On Dec 19, 2011, at 7:11 PM, Ian Lance Taylor wrote:
> Tristan Gingold writes:
>
>> 2011-12-19 Tristan Gingold
>>
>> * aclocal.m4: Assume strncmp works in cross case.
>> * configure: Regenerate.
>
> This is OK.
Thanks, committed.
Tristan.
> The following two patches are necessary to allow a i?86-linux
> --enable-targets=all bootstrap to complete, but have remained
> unreviewed
> for a week:
>
> [ada] Support 64-bit libgnat multilib on i?86-linux
> http://gcc.gnu.org/ml/gcc-patches/2011-12/msg01009.html
This part is OK
Hi,
the alpha/vms backend emitted the trampoline entry point too often, as
functions declared directly within a translation unit weren't considered as
top-level functions.
Fixed by this patch, applied on trunk.
Tristan.
2011-12-20 Tristan Gingold
* config/alpha/alpha.c (alpha_star
Hi,
currently alpha/vms backend emits a trampoline entry point for all nested
functions. This is a waste of code space, as although nested functions are
very common in Ada, address of nested functions are only seldom taken.
The fact that the address of a function is taken seems only be availab
PR libstdc++/51365
* include/std/tuple (_Tuple_impl): Check __is_final as well as
is_empty.
* testsuite/20_util/tuple/51365.cc: New.
This uses __is_final for std::tuple, it still needs to be used for
containers, shared_ptr and futures.
Tested x86_64-linux, committe
On Mon, 19 Dec 2011, Patrick Marlier wrote:
> On 12/16/2011 03:54 AM, Richard Guenther wrote:
> > On Thu, 15 Dec 2011, Patrick Marlier wrote:
> >
> > > In PR51280, LTO does ICE because the object file uses TM builtin but the
> > > TM is
> > > not enabled.
> > > In the patch, it displays a error m
On Mon, Dec 19, 2011 at 10:39 PM, Richard Sandiford
wrote:
> The recent(ish) improvements to widening multiplication support have
> disabled madd and msub for fixed-point types. The problem is that the
> optab is now chosen based on:
>
> optype = build_nonstandard_integer_type (from_mode, from_u
On Tue, Dec 20, 2011 at 9:37 AM, Tristan Gingold wrote:
> Hi,
>
> the alpha/vms backend emitted the trampoline entry point too often, as
> functions declared directly within a translation unit weren't considered as
> top-level functions.
>
> Fixed by this patch, applied on trunk.
>
> Tristan.
>
On Dec 20, 2011, at 10:57 AM, Richard Guenther wrote:
> On Tue, Dec 20, 2011 at 9:37 AM, Tristan Gingold wrote:
>> Hi,
>>
>> the alpha/vms backend emitted the trampoline entry point too often, as
>> functions declared directly within a translation unit weren't considered as
>> top-level funct
Hi,
this ICE - which leads to worse diagnostics - seems caused by a trivial
typo: to mean value initialization diagnose_non_constexpr_vec_init is
passing zoid_zero_node instead of void_type_node to build_vec_init_elt.
My guess is also confirmed by the beginning of the other caller,
build_vec_
On 12/19/2011 11:06 PM, Richard Sandiford wrote:
I've normally tried to avoid self-approving rtl optimiser stuff,
but since this is a simple extension of the existing MEM handling,
and since the mode_rep_extended test effectively makes it MIPS-specific,
I hope it's OK to make an exception here.
Eric Botcazou writes:
>> 2011-12-13 Rainer Orth
>>
>> * gcc-interface/Makefile.in (%86 linux%):
>> (LIBGNAT_TARGET_PAIRS_32): Split off from LIBGNAT_TARGET_PAIRS.
>> (LIBGNAT_TARGET_PAIRS_64): New.
>> (LIBGNAT_TARGET_PAIRS): Add either depending on multilib.
>
> OK, thanks.
This patch updates s-stposu.ads and s-stposu.adb to reflect the final version
of the Ada 2012 Reference Manual.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-12-20 Hristian Kirtchev
* sem_res.adb (Resolve_Allocator): Warning on allocation
of tasks on a subpool and rew
The RM forbids local names from being renamings, hence the
attempt to specify an attribute such as size or alignment
for a renaming should be illegal. We detected this for the
case of an address clause but missed many other cases.
