Hi Paul, hi all,
Paul, thanks for the review. I have commited the patch for 64787 as r221621.
Regards,
Andre
gcc/fortran/ChangeLog
2015-03-24 Andre Vehreschild ve...@gmx.de
PR fortran/64787
PR fortran/57456
PR fortran/63230
* class.c
On Mon, Mar 23, 2015 at 08:50:27PM -0600, Martin Sebor wrote:
PR target/63354
* gcc/config/rs6000/linux64.h (ARGET_KEEP_LEAF_WHEN_PROFILED):
Define.
^ typo
It's fixed in version 2 of the patch posted here:
Dear Paul, Dear Mikael, hi all,
thanks for reviewing. I have just committed the patches for:
[Patch, Fortran, pr55901, v1] [OOP] type is (character(len=*)) misinterpreted
as array, and
[Patch, Fortran, v1] Cosmetics and code simplify
as r221627.
Regards,
Andre
2015-03-24
The following fixes PR65519 - we were using gimple_build from
gimple_simplify which isn't a good idea as that doesn't properly
fail when materializing stmts with operands we don't want
(SSA_NAME_OCCURS_IN_ABNORMAL_PHI).
Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
Richard.
Hello,
On 23 Mar 19:02, Ilya Tocar wrote:
Hi,
I've renamed EXT_SSE_REG_P into EXT_REX_SSE_REG_P for consistency.
Ok for stage1?
Patch is OK for stage1.
--
Thanks, K
On 19 Mar 12:24, Ilya Tocar wrote:
Hi,
There were some discussion about x constraints being too conservative
for
This attribute is equivalent to (atyp!(address-expr)).all where atyp is a
general-access-to-typ type. Right now, only the front end changes are done.
The back end needs to adapt to this change too.
The following is a test which should compile and run silently
1. with System; use System;
Steven Bosscher stevenb@gmail.com wrote:
This patch merely papers over another issue, probably a missing
CROSSING_JUMP_P test.
Perhaps. Surely it has looked the current DBR is not so well for
crossing jumps and my fix might be a bit ad-hoc.
The first part of the patch could be rewritten
Hi.
In following patch, I've added missing delete call for all item summaries that
are
allocated within a function_summary container in case the container does not use
GGC memory allocation.
Can boostrap on ppc64le and no regression is seen on x86_64-linux-pc.
Ready for trunk?
Thanks,
Martin
On Tue, Mar 24, 2015 at 12:22:27PM +0300, Ilya Enkovich wrote:
2015-03-24 11:33 GMT+03:00 Jakub Jelinek ja...@redhat.com:
On Thu, Mar 19, 2015 at 11:29:44AM +0300, Ilya Enkovich wrote:
+ /* We might propagate instrumented function pointer into
+ not instrumented function and vice
The following fixes PR65517 - we need to mark loops for fixup if we
remove a path inside a loop.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2015-03-23 Richard Biener rguent...@suse.de
PR middle-end/65517
* tree-cfg.c
2015-03-24 11:33 GMT+03:00 Jakub Jelinek ja...@redhat.com:
On Thu, Mar 19, 2015 at 11:29:44AM +0300, Ilya Enkovich wrote:
+ /* We might propagate instrumented function pointer into
+ not instrumented function and vice versa. In such a
+ case we need to either fix function declaration
Dear Andre,
Dominique pointed out to me that the 'loc' patch causes a ICE in the
testsuite. It seems that 'loc' should provide the address of the class
container in some places and the address of the data in others. I will
put my thinking cap on tonight :-)
Cheers
Paul
On 23 March 2015 at
Hi!
Even if vect_build_slp_tree on child fails, it might have pushed some
nodes to SLP_TREE_CHILDREN (child) vector before returning false.
If we retry, we need to start with no grandchildren of course.
Bootstrapped/regtested on x86_64-linux and i686-linux, acked by Richard in
the PR, committed
On 03/24/2015 06:38 PM, Jan Hubicka wrote:
Hi.
In following patch, I've added missing delete call for all item summaries
that are
allocated within a function_summary container in case the container does not
use
GGC memory allocation.
Can boostrap on ppc64le and no regression is seen on
Hello,
here is a fix for the regression I introduced with my PR64952 patch.
The regression is a spurious shape mismatch error message coming from a
variable partial initialization through data statements.
