[PATCH] D24098: AMDGPU/SI: Don't run SLSR and NaryReassociate passes

[Tom Stellard via llvm-commits]

On Thu, Sep 01, 2016 at 07:37:01PM +0000, Tom Stellard via llvm-commits wrote:
> On Wed, Aug 31, 2016 at 05:15:40PM -0700, Daniel Berlin via llvm-commits wrote:
> > This is SLSR, not NaryReassociate.
> > 
> 
> The testcase in the SLSR bug was reduced from a very large test which was also
> reported to hang with the NaryReassociate pass.  I will try to come up with a reduced
> test case for NaryReassociate, though I suspect both problems are probably
> caused by the same SCEV bug.
> 

Hi,

I just filed a bug for NaryReassociate with a reduced testcase:
https://llvm.org/bugs/show_bug.cgi?id=30257

> -Tom
> 
> > 
> > On Wed, Aug 31, 2016 at 5:11 PM, Matt Arsenault <arsenm2 at gmail.com> wrote:
> > 
> > >
> > > On Aug 31, 2016, at 17:07, Daniel Berlin via llvm-commits <
> > > llvm-commits at lists.llvm.org> wrote:
> > >
> > > Drive-by comment: Do you have a bug number for the naryreassociate being
> > > slow problem?
> > >
> > > Searching bugzilla is not helping to find one
> > >
> > >
> > > https://llvm.org/bugs/show_bug.cgi?id=28721
> > >
> 
> > _______________________________________________
> > llvm-commits mailing list
> > llvm-commits at lists.llvm.org
> > http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits
> 
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at lists.llvm.org
> http://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-commits