The following should compile with the messages shown
with -gnatj60
This change prevents the compiler from hanging in the presence of a WITH
clause on a child unit that is an illegal generic instantiation. The following
compilation must be rejected:
$ gcc -c dist_monolithic_app.adb
worker-registry.ads:4:51: expect valid subtype mark to instantiate "Handle"
worker
New GNAT library packages that defines the MKS dimension system (SI system of
units) and provide IO routines for this dimension system.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-12-20 Vincent Pucci
* impunit.adb: s-dimkio, s-dimmks and s-dmotpr defined as GNAT
Def
When the Class attribute is applied in one package that has a limited view of
a tagged type with an untagged incomplete type in its own package, then when
another unit with a nonlimited with clause for the type's package is compiled,
the compiler issues an error complaining that the prefix must be
This patch corrects the machinery which identifies an object as being a
transient variable. Objects which denote Ada containers in the context of
iterator loops are not considered transients and now share the life time of the
related loop.
-- Source --
pragma Ada_2012;
w
This patch corrects the expansion of attribute 'Width when applied to a
dynamic enumeration type with suppressed names.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-12-20 Hristian Kirtchev
* exp_imgv.adb (Expand_Width_Attribute): Add a
type conversion from the enumer
This patch changes the default from -gnatw.l to -gnatw.L. That is, inherited
aspects are no longer listed by default.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-12-20 Bob Duff
* opt.ads (List_Inherited_Aspects): Default to False
(i.e. -gnatw.L is the default).
This patches handles complex cases of nested package instantiations, and
determines properly the placement of freeze nodes for instantiations of
generic units that are themselves declared in a previous instantiaion in the
same compilation unit.
The following must compile quietly:
gcc -c ppp.ad
Hello,
The testcase attached causes ICE when compiling with
-fmodulo-sched-allow-regmoves on ARM due to reg-moves created for the
definition of mode MODE_CC.
The following is a snippet from the ddg of the definition and use of vfpcc
which triggers the creation of the reg-move:
Node num: 1
(insn
OK.
Jason
On 17/12/11 00:30, Richard Henderson wrote:
> This follows the other logical patterns, andsi3 and iorsi3, which
> also accept reg_or_int_operand and split post-reload.
>
> For the atomic optabs, I have two choices: either rely on this
> post-reload splitting, or call arm_split_constant myself. M
On 17/12/11 00:26, Richard Henderson wrote:
> I noticed this while testing my atomic optabs patch with -mthumb.
> This is identical to what we actually output in the current sync code:
>
>> arm_output_asm_insn (emit, 0, operands, "cmp\t%%0, %%1");
>> if (is_di)
>> {
>>
Ian Lance Taylor writes:
> It's possible that this patch will once again break the Solaris and Irix
> support. I've tried to ensure that I didn't make any stupid errors, but
> I haven't done any actual testing. Sorry about any problems.
Now IRIX finally bootstrap again, I had to make two adjus
On 11/29/11 17:35, Mitchell, Mark wrote:
>>> So, I still think this patch is the best way to go forward, and it
>> does
>>> fix incorrect code generation. Would appreciate an OK.
>>
>> Ping.
>
> If you don't hear any objections within a week, please proceed.
Committed now after bootstrapping i686
Some patch within the last week has broken Tru64 UNIX Ada bootstrap:
s-taprop.adb:46:12: warning: no entities of "Os_Constants" are referenced
s-taprop.adb:58:04: warning: no entities of "OSC" are referenced
s-taprop.adb:594:35: "CLOCK_RT_Ada" not declared in "OS_Constants"
make[6]: *** [s-taprop.
Andreas Tobler pointed out that passing a string with leading dash as
argument to expr is non-portable. Checked in as obvious.
Andreas.
2011-12-20 Andreas Schwab
* warnings.m4 (ACX_PROG_CC_WARNING_OPTS): Avoid leading dash in
expr call.
diff --git a/config/warnings.m4 b/conf
The problem here is a verify_gimple ICE because a vector store's LHS is
not a gimple register, but the vector type is a gimple_reg_type.
if (!is_gimple_reg (lhs)
&& is_gimple_reg_type (TREE_TYPE (lhs)))
{
error ("invalid rhs for gimple memory store");
d
> Some patch within the last week has broken Tru64 UNIX Ada bootstrap:
>
> s-taprop.adb:46:12: warning: no entities of "Os_Constants" are
> referenced
> s-taprop.adb:58:04: warning: no entities of "OSC" are referenced
> s-taprop.adb:594:35: "CLOCK_RT_Ada" not declared in "OS_Constants"
> make[6]:
Revital Eres writes:
> +/* Return true if one of the definitions in INSN has MODE_CC. Otherwise
> + return false. */
> +static bool
> +def_has_ccmode_p (rtx insn)
> +{
> + df_ref *def;
> +
> + for (def = DF_INSN_DEFS (insn); *def; def++)
> +{
> + enum machine_mode mode = GET_MODE (D
Rainer Orth writes:
> Now IRIX finally bootstrap again, I had to make two adjustments to have
> libgo build there.