Before the patch at the time of the shape check, the initialization is
still unset as
On Tue, Mar 24, 2015 at 04:19:46PM +0100, Jakub Jelinek wrote:
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
Also tested with
../configure --with-build-config=bootstrap-lto
--enable-languages=c,c++,fortran,objc,obj-c++,go
make -j16; make -j16 -k check
on x86_64-linux, no
On Tue, Mar 24, 2015 at 10:54:25PM +0100, Martin Liška wrote:
--- a/gcc/symbol-summary.h
+++ b/gcc/symbol-summary.h
@@ -81,6 +81,12 @@ public:
m_symtab_insertion_hook = NULL;
m_symtab_removal_hook = NULL;
m_symtab_duplication_hook = NULL;
+
+/* Release all
On Tue, Mar 24, 2015 at 10:54:25PM +0100, Martin Liška wrote:
--- a/gcc/symbol-summary.h
+++ b/gcc/symbol-summary.h
@@ -81,6 +81,12 @@ public:
m_symtab_insertion_hook = NULL;
m_symtab_removal_hook = NULL;
m_symtab_duplication_hook = NULL;
+
+/* Release all summaries in
Adding a testcase so the bug can be closed.
I believe the segfault was fixed for 3.4.0 by
https://gcc.gnu.org/r67912
Tested x86_64-linux, committed to trunk.
commit c2ae41d5312dce3b4b81653efba477b232dd39f1
Author: Jonathan Wakely jwak...@redhat.com
Date: Tue Mar 24 14:31:58 2015 +
Hello.
Following patch should be final speed-up patch for IPA ICF. It just basically
enhances hash values of a symbol,
computed in a TU. Having finer hash values, we do not spend so much time in WPA
phase. It shows, we can reduce
utilization from 14% - 9% (_w/o_ WPA streaming out). Time report
Hi!
Without this patch, compilation of limits-fndefn.c with -flto
needs huge amounts of stack (more than 20 frames in backtrace).
This patch reworks it so that we use a vector worklist instead,
most of the DFS::DFS_write_tree function body has been moved
into DFS::DFS and DFS_write_tree now
Following Richard Biener's patch at
https://gcc.gnu.org/ml/gcc-patches/2015-03/msg01064.html (r221532),
gcc.target/aarch64/c-output-template-3.c fails with:
c-output-template-3.c: In function 'test':
c-output-template-3.c:7:5: error: impossible constraint in 'asm'
__asm__ (@ %c0 : : S
On Tue, Mar 24, 2015 at 3:06 PM, Jakub Jelinek ja...@redhat.com wrote:
On Tue, Mar 24, 2015 at 12:22:27PM +0300, Ilya Enkovich wrote:
2015-03-24 11:33 GMT+03:00 Jakub Jelinek ja...@redhat.com:
On Thu, Mar 19, 2015 at 11:29:44AM +0300, Ilya Enkovich wrote:
+ /* We might propagate
Hi,
as I said in the audit trail, this isn't a regression but the fix seems
safe enough to me even for mainline: it's matter of reverting a dumb,
supposedly cosmetic change of mine dating back to 2009. I mean to
apply it to 4_9 and 4_8 too. Tested x86_64-linux.
Thanks,
Paolo.
Mis-applied patch, committed.
* config/rl78/rl78.c (rl78_print_operand_1): Move 'p' test to
correct clause.
Index: config/rl78/rl78.c
===
--- config/rl78/rl78.c (revision 221648)
+++ config/rl78/rl78.c (working
Hi Jason.
I found that for locally scoped functions we were not emitting early
dwarf. I've removed the restriction that only emitted non
function-context functions to handle the case below. BTW, this
shouldn't be a (bloat) problem, as we are going to clean up unused DIEs
later (well, next
I thought we could do without this, but rest_of_decl_compilation is not
called for constant decls, so I've put back what the Fortran FE
originally had. This fixes a gdb regression.
Committed to branch.
Aldy
commit 0f6f7418836c533a49a5464828a0461aec437ebc
Author: Aldy Hernandez
Hello!
Attached patch introduces bootstrap-lto-noplugin bootstrap
configuration for hosts that do not support linker plugin (e.g. CentOS
5.11 with binutils 2.17). Also, the patch adds some additional
documentation to bootstrap-lto option.
config/ChangeLog:
2015-03-24 Uros Bizjak
I've been trying to squash DIEs that get generated too late, and here
are some minor fixes.
First, I'm now caching CONST_DECL DIEs. No sense recreating them in
late debug.
