On Mon, 2015-12-07 at 20:59 +0100, Bernd Schmidt wrote:
> On 12/07/2015 08:43 PM, Steve Ellcey wrote:
> > I am not sure about this. There is an earlier if statement in the loop
> > that does a 'return' instead of a break (or continue) and there is a
> > return in the 'else' part of the if that set
Hi!
This PR fix has been committed with a *.c testcase in g++.dg/tree-ssa/,
where *.c tests are ignored (need *.C).
Fixed as obvious, committed to trunk.
2015-12-07 Jakub Jelinek
PR middle-end/66726
* g++.dg/tree-ssa/pr66726.c: Renamed to...
* g++.dg/tree-ssa/pr66726.
Hi!
On the following testcase we ICE because we call dump_global_iord
on the artificial function with file cdtor, and that function uses
input_location, which can be anything at that spot, can be completely
unrelated to that function and can be (as in this case) UNKNOWN_LOCATION,
where LOCATION_FI
On Mon, 2015-12-07 at 12:30 -0700, Jeff Law wrote:
> On 12/07/2015 12:28 PM, Bernd Schmidt wrote:
> > On 12/07/2015 07:54 PM, Steve Ellcey wrote:
> >> if (must_annul)
> >> -used_annul = 1;
> >> +{
> >> + /* Frame related instructions cannot go into annulled delay
>
On 12/01/2015 01:02 AM, Gary Funck wrote:
This support is enabled via
the -dwarf-2-upc compilation switch. It is not enabled by default,
because some older versions of GDB will abort when encountering
the UPC-related DWARF extensions.
The way we usually handle extensions that cause old debugge
OK.
Jason
Applied, thanks.
Jason
OK.
Jason
On Mon, Dec 07, 2015 at 04:11:48PM +0100, Bernd Schmidt wrote:
> Let's document arguments; for the ones identical to read_cmdline_option an
> explicit pointer there is sufficient, but errors is new.
> This also needs an update to the function comment.
>
> Other than that I'm ok with this. This ar
As reported by Richard H, the -mstd-struct-return option of the SPARC compiler
is totally broken with optimization (and has probably been so since the merge
of the DF branch). Now it's also undocumented, which might explain why nobody
apparently noticed the breakage before Richard.
The problem
After shrink-wrapping has found the "tightest fit" for where to place
the prologue, it tries move it earlier (so that frame saves are run
earlier) -- but without copying any more basic blocks.
Unfortunately a candidate block we select can be inside a loop, and we
will still allow it (because the l
On Mon, Dec 07, 2015 at 11:30:18PM +, Segher Boessenkool wrote:
> --- a/gcc/shrink-wrap.c
> +++ b/gcc/shrink-wrap.c
> @@ -744,33 +744,64 @@ try_shrink_wrapping (edge *entry_edge, bitmap_head
> *bb_with,
> vec.quick_push (e->dest);
> }
>
> - vec.release ();
> -
>if (dump_fil
On Tue, Dec 08, 2015 at 12:59:54AM +0100, Jakub Jelinek wrote:
> Isn't this missing vec.release () for the pro == entry case?
Yes it is. Fixed in v3, thanks!
Segher
On 12/01/2015 11:58 PM, Eric Botcazou wrote:
> IA-64 (stack checking improvement):
> https://gcc.gnu.org/ml/gcc-patches/2015-11/msg01604.html
OK.
Jim
After shrink-wrapping has found the "tightest fit" for where to place
the prologue, it tries move it earlier (so that frame saves are run
earlier) -- but without copying any more basic blocks.
Unfortunately a candidate block we select can be inside a loop, and we
will still allow it (because the l
We decided to apply this to ARM/embedded-5-branch.
Best regards,
Thomas
> -Original Message-
> From: gcc-patches-ow...@gcc.gnu.org [mailto:gcc-patches-
> ow...@gcc.gnu.org] On Behalf Of Andre Vieira
> Sent: Wednesday, October 28, 2015 1:03 AM
> To: gcc-patches@gcc.gnu.org
> Subject: Re:
Various fixes for far memory addressing (and large programs in general).
Committed.
* config/rl78/constraints.md (Wfr): Change to be a non-memory
constraint.
* config/rl78/rl78-protos.h (rl78_one_far_p): Declare.
* config/rl78/rl78.c (rl78_one_far_p): Define.
Richi and I have been discussing revamping slightly how DOM handles
conditionals which it detects are always true or always false.
During gcc6 stage1 I added code to allow DOM to clean them up
immediately, primarily to avoid the waste of having the threader handle
those cases. It was also bel
First in the series. This merely refactors code from tree-ssa-sccvn.c
into domwalk.c so that other walkers can use it as they see fit.
There's an initialization function which sets all edges to executable.
There's a test function to see if a block is reachable.
There's a propagation functi
And testcases. One from the BZ. Two ICEs that showed up during
development, one case where we optimize better now than before, and one
case where we missed an optimization during development that's since
been fixed.
commit f5b74ee83944177f0a1b98ca577343e45aa35584
Author: Jeff Law
Date:
This patch tweaks tree-ssa-dom.c to use the new capability in the dom
walker. Additionally:
The code to remove jump threading paths now runs after the walk is
finished rather than when the conditional is optimized. The code which
optimizes conditionals replaces the condition with a true/fa
Hi,
This patch enables LOOP_VECTORIZED calls processing when debug counters are
used for vectorizer. Bootstrapped and regtested on x86_64-unknown-linux-gnu.
Patch was approved in tracker [1] and applied to trunk.
Thanks,
Ilya
[1] - https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68766#c3
--
gcc
101 - 122 of 122 matches
Mail list logo
