On Sat, 2014-03-08 at 11:15 +0100, Basile Starynkevitch wrote:
I am pinging again this documentation patch
http://gcc.gnu.org/ml/gcc-patches/2014-02/msg00074.html
(pinged at http://gcc.gnu.org/ml/gcc-patches/2014-02/msg01002.html on
febµ.17th 2014)
and also pinged at
Hi,
For float value, movsfcc/movdfcc is required by emit_conditional_move
called in ifcvt pass to expand if-then-else to fcsel insn.
Bootstrap and no make check regression with qemu-aarch64.
Is it OK for next stage1?
Thanks!
-Zhenqiang
ChangeLog:
2014-03-18 Zhenqiang Chen
Hi,
Here is a small patch to fix ICE in change_decl_assembler_name. I found this
problem working on my experimental branch which uses transparent alias chains
heavily and I don't have reproducer for the trunk.
Bootstrapped and checked on linux-x86_64.
Thanks,
Ilya
--
2014-03-17 Ilya
On Mon, 17 Mar 2014, Jakub Jelinek wrote:
Hi!
Apparently rest_of_decl_compilation only calls varpool_finalize_decl
if not in_lto_p, so this patch calls it explicitly after that call to
make sure with -flto we register the newly created vars with varpool as
well.
Additionally, the patch
On Tue, Mar 18, 2014 at 4:36 AM, Mike Stump mikest...@comcast.net wrote:
So, to support things like this:
(define_constants
(C1_TEMP_REGNUM PROLOGUE_SCRATCH_1)
(C1_TEMP2_REGNUM PROLOGUE_SCRATCH_2)
I need the rtl reader to do less checking. We we turn off int
On Tue, Mar 18, 2014 at 9:20 AM, Ilya Enkovich enkovich@gmail.com wrote:
Hi,
Here is a small patch to fix ICE in change_decl_assembler_name. I found this
problem working on my experimental branch which uses transparent alias chains
heavily and I don't have reproducer for the trunk.
Janne Blomqvist wrote:
* io/unix.c (set_close_on_exec): New function.
Since this patch, most Fortran tests fail on spu-elf since the system
libraries do not support fcntl on the SPU.
The patch below fixes this by using an autoconf check to verify
fcntl is present before using it, as is
On 17 Mar 22:18, Ulrich Drepper wrote:
On Mon, Mar 17, 2014 at 7:39 AM, Ilya Tocar tocarip.in...@gmail.com wrote:
undefined is similar in behavior to setzero, but it also clobbers
flags. Maybe just define it to setzero for now?
What do you mean by clobbers flags? Do you have an
Ulrich Weigand wrote:
Janne Blomqvist wrote:
* io/unix.c (set_close_on_exec): New function.
Since this patch, most Fortran tests fail on spu-elf since the system
libraries do not support fcntl on the SPU.
The patch below fixes this by using an autoconf check to verify
fcntl is present
Hi,
While looking at something else I realized that we had MEMORY_MOVE_COST
defined in the backend. However we also have the more recent target hook
defined for this through TARGET_MEMORY_MOVE_COST making it obvious to
remove this definition, given that the only use of the macro
Hi,
After control flow graph change made by
http://gcc.gnu.org/ml/gcc-patches/2014-02/msg01492.html, case
gcc.dg/tree-ssa/ssa-dom-thread-4.c is broken on logical_op_short_circuit
targets including cortex-m3/cortex-m0.
The regression reveals a missed opportunity in jump threading, which causes
a
The new gcc.dg/tls/pr58595.c testcase FAILs on Solaris 9:
FAIL: gcc.dg/tls/pr58595.c (test for excess errors)
Excess errors:
Undefined first referenced
symbol in file
___tls_get_addr /var/tmp//ccuBbAna.o
ld: fatal: Symbol
On Tue, Mar 18, 2014 at 11:19:52AM +0100, Rainer Orth wrote:
The new gcc.dg/tls/pr58595.c testcase FAILs on Solaris 9:
FAIL: gcc.dg/tls/pr58595.c (test for excess errors)
Excess errors:
Undefined first referenced
symbol in file
This is a follow-up to
http://gcc.gnu.org/ml/gcc-patches/2013-07/msg00959.html
which implemented the workaround for the data cache nullify issues on LEON3.
