On Wed, 19 Jul 2017, Jan Hubicka wrote:
> > Did you see the inline comments?
>
> I missed them, sorry :)
> This is updated patch I am testing.
Ok with adding
+ maybe_add_nop_after_section_switch ();
in pass_convert_to_eh_region_ranges::execute instead.
Richard.
> Index: except.c
> =
> Did you see the inline comments?
I missed them, sorry :)
This is updated patch I am testing.
Index: except.c
===
--- except.c(revision 250312)
+++ except.c(working copy)
@@ -2444,6 +2444,61 @@ emit_note_eh_region_end (rtx_i
On Wed, 19 Jul 2017, Jan Hubicka wrote:
> > On Wed, 19 Jul 2017, Jan Hubicka wrote:
> >
> > > > > I think we could just output from generic code - I think it can be
> > > > > done by
> > > > > final_scan_insn. I don't know however if we have a way to tell if the
> > > > > section
> > > > > star
> On Wed, 19 Jul 2017, Jan Hubicka wrote:
>
> > > > I think we could just output from generic code - I think it can be done
> > > > by
> > > > final_scan_insn. I don't know however if we have a way to tell if the
> > > > section
> > > > starts with a landing pad?
> > >
> > > Not sure either --
On Wed, 19 Jul 2017, Jan Hubicka wrote:
> > > I think we could just output from generic code - I think it can be done by
> > > final_scan_insn. I don't know however if we have a way to tell if the
> > > section
> > > starts with a landing pad?
> >
> > Not sure either -- some insn note / bb note?
> > I think we could just output from generic code - I think it can be done by
> > final_scan_insn. I don't know however if we have a way to tell if the
> > section
> > starts with a landing pad?
>
> Not sure either -- some insn note / bb note? Some flag on the label?
> At least the latter shoul
On Tue, 18 Jul 2017, Jan Hubicka wrote:
> > On Tue, 18 Jul 2017, Jan Hubicka wrote:
> >
> > > Hi,
> > > this patch fixes wrong code issue with BB partitioning where sometimes EH
> > > is not delivered. This is very old issue that affect all release branches
> > > with -fprofile-use, but recently
> On Tue, 18 Jul 2017, Jan Hubicka wrote:
>
> > Hi,
> > this patch fixes wrong code issue with BB partitioning where sometimes EH
> > is not delivered. This is very old issue that affect all release branches
> > with -fprofile-use, but recently surfaced as libstdc++ testsuite regression
> > becau
On Tue, 18 Jul 2017, Jan Hubicka wrote:
> Hi,
> this patch fixes wrong code issue with BB partitioning where sometimes EH
> is not delivered. This is very old issue that affect all release branches
> with -fprofile-use, but recently surfaced as libstdc++ testsuite regression
> because we now part
Hi,
this patch fixes wrong code issue with BB partitioning where sometimes EH
is not delivered. This is very old issue that affect all release branches
with -fprofile-use, but recently surfaced as libstdc++ testsuite regression
because we now partition functions based on static profile prediction.
10 matches
Mail list logo
