Hi,
I would like to apply backports of the TX support patches to 4.8
branch. We released 4.8 with EC12 support already but the support is
somewhat incomplete without having TX. With these patches we will
have full EC12 support in 4.8.
The patches apply without changes but I had to add some conf
On 31 July 2013 15:55, Sofiane Naci wrote:
> Hi,
>
> This patch is the first of a series of patches that aim to unify instruction
> classification between the ARM and AARCH64 backends.
>
> This patch updates the definition of the "type" attribute, used in the ARM
> backend, in the AARCH64 backend
On 31 July 2013 15:56, Sofiane Naci wrote:
> Hi,
>
> This patch is part of the ongoing work to unify instruction classification
> between the ARM and AARCH64 backends.
>
> This patch wires up the cortex-a53 pipeline description defined in the ARM
> backend to be used in the AARCH46 backend.
>
> OK
On 14 July 2013 19:43, Diego Novillo wrote:
> On Sun, Jul 14, 2013 at 2:08 AM, Chung-Lin Tang
> wrote:
>> Ping.
>
> Could you please repost the patch with its description? This thread
> is sufficiently old and noisy that I'm not even sure what the patch
> does nor why.
Chung-Lin Tang, can you
On 13/8/1 5:16 PM, Bernhard Reutner-Fischer wrote:
> On 14 July 2013 19:43, Diego Novillo wrote:
>> On Sun, Jul 14, 2013 at 2:08 AM, Chung-Lin Tang
>> wrote:
>>> Ping.
>>
>> Could you please repost the patch with its description? This thread
>> is sufficiently old and noisy that I'm not even su
Hi,
On 07/31/2013 10:01 PM, Peter Bergner wrote:
Hi Fabien,
Can you tell me what the status of the following patch that
removes the pow() overload from tr1 is? Specifically:
http://gcc.gnu.org/ml/gcc-patches/2012-11/msg01166.html
It seemed to have been approved and you were going to do
Hi,
gently pinging this small clean-up:
http://gcc.gnu.org/ml/gcc-patches/2013-06/msg00905.html
Thanks!
Paolo.
> Honza,
>
> After this patch, I see new testsuite failures on PowerPC related to
> profiling.
>
> FAIL: gcc.dg/tree-prof/ic-misattribution-1.c scan-ipa-dump profile
> "hist->count 1 hist->all 1"
> FAIL: gcc.dg/tree-prof/indir-call-prof.c scan-ipa-dump profile
> "Indirect call -> direct call.* a
On 07/29/13 11:05, Mikael Pettersson wrote:
Ramana Radhakrishnan writes:
> Hi,
>
> This fixes up the issues with PR target/19599 and the issues we've had
> around it.
...
> 2013-07-25 Ramana Radhakrishnan
>
> PR target/19599
> PR target/57731
> P
On 07/20/2013 01:32 AM, DJ Delorie wrote:
>> Every pattern that is using (subreg:SI (thing:PSI)) needs to be
>> explained on this list and given an explicit clearance. It really looks
>> like you're just papering over problems elsewhere.
>
> Most of them are just optimizations, but the problem
On 08/01/13 11:46, Vidya Praveen wrote:
Ping!
On Tue, Jul 23, 2013 at 10:21:52AM +0100, Vidya Praveen wrote:
Hello
gcc.dg/vect/vect-iv-5.c XPASSes for arm-*-* since gcc.dg/vect/*.c tests are
always run with -ffast-math for arm-*-*. This patch makes xfail conditional
for this test by adding eff
Thanks Jakub!
Committed revision 201402.
-Original Message-
From: Jakub Jelinek [mailto:ja...@redhat.com]
Sent: Thursday, July 04, 2013 4:46 PM
To: Gopalasubramanian, Ganesh
Cc: Uros Bizjak (ubiz...@gmail.com); gcc-patches@gcc.gnu.org
Subject: Re: [PATCH,i386] Default alignment for AMD B
Ping!
On Tue, Jul 23, 2013 at 10:21:52AM +0100, Vidya Praveen wrote:
> Hello
>
> gcc.dg/vect/vect-iv-5.c XPASSes for arm-*-* since gcc.dg/vect/*.c tests are
> always run with -ffast-math for arm-*-*. This patch makes xfail conditional
> for this test by adding effective target keyword !arm_neon_o
Hi,
When using busybox' awk to bootstrap, libgo's syscalls are generated
incorrectly.
I'm attaching the split() used by busybox' awk including output before
and after the patch for reference.
Please install / ok to install?
Bernhard Reutner-Fischer (1):
mksyscalls.awk: make split ERE more por
awk's split() ERE was splitting on (essentially) .
