On Thu, 20 Mar 2014, Martin Jambor wrote:
> Hi,
>
> in PR 60419 we end up with a call graph node for a thunk that has no
> callee because symtab_remove_unreachable_nodes has determined its body
> is not needed although its declaration is still reachable (more
> details in comment 11 in bugzilla)
On Thu, 20 Mar 2014, Martin Jambor wrote:
> Hi,
>
> On Thu, Mar 20, 2014 at 07:40:56PM +0100, Jakub Jelinek wrote:
> > On Thu, Mar 20, 2014 at 05:07:32PM +0100, Martin Jambor wrote:
> > > in the PR, verifier claims an edge is pointing to a wrong declaration
> > > even though it has successfully v
2014-03-21 1:33 GMT+01:00 Paul Pluzhnikov :
> Greetings,
>
> Attached patch deletes code to warn about repeated friend declaration.
Why not making this warning suppressable, instead of removing it ?
Shouldn't it fall under -W(no)-redundant-decls ?
--
Fabien
This patch mentions that -flto now generates slim objects. That's especially
relevant for static libraries as one can there run into surprises, if one
does not know about gcc-ar.
OK - or do you have a better suggestion?
Tobias
--- changes.html8 Mar 2014 20:45:54 - 1.63
+++ cha
On Fri, 21 Mar 2014, Tobias Burnus wrote:
> This patch mentions that -flto now generates slim objects. That's especially
> relevant for static libraries as one can there run into surprises, if one
> does not know about gcc-ar.
>
> OK - or do you have a better suggestion?
Please adjust slightly t
This fixes a missed optimization regarding to coverage counters.
In the past we failed to apply proper may-alias analysis and
disregarded those counters as being possibly aliased by pointers
by accident. The following patch adds a flag so we can mark
VAR_DECLs as to be not considered as aliased b
On Fri, Mar 21, 2014 at 11:28:14AM +0100, Richard Biener wrote:
>
> This fixes a missed optimization regarding to coverage counters.
> In the past we failed to apply proper may-alias analysis and
> disregarded those counters as being possibly aliased by pointers
> by accident. The following patch
Hi Mircea,
sorry for making you wait.
--
Roman Gareev
patch
Description: Binary data
ChangeLog_entry
Description: Binary data
> So the patch below makes it so that if HOST_WIDE_INT is wide enough then,
> depending on whether the range type is signed or not, add_AT_unsigned or
> add_AT_int is used. This is more efficient for small ranges. And makes it
> so that the value can be deduced from the DW_FORM by the consumer (whi
On Fri, Mar 21, 2014 at 10:40:26AM +0100, Richard Biener wrote:
> On Fri, 21 Mar 2014, Tobias Burnus wrote:
> > This patch mentions that -flto now generates slim objects. That's especially
> > relevant for static libraries as one can there run into surprises, if one
> > does not know about gcc-ar.
On Fri, 21 Mar 2014, Tobias Burnus wrote:
> On Fri, Mar 21, 2014 at 10:40:26AM +0100, Richard Biener wrote:
> > On Fri, 21 Mar 2014, Tobias Burnus wrote:
> > > This patch mentions that -flto now generates slim objects. That's
> > > especially
> > > relevant for static libraries as one can there r
Hi,
On Fri, Mar 21, 2014 at 09:41:24AM +0100, Richard Biener wrote:
> On Thu, 20 Mar 2014, Martin Jambor wrote:
>
> > Hi,
> >
> > in PR 60419 we end up with a call graph node for a thunk that has no
> > callee because symtab_remove_unreachable_nodes has determined its body
> > is not needed alth
On Thu, Mar 20, 2014 at 9:38 PM, Bill Schmidt
wrote:
> Hi,
>
> The original workaround for vector permute on a little endian platform
> includes subtracting each element of the permute control vector from 31.
