All cross-builds are "still" done as C. In C++ you don't need
the missing struct qualifier or the typedef in "typedef struct
gfc_expr ... gfc_expr;" (the struct declaration suffices) as
there's no separate struct namespace IIUC.
Doesn't this show a bug in the compatibility warning system, or
is t
OK.
Jason
On 01/02/2012 05:49 PM, Richard Henderson wrote:
On 01/03/2012 09:10 AM, Jakub Jelinek wrote:
Attached are two different patches, the first one puts the
CLEANUP_POINT_EXPR around the whole OMP_PARALLEL etc. stmt, the second
one wraps the individual clause expressions into CLEANUP_POINT_EXPR.
Bo
On Mon, Jan 2, 2012 at 3:30 PM, Richard Sandiford
wrote:
> Ayal Zaks writes:
>> + for (i = 0; i < ira_pressure_classes_num; i++)
>> + {
>> + enum reg_class pressure_class;
>> +
>> + pressure_class = ira_pressure_classes[i];
>> +
>> + if (max_reg_pressure[pressure_class] == 0)
>
Hello Everyone,
Just added myself under Write After Approval. I am cut and pasting
the patch below.
Here is the Changelog entry:
2012-01-02 Balaji V. Iyer
* MAINTAINERS (Write After Approval): Add myself.
Index: MAINTAINERS
On 01/03/2012 09:10 AM, Jakub Jelinek wrote:
> Attached are two different patches, the first one puts the
> CLEANUP_POINT_EXPR around the whole OMP_PARALLEL etc. stmt, the second
> one wraps the individual clause expressions into CLEANUP_POINT_EXPR.
>
> Both patches have been bootstrapped/regteste
Hi,
another old PR, about -Wredundant-decls (not in -Wall, nor in -Wextra,
thus safe bootstrap-wise).
The issue is that we are emitting a bogus warning for a declaration
followed by a specialization. The fix seems easy, just check !
DECL_TEMPLATE_SPECIALIZATION on newdecl; then however, don'
Hi!
When gimple_ic is changing an indirect call into
if (cond)
direct_call ();
else
(*indirect_call) ()
and the indirect_call is not noreturn, but direct_call is
noreturn, we ICE during checking after profile, because the noreturn
call still has lhs set and edges that it shouldn't have.
The f
Hi!
The f1 function in the testcase fails newly starting with
http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=181332 because there
is no CLEANUP_POINT_EXPR around OMP_PARALLEL/OMP_TASK/OMP_FOR whose
IF/FINAL/NUM_THREADS/SCHEDULE clause expression might need some cleanups.
But as the testcase show
On 01/03/2012 06:59 AM, Jakub Jelinek wrote:
> PR bootstrap/51725
> * cselib.c (add_mem_for_addr): Call canonical_cselib_val
> on mem_elt first.
Ok.
r~
On 12/20/2011 11:13 AM, Torvald Riegel wrote:
> On Mon, 2011-12-19 at 15:17 -0800, Richard Henderson wrote:
>> On 12/19/2011 02:58 PM, Torvald Riegel wrote:
>>> In the particular case (the validated loads technique used in
>>> method-gl.cc, load(), store(), and validate()), we actually do not need
On 01/03/2012 06:47 AM, Uros Bizjak wrote:
>if (d->testing_p)
> return true;
>
> + hi = shift < nelt ? d->op1 : d->op0;
> + lo = shift < nelt ? d->op0 : d->op1;
> +
> + shift %= nelt;
This bit only works for little-endian. It's why I had that assert
for shift range 1-63, for one thi
Hi!
As discussed in the PR, the problem here is that when add_mem_for_addr
calls new_elt_loc_list, it adds the new MEM loc to canonical_cselib_val's
locs, so when mem_elt isn't canonical, we enter into addr_list as well as
first_containing_mem chain some VALUE that doesn't contain any MEM locs.
Th
Hello!
Attached patch fixes a couple of thinkos in new ia64 vector permutation code.
2012-01-02 Uros Bizjak
PR target/51681
* config/ia64/ia64.c (expand_vec_perm_shrp): Use correct operands
for shrp pattern. Correctly handle and fixup shift variable.
2012-01-02 Uros
On 01/02/2012 10:49 AM, Richard Guenther wrote:
For the idea creating the DIE at the point we process the limbo DIE
yes. But would you consider doing that unconditional or only for LTO?
