Re: backward threading heuristics tweek

2016-08-16 Thread Jeff Law
On 08/15/2016 02:06 PM, Jan Hubicka wrote: So the threaded path lives fully inside loop1: 6->8->9->3->4->6 propagating that phi_inserted is 0 after the first iteration of the loop. This looks like useful loop peeling oppurtunity which does not garble loop structure. So perhaps threading paths

Re: backward threading heuristics tweek

2016-08-15 Thread Jan Hubicka
> >So the threaded path lives fully inside loop1: 6->8->9->3->4->6 propagating > >that phi_inserted is 0 after the first iteration of the loop. This looks > >like > >useful loop peeling oppurtunity which does not garble loop structure. So > >perhaps threading paths starting and passing loop

Re: backward threading heuristics tweek

2016-08-15 Thread Jeff Law
On 08/11/2016 06:35 AM, Jan Hubicka wrote: This also caused: FAIL: gcc.dg/tree-ssa/pr69270-3.c scan-tree-dump-times uncprop1 ", 1" 4 Failures: gcc.dg/tree-ssa/pr69270-3.c Bisected to: Author: hubicka Date: Sun Aug 7 10:50:16 2016 + *

Re: backward threading heuristics tweek

2016-08-12 Thread James Greenhalgh
On Thu, Aug 11, 2016 at 01:35:16PM +0200, Jan Hubicka wrote: > > On Mon, Jun 6, 2016 at 3:19 AM, Jan Hubicka wrote: > > > Hi, > > > while looking into profile mismatches introduced by the backward > > > threading pass > > > I noticed that the heuristics seems quite simplistics.

Re: backward threading heuristics tweek

2016-08-11 Thread Jan Hubicka
> This also caused: > > FAIL: gcc.dg/tree-ssa/pr69270-3.c scan-tree-dump-times uncprop1 ", 1" 4 > > Failures: > gcc.dg/tree-ssa/pr69270-3.c > > Bisected to: > > Author: hubicka > Date: Sun Aug 7 10:50:16 2016 + > > * tree-ssa-threadbackward.c: Include

Re: backward threading heuristics tweek

2016-08-11 Thread Jan Hubicka
> On Mon, Jun 6, 2016 at 3:19 AM, Jan Hubicka wrote: > > Hi, > > while looking into profile mismatches introduced by the backward threading > > pass > > I noticed that the heuristics seems quite simplistics. First it should be > > profile sensitive and disallow duplication when

Re: backward threading heuristics tweek

2016-08-08 Thread James Greenhalgh
On Sun, Aug 07, 2016 at 10:30:48AM -0700, Andrew Pinski wrote: > On Mon, Jun 6, 2016 at 3:19 AM, Jan Hubicka wrote: > > Hi, > > while looking into profile mismatches introduced by the backward threading > > pass > > I noticed that the heuristics seems quite simplistics. First it

Re: backward threading heuristics tweek

2016-08-07 Thread Andrew Pinski
On Mon, Jun 6, 2016 at 3:19 AM, Jan Hubicka wrote: > Hi, > while looking into profile mismatches introduced by the backward threading > pass > I noticed that the heuristics seems quite simplistics. First it should be > profile sensitive and disallow duplication when optimizing

Re: backward threading heuristics tweek

2016-08-05 Thread Jeff Law
On 06/06/2016 04:19 AM, Jan Hubicka wrote: Hi, while looking into profile mismatches introduced by the backward threading pass I noticed that the heuristics seems quite simplistics. First it should be profile sensitive and disallow duplication when optimizing cold paths. Second it should use

backward threading heuristics tweek

2016-06-06 Thread Jan Hubicka
Hi, while looking into profile mismatches introduced by the backward threading pass I noticed that the heuristics seems quite simplistics. First it should be profile sensitive and disallow duplication when optimizing cold paths. Second it should use estimate_num_insns because gimple statement