Another fixlet.
Tested on SPARC/Solaris, applied on the mainline and 4.8 branch.
2014-03-18 Eric Botcazou ebotca...@adacore.com
On Tue, Mar 18, 2014 at 10:34 AM, Ilya Tocar tocarip.in...@gmail.com wrote:
On 17 Mar 22:18, Ulrich Drepper wrote:
On Mon, Mar 17, 2014 at 7:39 AM, Ilya Tocar tocarip.in...@gmail.com wrote:
undefined is similar in behavior to setzero, but it also clobbers
flags. Maybe just define it to
On Sat, Mar 15, 2014 at 2:13 PM, Jakub Jelinek ja...@redhat.com wrote:
Hi!
Here is an updated patch for what Tobias has posted earlier:
http://gcc.gnu.org/ml/gcc-patches/2014-03/msg00043.html
While that version bootstrapped/regtested fine, most of the Fortran
tests ICEd, primarily because
On Tue, Mar 18, 2014 at 2:12 AM, Basile Starynkevitch
bas...@starynkevitch.net wrote:
On Sat, 2014-03-08 at 11:15 +0100, Basile Starynkevitch wrote:
I am pinging again this documentation patch
http://gcc.gnu.org/ml/gcc-patches/2014-02/msg00074.html
(pinged at
On Mon, 17 Mar 2014, Cong Hou wrote:
On Mon, Mar 17, 2014 at 6:44 AM, Richard Biener rguent...@suse.de wrote:
On Fri, 14 Mar 2014, Cong Hou wrote:
On Fri, Mar 14, 2014 at 12:58 AM, Richard Biener rguent...@suse.de wrote:
On Fri, 14 Mar 2014, Jakub Jelinek wrote:
On Fri, Mar 14,
On Mon, Mar 17, 2014 at 03:24:14PM -0400, Vladimir Makarov wrote:
It is complicated. There is no guarantee that it is used only for
addresses. I need some time to think how to fix it.
Meanwhile, you *should* commit the patch into the trunk because it
solves the real problem. And I can
Hi,
this patch fixes an ICE regarding COFF-debugging information. Problem is that
the parms isn't necessarily an incoming argument, and not necessarily
already set.
By checking this - as done already in dbxout - we can avoid to process
on invalid
parm-declarations.
ChangeLog
2014-03-18 Kai
Committed as obvious.
Richard.
2014-03-18 Richard Biener rguent...@suse.de
* doc/contrib.texi: Adjust my name.
Index: gcc/doc/contrib.texi
===
--- gcc/doc/contrib.texi(revision 208642)
+++ gcc/doc/contrib.texi
On 18/03/14 09:43, Ramana Radhakrishnan wrote:
Hi,
While looking at something else I realized that we had MEMORY_MOVE_COST
defined in the backend. However we also have the more recent target hook
defined for this through TARGET_MEMORY_MOVE_COST making it obvious to
remove this
Just found another patch in my local tree.
Committed as obvious.
Richard.
2014-03-18 Richard Biener rguent...@suse.de
* doc/lto.texi (-fresolution): Document.
Index: gcc/doc/lto.texi
===
--- gcc/doc/lto.texi
Another one.
Committed.
Richard.
2014-03-18 Richard Biener rguent...@suse.de
* doc/loop.texi: Remove section on the removed lambda framework.
Update loop docs with recent changes in preserving loop structure.
Index: gcc/doc/loop.texi
On Tue, 11 Mar 2014, Richard Biener wrote:
On Sat, 8 Mar 2014, Gerald Pfeifer wrote:
Thanks for the time and diligence writing this up, Richi!