>
> * go/syscall/wait.c doesn't compile:
>
> /vol/gcc/src/hg/trunk/local/libgo/go/syscall/wait.c:11:0: error:
> "__EXTENSIONS__" redefined [-Werror]
> /vol/gcc/src/hg/trunk/local/li
Rainer Orth writes:
> * There's a redefinition of IPMreq now:
>
> sysinfo.go:3089:6: error: redefinition of 'IPMreq'
> /vol/gcc/src/hg/trunk/local/libgo/go/syscall/socket_irix.go:81:6: note:
> previous definition of 'IPMreq' was here
>
> For some reason, sysinfo.go now gets the right definitio
I realized that I didn't remove uses of _GNU_SOURCE from libgo/runtime.
They are no longer necessary now that it is defined on the command line
for all libgo .c files. Bootstrapped on x86_64-unknown-linux-gnu.
Committed to mainline.
Ian
diff -r cf7ab235f776 libgo/runtime/runtime.h
--- a/libgo/ru
Hello Everyone,
This patch affects the C-Compiler in the Cilkplus branch. This patch will
add a test case for the N-Dimension array notation feature.
Thanking You,
Yours Sincerely,
Balaji V. iyer.diff --git a/gcc/testsuite/ChangeLog.cilk b/gcc/testsuite/ChangeLog.cilk
index 6fef5a8..e4df2
Hello,
> FWIW, an alternative might be to test have_regs_of_mode[(int) mode].
> That says whether there are any allocatable (non-fixed) registers
> of the given mode.
Thanks, I'll prepare a new version of the patch using have_regs_of_mode.
Revital
>
> Richard
Hi
This patch make -D and -U work in the spec language, bug pr48524.
Tested on x86_64-unknown-linux-gnu snapshot 4.7-20111217
Magnus.
2011-12-18 Magnus Granberg
pr48524
* gcc/gcc.c (switch_matches) Support switches with
separated
Now that C11 has been published, this patch implements options
-std=c11, -std=iso9899:2011 and -std=gnu11 and deprecates -std=c1x and
-std=gnu1x. The intended final __STDC_VERSION__ value, 201112L, is
also implemented (the editor forgot to update the 201ymmL placeholders
before sending the documen
> "Ramana" == Ramana Radhakrishnan writes:
Ramana> Otherwise looks ok to me .
Sergio was asked me to ping this for him, but instead I re-read the
entire thread and I think that this was an approval conditional on him
making the request changes (just to ChangeLog), which he did.
So, I am che
Hi,
As a follow up of c++/14258, this one is about undiagnosed using
declarations whose nested-name-specifier is not a base.
The diagnostic was only emitted at instantiation time, whereas it
could be emitted before.
I just removed a test to perform the verification before
instantiation, it seems t
The following patch solves PR49865. The problem is described on
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49865
The patch changes code a bit only for gcc, crafty, and mesa on x86.
There is no visible change in the performance.
Only code for crafty is changed on x86-64 and improvement in the
Ping for this patch to add a prepare_pch_save target hook:
http://gcc.gnu.org/ml/gcc-patches/2011-12/msg00273.html
It fixes many PCH failures for MIPS16 (which are a regression from
earlier releases).
The patch gets rid of some locally-cached state before writing out the
PCH file. We have t
Anyone?
On Wed, Dec 14, 2011 at 1:10 PM, Brendan Conoboy wrote:
> On Mon, 12 Dec 2011 08:24:41 +0800, Doug Kwan wrote:
>> Sorry about my oversight. I am on vacation now until Dec 19. I don't
>> have good internet access now and I will backport this to upstream 4.6
>> after I come back if the 4.6
On Tue, Dec 20, 2011 at 02:35:29PM -0800, Brendan Conoboy wrote:
> Anyone?
This is ok for 4.6 if it has been sufficiently tested.
Jakub
> So, the patch causing this was reverted and bootstrap worked again(?),
> and now it is broken again with the same failure mode:
>
> cc1: error: unrecognized command line option "-Wno-narrowing"
>
> Actually, not the same, now it is invoking system gcc instead of
> system g++:
>
> gcc -I/scra
On Tue, 6 Dec 2011, Han Shen wrote:
> +gcc_gxx_include_dir_add_sysroot=0
> +if test "${with_sysroot+set}" = set; then :
> + gcc_gxx_without_sysroot=`expr "${gcc_gxx_include_dir}" :
> "${with_sysroot}"'\(.*\)'`
> + if test "${gcc_gxx_without_sysroot}"; then :
No need for the " :" after "then" i
On Tue, 20 Dec 2011, Magnus Granberg wrote:
> This patch make -D and -U work in the spec language, bug pr48524.
