> The following patch fixed the bootstrap for me, will commit it
> as obvious once also regtesting finishes.
>
> 2011-08-29 Jakub Jelinek
>
> * gcc-interface/Makefile.in (../stamp-gnatlib1-$(RTSDIR)): Copy
> tsystem.h into $(RTSDIR) instead of rts.
Patch is certainly fine, thanks
Hi,
This is a ping.
Is the patch ok for trunk?
Thanks, K
On Fri, Aug 26, 2011 at 5:52 PM, H.J. Lu wrote:
> On Fri, Aug 26, 2011 at 6:45 AM, Kirill Yukhin
> wrote:
>> Hi guys,
>> Thanks for your objections.
>>
>> HJ, I scanned all AVX2 tests. So, every tests has at least which
>> distinguishe
The lambda capture path in finish_id_expression wasn't calling mark_used
like the other paths.
Tested x86_64-pc-linux-gnu, applying to trunk and 4.6.
commit 0aceb42a8ba7bde23447d6ae74c8615721a40952
Author: Jason Merrill
Date: Mon Aug 29 11:23:56 2011 -0400
PR c++/50224
* semantics.
DR 994 added default arguments to the set of places that a
brace-enclosed initializer list can appear in C++11. This patch
implements that.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 5ba9be05f72a270674e434e447d50c4b8f6d30f8
Author: Jason Merrill
Date: Mon Aug 29 00:54:22 2011 -04
The DFP classes are treated as transparent aggregates, so they get
mangled and passed like their first field. In the testcase it is
defined to have a base class, so the first field is an artificial base
field, which doesn't make any sense.
Tested x86_64-pc-linux-gnu, applying to trunk.
commi
This patch adds new tests for overload resolution. In contrast to
prior tests, we want these tests to have assembly differences because
PPH binds overloads at header compilation time, not based on accident
of include order. The test control files change as a consequence.
Changed the compiler com
"H.J. Lu" writes:
> On Mon, Aug 29, 2011 at 12:44 PM, Iyer, Balaji V
> wrote:
>> Hello Everyone,
>>
>> Here is a link for a new patch (http://software.intel.com/file/38290). This
>> patch is for the "cilkplus" branch and includes the following modifications:
>>
>> 1) Merges changes in the gcc
Hi
This is a patch for the explicit vector shuffling we have discussed a
long time ago here:
http://gcc.gnu.org/ml/gcc-patches/2010-08/msg01092.html
The new patch introduces the new tree code, as we agreed, and expands
this code by checking the vshuffle pattern in the backend.
The patch at the m
Recently, we ran "make check" on the GCC built from the GUPC branck
and compared it against "make check" run against the GCC trunk version
of the most recent merge with the trunk. The following failure
was detected on the GUPC branch.
cc1: internal compiler error: in print_specific_help, at opts
The attached patches are several changes required for a new TM method,
gl_wt (global lock, write-through), which is added by the last patch
patch1: Add TM-method-specific begin code. All time-based TMs need to
know at which point in time they start working. Initializing lazily on
the first txnal l
This patch to the Go frontend lowers calls to bound method expressions,
rather than handling them when converting to GENERIC. Bootstrapped and
ran Go testsuite on x86_64-unknown-linux-gnu. Committed to mainline.
Ian
diff -r 5097a3ef0927 go/expressions.cc
--- a/go/expressions.cc Mon Aug 29 15:00
The Go frontend was lowering expressions in blocks twice: once the usual
way, and once by traversing all subexpressions of a block statement.
The main effect of this was to waste compilation time. This patch fixes
it. Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu.
Committed to mai
>> The patch was regtested on x86_64-unknown-linux-gnu. Ok for trunk?
>
> OK. Looks rather obvious.
Thanks. Committed as r178262.
