Aldy Hernandez wrote:
PR tree-optimization/52558
* cfg.c (alloc_aux_for_edge): Fix comment.
(alloc_aux_for_edge): Remove static.
* basic-block.h (alloc_aux_for_edge): Protoize.
* tree-ssa-loop-im.c (execute_sm_if_changed): New.
(execute_sm_if_change
> From: Hans-Peter Nilsson
> Date: Fri, 1 Jun 2012 07:58:13 +0200
> - the gUSA atomic method in config/sh/sh.md is really cute, go
> see it.
Doh, I mean config/sh/sync.md.
brgds, H-P
Unfortunately for subtargets where atomic accesses matter, the
LL/SC scheme only works for aligned data (i.e. excluding plain
cris-elf which is hardly usable), and the CRIS ABI doesn't
mandate any alignment (except for functions, mandated 16-bit
aligned by the ISA). I guess I could have emitted ca
On Fri, Jun 1, 2012 at 12:05 AM, Tobias Burnus wrote:
> This patch tries to silence most warnings reported for MinGW64 by
> Nightstrike at http://pastebin.mozilla.org/1652586
>
> One item is a true bug. I didn't fix the coarray-related warnings as the ABI
> will still change. And I ignored the las
Hi,
When I modified GCC to change the majority of bitfield accesses
which were done via component ref to BIT_FIELD_REF, SRA messes up
because when it does the replacement it still tries to use the
BIT_FIELD_REF except it never places the old value in the struct for
the BIT_FIELD_REF to work corre
Hi,
this patch uses the aggregate jump functions collected by the previous
one to do indirect inlining. It can do stuff we could not do before
(see the testcase) and also removes the special constant jump
functions representing C++ function member pointers and represents
that information with ord
Hi,
this patch adds the capability to build aggregate jump functions.
These consist of:
1) information what known compile time IPA-invariants are at various
offsets of an aggregate passed to a callee either by reference
(when the parameter is pointer) or by value (when it is an
aggregate
Hi,
this patch uses the aggregate jump functions created by a previous
patch to determine benefits of inlining a particular call graph edge.
It does so by fairly straightforward way. It a flag to struct
condition to specify it is actually an aggregate value at an offset,
also newly stored in the
Hi,
this patch is basically a proof-of-concept aiming at alleviating the
following code found in Fortran functions when they look at the
contents of array descriptors:
:
stride.156_7 = strain_tensor_6(D)->dim[0].stride;
if (stride.156_7 != 0)
goto ;
else
goto ;
:
:
Hi,
the first patch mainly introduces access methods to read parts of jump
functions. I am no particular fan of these but the are now more
widely spread than originally and above all checking asserts verifying
that the correct part of the union is read should really be useful.
I have also unifie
Hi,
this patch set extends the current jump functions we use in
ipa-prop.[ch], ipa-inline-analysis.c and ipa-cp.c to also carry
information about aggregates passed across function calls in a
parameter, regardless whether by value or only by reference.
The long term goal is that these description
The status table doesn't mention the fact we're missing the new const
char* constructors from all the exception classes.
* doc/xml/manual/status_cxx2011.xml: Correct C++11 status table.
Tested with make doc-xml-validate-docbook doc-html, committed to trunk.
commit 7cb382e8086a3769ae408571
2012-06-01 Manuel López-Ibáñez
Jonathan Wakely
PR c++/50134
* doc/invoke.texi (Warning Options): Explain purpose and differences
between -Wmissing-prototypes and -Wmissing-declarations.
This patch tries to improve the docs to clarify the intent of these op
Reviewers: xur, davidxl, iant2, Diego Novillo,
Message:
The relevant bug is this:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53546.
Please take a look and provide feedback.
Thanks,
Description:
2012-05-31 Ahmad Sharif
* gcc/dwarf2asm.c: (dw2_output_indirect_constant_1): Copy TREE
This patch fixes target-header-probing for cross-builds that
aren't using sysroot (just --target and --prefix), as in the
difference for crisv32-axis-linux-gnu with an eglibc-2.15-based
port:
--- cllog.prev Thu May 31 22:00:10 2012
+++ cllog Thu May 31 22:00:09 2012
@@ -1276,7 +1276,7 @@ ch
> I'm sure it works, but function calls are quite expensive on sparc.
