Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-07-08 Thread Richard Biener
On July 8, 2016 4:23:31 PM GMT+02:00, "Martin Liška" wrote: >On 07/07/2016 04:15 PM, Richard Biener wrote: >> I think it's fine though the inliners initializer handling looks >> incredibly fragile to me ;) >> >> Richard. > >OK, installed in trunk. May I install the patch to all

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-07-08 Thread Martin Liška
On 07/07/2016 04:15 PM, Richard Biener wrote: > I think it's fine though the inliners initializer handling looks > incredibly fragile to me ;) > > Richard. OK, installed in trunk. May I install the patch to all active branches? Reg works for all of them.

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-07-07 Thread Richard Biener
On Thu, Jul 7, 2016 at 4:01 PM, Martin Liška wrote: > On 07/01/2016 12:15 PM, Richard Biener wrote: >> IMHO using fold-convert in this case is bogus and ideally the testcase >> should have been diagnosed. >> >> fold_convertible_p has a comment >> >> /* Returns true, if ARG is

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-07-07 Thread Martin Liška
On 07/01/2016 12:15 PM, Richard Biener wrote: > IMHO using fold-convert in this case is bogus and ideally the testcase > should have been diagnosed. > > fold_convertible_p has a comment > > /* Returns true, if ARG is convertible to TYPE using a NOP_EXPR. * > > but clearly it isn't generating

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-07-01 Thread Richard Biener
On Thu, Jun 23, 2016 at 1:14 PM, Eric Botcazou wrote: >> The gimplifier has been changed recently to use anonymous SSA_NAMEs instead >> of temporary decls. > > But the PR is a regression present since GCC 4.7... > >> And the gimplifier uses save_expr (which is a gimplifier

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-06-23 Thread Eric Botcazou
> The gimplifier has been changed recently to use anonymous SSA_NAMEs instead > of temporary decls. But the PR is a regression present since GCC 4.7... > And the gimplifier uses save_expr (which is a gimplifier function BTW) on > both not gimplified at all as well as partially gimplified trees.

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-06-23 Thread Jakub Jelinek
On Thu, Jun 23, 2016 at 12:41:53PM +0200, Eric Botcazou wrote: > > This is candidate patch for the PR, which do not create SAVE_EXPR trees for > > already assigned SSA_NAMEs. > > > > Patch survives reg on x86_64-linux-gnu. > > > > Thoughts? > > This looks like a layering violation, save_expr is

Re: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-06-23 Thread Eric Botcazou
> This is candidate patch for the PR, which do not create SAVE_EXPR trees for > already assigned SSA_NAMEs. > > Patch survives reg on x86_64-linux-gnu. > > Thoughts? This looks like a layering violation, save_expr is a GENERIC thing so invoking it on an SSA_NAME is weird. How does this

[PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606).

2016-06-23 Thread Martin Liška
d, 22 Jun 2016 18:07:55 +0200 Subject: [PATCH] Do not emit SAVE_EXPR for already assigned SSA_NAMEs (PR71606). gcc/ChangeLog: 2016-06-22 Martin Liska <mli...@suse.cz> PR middle-end/71606 * tree.c (save_expr): Do not generate SAVE_EXPR if the argument is already an assigned SSA_NAME.