> Because the upper 3 bits of each element are unimportant, this was
> implemented as sub
On 19/03/14 16:53, Ian Bolton wrote:
> This is a follow-on patch to one already committed:
> http://gcc.gnu.org/ml/gcc-patches/2014-02/msg01128.html
>
> It implements patterns to simplify our RTL as follows:
>
> OR (Not:DI (A:DI), ZeroExtend:DI (B:SI))
> --> the top half can be done with a MVN
On Fri, Mar 21, 2014 at 1:58 AM, Fabien Chêne wrote:
> Why not making this warning suppressable, instead of removing it ?
> Shouldn't it fall under -W(no)-redundant-decls ?
Thanks. I'll revise the patch to do that.
--
Paul Pluzhnikov
This patch fixes a bug whereby a vector like V8QImode {1,0,1,0,1,0,1,0} can
result in an instruction like
movi v1.4h, 0x1
whereas on bigendian this should be
movi v1.4h, 0x1, lsl 8
Regression tested on aarch64_be-none-elf: no changes in libstdc++, newlib; no
regressions in gcc or g++ and FAI
On 03/20/2014 07:56 PM, Jakub Jelinek wrote:
When we were discussing the design last year, my strong preference was that
either this lives in some other crt object that mkoffload/linker plugin adds
to link, or that it would be completely mkoffload synthetized.
mkoffload is only concerned with g
Let's assert errorcount|sorrycount before returning in the !identifier
case. OK with that change.
Jason
On Fri, Mar 21, 2014 at 04:13:45PM +0100, Bernd Schmidt wrote:
> On 03/20/2014 07:56 PM, Jakub Jelinek wrote:
> >When we were discussing the design last year, my strong preference was that
> >either this lives in some other crt object that mkoffload/linker plugin adds
> >to link, or that it would b
On 03/21/2014 04:20 PM, Jakub Jelinek wrote:
And, what is the exact reason why you are using protected visibility rather
than hidden?
Also, supposedly if you've used section names without . in them, the linker
itself would provide the symbols automatically and you wouldn't actually
need begin/end
Greetings,
To fix PR59295, this patch moves (generally useless) warning about
repeated / redundant friend declarations under -Wredundant-decls.
Tested on Linux/x86_64 with no regressions.
Ok for trunk once it opens in stage 1?
Thanks,
--
2014-03-21 Paul Pluzhnikov
PR c++/59295
The following patch has lived on mainline for 6 months and has
not generated any issues there. We've also been using it on
our 4.8 based IBM branch with no problems either, so I'd like to
ask for permission to backport this fix to the FSF 4.8 branch.
This will bring GCC into compliance with CLANG
Hi,
I'm committing this as obvious.
Thanks,
Paolo.
2014-03-21 Paolo Carlini
* g++.dg/cpp1y/pr60033.C: Use target c++1y.
* g++.dg/cpp1y/pr60393.C: Likewise.
Index: g++.dg/cpp1y/pr60033.C
===
>
> This fixes PR59543 (confirmed by Jakub for the testcase at least)
> by not dropping debug stmts during WPA phase.
>
> LTO profiled-bootstrapped on x86_64-unknown-linux-gnu, applied.
>
> Honza - you can always come up with a better fix for 4.10.
I guess this may work well. Other option (as m
While looking at an unrelated issue, I noticed that the gcc/macro_list
file is empty. In the build logs, I see
echo | /var/gcc/regression/trunk/10-gcc/build/./gcc/xgcc
-B/var/gcc/regression/trunk/10-gcc/build/./gcc/ -E -dM - | \
sed -n -e 's/^#define \([^_][a-zA-Z0-9_]*\).*/\1/p' \
This patch guards autofdo annotation coverage recording with a flag.
Test on-going.
OK for google-4_8 if test passes?
Thanks,
Dehao
Index: gcc/auto-profile.c
===
--- gcc/auto-profile.c (revision 208753)
+++ gcc/auto-profile.c (work
ok to remove the empty directory gcc/testsuite/g++.dg/cpp0x/regress on the
trunk?