I thought the whole business of the register windows was aimed at making them
quite cheap. :-) This one isn't performance critical anyway, so the fix is
fine I think (without the terminal 's' in "Implements" though).
--
Eri
Renames to avoid builtins on clang.
(This patch of code is actually the punchline to the joke, "how many
template aliases are really required if you are going for no line
breaks and using 80 columns while using policy based designs.")
tested x86/linux
-benjamin2012-05-31 Benjamin Kosnik
P
From: Steven Bosscher
Date: Fri, 1 Jun 2012 01:00:37 +0200
> Besides, at least as far as I understand, it's sort-of a "design goal"
> for gcc to move from target macros to target hooks. So you're going to
> have to deal with some more calls anyway.
I understand, I'm just disappointed :-)
I'll t
Hi Diego,
This is a backport of this patch to fix an ICE on PowerPC64.
http://gcc.gnu.org/ml/gcc-patches/2011-03/msg00829.html
-Doug
2012-05-31 Doug Kwan
Backport r171031 from upstream trunk in gcc/.
2011-03-16 Alan Modra
PR target/45844
On Fri, Jun 1, 2012 at 12:05 AM, David Miller wrote:
> From: Steven Bosscher
> Date: Thu, 31 May 2012 23:42:26 +0200
>
>> Could you please give the attached patch a try?
>> Sorry for the inconvenience!
>
> I'm sure it works, but function calls are quite expensive on
> sparc.
There are four elimi
The R4700 is identical to the R4600 except for the integer and
floating-point multiplication costs.
See page 4 of http://datasheets.chipdb.org/IDT/MIPS/79RV4700.pdf
2012-03-24 Matt Turner
gcc/
* config/mips/4600.md (r4700_imul_si): New.
(r4700_imul_di): New.
(r
On Thu, May 31, 2012 at 5:35 PM, Richard Sandiford
wrote:
> Matt Turner writes:
>> On Sat, Feb 25, 2012 at 3:11 AM, Richard Sandiford
>> wrote:
>>> Matt Turner writes:
The r4600_imul and r4600_idiv reservations were correct for si, but
there were no *_di reservations.
See pa
Hello,
The attached patch addresses PR 53512.
Tested with
make all-gcc
make info dvi pdf
Previous default behavior for SH4A:
make check-gcc RUNTESTFLAGS="sh.exp=sh4a-sincosf*
--target_board=sh-sim/-m4a-single/-ml"
check-gcc RUNTESTFLAGS="sh.exp=sh4a-fsrra*
--target_board=sh-sim/-m4a-single/-ml"
Oleg Endo wrote:
> Original patch:
> http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00210.html
>
> 'Tested' with..
>
> ../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
> --enable-languages=c,c++ --enable-multilib --disable-libssp
> --disable-nls --disable-werror --enable-lto --with-
Oleg Endo wrote:
> Original patch:
> http://gcc.gnu.org/ml/gcc-patches/2012-03/msg01393.html
>
> 'Tested' with..
>
> ../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
> --enable-languages=c,c++ --enable-multilib --disable-libssp
> --disable-nls --disable-werror --enable-lto --with-
Oleg Endo wrote:
> Original patch:
> http://gcc.gnu.org/ml/gcc-patches/2012-03/msg01695.html
>
> 'Tested' with..
>
> ../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
> --enable-languages=c,c++ --enable-multilib --disable-libssp
> --disable-nls --disable-werror --enable-lto --with-
Oleg Endo wrote:
> Original patch:
> http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00209.html
>
> 'Tested' with..
>
> ../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
> --enable-languages=c,c++ --enable-multilib --disable-libssp
> --disable-nls --disable-werror --enable-lto --with-
Oleg Endo wrote:
> Original patch:
> http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00208.html
>
> 'Tested' with..