Unconditional. Anything that already passed the assert should be
unaffected by the change.
I can certai
Eric Botcazou writes:
>> You are basically (but non-optimally) locally re-implementing
>> what expr.c:get_object_or_type_alignment does, for the
>> bare MEM_REF case (you don't consider offsets that
>> do not change the alignment, nor alignment information
>> present on the SSA name).
>
> Adjusted
2012/1/2 Georg-Johann Lay :
> This is fix for ISR prologue that "cleared" zero reg with
>
> mov __zero_reg__,__zero_reg__
>
> The right way is
>
> clr __zero_reg__
>
> of course. As CLR does change cc0 notice_update_cc needs to be adapted.
>
> Passes testsuite. Moreover, lightly tested on ISR sou
On Tue, 2011-12-20 at 01:13 +0100, Torvald Riegel wrote:
> On Mon, 2011-12-19 at 15:17 -0800, Richard Henderson wrote:
> > On 12/19/2011 02:58 PM, Torvald Riegel wrote:
> > > In the particular case (the validated loads technique used in
> > > method-gl.cc, load(), store(), and validate()), we actua
On Wed, Dec 21, 2011 at 6:43 AM, Jakub Jelinek wrote:
> 2011-12-21 Jakub Jelinek
>
> * make-relative-prefix.c (make_relative_prefix_1): Avoid
> stack overflow if PATH contains just a single entry and
> HOST_EXECUTABLE_SUFFIX needs to be used.
>
> PR driver/48306
>
Attached is Richard Henderson's patch that marks calls that start a
transaction as returns-twice. In turn, these calls will be treated like
a call to setjmp.
This was motivated by the miscompilation of one of the STAMP
applications (Genome), where a stack slot was used as temp storage for a
CPU r
DR 1359 points out that the rules for constexpr constructors require all
data members to be initialized, which is wrong for unions. This patch
implements the obvious resolution.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 82927de9eb3327c2e65be4f56aec991e36c44fa9
Author: Jason Merrill
Attached is a suggested fix for a long-standing C pre-processor bug.
Ref: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33919
Ref: http://gcc.gnu.org/ml/gcc/2004-10/msg00534.html
The patch implements the approach suggested by Harald van Dijk
in the cited bug report.
I am not familiar with the subtl
Hi,
this seem to have caused a bootstrap failure on s390x: PR51735
/home/andreas/git/gcc-head/gcc/tree-ssa-pre.c: In function ‘bool
insert_aux(basic_block)’:
/home/andreas/git/gcc-head/gcc/tree-ssa-pre.c:3791:1: internal compiler error:
Segmentation fault
Bye,
-Andreas-
On 11/23/2011 11:10 AM,
On Mon, Jan 02, 2012 at 05:41:28PM +0100, Dodji Seketeli wrote:
> I'd like to apply this patch to 4.5 and 4.6 as I find the jumpy stepping
> behaviour very annoying while using 4.6 myself.
>
> I have bootstrapped and tested the patch on these branches
> for x86_64-unknown-linux-gnu successfully.
>
Jason Merrill writes:
> OK.
I'd like to apply this patch to 4.5 and 4.6 as I find the jumpy stepping
behaviour very annoying while using 4.6 myself.
I have bootstrapped and tested the patch on these branches
for x86_64-unknown-linux-gnu successfully.
Would this be OK?
For the record, here is
This is fix for ISR prologue that "cleared" zero reg with
mov __zero_reg__,__zero_reg__
The right way is
clr __zero_reg__
of course. As CLR does change cc0 notice_update_cc needs to be adapted.
Passes testsuite. Moreover, lightly tested on ISR source (there is no ISR test
case in test suit
When I was looking at some LTO-related test failures last week (see mail
on gcc@), I found that the many punctuation and grammatical errors in
the -flto documentation made this section of the GCC manual hard to
read. I made a pass over it to clean up the worst problems -- since the
changes don
On Mon, 2 Jan 2012, Jason Merrill wrote:
> It seems like using get_context_die instead of lookup_decl_die might do the
> trick.
For the idea creating the DIE at the point we process the limbo DIE
yes. But would you consider doing that unconditional or only for LTO?