Cheers,
Janus
>> 2011-08-29 Janus Weil
>>
>> PR fortran/50225
>> * trans-decl.c (gfc_generate_function_code): Nullify polymorphic
>> allocatab
David asked me to reverse the default for the -msave-toc-indirect option, since
it is buggy in a few cases, and in other places causes slowdowns if the
function has an early exit before calling the indrect function. I also
documented the switch.
I did the usual bootstrap and make check with no re
Am 29.08.2011 22:28, schrieb Janus Weil:
Hi all,
here is a relatively simple patch which fixes the PR in the subject
line. The problem was: For functions with polymorphic allocatable
result, the result variable was not nullified at the start.
The patch was regtested on x86_64-unknown-linux-gnu.
On 8/29/2011 4:50 PM, Paolo Carlini wrote:
.. also, you forgot to add 2011 to the Copyright years.
Paolo.
In the GNAT development environment we have an SVN style
checking filter, and this is one of the things it checks
for so we prevent any checkin missing the current year in
the copyright no
Hi!
As the following testcase shows, sometimes lp->landing_pad may be a
NOTE_DELETED_LABEL note and in that case
find_rarely_executed_basic_blocks_and_crossing_edges crashes.
for_each_eh_label in except.c does:
rtx lab = lp->landing_pad;
if (lab && LABEL_P (lab))
(*
.. also, you forgot to add 2011 to the Copyright years.
Paolo.
On 08/29/2011 09:58 PM, Kai Tietz wrote:
Done and added missing _GLIBCXX_HAVE_ for all checks. Committed at
revision 178257.
Something seems wrong here, in terms of commas:
too_many_links = EMLINK
//too_many_symbolic_link_levels = ELOOP,
#ifdef _GLIBCXX_HAVE_EOVERFLOW
Hi!
One of the Ada changes committed today, most likely
http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=178176
(couldn't find it posted on gcc-patches) broke x86_64-linux
bootstrap, with errors like:
ctrl_c.c:34:21: fatal error: tsystem.h: No such file or directory
compilation terminated.
make[9]
Hi all,
here is a relatively simple patch which fixes the PR in the subject
line. The problem was: For functions with polymorphic allocatable
result, the result variable was not nullified at the start.
The patch was regtested on x86_64-unknown-linux-gnu. Ok for trunk?
Cheers,
Janus
2011-08-29
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/29/11 03:03, Jakub Jelinek wrote:
> Hi!
>
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01767.html PR
> middle-end/48722 Fix RTL sharing problem with
> CALL_INSN_FUNCTION_USAGE
>
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg00428.html Adjust
On Mon, Aug 29, 2011 at 12:44 PM, Iyer, Balaji V
wrote:
> Hello Everyone,
>
> Here is a link for a new patch (http://software.intel.com/file/38290). This
> patch is for the "cilkplus" branch and includes the following modifications:
>
> 1) Merges changes in the gcc master (SHA1:
> f326eb816922b
2011/8/29 Paolo Carlini :
> On 08/29/2011 10:54 AM, Pedro Alves wrote:
>>
>> While these two didn't get the HAVE_... prefix. Was that intended?
>
> To be clear: make sure before committing that *all* use _GLIBCXX_HAVE_*
>
> Paolo.
Done and added missing _GLIBCXX_HAVE_ for all checks.
Committed at
On Mon, 29 Aug 2011, Uros Bizjak wrote:
> On Mon, Aug 29, 2011 at 4:31 PM, Richard Guenther wrote:
>
> > This patch makes a conversion optab from the direct optabs vcond
> > and vcondu. This allows to specify different modes for the
> > actual comparison and the value that is selected.
> >
> >
Hello Everyone,
Here is a link for a new patch (http://software.intel.com/file/38290). This
patch is for the "cilkplus" branch and includes the following modifications:
1) Merges changes in the gcc master (SHA1:
f326eb816922bc183133c09b25564d550ab9a282).