>
> ../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
> --enable-languages=c,c++ --enable-multilib --disable-libssp
> --disable-nls --disable-werror --enable-lto --with-
From: Steven Bosscher
Date: Thu, 31 May 2012 23:42:26 +0200
> Could you please give the attached patch a try?
> Sorry for the inconvenience!
I'm sure it works, but function calls are quite expensive on
sparc.
On Thu, May 31, 2012 at 11:15 PM, David Miller wrote:
>
> Removing output.h from reload1.c broke the sparc build because
> sparc's define of INITIAL_ELIMINATION_OFFSET (which is used in
> reload1.c) references current_function_is_leaf.
I was afraid of some fall-out. This is why target macros must
Matt Turner writes:
> On Sat, Feb 25, 2012 at 3:11 AM, Richard Sandiford
> wrote:
>> Matt Turner writes:
>>> The r4600_imul and r4600_idiv reservations were correct for si, but
>>> there were no *_di reservations.
>>>
>>> See page 4 of
>>> http://www.sgistuff.net/hardware/other/documents/R4600_P
Tejas Belagod writes:
> In the currect scheme, when multiple code/mode iterators are in an rtx
> pattern,
> they are expanded for each combination of iterator values in
> apply_iterator_traverse () and a repeated traversal of the expanded rtx's for
> each iterator achieves the 'cross-product'
Removing output.h from reload1.c broke the sparc build because
sparc's define of INITIAL_ELIMINATION_OFFSET (which is used in
reload1.c) references current_function_is_leaf.
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg01393.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
This patch tries to silence most warnings reported for MinGW64 by
Nightstrike at http://pastebin.mozilla.org/1652586
One item is a true bug. I didn't fix the coarray-related warnings as the
ABI will still change. And I ignored the last ones as rstride is set.
(The compiler cannot see that "dim
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg01695.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00210.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00210.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00209.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
Original patch:
http://gcc.gnu.org/ml/gcc-patches/2012-03/msg00208.html
'Tested' with..
../gcc-4_7-branch/configure --target=sh-elf --prefix=/usr/local
--enable-languages=c,c++ --enable-multilib --disable-libssp
--disable-nls --disable-werror --enable-lto --with-newlib --with-gnu-as
--with-gnu-ld
As discussed earlier, I have moved the flag_tm check inside of
block_in_transaction.
Committed to trunk.
* tree-ssa-loop-im.c (execute_sm): Do not check flag_tm.
* gimple.h (block_in_transaction): Check for flag_tm.
Index: tree-ssa-loop-im.c
=
The comment mentions PCH in connection with deferred seting of
DECL_ASSEMBLER_NAME; off the top of my head it occurs to me that that
might be connected with anonymous unions, which need to have different
linkage names in different translation units.
Jason
The new mklog script prints three spaces after the date part, which
is wrong. Thus fixed by adjusting the split pattern. Tested manually.
2012-05-31 Marek Polacek
* mklog: Prevent printing three spaces after the date.
--- gcc/contrib/mklog.mp2012-05-31 22:03:07.492457065 +02
As Georg-Johann Lay wrote:
> So here is the next version of the patch, returning to printf again
> :-)
"works for me"
--
cheers, J"org .-.-. --... ...-- -.. . DL8DTL
http://www.sax.de/~joerg/NIC: JW11-RIPE
Never trust an operating system you don't hav
On 05/29/12 06:13, Richard Guenther wrote:
On Mon, 21 May 2012, Aldy Hernandez wrote:
On 05/16/12 07:53, Richard Guenther wrote:
On Mon, 7 May 2012, Aldy Hernandez wrote:
(flag_tm&& loop_preheader_edge (loop)->src->flags& BB_IN_TRANSACTION)
can you encapsulate this into a predicate? Li
From: Eric Botcazou
Date: Thu, 31 May 2012 15:06:41 +0200
>> + do {
>> \
>> +UDItype __carry = 0;\
>> +__asm__ ("addcc\t%r5,%6,%1\n\t" \
>> +
On 12-05-31 14:57 , Steven Bosscher wrote:
Hello,
Almost all files include output.h because it defines dump_file. IMHO
output.h should only be included in files that actually output
something to asm_out_file. Therefore wanted to I move dump_file to
some other include file. I ended up with system
> Testing:
> * Crossbuild for target arm-none-eabi with cpu cortex-a9 neon softfp and
> tested in three configuration: -marm (default), -mthumb, -mapcs-frame. No
> regression on qemu.