Double checked with mawk 1.3.3, GNU Awk 4.0.1, busybox awk
that they still produce identical output.
libgo/ChangeLog (???)
2013-08-01 Bernhard Reutner-Fischer
* go/syscall/mksyscall.awk (split): Fix ere argument.
Signed-off-by: Bernh
> >> +;; Rd = (eq (reg1) (reg2/imm))// optimize for size on Thumb2
> >> +;;subs T1, Reg1, reg2
> >> +;;negs Rd, T1
> >> +;;adcs Rd, Rd, T1
> >>
> >> Only the second operation has to be flag setting. A later pass will
> >> convert the first and third instructions to fl
Hi all,
I noticed that the minmax_arithsi_non_canon splitter had a case where it would
generate a MINUS rtx with an immediate. The canonical form is PLUS with
negated immediate. This patch fixes that.
Tested arm-none-eabi on qemu with ARMv7-a, ARMv8-a, with and without -mthumb.
Ok for trunk?
Th
Hi,
this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into function
sections during WPA. Even with some work to release unused ones, there are 4M
of PARM_DECLs and 2M of RESULT_DECLs streamed during LTO (for 6M of
function_decls) making them one of the most common nodes.
This patch
Hi,
current functions that appear as abstract origin of other function gets
abstract_and_needed flag set. This flag is then not maintained in any way,
but it is used by cgraph_remove_node to avoid removing block tree dneeded later
by dwarf2out.
This patch makes the tracking more explicit. Unreach
Jan Hubicka wrote:
>Hi,
>this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into
>function
>sections during WPA. Even with some work to release unused ones, there
>are 4M
>of PARM_DECLs and 2M of RESULT_DECLs streamed during LTO (for 6M of
>function_decls) making them one of the most
Hi,
On 08/01/2013 03:11 PM, Jan Hubicka wrote:
+ replace_info->new_tree = fold_build1 (NOP_EXPR, req_type,
replace_info->new_tree);
the tree-inline.c changes don't seem formatted to 80 columns.
Paolo.
On 01/08/13 14:02, Kyrylo Tkachov wrote:
+;; Rd = (eq (reg1) (reg2/imm)) // optimize for size on Thumb2
+;; subs T1, Reg1, reg2
+;; negs Rd, T1
+;; adcs Rd, Rd, T1
Only the second operation has to be flag setting. A later pass will
convert the first and third instructions to flag
On 01/08/13 14:05, Kyrylo Tkachov wrote:
Hi all,
I noticed that the minmax_arithsi_non_canon splitter had a case where it would
generate a MINUS rtx with an immediate. The canonical form is PLUS with
negated immediate. This patch fixes that.
Tested arm-none-eabi on qemu with ARMv7-a, ARMv8-a, w
Fully tested under x86_64. (make bootstrap && make -k check).
Next, I'll try to refactor _Grep_matcher using templates instead of
virtual functions, to make implementing back-reference easier.
Thanks!
--
Tim Shen
bracket.patch
Description: Binary data
changelog
Description: Binary data
Hi,
Now when we have abstract origins tracked, this patch makes DECL_ARGUMENTS and
DECL_RESULT to be removed from FUNCTION_DECLs that are never passed to symbol
table. This reduces LTO streaming effort (by about 1/3rd of PARM_DECls)
Bootstrapped/regtested ppc64-linux, will commit it after further
> Jan Hubicka wrote:
> >Hi,
> >this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into
> >function
> >sections during WPA. Even with some work to release unused ones, there
> >are 4M
> >of PARM_DECLs and 2M of RESULT_DECLs streamed during LTO (for 6M of
> >function_decls) making them
Hi,
On Thu, Aug 01, 2013 at 03:11:36PM +0200, Jan Hubicka wrote:
> Hi,
> this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into function
> sections during WPA. Even with some work to release unused ones, there are 4M
> of PARM_DECLs and 2M of RESULT_DECLs streamed during LTO (for 6M
Hi,
On Thu, Aug 01, 2013 at 03:59:01PM +0200, Richard Biener wrote:
> Jan Hubicka wrote:
> >Hi,
> >this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into
> >function
> >sections during WPA. Even with some work to release unused ones, there
> >are 4M
> >of PARM_DECLs and 2M of RESUL
> Hi,
>
> On Thu, Aug 01, 2013 at 03:11:36PM +0200, Jan Hubicka wrote:
> > Hi,
> > this is preparation work to move DECL_ARGUMENTS and DECL_RESULT into
> > function
> > sections during WPA. Even with some work to release unused ones, there are
> > 4M
> > of PARM_DECLs and 2M of RESULT_DECLs str
Patch 3 of 3 split out from the patch I sent in May that fixes problems with
-freorder-blocks-and-partition, with changes/fixes discussed in that thread.