2) Adds functionality to allow _Cilk_sp
On Mon, Aug 29, 2011 at 4:31 PM, Richard Guenther wrote:
> This patch makes a conversion optab from the direct optabs vcond
> and vcondu. This allows to specify different modes for the
> actual comparison and the value that is selected.
>
> All targets but i386 are trivially converted by
> s/vco
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/29/11 10:55, Vladimir Makarov wrote:
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01349.html
OK.
jeff
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/
iQEcBAEBAgAG
http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01349.html
On Mon, Aug 29, 2011 at 06:38:11PM +0200, Bernd Schmidt wrote:
> On 08/29/11 18:37, H.J. Lu wrote:
> > On Mon, Aug 29, 2011 at 9:36 AM, H.J. Lu wrote:
>
> >> Doesn't Google have a patch to change it to __x86.get_pc_thunk.bx?
> >>
> >
> > http://gcc.gnu.org/ml/gcc-patches/2011-04/msg02422.html
>
On 08/29/11 18:37, H.J. Lu wrote:
> On Mon, Aug 29, 2011 at 9:36 AM, H.J. Lu wrote:
>> Doesn't Google have a patch to change it to __x86.get_pc_thunk.bx?
>>
>
> http://gcc.gnu.org/ml/gcc-patches/2011-04/msg02422.html
Cool. I guess I'll approve and commit this version, unless someone sees
a reas
On Mon, Aug 29, 2011 at 9:36 AM, H.J. Lu wrote:
> On Mon, Aug 29, 2011 at 9:11 AM, Bernd Schmidt
> wrote:
>> We currently generate
>>
>> __i686.get_pc_thunk.bx:
>> movl (%esp), %ebx
>> ret
>> in PIC binaries. This can cause problems if the assembly output ends up
>> in a .S file
On Mon, Aug 29, 2011 at 9:11 AM, Bernd Schmidt wrote:
> We currently generate
>
> __i686.get_pc_thunk.bx:
> movl (%esp), %ebx
> ret
> in PIC binaries. This can cause problems if the assembly output ends up
> in a .S file which is then compiled again: __i686 is a predefined macro
>
On 07/29/2011 07:55 PM, John David Anglin wrote:
Ping?
Hi Dave,
ltmain.sh is a generated file, so this patch is not correct. Perhaps
something like (pasted in mail, so wrapped):
diff --git a/libltdl/config/ltmain.m4sh b/libltdl/config/ltmain.m4sh
index 9358ec5..bd5736c 100644
--- a/libltdl/
We currently generate
__i686.get_pc_thunk.bx:
movl(%esp), %ebx
ret
in PIC binaries. This can cause problems if the assembly output ends up
in a .S file which is then compiled again: __i686 is a predefined macro
and expands to "1". This happens in glibc when compiling csu/crti.S
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 08/26/11 08:47, Bernd Schmidt wrote:
> In rest_of_reorder_blocks, we avoid reordering if
> !optimize_function_for_speed_p. However, we still call
> insert_section_bounary_note, which can cause problems because now, if
> we have a sequence of HOT-C
The newly introduced target 'melt-tiny-tests' was buggy, this commit
fixes the issue by:
- splitting target in two targets, introducing the
'melt-sayhello.melt' target which produces the MELT code
corresponding, with a dependency on
$(melt_default_modules_list).modlis.
- adding a depende
Le 29/08/2011 17:22, Alexandre Lissy a écrit :
> Le 29/08/2011 17:08, Alexandre Lissy a écrit :
>> Le 29/08/2011 16:57, Alexandre Lissy a écrit :
>>> Le 29/08/2011 16:44, Alexandre Lissy a écrit :
Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
> On Mon, Aug 29, 2011 at 12:21:12PM +020
Le 29/08/2011 17:08, Alexandre Lissy a écrit :
> Le 29/08/2011 16:57, Alexandre Lissy a écrit :
>> Le 29/08/2011 16:44, Alexandre Lissy a écrit :
>>> Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
> Files containing the
Hi Paolo,
> Interesting. I suppose you are going to inform the SunStudio engineer
> that some of the above has to be adjusted for gcc 4.7+ (otherwise,
> please let me know, I can probably do it via Stephen Clamage)
that's the plan. I had some discussions about those issues when I
brought up th
On Mon, 29 Aug 2011, Richard Guenther wrote:
>
> This patch makes a conversion optab from the direct optabs vcond
> and vcondu. This allows to specify different modes for the
> actual comparison and the value that is selected.