> * Crossbuild for target arm-none-eabi thumb2 with cpu cortex-m3. No
> regression on qemu.
> * Crossbuild for targ
Add new skeleton ChangeLog file generator to contrib/
This script analyzes a .diff file and generates a skeleton ChangeLog
entry for it. It tries fairly hard to find function names, so it's
important to use the -p switch with diff.
I wrote the initial code a while ago. Cary added a bunch of
Finally, I decided to document the new -msp8 option. It's always bad idea to
have undocumented stuff...
http://gcc.gnu.org/viewcvs?view=revision&revision=188070
Also flushed the 4.7 documentation to be the same as trunk:
http://gcc.gnu.org/viewcvs?view=revision&revision=188071
Johann
On 12-05-31 11:58 , DJ Delorie wrote:
OK to commit this amended patch?
Ok.
Thanks. Committed.
Do we have a build-with-c++ FAQ page anywhere? /me thinks it
will be useful soon ;-)
Not yet. If there are any FAQs, I propose adding them to
http://gcc.gnu.org/wiki/cxx-conversion and then mo
On May 31, 2012, at 6:42 AM, Dominique Dhumieres wrote:
>> This is really stretching my testsuite knowledge. Maybe add
>>
>> /* { dg-additional-options "-mdynamic-no-pic" { target *-*-darwin* } } */
>
> Using
>
> /* { dg-options "-fno-inline -fomit-frame-pointer" } */
> /* { dg-additional-optio
Actually, for number of threads >=3 this patch is better:--- old/libstdc++-v3/include/bits/basic_string.h 2012-05-30 17:03:15.974035000 -0400
+++ new/libstdc++-v3/include/bits/basic_string.h 2012-05-31 13:06:39.903442000 -0400
@@ -224,14 +224,20 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
_M_refdata() th
For some time now, I've wanted to add configuration support to add additional
directories for the compiler to search automatically, without having to edit
local files in my directory.
I have some questions about these patches, that I will put after the
explanation of the functionality.
One of the
This patch replaces the large number of spellings of "2 *
HOST_BITS_PER_WIDE_INT" with the existing preprocessor symbol
HOST_BITS_PER_DOUBLE_INT.
Even though this is a large patch, it is trivial and so unless i hear
some objections, i will check it in soon.
the patch has been bootstrapped on
Hello,
A smaller patch this time, but bigger impact: GIMPLE_H does not have
to depend on TARGET_H. Only a few files need TARGET_H as well as
GIMPLE_H, and most of them already included TARGET_H anyway. The ones
that don't, are fixed with this patch.
Bootstrapped&tested on powerpc64-unknown-linux-
On Thu, 31 May 2012, tlknv wrote:
I would like to propose a patch to
libstdc++-v3/include/bits/basic_string.h that solves the problem
described in the bug 21334 (
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21334 ).
Hello,
correct me if I am wrong, but this patch doesn't aim to fix the issu
> OK to commit this amended patch?
Ok. Do we have a build-with-c++ FAQ page anywhere? /me thinks it
will be useful soon ;-)
This removes defering DECL_ASSEMBLER_NAME computation inside dwarf2out.c,
something which is not possible when LTO is enabled during compilation
as we free up tree fields after we think the frontend is finished - but
DECL_ASSEMBLER_NAME computation involves a langhook and frontend specific
data.
This patch makes __aarch64_sync_cache_range a LIB2ADD and removes it
from lib1funcs.S. Since it is the only function in lib1funcs.S, that
file can be removed. It also changes the functionality of
__aarch64_sync_cache_range to use userland instructions instead of an
exception.
This should be appli
This backports a fix for dumping.