I can certainly give it a shot
As discussed in the audit trail, this is DR 325 as it affects NSDMI
rather than default arguments. So to fix it I re-used the same code for
dealing with the default argument case.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit bd815e955cfe15f7ecd36366bc79062278e8f2e5
Author: Jason Merri
On 01/02/2012 04:06 PM, Jason Merrill wrote:
On 01/02/2012 08:25 AM, Paolo Carlini wrote:
Bad, because currently many
circumstances in which we do TREE_TYPE (init) = type are in fact benign:
I don't think they are benign. Even when types are equivalent, we
need to convert between them, as seve
On 01/02/2012 08:25 AM, Paolo Carlini wrote:
Bad, because currently many
circumstances in which we do TREE_TYPE (init) = type are in fact benign:
I don't think they are benign. Even when types are equivalent, we need
to convert between them, as several things in the back end rely on
pointer
It seems like using get_context_die instead of lookup_decl_die might do
the trick.
Jason
I am testing the following patch to fix PR51730.
Richard.
2012-01-02 Richard Guenther
PR middle-end/51730
* fold-const.c (fold_comparison): Properly canonicalize
tree offset and HOST_WIDE_INT bit position.
* gcc.dg/fold-compare-6.c: New testcase.
Index: gcc/
Hi Tobias and Steve,
Same here: Not tested, but it looks OK.
Committed (after Steve's mail already). Thanks a lot for the reviews!
Thomas
Ping for this reload patch:
http://gcc.gnu.org/ml/gcc-patches/2011-12/msg00266.html
or perhaps the original:
http://gcc.gnu.org/ml/gcc-patches/2011-10/msg00663.html
which fixes some wrong-code regressions on mips64-linux-gnu.
Richard
Ping for this prepare_shrink_wrap patch:
http://gcc.gnu.org/ml/gcc-patches/2011-12/msg00274.html
It fixes some wrong-code regressions on MIPS16.
Richard
g++.dg/torture/pr37922.C -O3 -fomit-frame-pointer -funroll-loops was
failing for n32 and n64 on mips64-linux-gnu because we were renaming
the global pointer register to $25, even though $25 is needed by the
sequence that sets up the global pointer. This in turn was caused by
a missing earlyclobber
Back in:
http://gcc.gnu.org/ml/gcc-patches/2007-08/msg00944.html
MIPS was changed to prefer block copies over store_by_pieces for all
sizes of constant. I was a little doubtful about this change at the time,
and perhaps I should have been less ready to accept it. The current failure
of gfor
On Mon, Jan 2, 2012 at 2:29 PM, Jakub Jelinek wrote:
> I've noticed that my gather vectorization patch unfortunately regressed code
> quality of
> gcc.target/i386/avx2-i64gatherd256-2.c
> gcc.target/i386/avx2-i64gatherd256-3.c
> gcc.target/i386/avx2-i64gatherd256-4.c
> gcc.target/i386/avx2-i64gat
This fixes various instances of ICEs in dwarf2out_finish when
processing the limbo DIE list. With LTO we can end up with
limbos with RECORD_TYPE context for example when we partitioned
away record-local statics and nothing references the record
type itself.
In general it seems safe to not expose
Ayal Zaks writes:
> + for (i = 0; i < ira_pressure_classes_num; i++)
> +{
> + enum reg_class pressure_class;
> +
> + pressure_class = ira_pressure_classes[i];
> +
> + if (max_reg_pressure[pressure_class] == 0)
> + continue;
> +
> + if (dump_file)
> + fprintf (dump_
Hi!
I've noticed that my gather vectorization patch unfortunately regressed code
quality of
gcc.target/i386/avx2-i64gatherd256-2.c
gcc.target/i386/avx2-i64gatherd256-3.c
gcc.target/i386/avx2-i64gatherd256-4.c
gcc.target/i386/avx2-i64gatherps256-3.c
gcc.target/i386/avx2-i64gatherps256-4.c
tests. T
Hi,
On Mon, Jan 2, 2012 at 3:24 AM, Paolo Carlini wrote:
On 01/02/2012 02:49 AM, Jason Merrill wrote:
On 01/01/2012 08:10 PM, Paolo Carlini wrote:
The analysis is confirmed by the fact that the rather heavy handed
patchlet which I'm attaching, which uses copy_node to avoid the
corruption, "wo
OK.
Jason
Hi!
As described in the PR, this is an attempt to avoid false positive
-Warray-bounds warnings by adding some extra ASSERT_EXPRs in vrp,
so that it can better optimize the code.