>
> All targets but i386 are trivially converted by
> s/vcond/vcond
Le 29/08/2011 16:57, Alexandre Lissy a écrit :
> Le 29/08/2011 16:44, Alexandre Lissy a écrit :
>> Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
>>> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
Files containing the MELT translated to C code were not loaded correctly
>
On 08/23/11 12:12, Richard Sandiford wrote:
> Ping for:
>
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01181.html
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01182.html
Ok.
Bernd
2011/8/26 Georg-Johann Lay :
> Georg-Johann Lay wrote:
>> progmem_section is a section to put jump tables in.
>>
>> This patch puts jump tables in individual sections if
>> -ffunction-section is on and does some more cleanup around
>> that, i.e. implement TARGET_ASM_FUNCTION_RODATA_SECTION hook.
>>
I looked at the ISL sources and saw that they generally had license
notices of the form:
* Use of this software is governed by the GNU LGPLv2.1 license
For use in GCC I think it needs to have the "any later version" clause to
allow distribution under later versions of the LGPL as well, so that
Le 29/08/2011 16:44, Alexandre Lissy a écrit :
> Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
>> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>>> Files containing the MELT translated to C code were not loaded correctly
>>> due to bad search path when working with an insta
Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>> Files containing the MELT translated to C code were not loaded correctly
>> due to bad search path when working with an installed plugin. Present
>> commit fixes this by:
>> -
Formal verification relies on a computation of effects based on the local
effects generated in ALFA sections in ALI files. These effects should mention
explicitly all read/writes to memory through explicit dereferences. This is
now done with a special "HEAP" variable.
Tested on x86_64-pc-linux-gnu
This patch makes the compiler more robust in certain cases where "with X"
refers to a library-level instance that got errors that caused "instantiation
abandoned". Previously, the compiler would sometimes go into an infinite loop.
The following test should get errors:
gcc -c main.adb
p-i.ads:4:42:
This patch makes a conversion optab from the direct optabs vcond
and vcondu. This allows to specify different modes for the
actual comparison and the value that is selected.
All targets but i386 are trivially converted by
s/vcond/vcond/. The i386 port is enhanced
to support a OP b ? c : d as (
The equivalence defined in AI05-0071-1 for formal subprogram matching
is extended such that it applies to explicit as well as default actual
subprograms. The following test now compiles without errors.
package Pack1 is
type Root is tagged record
F1 : Integer;
end record;
proced
The semantics of operation Move for the bounded forms must match the semantics
of the unbounded forms. The post-condition is defined (for all containers) such
that when Move completes, the Source container is empty. The fix is to clear
the Source container after its elements have been assigned to T
This patch causes exceptions propagated by the termination handler to be
ignored, as required by the Ada RM. The following test should run to completion
silently:
with Ada.Task_Identification;
with Ada.Task_Termination; use Ada.Task_Termination;
with Ada.Exceptions;
package Task_Term_Exc is
task
If an object of type Cursor does not have an explicit initialization
expression, then it must default-initialize to the same value as constant
No_Element. The Node component of type Cursor has scalar type Count_Type, so it
requires an explicit initialization expression of its own declaration, in or
Make_Independent decrements Wait_Count, which caused a race condition -- if the
timing is just wrong, it causes Make_Passive to find Wait_Count = 0, which
eventually causes the program to hang. This patch prevents hanging by
decrementing Wait_Count only if nonzero.