Built and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2012-05-31 Richard Guenther
PR middle-end/53541
* tree-pretty-print.c (dump_generic_node): Guard against
NULL_TREE TREE_TYPE when dumping MEM_REF offset type.
Index: gc
(resending to the mailing list because due to html content)
On 05/31/2012 04:43 PM, Jason Merrill wrote:
Does the C front end warn about this mismatch?
I just tried the first test of g++.old-deja/g++.other/cond5.C and the C
front-end does *not* warn neither by default, neither with -Wall.
This fixes ICE on any source compiled with -mint8.
Missing definition of UINT16_TYPE (defined to 0) crashed
the compiler when it tries to build wchar stuff.
As mentioned in the PR, I chose to add the stdint stuff
as a new file.
The -mint8 part is only lightly tested because there is no
test suite
Does the C front end warn about this mismatch? Do we warn about
mismatch in comparisons?
Jason
Hi All,
I would like to propose a patch to libstdc++-v3/include/bits/basic_string.h
that solves the problem described in the bug 21334 (
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=21334 ).
Briefly:
Developers often use non constant methods begin(), end(), operator [], at() for
constant operatio
This re-organizes builtin detection and generation properly into
an analysis and transform phase (noting that when we fail late
we will generate wrong code at the moment - eventually non-fatal,
but at least it will have duplicate work in the left-over loop).
With this in place adding more kinds o
On 05/31/2012 03:47 AM, Georg-Johann Lay wrote:
* config/avr/t-avr: Correct avr-mmcu.texi dependencies.
(s-avr-mmcu-texi): Don't sed on gen-avr-mmcu-texi output.
* config/avr/gen-avr-mmcu-texi.c: Sort MCUs.
* doc/avr-mmcu.texi: Regenerate.
Ok.
Thanks for your pa
As a result of the previous changes, epilogue_insns pattern can only be
generated in Thumb1. After removing other cases in define_insn for
epilogue_insns, the function arm_output_epilogue becomes dead code and can
be eliminated, along with all its helper functions.
ChangeLog:
gcc
2012-05-31 Ia
On Thu, May 31, 2012 at 6:50 AM, Jakub Jelinek wrote:
> On Thu, May 31, 2012 at 06:35:19AM -0700, H.J. Lu wrote:
>> This patch makes __WCHAR_TYPE__/wchar_t the same for gcc -m32 on
>> Linux/i386 and Linux/x86-64. OK for trunk?
>
> That looks wrong. For Linux/i386, the 32-bit only compiler should
Generate RTL for return in Thumb2 mode. Used by expand of return insn.
ChangeLog:
gcc
2012-05-31 Ian Bolton
Sameera Deshpande
Greta Yorsh
* config/arm/arm-protos.h (thumb2_expand_return): New declaration.
* config/arm/arm.c (thumb2_expand_return):
Add a new parameter to the function output_return_instruction to handle
simple cases of return when no epilogue needs to be printed out.
ChangeLog:
gcc
2012-05-31 Ian Bolton
Sameera Deshpande
Greta Yorsh
* config/arm/arm-protos.h (output_return_instruction
This abstracts the notion of a partition properly so we can add
more information to it in followup patches (and not re-compute
everything all the time).
Bootstrap and regtest running on x86_64-unknown-linux-gnu.
Richard.
2012-05-31 Richard Guenther
* tree-loop-distribution.c (struct
The main function for epilogue RTL generation, used by expand epilogue
patterns.
ChangeLog:
gcc
2012-05-31 Ian Bolton
Sameera Deshpande
Greta Yorsh
* config/arm/arm-protos.h (arm_expand_epilogue): New declaration.
* config/arm/arm.c (arm_expand_epi
Helper function for epilogue expansion. Emit RTL for APCS frame epilogue
(when -mapcs-frame command line option is specified).
This function is used by a later patch.
For APCS frame epilogue, the compiler currently generates LDM with SP as
both the base register
and one of the destination register
New define insn pattern for epilogue with floating point registers (DFmode)
and a new function that emits RTL for this pattern. This function is a
helper for epilogue extension. It is used by a later patch.
