The following patch prevents mistaken ASHIFT exchange found by
Richard Sandiford and removes check early_clobber matching operand.
The patch was successfully bootstrap on x86-64.
Committed as rev. 192742.
2012-10-23 Vladimir Makarov
* lra-constraints.c (extract_loc_address_regs):
Uros reported that GCC after LRA submitting has a new test failure on
x86. The reason was in wrong update live info in EBB containing empty
BB as the last EBB block. For such case live_out of the last EBB was
undefined and actually a garbage. It resulted in triggering a code
which checks t
On 10/21/2012 05:27 AM, Richard Sandiford wrote:
Hi Vlad,
Richard, sorry for long delay with the answer. I was really busy all
these days trying to fix a lot of GCC testsuite failures (for some ones
it was extremely difficult to find failure reasons) and push LRA into trunk.
Vladimir Makaro
The patch is a patch posted for trunk on
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg02085.html
Committed as rev. 192744.
Index: lra-constraints.c
===
--- lra-constraints.c (revision 192719)
+++ lra-constraints.c (working copy)
@
On 10/23/12, Kenneth Zadeck wrote:
> On 10/23/2012 02:38 PM, Lawrence Crowl wrote:
>> On 10/23/12, Kenneth Zadeck wrote:
>>> On 10/23/2012 10:12 AM, Richard Biener wrote:
+ inline bool minus_one_p () const;
+ inline bool zero_p () const;
+ inline bool one_p () const;
+ inl
When we try to thread across a back edge in the CFG we have a check to
ensure that we don't use temporary equivalences which are invalidated by
traversal of the back edge to simplify the final conditional.
About a year ago I added code to pick up secondary threading
opportunities after an in
On Tue, Oct 23, 2012 at 02:35:24PM -0600, Jeff Law wrote:
> +/* Return TRUE if the statement at the end of e->dest depends on
> + the output of any statement in BB. Otherwise return FALSE.
> +
> + This is used when we are threading a backedge and need to ensure
> + that temporary equivalenc
As reported by Ian and Peter in the audit trail, the check I added to detect
whether -static-libstdc++ is supported by g++ doesn't work because the option
is silently rejected by versions prior to 4.5.
The attached patch forces an error for these versions so as to make the check
always fail. T
On 10/23/2012 02:50 PM, Jakub Jelinek wrote:
On Tue, Oct 23, 2012 at 02:35:24PM -0600, Jeff Law wrote:
+/* Return TRUE if the statement at the end of e->dest depends on
+ the output of any statement in BB. Otherwise return FALSE.
+
+ This is used when we are threading a backedge and need t
On Tue, Oct 23, 2012 at 03:21:59PM -0600, Jeff Law wrote:
> On 10/23/2012 02:50 PM, Jakub Jelinek wrote:
> >>+static bool
> >>+cond_arg_set_in_bb (edge e, basic_block bb, int n)
> >>+{
> >>+ ssa_op_iter iter;
> >>+ use_operand_p use_p;
> >>+ gimple last = gsi_stmt (gsi_last_bb (e->dest));
> >
>
Committed as obvious.
* tree-ssa-threadedge.c (cond_arg_set_in_bb): Remove unused
debugging argument.
Index: tree-ssa-threadedge.c
===
--- tree-ssa-threadedge.c (revision 192745)
+++ tree-ssa-threadedge.c
On 2012-10-24 01:50, Torvald Riegel wrote:
> Clarify ABI requirements for data-logging functions.
>
> * libitm.texi: Clarify ABI requirements for data-logging functions.
Ok.
r~
From: Vladimir Makarov
Date: Tue, 23 Oct 2012 11:46:34 -0400
> Hi, I was going to merge LRA into trunk last Sunday. It did not
> happen. LRA was actively changed last 4 weeks by implementing
> reviewer's proposals which resulted in a lot of new LRA regressions on
> GCC testsuite in comp
On 10/23/2012 04:25 PM, Lawrence Crowl wrote:
On 10/23/12, Kenneth Zadeck wrote:
On 10/23/2012 02:38 PM, Lawrence Crowl wrote:
On 10/23/12, Kenneth Zadeck wrote:
On 10/23/2012 10:12 AM, Richard Biener wrote:
+ inline bool minus_one_p () const;
+ inline bool zero_p () const;
+ inline boo
On 2012-10-24 01:48, Torvald Riegel wrote:
> Ask dispatch whether it requires serial mode.
>
> * retry.cc (gtm_thread::decide_begin_dispatch): Ask dispatch whether
> it requires serial mode instead of assuming that for certain
> dispatchs.