Matthias
On Fri, 2014-03-21 at 11:30 -0500, Peter Bergner wrote:
> The following patch has lived on mainline for 6 months and has
> not generated any issues there. We've also been using it on
> our 4.8 based IBM branch with no problems either, so I'd like to
> ask for permission to backport this fix to the
On Wed, Mar 19, 2014 at 02:39:22PM -0500, Bill Schmidt wrote:
> Hi,
>
> The main patch for this series was too large for the mailer to accept.
> Sorry about that. This piece is all powerpc-related and seems to have
> been delivered to David ok. If anyone else wants a copy of the patch,
> please
On 20/03/14 20:29 +, Jonathan Wakely wrote:
Everything passes for "make check check-debug", so I plan to commit it
tomorrow, unless anyone points out a problem.
Committed to trunk.
PR libstdc++/60587
* include/debug/functions.h (_Is_contiguous_sequence): Define.
(
On 03/20/2014 06:38 PM, Bill Schmidt wrote:
> - rtx splat = gen_rtx_VEC_DUPLICATE (V16QImode,
> - gen_rtx_CONST_INT (QImode, -1));
> + rtx notx = gen_rtx_NOT (V16QImode, force_reg (V16QImode, x));
> + rtx andx = gen_rtx_AND (V16QImode, notx, notx)
On Mar 20, 2014, at 1:48 PM, Steve Kargl
wrote:
> On Thu, Mar 20, 2014 at 08:24:49PM +0100, Tobias Burnus wrote:
>> gfortran's modules are since GCC 4.9 zipped. There are two functions,
>> which test for the existence and absent of strings in the .mod files.
>
>> PR fortran/60599
>> *
On Mar 21, 2014, at 10:46 AM, Matthias Klose wrote:
> ok to remove the empty directory gcc/testsuite/g++.dg/cpp0x/regress on the
> trunk?
Ok.
Hi!
While the cpp_force_token_locations/cp_stop_forcing_token_locations
pair forces BUILTINS_LOCATION upon tokens, the change introducing them
removed cb_file_change/linemap_add, which is needed e.g. for proper
line numbers of builtin defines in -g3 .debug_macro/.debug_macinfo.
Fixed by reverting
Hi!
As MINUS_EXPR is not commutative, we really can't swap op0 with op1
for testing whether subtraction overflowed, that is only possible for
PLUS_EXPR, for MINUS_EXPR we really have to know if op1 is constant
or negative or non-negative and have to compare result with op0 depending on
that.
Boot
> This patch mentions that -flto now generates slim objects. That's especially
> relevant for static libraries as one can there run into surprises, if one
> does not know about gcc-ar.
>
> OK - or do you have a better suggestion?
>
> Tobias
>
>
> --- changes.html8 Mar 2014 20:45:54 -000
Hi!
Prior to r203634 we were comparing TARGET_64BIT with
ix86_isa_flags & OPTION_MASK_ISA_64BIT, which is the same thing for
TARGET_BI_ARCH, otherwise the former is hardcoded constant.
But with r203634, the condition was changed and is now always false
and so e.g. for 32-bit non-multilib i?86 gcc
> On Thu, 20 Mar 2014, Martin Jambor wrote:
>
> > Hi,
> >
> > On Thu, Mar 20, 2014 at 07:40:56PM +0100, Jakub Jelinek wrote:
> > > On Thu, Mar 20, 2014 at 05:07:32PM +0100, Martin Jambor wrote:
> > > > in the PR, verifier claims an edge is pointing to a wrong declaration
> > > > even though it ha
Hi!
--with-build-config=bootstrap-ubsan on i686-linux showed
most often these two issues.
The first one is that 32-bit signed time_t is multiplied by 1000, which
overflows the int type. The result is then cast to unsigned, so even on
64-bit hosts we don't care about the upper bits.
The other ch
ping ^2...