See http://gcc.gnu.org/ml/gcc-patches/2013-05/threads.html#00388 for context.
This patch sanitizes the partitioning to address issues such as e
Hi,
As originally signalled here:
http://gcc.gnu.org/ml/gcc-patches/2012-02/msg00140.html
here is a change to prevent MIPS16 stub calls from being made through the
PLT by making them hidden. This is needed to avoid $2 and possibly $3
from being clobbered by MIPS16 code in the PLT. Originall
On Mon, 2013-07-29 at 15:41 -0600, Jeff Law wrote:
> On 07/26/2013 09:04 AM, David Malcolm wrote:
> > This patch is the hand-written part of the conversion of passes from
> > C structs to C++ classes. It does not work without the subsequent
> > autogenerated part, which is huge.
> [ ... ]
> With t
On 07/28/2013 12:03 PM, Maxim Kuvyrkov wrote:
> Richard, did you and Red Hat intend to license config/ia64/unwind-ia64.h
> under GPL-3.0-with-GCC-exception?
>
> DJ, did you and Red Hat intend to license config/mips/vr4120-div.S under
> GPL-3.0-with-GCC-exception?
Yes, Red Hat intended to licens
This patch does more than just remove the hardcoded assumptions about
pass sizes - as noted in
http://gcc.gnu.org/ml/gcc-patches/2013-08/msg00041.html
it also is needed to ensure that per-pass dumpfiles get the correct
switch names.
So the short version is that patch 6 is needed for patches 3-5 to
On Wed, Jul 31, 2013 at 02:52:39PM -0400, Jason Merrill wrote:
> On 07/31/2013 01:33 PM, Marek Polacek wrote:
> >There are still at least two issues though, which is why
> >bootstrap with -fsanitize=undefined fails:
> >
> >http://gcc.gnu.org/ml/gcc-patches/2013-07/msg01480.html
>
> This looks like
On Thu, 2013-08-01 at 13:55 -0400, David Malcolm wrote:
[...snip...]
> OK for trunk? (on top of the other patches, of course; see notes in
> http://gcc.gnu.org/ml/gcc-patches/2013-08/msg00041.html in how I've
> tested this).
...with "pipeline" renamed to "pass_manager", of course.
This patch adds builtin support for the new 32-bit Altivec multiply
instructions that were added in ISA 2.07 (ie, POWER8).
This passed bootstrap and regtesting with no errors. Ok for mainline?
P.S. I will be working on a followup patch sometime later that will attempt
to generate these new
> + if (simd
> + /*
> + || (fd->sched_kind == OMP_CLAUSE_SCHEDULE_STATIC
> + && !fd->have_ordered)*/)
Debugging leftovers or what?
> + /* Enforce simdlen 1 in simd loops with data sharing clauses referencing
> + variable sized vars. That is unnecessarily hard to
On Thu, 2013-08-01 at 11:33 +0200, Paolo Carlini wrote:
> On 07/31/2013 10:01 PM, Peter Bergner wrote:
> > Can you tell me what the status of the following patch that
> > removes the pow() overload from tr1 is? Specifically:
> >
> > http://gcc.gnu.org/ml/gcc-patches/2012-11/msg01166.html
> >
Hello!
This patch generalizes register class check, no other functional changes.
2013-08-01 Uros Bizjak
* config/i386/i386.h (MAYBE_NON_Q_CLASS_P): New.
* config/i386/i386.c (ix86_secondary_reload): Use INTEGER_CLASS_P and
MAYBE_NON_Q_CLASS_P where appropriate.
Tested on x86_64-p
Quick Reminder; The vtable verification feature (controlled by a
flag) is designed to detect, at run time, if/when the vtable pointer
in a C++ object has been corrupted, before allowing virtual calls
through that pointer. If pointer corruption is detected, execution of
the program is halted.
I
Attached patch applied.
Compare to the proposed one I had to:
- Disable the new check for __gnu_debug::basic_string<>, it is supported
following Standard words. To do so I had to slithly review how
_GLIBCXX_DEBUG_PEDANTIC was managed.
- Add check on forward_list::insert_after
2013-08-01 Fran
Nice to see!
> I have created (with some help) a git branch on gcc.gnu.org to contain
> the vtable verification feature work. This work is now well
> integrated with GCC trunk, and the sources are in a good state for
> future work. I believe all previous review comments have been
> addressed.
On 07/26/2013 05:04 AM, David Malcolm wrote:
> This is an example of converting the "gate" and "execute" functions of
> a pass into C++ virtual functions, so that in the next patch we can move
> a variable into member data of the opt_pass subclass.