On Thu, 6 Mar 2014, Richard Biener wrote:
-files; if @option{-flto} is not passed to the linker, no
-interprocedural optimizations are
Hello Diego all,
Here is a slightly improved patch to follow Diego's comments on
http://gcc.gnu.org/ml/gcc-patches/2014-03/msg00863.html
(since Diego improved the wording extracted from comments
in gcc/plugin.def I am also patching the comments in that file).
### gcc/Changelog entry
Richard Biener rguent...@suse.de writes:
+@item -fresolution=@var{file}
+@opindex fresolution
+This option specifies the linker resolution file. This option is
+only meaningful in conjunction with @option{-fwpa} and as option
+to pass through to the LTO linker pluign.
I found this patch in my local tree which avoids copying a bitmap
by doing
bitmap_and_compl (tmp, op2, dense_invalidated);
instead of
bitmap_copy (tmp, op2);
bitmap_and_compl_into (tmp, dense_invalidated);
which should besides speed also eventually reduce peak memory
On Tue, 18 Mar 2014, Rainer Orth wrote:
Richard Biener rguent...@suse.de writes:
+@item -fresolution=@var{file}
+@opindex fresolution
+This option specifies the linker resolution file. This option is
+only meaningful in conjunction with @option{-fwpa} and as option
+to pass through
Hi Thomas!
This patch introduces support of OpenACC loop directive (and combined
directives) in C front-end up to GENERIC. Currently no clause is allowed.
This patch is necessary to finish implementation of OpenACC 1.0 in
fortran front-end. As you know, OpenACC fortran implementation does
Hello,
I found ICE in Chromium compiled with LTO. There's a call that is
proved by ipa-devirt as __builtin_unreachable; same decision is done by
gimple-fold and this call is replaced by GIMPLE_CALL and GIMPLE_ASSIGN
(in this order). After that condition for
OK with:
+Pragmas registered with @code{c_register_pragma_with_expansion} or
+@code{c_register_pragma_with_expansion_and_data} are supporting
+preprocessor expansions. For an example of using such a pragma:
s/are supporting/support/
s/For an example of using such a pragma/For example/
Diego.
On Tue, Mar 18, 2014 at 01:32:57PM +0100, Richard Biener wrote:
2014-03-18 Richard Biener rguent...@suse.de
* df-problems.c (df_rd_confluence_n): Avoid bitmap_copy
by using bitmap_and_compl instead of bitmap_and_compl_into.
(df_rd_transfer_function): Likewise.
Ok,
On Tue, Mar 18, 2014 at 1:38 PM, Martin Liška mli...@suse.cz wrote:
Hello,
I found ICE in Chromium compiled with LTO. There's a call that is proved
by ipa-devirt as __builtin_unreachable; same decision is done by gimple-fold
and this call is replaced by GIMPLE_CALL and GIMPLE_ASSIGN (in
Hi!
2014-03-18 Martin Liska mli...@suse.cz
* cgraph.c (cgraph_update_edges_for_call_stmt_node): added case when
gimple call statement is updated.
Capital letter after :
* gimple-fold.c (gimple_fold_call): changed order for GIMPLE_ASSIGN
and
Likewise here.
Hi!
In gomp-4_0-branch's r208648, I have just committed some ;-) support for
OpenACC num_gangs, num_workers, vector_length clauses (OpenACC parallel
directive), that is, just passing them through the compiler -- now,
libgomp just needs to do something useful with that information.
commit
Hi,
this patch skips anon2.C and anon3.C test for mingw target. Issue
here is that weak under pe-coff is different to ELF-targets and
therefore test doesn't apply for
FAIL: g++.dg/abi/anon2.C -std=c++11 scan-assembler
.weak(_definition)?[ \t]_?_ZN2N11D1C3fn1ENS0_1BE
FAIL: g++.dg/abi/anon2.C
Thank you for feedback,
new changelog:
2014-03-18 Martin Liska mli...@suse.cz
* cgraph.c (cgraph_update_edges_for_call_stmt_node): Added case
when
gimple call statement is update.