> Tested on x86_64-unknown-linux-gnu snapshot 4.7-20111217
Thanks for your contributions. As you've contributed before, this patch
brings things to about the point where a copyright a
When the gold linker sees a call from -fsplit-stack code to code
compiled without -fsplit-stack, it arranges for the function to call
__morestack_non_split instead of __morestack. The __morestack_non_split
function requests a larger stack, one that is large enough to run code
which doesn't know ab
On Tue, Dec 20, 2011 at 2:37 PM, Jakub Jelinek wrote:
> This is ok for 4.6 if it has been sufficiently tested.
Okay! Is there any testing you'd like to see beyond the
aforementioned success with arm and x86_64 linux?
-Brendan
As seen on ia64 with some of our explicit test cases for V4SI-alike operations.
On ia64 the actual variables get instantiated in TImode, not BLKmode as bits of
this code had been assuming. By checking !vector_mode rather than for BLKmode
directly, we can decompose the V4SI shift into two V2SImode
Some improvements to the previous versions, which include more
permutation patterns actually generated by the vectorizer in the
course of the testsuite.
Tested on ia64-linux and committed.
commit b155a608c2332b4bf64bcfe07ed059178044f97a
Author: rth
Date: Wed Dec 21 01:03:00 2011 +
ia6
Hi,
this is a rejects-valid with constexpr & noexcept, noticed by Daniel
(and myself time ago). I find it pretty annoying. Anyway, the issue is,
we reject:
constexpr bool ok() noexcept
{
typedef int type;
return true;
}
constexpr auto x = ok();
because of the noexcept. What happens is t
This seems problematic to me; it could be that a dependent scope ends up
matching a non-dependent base in the end, i.e.
struct A { int x; };
template
struct B: A
{
using T::x;
};
B b;
this code is silly, but I think well-formed, and that your patch will
break it.
I think that just chang
Reviewers: xur, bjanakiraman_google.com, jingyu,
Message:
Please review this backport.
Please review this at http://codereview.appspot.com/5501051/
Affected files:
Mgcc-4_6-mobile
M gcc-4_6-mobile/gcc/ChangeLog.google-4_6
M gcc-4_6-mobile/gcc/common.opt
M gcc-4_6-mobi
On Wed, Dec 21, 2011 at 12:46 AM, Bernd Schmidt wrote:
> On 11/29/11 17:35, Mitchell, Mark wrote:
So, I still think this patch is the best way to go forward, and it
>>> does
fix incorrect code generation. Would appreciate an OK.
>>>
>>> Ping.
>>
>> If you don't hear any objections within
This should be in 4.6 too
- Joey
On Wed, Dec 21, 2011 at 12:46 AM, Bernd Schmidt wrote:
> On 11/29/11 17:35, Mitchell, Mark wrote:
So, I still think this patch is the best way to go forward, and it
>>> does
fix incorrect code generation. Would appreciate an OK.
>>>
>>> Ping.
>>
>> If y
Adding shenhan.
http://codereview.appspot.com/5501051/
Hi,
On Tue, Aug 2, 2011 at 6:32 PM, Richard Henderson wrote:
> I got Jeff Law to review the reload change on IRC
> and committed the composite patch.
>
> Tested on x86_64, i586, avr, and h8300. Most other
> tier1 targets ought not be affected, as this patch
> only applies to ACCUMULATE_OUTGOING_
LGTM++
On Tue, Dec 20, 2011 at 8:22 PM, wrote:
>
> Adding shenhan.
>
> http://codereview.appspot.com/5501051/
On 20.12.11 23:35, Eric Botcazou wrote:
So, the patch causing this was reverted and bootstrap worked again(?),
and now it is broken again with the same failure mode:
cc1: error: unrecognized command line option "-Wno-narrowing"
Actually, not the same, now it is invoking system gcc instead of
> -Original Message-
> From: Richard Henderson [mailto:r...@redhat.com]
> Sent: Tuesday, November 22, 2011 9:55 AM
> To: Jiangning Liu
> Cc: gcc-patches@gcc.gnu.org; 'Richard Guenther'
> Subject: Re: [RFC] Use REG_EXPR in back-end (introduced by optimization
> to conditional and/or in ARM
On Tue, Dec 20, 2011 at 04:04:16PM -0800, Brendan Conoboy wrote:
> On Tue, Dec 20, 2011 at 2:37 PM, Jakub Jelinek wrote:
> > This is ok for 4.6 if it has been sufficiently tested.
>
> Okay! Is there any testing you'd like to see beyond the
> aforementioned success with arm and x86_64 linux?
Tha
63 matches
Mail list logo