No test is available, because it'
The formal verification back-end expects declaration to be present for Itypes,
even if they are not attached to the tree. Add such declarations in the case
of Itypes introduced for index/component types of arrays and anonymous array
types in object declaration.
Tested on x86_64-pc-linux-gnu, commi
It was not possible to use an Ada reserved word such as Interface as an
identifier, when the default Ada version is Ada 2005. This patch corrects
this.
The following project file should be accepted by all project aware tools
such as gnatmake:
project Interface is
end Interface;
Tested on x86_64-p
If the domain of iteration is a derived container type, the aspect Default_
Iterator is inherited. As for other calls to inherited operations, the actual
must be view-converted to the type of the formal to be a valid argument.
The following must compile quietly:
gcc -c -gnat12 repro.adb
---
w
There are two overloadings of Next. One is a selector function to return the
next node in the current bucket, the other navigates to the next element in the
container. The implementation of Generic_Iteration needs to call the former, as
it interrogates the items in each bucket, before moving on the
The compiler may crash processing a constant object declaration whose
expression is a nested aggregate. After this patch the following test
compiles without errors.
package Test_Nested_Aggregate is
type T_Comp (Validity : Boolean := False) is record
null;
end record;
type T_Root is
This patch adds new debugging routines (pp to print a node, node list, name_id,
etc., and ppp to print a tree). No change in functionality; no test available.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Bob Duff
* treepr.ads: Improve debugging facilities. pn(x) no lo
Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>> Files containing the MELT translated to C code were not loaded correctly
>> due to bad search path when working with an installed plugin. Present
>> commit fixes this by:
>> -
This patch fixes a spurious error on an exponentation operand that has a
universal interpretation, when the parent node is overloaded and one of its
interpretations is a fixed-point type.
The following must compile quietly:
gcc -c adt.adb
with Sim;
package adt is
task type Cust;
end;
pack
Instance bodies nested within generic bodies are always built, so that semantic
checks can be performed on them. Finalization actions on entities within these
instances must not be generated because expansion will have been disabled at
various points and because finalizers are only relevant in exec
It is not possible to specify "-largs" as a switch in package linker of
the main project file, to link with an external library libargs. This
patch corrects this. Now any option found in the switches of packages
Compiler, Binder or Linker is taken as is into account.
Tested on x86_64-pc-linux-gnu,
On Tue, May 10, 2011 at 2:55 AM, Richard Guenther wrote:
>
> This moves fold-consts conversion folding to forwprop, our tree
> pattern matcher & optimizer. In addition to what fold-const
> does this makes us canonicalize unsigned truncations to use
> bitwise and of a mask, this form is more readi
Hi,
tested x86_64-linux, committed to mainline.
Thanks,
Paolo.
///
2011-08-29 Paolo Carlini
PR libstdc++/50118
* include/bits/stl_list.h (_List_base<>::
_List_base(const allocator_type&)): Remove.
(_List_base<>::_List_base(const _Node_
Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>> Files containing the MELT translated to C code were not loaded correctly
>> due to bad search path when working with an installed plugin. Present
>> commit fixes this by:
>> -
If a package in a limited_with clause contains an incomplete type and its
completion, both of them have a limited view, and these may be visible at the
same time though intervening packages. These two limited views must be treated
as identical in conformance tests.
The following must build quietly;
As the length of any logical is limited to 255 bytes.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* gnatcmd.adb (GNATCmd): On OpenVMS, truncate the length of
GNAT_DRIVER_COMMAND_LINE to 255.
Index: gnatcmd.adb
=
There are two overloadings of Next. One is a selector function to return the
next node in the current bucket, the other navigates to the next element in the
container. The implementation of Generic_Equal needs to call the former, as it
interrogates the items in each bucket.
Also, the buckets array
The expanded code for an instantiation may contain apparent ambiguities when
two subprograms that are generic formals or that depend on formal types become
homonyms after formals are replaced with actuals. In some cases, the ambiguity
can be resolved by favoring the innermost interpretation, when b
Hi,
> Hi Paolo,
>
>>> I guess the patch is ok now?