* gimple-fold.c (gimple_fold_call): Changed order for
GIMPLE_ASSIGN and
Hi!
On Tue, 18 Mar 2014 16:37:24 +0400, Ilmir Usmanov i.usma...@samsung.com wrote:
This patch introduces support of OpenACC loop directive (and combined
directives) in C front-end up to GENERIC. Currently no clause is allowed.
Thanks! I had worked on a simpler patch, not yet dealing with
This is another patch (well, I've polished it a bit) that was sitting
in my local tree for some time. I've not enabled the ggc-common.c
code (I merely want to get rid of the false possitives).
Queued for 4.10 unless I'm told otherwise.
Richard.
2014-03-18 Richard Biener rguent...@suse.de
On Tue, Mar 18, 2014 at 2:29 PM, Martin Liška mli...@suse.cz wrote:
Thank you for feedback,
Ok if it passes bootstrap / regtest.
Thanks,
Richard.
new changelog:
2014-03-18 Martin Liska mli...@suse.cz
* cgraph.c (cgraph_update_edges_for_call_stmt_node): Added case when
Current, indirect function call prevents tail-call optimization on AArch64.
This patch adapt the fix for PR arm/19599 to AArch64.
Is it ok for next stage 1?
Thanks.
-- Jiong
gcc/
* config/aarch64/predicates.md (aarch64_call_insn_operand): New
predicate.
*
cpplib-4.9-b20140202.de.po.gz
Description: Binary data
The Translation Project robot, in the
name of your translation coordinator.
coordina...@translationproject.org
Hello, gentle maintainer.
This is a message from the Translation Project robot.
A revised PO file for textual domain 'cpplib' has been submitted
by the German team of translators. The file is available at:
http://translationproject.org/latest/cpplib/de.po
(This file,
Committed as r208649.
--
Ilmir.
2014-02-13 9:46 GMT+00:00 Alex Velenko alex.vele...@arm.com:
Hi,
This patch re-implements vreinterpret intrinsics to directly call a cast.
The aim is to forward as much information to front-end as possible.
This patch had a full LE and BE regression run with no regressions.
Is patch good to
This patch introduces a new class toplev and changes toplev_main and
toplev_finalize to be methods of this class. Additionally, now the
timevars are automatically stopped when the object is destroyed. This
cleans up compile a bit and makes it simpler to reuse the toplev
logic in other code.
---
This introduces a new scoped_timevar class. It pushes a given timevar
in its constructor, and pops it in the destructor, giving a much
simpler way to use timevars in the typical case where they can be
scoped.
---
gcc/ChangeLog.jit | 4
gcc/jit/ChangeLog.jit | 4
I wanted to do something like playback::context::compile, but in my
project I can't really reuse all the JIT code -- really I just wanted
to be able to use toplev_main and toplev_finalize.
Looking into the code, though, I saw a few spots that could be cleaned
up a little, so I wouldn't have to
Patch passes bootstrap and regtest.
I fixed indentation according to discussion with Jakub.
OK for trunk?
Thanks,
Martin
On 03/18/2014 02:55 PM, Richard Biener wrote:
On Tue, Mar 18, 2014 at 2:29 PM, Martin Liška mli...@suse.cz wrote:
Thank you for feedback,
Ok if it passes bootstrap /
On Tue, Mar 18, 2014 at 3:59 PM, Martin Liška mli...@suse.cz wrote:
Patch passes bootstrap and regtest.
I fixed indentation according to discussion with Jakub.
OK for trunk?
Ok.
Thanks,
Richard.
Thanks,
Martin
On 03/18/2014 02:55 PM, Richard Biener wrote:
On Tue, Mar 18, 2014 at
On Tue, Mar 18, 2014 at 7:13 AM, Richard Biener
richard.guent...@gmail.com wrote:
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_undefined_ps (void)
{
__m512 __Y = __Y;
return __Y;
}
This provokes no warnings (as you wrote) and it
On Tue, Mar 18, 2014 at 10:00:09AM +0100, Richard Biener wrote:
Isn't varpool_finalize_decl the canonical interface anyway?