ChangeLog:
gcc
2012-05-31 Ian Bolton
Sameera Deshpande
Greta
This patch adds new define_insn patterns for epilogue with integer
registers.
The patterns can handle pop multiple with writeback and return (loading into
PC directly).
To handle return, the patterns use a new special predicate
pop_multiple_return, that uses ldm_stm_operation_p function from a pre
This applies some TLC to loop distribution mainly to make pattern
recognition more sane and not so costly. This first patch revisits
the fix for PR45948 which turned out not effective later and
effectively fixed by another patch, too (see the ??? marker for
how that patch was overly conservative)
This patch updates ldm_stm_operation_p to check for loads that if SP is in
the register list, then the base register is SP. It guarantees that SP is
reset correctly when an LDM instruction is interrupted. Otherwise, we might
end up with a corrupt stack.
ChangeLog:
gcc
2012-05-31 Greta Yorsh
On Thu, May 31, 2012 at 06:35:19AM -0700, H.J. Lu wrote:
> This patch makes __WCHAR_TYPE__/wchar_t the same for gcc -m32 on
> Linux/i386 and Linux/x86-64. OK for trunk?
That looks wrong. For Linux/i386, the 32-bit only compiler should be
the standard, rather than x86_64 -m32 if they differ.
So,
This sequence of patches adds support for epilogue generation in RTL.
This is the first part of Sameera's work on ARM prologue/epilogue. Sameera
Deshpande posted it for review in December 2011, having addressed all
previous comments: http://gcc.gnu.org/ml/gcc-patches/2011-12/msg00049.html.
The lat
> This is really stretching my testsuite knowledge. Maybe add
>
>/* { dg-additional-options "-mdynamic-no-pic" { target *-*-darwin* } } */
Using
/* { dg-options "-fno-inline -fomit-frame-pointer" } */
/* { dg-additional-options "-mdynamic-no-pic" { target *-*-darwin* } } */
works for me on powe
Hi,
This patch makes __WCHAR_TYPE__/wchar_t the same for gcc -m32 on
Linux/i386 and Linux/x86-64. OK for trunk?
Thanks.
H.J.
2012-05-31 H.J. Lu
PR target/53539
* config/i386/gnu-user.h (WCHAR_TYPE): Use "int".
(WCHAR_TYPE_SIZE): Set to 32.
diff --git a/gcc/con
On Thu, May 31, 2012 at 02:16:32PM +0200, Dominique Dhumieres wrote:
> (the final patch will require the suitable dg directives;-).
This is really stretching my testsuite knowledge. Maybe add
/* { dg-additional-options "-mdynamic-no-pic" { target *-*-darwin* } } */
--
Alan Modra
Australia Dev
> Eric, while looking at soft-fp code generated in glibc I noticed that
> for v9 on 32-bit we end up doing software multiplies and divides :-/
>
> I also noticed that the two-limb addition and subtraction could be
> done using a branchless sequence on 64-bit.
>
> Any objections?
None on principle,
Hi,
On Thu, 31 May 2012, Richard Guenther wrote:
> > Bootstrapped (with memset recognition enabled by default) and tested
> > on x86_64-unknown-linux-gnu with the aforementioned issues.
>
> The following fixes it by simply always adding
> -fno-tree-loop-distribute-patterns to builtins.exp.
>
On Thu, May 31, 2012 at 12:28 PM, Tobias Burnus wrote:
> Dear all,
>
> gfortran was producing the following code:
>
> res = realloc (mem, size)
> if (size == 0)
> res = NULL;
> ...
> if (res != 0)
> free (res)
>
> The problem is that "realloc (..., 0)" does not have to produce a NULL
> p
On Thu, 31 May 2012, Eric Botcazou wrote:
> > This fixes PR53501, fold_plusminus_mult_expr does not expect that
> > operands have a sign-conversion stripped. So don't call it with
> > such arguments.
> >
> > Bootstrap and regtest pending on x86_64-unknown-linux-gnu.
> >
> > Richard.