Dehao
On Mon, Feb 10, 2014 at 8:35 AM, Dehao Chen wrote:
> ping...
>
> Dehao
>
> On Fri, Jan 24, 2014 at 1:54 PM, Dehao Chen wrote:
>> Thanks, test updated:
>>
>> Index: gcc/testsuite/gcc.dg/predict-8.c
>> ===
>> --- gcc
Hi,
This patch updates node's inline summary after edge_summary is
updated. Otherwise it could lead to incorrect inline summary.
Bootstrapped and gcc regression test on-going.
OK for trunk?
Thanks,
Dehao
gcc/ChangeLog:
2014-03-21 Dehao Chen
*ipa-inline.c (early_inliner): updates overall s
On 03/21/2014 01:38 PM, Jakub Jelinek wrote:
> PR target/60610
> * config/i386/i386.h (TARGET_64BIT_P): If not TARGET_BI_ARCH,
> redefine to 1 or 0.
> * config/i386/darwin.h (TARGET_64BIT_P): Redefine to
> TARGET_ISA_64BIT_P(x).
Ok.
r~
Hi
Here is a patch to fix _Hashtable Standard extension type which is
almost unusable at the moment if instantiated with anything else that
the types used for the std unordered containers that is to say
__detail::_Default_ranged_hash and __detail::_Mod_range_hashing.
It is a really s
Committed revision 208759.
Index: io/transfer.c
===
--- io/transfer.c (revision 208755)
+++ io/transfer.c (working copy)
@@ -2674,7 +2674,8 @@ data_transfer_init (st_parameter_dt *dtp, int read
if (dtp->u.p.current_unit
On 21/03/14 22:39 +0100, François Dumont wrote:
Hi
Here is a patch to fix _Hashtable Standard extension type which
is almost unusable at the moment if instantiated with anything else
that the types used for the std unordered containers that is to say
__detail::_Default_ranged_hash and __de
On Fri, Mar 21, 2014 at 10:13 PM, Dehao Chen wrote:
> ping ^2...
Assuming this concerns
http://gcc.gnu.org/ml/gcc-patches/2014-01/msg01460.html and
follow-ups.
OK.
Ciao!
Steven
On Fri, 21 Mar 2014, Jakub Jelinek wrote:
> Hi!
>
> While the cpp_force_token_locations/cp_stop_forcing_token_locations
> pair forces BUILTINS_LOCATION upon tokens, the change introducing them
> removed cb_file_change/linemap_add, which is needed e.g. for proper
> line numbers of builtin defines
On 21/03/14 22:59 +, Jonathan Wakely wrote:
On 21/03/14 22:39 +0100, François Dumont wrote:
It is a really safe patch so I would propose it for current
trunk but at the same time it only impacts a Standard extension and
it hasn't been reported by anyone so just tell me when to apply it.
On Thu, 2014-03-20 at 08:53 -0600, Tom Tromey wrote:
> This patch initializes input_location at the same spot where the line
> table is initialized. Without this, it's possible to crash when
> emitting a diagnostic in a reinvocation of the compiler, because
> input_location refers to a location th
On Fri, Feb 28, 2014 at 9:13 AM, Teresa Johnson wrote:
Here's the new patch. The only changes from the earlier patch are in
handle_missing_profiles, where we now get the counts off of the entry
and call stmt bbs, and in tree_profiling, where we call
handle_missing_profiles earl
Minor dumping fix.
Bootstrapped and tested on x86_64-unknown-linux-gnu. Ok for stage 1?
Thanks,
Teresa
2014-03-21 Teresa Johnson
* ipa-utils.c (ipa_print_order): Use specified dump file.
Index: ipa-utils.c
===
--- ipa-ut
We were failing to give an error for B::foo not having a definition to
deduce the return type from; an easy way to avoid the ICE is to promote
the existing permerror about 'virtual auto' to a full error.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 62fa8dfdd93b964246d06f3d5b41a3c265950
52 matches
Mail list logo