Thus, you don't need to call rest_of_decl_compilation AFAIK - the
varpool machinery will do that.
...
Should be enough to check if int128_integer_type_node is not NULL.
Hi!
Apparently with -std=c++11 and higher, if in a template we call
finish_expr_stmt on a DECL_EXPR (as cp_parser_omp_declare_reduction_exprs
does) and the omp_priv var in there is not type dependent, it ICEs during:
else if (!type_dependent_expression_p (expr))
convert_to_void
On Tue, Mar 18, 2014 at 4:03 PM, Ulrich Drepper drep...@gmail.com wrote:
On Tue, Mar 18, 2014 at 7:13 AM, Richard Biener
richard.guent...@gmail.com wrote:
extern __inline __m512
__attribute__ ((__gnu_inline__, __always_inline__, __artificial__))
_mm512_undefined_ps (void)
{
__m512 __Y =
On Tue, Mar 18, 2014 at 4:09 PM, Richard Biener
richard.guent...@gmail.com wrote:
On Tue, Mar 18, 2014 at 4:03 PM, Ulrich Drepper drep...@gmail.com wrote:
On Tue, Mar 18, 2014 at 7:13 AM, Richard Biener
richard.guent...@gmail.com wrote:
extern __inline __m512
__attribute__ ((__gnu_inline__,
On Tue, Mar 18, 2014 at 11:11 AM, Richard Biener
richard.guent...@gmail.com wrote:
Btw, without this zeroing (where zero is also undefined) this may
be an information leak and thus possibly a security issue? That is,
how is _mm_undefined_pd () specified?
People aren't accidentally using the
On Fri, Mar 14, 2014 at 11:45:48PM -0300, Alexandre Oliva wrote:
This bug report had various testcases that had to do with full loop
unrolling with non-automatic iterators and fixed boundaries, which
resulted in duplicating debug stmts in the loop for each iteration. In
some cases, the
This fixes a 4.8/4.9 regression where we move from an lvalue when
constructing a packaged_task.
Tested x86_64-linux, committed to trunk and 4.8 branch.
commit b07f31e30bc3ca1b91e199c37466824bb7c3bafa
Author: Jonathan Wakely jwak...@redhat.com
Date: Tue Mar 18 13:23:52 2014 +
PR
Hi,
I'm adding the below testcases for bugs already fixed in mainline and 4.8.x.
Thanks,
Paolo.
//
2014-03-18 Paolo Carlini paolo.carl...@oracle.com
PR c++/54250
* g++.dg/cpp0x/lambda/lambda-ice12.C: New.
Index: g++.dg/cpp0x/constexpr-ice14.C
On 17 Mar 16:00, Thomas Schwinge wrote:
GOMP_4.0 {
global:
+ GOMP_offload_register;
GOMP_barrier_cancel;
GOMP_cancel;
GOMP_cancellation_point;
Now that the GOMP_4.0 symbol version is being used in GCC trunk, and will
be in the GCC 4.9 release, can we still add new
Committed to branch dmalcolm/jit:
gcc/testsuite/
* jit.dg/test-combination.c: Add test-arrays.c and test-volatile.c.
Add comment about test-error-*.c. Remove comment about
test-failure.c, which was removed in
96b218c9a1d5f39fb649e02c0e77586b180e8516.
Hi!
While most of the sanitization is done in the FEs (c-family/),
some is done in the middle-end (ubsan pass) or when folding
__builtin_unreachable ().
This patch fixes ICE with that, bootstrapped/regtested on x86_64-linux and
i686-linux, committed to trunk as obvious.