For
if (x >> cst1 == cst2)
we can assert that (x - ((unsigned) cst2 << cst1) < (1U << cst1))
and e.g. for
if (x >> cst1
Committed the following change because -mbranch-cost was declared twice in
avr.opt:
http://gcc.gnu.org/viewcvs?view=revision&revision=182798
Johann
The below passed bootstrap and testing, OK for trunk?
Thanks.
Dodji Seketeli a écrit:
> Jason Merrill writes:
>
>> On 12/16/2011 11:40 AM, Dodji Seketeli wrote:
>>> /* It's OK to skip a member with a trivial constexpr ctor.
>>> A constexpr ctor that isn't trivial should
Denis Chertykov wrote:
> 2011/12/19 Georg-Johann Lay :
>> Joseph S. Myers wrote:
>>> On Sun, 18 Dec 2011, Georg-Johann Lay wrote:
>>>
> This new file needs to have the standard copyright and license notices.
> It's desirable to generate such notices in the output files as well.
What is
Committed as r182796. Thanks for the review Tobias. (BTW - I
normally use -cp for my .diffs. I don't know what happened this time
:-) )
PR46262 is completely fixed. The "lorentz" example of Ralph Kube
needs the commented out allocate restoring at lorentz.f03:34.
Similalrly, Arjen's 'particles'
On Fri, Dec 30, 2011 at 08:39:51PM +0400, Dmitry Melnik wrote:
> This series of patches solves few issues we found with Thumb-2
> conditional insns. These fixes include:
>
> Do you think some of this patches are OK for trunk?
Note that none of the included patches look like stage3 material (and
On Fri, Dec 23, 2011 at 9:07 PM, Olivier Hainque wrote:
> A minor update to provide a more precise ChangeLog:
>
>> * system.h: #undef fopen and freopen unconditionally.
>
Ok.
Thanks,
Richard.
> 2011-12-23 Olivier Hainque
>
> * system.h: Prior to #define, #undef fopen and freopen
On Fri, Dec 23, 2011 at 6:29 PM, Olivier Hainque wrote:
> Hello,
>
> Past http://gcc.gnu.org/ml/gcc-patches/2011-12/msg01691.html,
> bootstrap still fails on AIX, from
>
> gcc/collect2.c:1484:25: error: to be safe all intermediate pointers in cast
> from
> 'char **' to 'const char **' must be '
On Fri, Dec 30, 2011 at 5:58 PM, Dmitry Melnik wrote:
> After Thumb-2's peephole2 adds flag clobbering on suitable insns in order to
> generate 16-bit encoding for them, if-conversion can't transform these insns
> into cond_execs. In theory, if the instruction were converted to
> conditional form
On Mon, Jan 2, 2012 at 1:06 PM, Eric Botcazou wrote:
>> Note that you'll get ICEs whenever TYPE_CANONICAL of some
>> aggregate type is NULL (thus, TYPE_STRUCTURAL_EQUALITY), so this
>> seems inherently fragile (similar to using TYPE_CANONICAL == TYPE_CANONICAL
>> here, we'd break the TYPE_STRUCTUR
On Mon, Jan 2, 2012 at 11:37 AM, Jakub Jelinek wrote:
> Hi!
>
> http://gcc.gnu.org/ml/gcc-patches/2011-12/msg01539.html
> PR driver/48306, P2
> - libiberty fix for gcc driver to find paths even when
> $PATH contains some gcc subdirectories
>
> http://gcc.gnu.org/ml/gcc-patches/2011-11/msg01451
On Mon, Jan 2, 2012 at 3:24 AM, Paolo Carlini wrote:
> On 01/02/2012 02:49 AM, Jason Merrill wrote:
>>
>> On 01/01/2012 08:10 PM, Paolo Carlini wrote:
>>>
>>> The analysis is confirmed by the fact that the rather heavy handed
>>> patchlet which I'm attaching, which uses copy_node to avoid the
>>>
> Note that you'll get ICEs whenever TYPE_CANONICAL of some
> aggregate type is NULL (thus, TYPE_STRUCTURAL_EQUALITY), so this
> seems inherently fragile (similar to using TYPE_CANONICAL == TYPE_CANONICAL
> here, we'd break the TYPE_STRUCTURAL_EQUALITY case again).
Could you elaborate? The patch
libgcc/ChangeLog
2012-01-02 Jonathan Wakely
PR bootstrap/51006
* enable-execute-stack-mprotect.c (getpagesize): Do not define
for NetBSD.
