure if I should just write a PR about it a let someone who knows the code look at it instead. Anyway, for the bug to trigger I need to run the following passes in the same opt invocation: -sroa -instcombine -simplifycfg -instcombine -gvn The problem seems to be that GVN::PerformLoadPRE triggers and I see a GVN REMOVING PRE LOAD: %_tmp79 = load i16, i16* %_tmp78, align 2 printout. If I instead first run -sroa -instcombine -simplifycfg -instcombine and then -gvn I don't get the GVN REMOVING PRE LOAD printout, and the resulting code looks ok to me. Is this expect...

...t it a let someone who knows the code look at it >> instead. >> >> Anyway, for the bug to trigger I need to run the following passes in the >> same opt invocation: >> -sroa -instcombine -simplifycfg -instcombine -gvn >> >> The problem seems to be that GVN::PerformLoadPRE triggers and I see a >> >> GVN REMOVING PRE LOAD: %_tmp79 = load i16, i16* %_tmp78, align 2 >> >> printout. >> >> If I instead first run >> >> -sroa -instcombine -simplifycfg -instcombine >> >> and then >> >> -gvn >>...

(minor inline additions) On 09/13/2018 01:51 AM, Chandler Carruth wrote: > Haven't had time to dig into this, but wanted to add +Alina Sbirlea > <mailto:asbirlea at google.com> to the thread as she has been working on > promotion and other aspects of LICM for a long time here. Thanks! > On Wed, Sep 12, 2018 at 11:41 PM Philip Reames > <listmail at philipreames.com

Hi, Daniel: Thanks a lot for detailed background information, we are willing to provide the right fix, however it will take time, do you mind if you forward me the discussion you had 5 months ago? I may not be able to access it since I only joined ellvmdev list this week. I did some performance measurement last night, some of our critical benchmark degraded up to 30% with your patch, so we have