2014-03-18 Jakub
On Tue, 2014-03-18 at 08:53 -0400, Diego Novillo wrote:
OK with:
+Pragmas registered with @code{c_register_pragma_with_expansion} or
+@code{c_register_pragma_with_expansion_and_data} are supporting
+preprocessor expansions. For an example of using such a pragma:
s/are supporting/support/
Hi,
the attached is slightly larger than my last ones but, assuming the
analysis is correct, should be pretty safe, it avoids a segfault: in
fact, for a pure virtual called from a NSDMI, current_function_decl is
null and the existing warning about *structors breaks. I'm reading
12.7/4 and I
Patch got approved in bug-report for trunk.
Applied at revision 208663 to trunk.
Kai
Brown bag time on this one. While I did build and test both x86_64 and i686
separately, I apparently only missed the regression in test results.
Anyway, a simple fix for the problem: don't enable the pattern when it's not
supposed to be enabled due to excess precision. Moving the _387 pattern
On Tue, 2014-03-18 at 08:57 -0600, Tom Tromey wrote:
This patch introduces a new class toplev and changes toplev_main and
toplev_finalize to be methods of this class. Additionally, now the
timevars are automatically stopped when the object is destroyed. This
cleans up compile a bit and makes
On Tue, 2014-03-18 at 08:57 -0600, Tom Tromey wrote:
This introduces a new scoped_timevar class. It pushes a given timevar
in its constructor, and pops it in the destructor, giving a much
simpler way to use timevars in the typical case where they can be
scoped.
---
gcc/ChangeLog.jit |
Hi!
With -fno-tree-dce the scalar MASK_LOAD isn't removed from the IL and we ICE
on it during expansion (as we support only the vector loads, if those aren't
supported, MASK_LOAD is either not created by if-conversion at all, or
vectorization refuses to vectorize the loop and thus it is cfg
* Merge from the trunk: r208408 bis r208666.
Committed to the Fortran-caf branch, Rev. 208667.
Tobias
On 03/18/2014 04:59 AM, Jakub Jelinek wrote:
On Mon, Mar 17, 2014 at 03:24:14PM -0400, Vladimir Makarov wrote:
It is complicated. There is no guarantee that it is used only for
addresses. I need some time to think how to fix it.
Meanwhile, you *should* commit the patch into the trunk
It seems that the patch was tested only in the C testsuite, but the
changes to the testcase fail for C++. So I committed the following,
which is what my patch should have done in the first place, as
obvious.
I hope this is OK with you.
Cheers,
Manuel.
Index: gcc/testsuite/gcc.dg/cast-qual-3.c
On Mar 18, 2014, Jakub Jelinek ja...@redhat.com wrote:
--- a/gcc/function.c
+++ b/gcc/function.c
@@ -4498,6 +4498,8 @@ allocate_struct_function (tree fndecl, bool abstract_p)
cfun = ggc_alloc_cleared_function ();
+ SET_BUILD_DEBUG_STMTS (cfun, flag_var_tracking_assignments);
Dunno how
gcc/ChangeLog.jit | 14 +
David I see the Changelog.jit file listed here...
gcc/diagnostic.c | 2 +-
gcc/jit/ChangeLog.jit | 5 +
David ...and here, but I don't see the content below, or within the header of
David the email. Is this available somewhere?
Sorry
On Tue, Mar 18, 2014 at 08:57:44AM -0600, Tom Tromey wrote:
This introduces a new scoped_timevar class. It pushes a given timevar
in its constructor, and pops it in the destructor, giving a much
simpler way to use timevars in the typical case where they can be
scoped.
thanks for doing this.
PR c++/60573
* parser.c (synthesize_implicit_template_parm): Handle the fact that
nested class member declarations erroneously appearing in an enclosing
class contain an addition scope level for the class being defined.
PR c++/60573
*
Among other issues, PR 60563 complains that g++.dg/ext/sync-4.c fails on
Darwin because both the system header files and the test define the
typedef int64_t. This patch should avoid that problem by renaming the
typedef. Tested on x86_64-unknown-linux-gnu. Committed to mainline.
Ian
78 matches
Mail list logo