This removes the definition of getpagesize which is always present on BSD.
Tested x86_64-unknown-netbsd5.1 and already approved pri
On Sat, Dec 31, 2011 at 12:33 AM, Eric Botcazou
wrote:
> This is the bootstrap failure of the Ada compiler on MIPS/IRIX, a recent
> regression present on mainline and 4.6 branch. The stage 2 compiler
> miscompiles the stage 3 compiler (sem_type.adb:Disambiguate) because of an
> oversight in the f
I've opened PR target/51729 to track the long-standing failures of
gcc.target/mips/dspr2-MULT{,U}.c. The problem used to be in the final
two scan-assembler tests, but in the last few months, the -ffixed-hi and
-ffixed-lo options tripped some cost-consistency checking in IRA.
The original scan-asse
Thomas Koenig wrote:
--- dependency.c(Revision 182430)
+++ dependency.c(Arbeitskopie)
@@ -245,7 +245,9 @@ gfc_dep_compare_functions (gfc_expr *e1, gfc_expr
* 0 if e1 == e2
* -1 if e1< e2
* -2 if the relationship could not be determined
- * -3 if e1 /= e2, but we
Hello Mikael,
Mikael Morin wrote:
Regression tested on x86_64-unknown-linux-gnu. OK for 4.7/4.6/4.5[/4.4] ?
OK - thanks for the comprehensive patch explanation and for the patch
itself.
+ else
+ {
+ /* Otherwise, evaluate the argument out of the loop and pass
As Jakub says in the PR, the fix for PR 46309 indirectly depends on
BRANCH_COST, and the PR is still open because of that. For the time
being, it seems better to avoid the failure for MIPS targets by
tuning for a processor with a high branch cost.
Tested on mips64-linux-gnu and applied.
Richard
MIPS uses an .rdata pseudo-op rather than an .rodata section for
read-only data. Tested on mips64-linux-gnu and applied.
I'm tempted to remove this bit of historical baggage now that we
require fairly recent binutils, but that's for another day.
Richard
gcc/testsuite/
* g++.dg/cpp0x/co
The recent io_real_boz_[345].f90 tests were failing for 32-bit MIPS
GNU/Linux, which doesn't support 16-byte reals. Fixed with the patch
below. Tested on mips64-linux-gnu (where the test continue to pass for
64-bit multilibs) and applied as obvious.
Richard
gcc/testsuite/
* gfortran.dg
Committed.
Richard.
2012-01-02 Richard Guenther
PR bootstrap/51686
* Makefile.def (install-strip-gcc): Depend on install-strip-lto-plugin.
* Makefile.in: Regenerate.
Index: Makefile.def
===
--- Makefile.
g++.dg/opt/devirt2.C was failing for mips64-linux-gnu because each
call to xyzzy was accompanied by a R_MIPS_JALR .reloc statement that
also mentioned xyzzy.
This wouldn't happen if the compiler was built to use PLTs by default,
so rather than look for 4 copies of xyzzy, it seemed better to add
-m
Committed.
Richard.
2012-01-02 Richard Guenther
PR other/51679
* invoke.texi (fassociative-math): Remove spurious paranthesis.
Index: gcc/doc/invoke.texi
===
--- gcc/doc/invoke.texi (revision 182780)
+++ gcc/doc
Hi!
http://gcc.gnu.org/ml/gcc-patches/2011-12/msg01539.html
PR driver/48306, P2
- libiberty fix for gcc driver to find paths even when
$PATH contains some gcc subdirectories
http://gcc.gnu.org/ml/gcc-patches/2011-11/msg01451.html
- the passes.c and reload1.c memory leak fixes (opts-common
> Committed to trunk as every year...
You might want to add this one to your list. Applied to trunk.
2012-01-02 Eric Botcazou
* gnatvsn.ads (Current_Year): Bump to 2012.
--
Eric Botcazou
Index: gnatvsn.ads
===
--- gn
On Dec 30, 2011, at 3:20 PM, Richard Guenther wrote:
> On Sat, Dec 24, 2011 at 8:00 PM, Steven Bosscher
> wrote:
>> On Tue, Dec 20, 2011 at 9:46 AM, Tristan Gingold wrote:
>>> Hi,
>>>
>>> currently alpha/vms backend emits a trampoline entry point for all nested
>>> functions. This is a wast
71 matches
Mail list logo