>
> gcc/testsuite/
>
> * gcc.dg/plugin/one
On 08/01/2013 07:55 AM, David Malcolm wrote:
> On Fri, 2013-07-26 at 11:04 -0400, David Malcolm wrote:
>> > gcc/
>> >
>> >Rewrite how instances of passes are cloned to remove assumptions
>> >about their sizes (thus allowing pass subclasses to have
>> >additional data fields, albeit non
On 07/26/2013 05:04 AM, David Malcolm wrote:
> gcc/testsuite/
>
> Example of converting global state to per-pass state.
>
> * gcc.dg/plugin/one_time_plugin.c (one_pass::execute): Convert
> global state "static int counter" to...
> (one_pass::counter): ...this instance data
On 07/26/2013 05:04 AM, David Malcolm wrote:
> + "/* Types with a \"gcc::\" prefix have the prefix stripped\n"
> + " during gengtype parsing. Provide a \"using\" directive\n"
> + " to ensure that the fully-qualified types are found. */\n"
I'd rather not use the wor
On 07/26/2013 05:04 AM, David Malcolm wrote:
> +/* Functions relating to the garbage collector. */
> +void
> +gcc::context::gt_ggc_mx ()
> +{
> + /* Currently a no-op. */
> +}
> +
> +void
> +gcc::context::gt_pch_nx ()
> +{
> + /* Currently a no-op. */
> +}
> +
> +void
> +gcc::context::gt_pch_n
On 07/26/2013 05:04 AM, David Malcolm wrote:
> (opt_pass::gt_ggc_mx): New.
> (opt_pass::gt_pch_nx): New.
> (opt_pass::gt_pch_nx_with_op): New.
> (gt_ggc_mx (opt_pass *)): New.
> (gt_pch_nx (opt_pass *)): New.
> (gt_pch_nx_opt_pass): New.
> (pipeline::operat
On 07/26/2013 07:21 PM, Joseph S. Myers wrote:
On Fri, 26 Jul 2013, Andrew MacLeod wrote:
This patch adds an atomic type qualifier to GCC. It can be accessed via
__attribute__((atomic)) or in C11 mode via the _Atomic keyword.
Why the attribute - why not just the keyword?
* When C11 refers
2013/8/1 Paolo Carlini :
> Hi,
>
> gently pinging this small clean-up:
>
> http://gcc.gnu.org/ml/gcc-patches/2013-06/msg00905.html
>
> Thanks!
> Paolo.
It was supposed to provide symmetry and a good place to
put any "cleanup" code we might want to run, but that
never materialized, and now that
Ping?
Hi,
Here is the patch, Tested by profiledbootstrap. Ok for google gcc-4.8?
thanks, Dinar.
profiledbootstrap-fix1.patch
Description: Binary data
Sorry for the delay. The patch is ok and I have committed it to the
google branch.
thanks,
David
On Thu, Aug 1, 2013 at 4:51 PM, Dinar Temirbulatov wrote:
> Ping?
> Hi,
> Here is the patch, Tested by profiledbootstrap. Ok for google gcc-4.8?
> thanks, Dinar.
On Thu, 2013-08-01 at 14:42 -0500, Peter Bergner wrote:
> Jakub & richi,
>
> This bug also affects GCC 4.8. Is it appropriate to backport this patch
> to the FSF 4.8 branch? I can bootstrap and regtest it there too before
> committing if you think it's ok.
FYI, I tested the patch on the FSF 4.8
This patch broke bootstrap on AIX and probably many other targets.
In file included from /tmp/20130801/powerpc-ibm-aix7.1.0.0/libstdc++-v3/include/
debug/safe_sequence.h:34:0,
from /nasfarm/edelsohn/src/src/libstdc++-v3/src/c++11/debug.cc:
26:
/tmp/20130801/powerpc-ibm-aix7.1.0.0
On Thu, Aug 1, 2013 at 9:42 PM, David Edelsohn wrote:
> This patch broke bootstrap on AIX and probably many other targets.
>
> In file included from
> /tmp/20130801/powerpc-ibm-aix7.1.0.0/libstdc++-v3/include/
> debug/safe_sequence.h:34:0,
> from
> /nas
> > What I really need is an int20_t type in the core of gcc, so I can set
> > Pmode to *that*, to avoid the SImode stuff completely. But that's a
> > core change, not a target change.
>
> Sometimes you have to make core changes for a new port. This sounds
> like something that really should be
On 08/02/2013 05:02 AM, Gabriel Dos Reis wrote:
On Thu, Aug 1, 2013 at 9:42 PM, David Edelsohn wrote:
This patch broke bootstrap on AIX and probably many other targets.
In file included from /tmp/20130801/powerpc-ibm-aix7.1.0.0/libstdc++-v3/include/
debug/safe_sequence.h:34:0
58 matches
Mail list logo