>> Yes. Nice that Jon replied in the meanwhile and clarified the undefined
>> behavior issue: for the time being I think we can keep the __cplusplus
>> checks, should also help during this testing period. We can certainly clean
>> up the thing l
Project dependencies are needed in various cases, for instance because
the linker needs to find all the list of non-Ada objects when building
a global archive in gprbuild. But these project dependencies need to be
explicitly computed initially, and this wasn't done for aggregated
projects.
Tested
Alfa mode for formal verification currently does no deal with tasking and
tagged types. Suppress expansion of these features to simplify input to
formal verification back-end.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Yannick Moy
* exp_ch13.adb (Expand_N_Freeze_Enti
Le 29/08/2011 15:15, Basile Starynkevitch a écrit :
> On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
>> Files containing the MELT translated to C code were not loaded correctly
>> due to bad search path when working with an installed plugin. Present
>> commit fixes this by:
>> -
On Mon, Aug 29, 2011 at 12:21:12PM +0200, Alexandre Lissy wrote:
> Files containing the MELT translated to C code were not loaded correctly
> due to bad search path when working with an installed plugin. Present
> commit fixes this by:
> - using a melt find to search in current dir
> - ensuring t
This new implementation avoids dependencies on BLAS and LAPACK.
The advantages are simplicity (no external requirements), more certainty
about accuracy and availability for all floating point types. Obvious
disadvantages are lack of target-specific optimizations that may affect
both accuracy, less
This change prevents Put_SCOs from emitting a spurious blank line in the
ALI file in case the length of a statements line is congruent to 0 modulo 6.
The following compilation must not produce any blank line in the middle of
C lines.
$ gcc -c -gnateS blank_line_in_scos.adb
package Blank_Line_In_SC
No functional change.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* raise-gcc.c (get_ip_from_context): New function. Factorize code.
Index: raise-gcc.c
===
--- raise-gcc.c (revision 17820
This exception stack was once used by VMS, but isn't anymore.
No functional change.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* s-taprop-vms.adb (Get_Exc_Stack_Addr): Remove.
(Initialize_TCB): Remove Exc_Stack_Ptr initialization.
(Fin
Hi Paolo,
>> I guess the patch is ok now?
>
> Yes. Nice that Jon replied in the meanwhile and clarified the undefined
> behavior issue: for the time being I think we can keep the __cplusplus
> checks, should also help during this testing period. We can certainly clean
> up the thing later. Mayb
Compilation was stopping on errors in Alfa mode due to incorrect generation of
a type, and inconsistent treatment of packing.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Yannick Moy
* freeze.adb (Freeze_Record_Type): Ignore packing in Alfa mode, like
in CodePe
In Ada 2012 it is possible to inherit non-conformant homographs, but they
can't be called or overridden. The following test now compiles silently.
package Pack1 is
type Int1 is interface;
procedure Op (X : in Int1) is null;
end Pack1;
package Pack2 is
type Int2 is interface;
proce
This patch fixes a typo in the expansion of an allocator for an unconstrained
packed array, when the expression is a function call.
The execution of tester below must yield:
in next_frame
in next_frame
1
2
---
with Media;
with ada.text_IO;
procedure Tester is
Case_1 :Media.Image :=
Two components (Private_Data and Cleanup_Flag) are now removed because they
were unused.
No functionnal change.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* a-exstat.adb (String_To_EO): Do no set Cleanup_Flag.
* a-exexda.adb (Set_Exception_C_
On Mon, Aug 29, 2011 at 2:09 PM, Richard Guenther
wrote:
> On Mon, Aug 22, 2011 at 9:50 PM, Uros Bizjak wrote:
>> On Mon, Aug 22, 2011 at 5:34 PM, Richard Guenther
>> wrote:
>>
In this case it is simple to analyse that a is a comparison, but you
cannot embed the operations of a into VE
Remove all its parameters as they were unused.