> * dispatch.h (abi_dispat
On 10/23/2012 03:22 PM, Jakub Jelinek wrote:
On Tue, Oct 23, 2012 at 03:21:59PM -0600, Jeff Law wrote:
On 10/23/2012 02:50 PM, Jakub Jelinek wrote:
+static bool
+cond_arg_set_in_bb (edge e, basic_block bb, int n)
+{
+ ssa_op_iter iter;
+ use_operand_p use_p;
+ gimple last = gsi_stmt (gsi_las
> I should probably have responded to this earlier, sorry. I'm not sure
> which part you mean, so here's an attempt at justifying the whole block:
>
> 1) WORDS_BIG_ENDIAN is deliberately ignored:
>
> /* The following line once was done only if WORDS_BIG_ENDIAN,
>but I think that is
On Tue, Oct 23, 2012 at 03:34:46PM -0600, Jeff Law wrote:
> >I think it should be backported to 4.7, perhaps with a few days delay after
> >the
> >trunk commit.
> Do we even have debug statements after control flow statements?
They shouldn't be there, so if you just give up the same way for gsi_s
On Oct 23, 2012, at 6:52 AM, Dominique Dhumieres wrote:
> Following the changes in [PATCH] Add option for dumping to stderr
> (issue6190057)
> the tests gcc.dg/vect/vect-8[23]_64.c fails on powerpc*-*-*.
> This patch adjust the dump files and has been tested on powerpc-apple-darwin9.
Ok.
> gcc/
On 18 October 2012 00:24, Gabriel Dos Reis
wrote:
> On Wed, Oct 17, 2012 at 6:26 AM, Manuel López-Ibáñez
> wrote:
>> On 17 October 2012 11:55, Dodji Seketeli wrote:
>>> Hello Manuel,
>>>
>>> Let's CC Gaby on this one as well.
>>>
>>> Manuel López-Ibáñez writes:
>>>
The problem is that the
On Tue, Oct 23, 2012 at 2:11 PM, Eric Botcazou wrote:
>
> 2012-10-23 Eric Botcazou
>
> PR bootstrap/54820
> * configure.ac (have_static_libs): Force 'no' for GCC version < 4.5.
> * configure: Regenerate.
This is OK.
Thanks.
Ian
On Oct 23, 2012, at 6:33 AM, Dominique Dhumieres wrote:
> The test libgomp.graphite/force-parallel-6.c is not valid as it tries
> to write Y[2*N] for Y defined as
Ok. [ Could someone check it in for Dominique? ]
> 2012-10-23 Dominique d'Humieres
>
> * testsuite/libgomp.graphite/force-
Patch to update the MAINTAINERS file following the merge of the aarch64
port.
R.
2012-10-23 Richard Earnshaw
* MAINTAINERS (aarch64): Add Marcus and myself.
Index: MAINTAINERS
===
--- MAINTAINERS (revision 192746)
+++ MA
Gold on powerpc64 doesn't support old ABI objects, but libffi contains
old ABI assembly. This patch modifies those files to support both old
and new ABI, and adds a builtin define to powerpc64 gcc that can be
used to select between the ABIs in assembly. I figure a define is
generally useful, and
On Wed, Oct 24, 2012 at 12:32 AM, Richard Earnshaw wrote:
> Patch to update the MAINTAINERS file following the merge of the aarch64
> port.
Congrats on the new port!
Will you also add an announcement of this to the news page (home page)
and to gcc-4.8/changes.html?
Ciao!
Steven
On Tue, Oct 23, 2012 at 03:28:37PM -0600, Jeff Law wrote:
>
> Committed as obvious.
>
>* tree-ssa-threadedge.c (cond_arg_set_in_bb): Remove unused
> debugging argument.
Could you please remove the third argument in the calls to cond_arg_set_in_bb
as well?
/MF
> Index: tree-ssa-
On 23/10/12 23:36, Steven Bosscher wrote:
On Wed, Oct 24, 2012 at 12:32 AM, Richard Earnshaw wrote:
Patch to update the MAINTAINERS file following the merge of the aarch64
port.
Congrats on the new port!
Will you also add an announcement of this to the news page (home page)
and to gcc-4.8/cha
On 12-10-23 5:28 PM, David Miller wrote:
From: Vladimir Makarov
Date: Tue, 23 Oct 2012 11:46:34 -0400
Hi, I was going to merge LRA into trunk last Sunday. It did not
happen. LRA was actively changed last 4 weeks by implementing
reviewer's proposals which resulted in a lot of new LRA
Hi,
I'm pinging this patchlet:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg01013.html
For sure not an high priority issue, neither I can say to fully
understand whether in C++ we can and should have the exact same
semantics of the __transparent_union__ attribute in C, but I think that
i
The trunk seems to be broken at r192749 due to this patch.