Second, I've limited gen_generic_params_dies() to early dwarf. I don't
think we get any additional location info or
Hi all,
I have worked on the comments Mikael gave me. I am now checking for
class_pointer in the way he pointed out.
Furthermore did I *join the two parts* of the patch into this one, because
keeping both in sync was no benefit but only tedious and did not prove to be
reviewed faster.
Paul,
From 1943aa293c1ff7622cf9090a834d7bd9dfaaf086 Mon Sep 17 00:00:00 2001
From: mliska mli...@suse.cz
Date: Mon, 23 Mar 2015 15:36:11 +0100
Subject: [PATCH] IPA ICF: enhance hash value calculated in TU
gcc/ChangeLog:
2015-03-23 Jan Hubicka hubi...@ucw.cz
Martin Liska
Hi,
after the issue with duplicated alternatives exhibited by PR64208, I
checked the arm.md file and found that *arm_subsi3_insn has a
duplication where alt 4 is (r,rI,r) and alt 6 is (r,r,r), this results
in emitting an rsb instruction instead of a sub one, but it has also
an impact on
This testcase was breaking because is_same::operator() is only used in
constexpr evaluation, so cgraph decides it isn't used and throws it
away. Then mangling tries to use it for constexpr evaluation while
re-instantiating FA under get_mostly_instantiated_function_type, and
sadness ensues.
PR 65462 points out a problem with the way that the gccgo version of
go get handles dependencies, a problem due to the fact that with gccgo
the source code of the standard packages is not normally available.
This patch from Lynn Boger fixes the problem. Bootstrapped and ran Go
testsuite on
Dear Iain,
Iain Sandoe wrote:
a couple of minor nits that Dominique and I spotted while discussing
this :
Thanks for the suggestion. I've committed it as Rev. 221615.
Tobias
Index: gcc/fortran/ChangeLog
===
---
PR 65417 points out that the gccgo debug/elf package does not support
PPC32 relocations, which means that the cgo tool does not work
correctly. This patch fixes that. This is a backport of
http://golang.org/cl/7590 from the master library sources.
Bootstrapped and ran Go testsuite on
On Thu, Mar 19, 2015 at 11:29:44AM +0300, Ilya Enkovich wrote:
+ /* We might propagate instrumented function pointer into
+ not instrumented function and vice versa. In such a
+ case we need to either fix function declaration or
+ remove bounds from call statement. */
+ if
On Tue, Jan 27, 2015 at 12:52 AM, Kaz Kojima wrote:
This patch is to fix 2 issues found in dbr_schedule when trying to
fix PR target/64761. The first is relax_delay_slots removes
the jump insn in the insns like below:
(jump_insn/j 74 58 59 (set (pc) (label_ref:SI 29)) ...)
(barrier 59 74
On Sun, Mar 15, 2015 at 11:27 PM, Tim Shen tims...@google.com wrote:
Here's the simple version of it.
Ping?
--
Regards,
Tim Shen
I'm assuming that you mean a cross-compile (which is a stage1 without -werror).
If you look in the build output (I just repeated this on x86_64-darwin12 X
powerpc-darwin9) you'll see:
/GCC/gcc-trunk/gcc/config/rs6000/rs6000.c:24404:1: warning: ‘bool
rs6000_keep_leaf_when_profiled()’ defined
Hi.
In following patch, I've added missing delete call for all item summaries
that are
allocated within a function_summary container in case the container does not
use
GGC memory allocation.
Can boostrap on ppc64le and no regression is seen on x86_64-linux-pc.
Ready for trunk?
Hmmm. This is not the right fix: the tests Richard fixed, were failing because
of lack of constant propagation and DCE at compile-time, which then didn't
eliminate the call to link_error. The AArch64 test is failing because this from
aarch64/constraints.md:
(define_constraint S
A constraint
Dear Dominique,
Dominique Dhumieres wrote:
The test gfortran.dg/coarray/coindexed_3.f90 compiles without error,
see https://gcc.gnu.org/ml/gcc-testresults/2015-03/msg02446.html.
Ups, I somehow missed that files under coarray/ are run with both
-fcoarray=single and =lib; the error (rightly!)
Hi Martin,
On 24 Mar 2015, at 02:50, Martin Sebor wrote:
On 03/21/2015 01:48 PM, Iain Sandoe wrote:
2015-03-13 Anton Blanchard an...@samba.org
PR target/63354
* gcc/config/rs6000/linux64.h (ARGET_KEEP_LEAF_WHEN_PROFILED): Define.
^ typo
44 matches
Mail list logo