No functionnal changes.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* a-exexpr.adb (Propagate_Exception): Remove all the parameters as
they were unused.
* a-exexpr-gcc.adb (Propagate_Except
The procedure Setup_Exception did nothing and is now removed.
No functionnal change.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* a-exexpr.adb (Setup_Exception): Removed.
* a-exexpr-gcc.adb (Setup_Exception): Removed.
* a-except.adb (E
On Mon, Aug 29, 2011 at 1:59 PM, Christian Bruel wrote:
>
>
> On 08/27/2011 02:04 AM, Jan Hubicka wrote:
>>>
>>> Hello,
>>>
>>> Could I have a review for the trivial patch posted in
>>> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html
>>>
>>> -fprofile-use sets flag_branch_probabilities.
>>
On Mon, Aug 22, 2011 at 9:50 PM, Uros Bizjak wrote:
> On Mon, Aug 22, 2011 at 5:34 PM, Richard Guenther
> wrote:
>
>>> In this case it is simple to analyse that a is a comparison, but you
>>> cannot embed the operations of a into VEC_COND_EXPR.
>>
>> Sure, but if the above is C source the fronten
On 08/27/2011 02:04 AM, Jan Hubicka wrote:
Hello,
Could I have a review for the trivial patch posted in
http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01123.html
-fprofile-use sets flag_branch_probabilities.
But we should also be able to use -fbranch-probabilities on its own
using the informat
On Fri, Aug 26, 2011 at 8:54 PM, Aldy Hernandez wrote:
> This is a "slight" update from the last revision, with your issues addressed
> as I explained in the last email. However, everything turned out to be much
> tricker than I expected (variable length offsets with arrays, bit fields
> spanning
On 08/18/2011 11:23 AM, Richard Guenther wrote:
Yeah, well. That's really a question for language lawyers;) I agree
that it would be nice to have mask ? val0 : val1 behave "the same"
for scalars and vectors. The question is whether for vectors you
define it on the bit-level (which makes it eq
On 08/29/11 11:03, Jakub Jelinek wrote:
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg01767.html
> PR middle-end/48722
> Fix RTL sharing problem with CALL_INSN_FUNCTION_USAGE
Looks OK.
Bernd
With this new feature, foreign exceptions can be caught in 'when others'
handlers and will correctly propagate through Ada frames (including at-end
handlers).
No pure Ada test.
Tested on x86_64-pc-linux-gnu, committed on trunk
2011-08-29 Tristan Gingold
* a-exexpr-gcc.adb (Unwind_Exc
On Sat, Aug 27, 2011 at 3:39 AM, Artem Shinkarov
wrote:
> Hi
>
> Here is a patch with vector comparison only.
> Comparison is expanded using VEC_COND_EXPR, conversions between the
> different types inside the VEC_COND_EXPR are happening in optabs.c.
I have split out the middle-end infrastructure
Ada 2012 added a Synchronous_Barrier type to allow many tasks to be blocked
and be released at once, similar to the functionality provided by POSIX
barriers. See AI-0174 for details.
The following test case should compile quietly in Ada2012 mode and execute
without any output message:
pragma Task
The check that raises Constraint_Error for arguments
that equal the machine number nearest Pi / 2.0 must
be removed, as the tangent of this machine number is
well-defined and does not overflow.
-- Test case, should compile and execute quietly
with Ada.Numerics.Long_Long_Elementary_Functions;
use A
On Mon, 29 Aug 2011, Jakub Jelinek wrote:
> http://gcc.gnu.org/ml/gcc-patches/2011-08/msg00428.html
> Adjust gthr-posix.h so that g++ -C -E works with STL headers
>
OK in the absence of libgcc maintainer objecti
1 - 100 of 139 matches
Mail list logo