> >
> > 2012-
> This fixes PR53501, fold_plusminus_mult_expr does not expect that
> operands have a sign-conversion stripped. So don't call it with
> such arguments.
>
> Bootstrap and regtest pending on x86_64-unknown-linux-gnu.
>
> Richard.
>
> 2012-05-30 Richard Guenther
>
> PR middle-end/53501
>
> Please try out this patch on Darwin. Bootstrapped and regression
> tested powerpc-linux.
I have applied the patch to r188026 and updated the build.
As patched the test gcc.target/powerpc/savres.c now fails with
FAIL: gcc.target/powerpc/savres.c (test for excess errors)
Excess errors:
ld_classi
On 12-05-30 13:42 , DJ Delorie wrote:
BUILD_CFLAGS= @BUILD_CFLAGS@ -DGENERATOR_FILE
+BUILD_CXXFLAGS = $(INTERNAL_CFLAGS) $(CXXFLAGS) -DGENERATOR_FILE
Why are these so different?
The rest seem OK
Ah, I missed adding BUILD_CXXFLAGS to configure.ac. Thanks.
OK to commit this amended patch
On 12-05-31 02:27 , Doug Kwan wrote:
2012-05-30 Doug Kwan
* contrib/testsuite-management/powerpc64-grtev2-linux-gnu.xfail:
Mark TLS tests as flaky. These do not work well with QEMU.
Exempt g++.dg/cpp0x/variadic-value1.C like we did for x86.
* contrib/testsuite-
Pointed out by people working on targets with ptr-size != sizetype size.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
Richard.
2012-05-31 Richard Guenther
* tree-data-ref.c (dr_analyze_innermost): Properly convert
the MEM_REF offset to sizetype.
Index: gcc/
This fixes PR48493 by backporting a one-liner - we should not go
the movmisalign path for destinations that are not memory.
Bootstrapped and tested on x86_64-unknown-linux-gnu and on
mips by Andrew, installed.
Richard.
2012-05-31 Richard Guenther
PR middle-end/48493
* expr.c
On Wed, 30 May 2012, Richard Guenther wrote:
>
> The patch below extents memset recognition to cover a few more
> non-byte-size store loops and all byte-size store loops. This exposes
> issues with our builtins.exp testsuite which has custom memset
> routines like
>
> void *
> my_memset (void *
Richard Sandiford wrote:
Tejas Belagod writes:
Hi Richard,
Thanks for your comments. Some questions inline below.
Richard Sandiford wrote:
Marcus Shawcroft writes:
This patch adds an implementation of integer iterators.
Nice. A few comments from an onlooker (on top of what Stephen said).
Joerg Wunsch wrote:
> As Georg-Johann Lay wrote:
>
>> If it's appropriate I would also set svn:mime-type to something
>> like application/foo but that seems bit odd.
>
> Rather set the svn:eol-style attribute to "LF" instead, I'd say.
>
> But wait a moment, why not setting the svn:eol-style to "
Hi,
when I fixed PR16603 (for 4.7.0) I didn't anticipate that we would warn
more easily about mismatching enum types for user code using conditional
expressions to define enumerators basing on other enumerators of the
same open enum, like the testcase in this PR shows. Generally speaking,
IMH
As Georg-Johann Lay wrote:
> If it's appropriate I would also set svn:mime-type to something
> like application/foo but that seems bit odd.
Rather set the svn:eol-style attribute to "LF" instead, I'd say.
But wait a moment, why not setting the svn:eol-style to "native", and
the generator utility
Hi,
This patch re-factors code models option selection in the AArch64 port:
. Renaming variables such as mem_model to cmodel, for better clarity.
. Using the generic support for enumerated option arguments.
. Fixing touched code layout and formatting issues.
Thanks
Sofiane
-
ChangeLog:
Richard Henderson wrote:
>> +# The avr-mmcu.texi we want to compare against / check into svn should
>> +# have unix-style line endings. To make this work on MinGW, remove \r.
>> +# \r is not portable to Solaris tr, therefore we have a special case
>> +# for ASCII. We use \r for other encodings l
1 - 100 of 108 matches
Mail list logo