../../srctrunk/gcc/tree-ssa-threadedge.c: In function ‘void
thread_across_edge(gimple_statement_d*, edge_def*, bool,
vec_t**, tree_node* (*)(gimple_statement_d*,
gimple_statement_d*))’:
../../srctrunk/gcc/tree-ssa-threadedge.c:583: error:
The following trivial patch seems to fix it.
Index: tree-ssa-threadedge.c
===
--- tree-ssa-threadedge.c (revision 192749)
+++ tree-ssa-threadedge.c (working copy)
@@ -743,7 +743,7 @@
safe to thread this edge. */
if (e->flags
Committed in r192750.
Thanks,
Sharad
On Tue, Oct 23, 2012 at 2:46 PM, Mike Stump wrote:
> On Oct 23, 2012, at 6:52 AM, Dominique Dhumieres wrote:
>> Following the changes in [PATCH] Add option for dumping to stderr
>> (issue6190057)
>> the tests gcc.dg/vect/vect-8[23]_64.c fails on powerpc*-*-
On Tue, 23 Oct 2012, Manuel López-Ibáñez wrote:
> The problem is how to represent that Wformat-y2k is enabled by
> -Wformat=X with X >= 2, while Wformat-zero-length is enabled by X >=1.
>
> One possiblity is to allow to specify a condition directly:
I guess that's reasonable.
--
Joseph S. Myer
The obvious fix.
commit 36d5e6eace0697b2f8613ab5c24bc23ea359d347
Author: law
Date: Wed Oct 24 00:43:24 2012 +
* tree-ssa-threadedge.c (thread_across_edge): Remove unused
parameter in call to cond_arg_set_in_bb.
git-svn-id: svn+ssh://gcc.gnu.org/svn/gcc/trunk
Quoting Richard Biener :
On Tue, Oct 16, 2012 at 9:35 PM, Joern Rennecke
wrote:
..
Well, we could split it anyway, and give ports without the need for
multiple length attributes the benefit of the optimistic algorithm.
I have attached a patch that implements this.
Looks reasonable to me, t
From: Vladimir Makarov
Date: Tue, 23 Oct 2012 19:04:03 -0400
> I am not sure that anything except x86/x86-64 will work now on the
> branch. There were too many changes on the branch and I tested only
> x86/x86-64. I'll start testing the rest of targets on the branch
> next week when LRA is sett
On Tue, Oct 23, 2012 at 2:52 AM, Richard Biener
wrote:
> On Mon, Oct 22, 2012 at 8:31 PM, Easwaran Raman wrote:
>> On Mon, Oct 22, 2012 at 12:59 AM, Richard Biener
>> wrote:
>>> On Fri, Oct 19, 2012 at 12:36 AM, Easwaran Raman wrote:
Hi,
During expression reassociation, statement
From: David Miller
Date: Tue, 23 Oct 2012 21:44:05 -0400 (EDT)
> The first issue sparc runs into is that it does not define it's
> extra constraints properly. In particular 'T' and 'W' must use
> define_memory_constraint.
>
> Otherwise the EXTRA_MEMORY_CONSTRAINT logic in process_alt_operands()
Prerequisites to allow the port to build properly: the lock_length attribute:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg01890.html
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg02120.html
And from Easwaran Raman :
PR middle-end/54957
* optabs.c (emit_cmp_and_jump_insn_1): Remo
gcc:
2012-10-22 Saurabh Verma
Ramana Radhakrishnan
Joern Rennecke
Muhammad Khurram Riaz
Brendan Kehoe
Michael Eager
* config/arc: New directory.
arc.md.tar.xz
Description: application/xz
libgcc:
2012-10-18 Joern Rennecke
Brendan Kehoe
* libgcc/config/arc: New directory.
arc-libgcc.tar.xz
Description: application/xz
2012-10-22 Saurabh Verma
Ramana Radhakrishnan
Joern Rennecke
Muhammad Khurram Riaz
Brendan Kehoe
Michael Eager
* config/arc, common/config/arc: New directories.
arc-rest.tar.xz
Description: application/xz
On Tue, Oct 23, 2012 at 6:34 PM, Alan Modra wrote:
> Gold on powerpc64 doesn't support old ABI objects, but libffi contains
> old ABI assembly. This patch modifies those files to support both old
> and new ABI, and adds a builtin define to powerpc64 gcc that can be
> used to select between the AB
Back in May 2004 Richard Henderson ensured that using a '+' constraint
in an extended asm statement would work with an 'm' constraint.
http://gcc.gnu.org/ml/gcc-patches/2004-05/msg00438.html
Unfortunately he did not update the documentation. I verified that
Richard's code is still in the compile
This patch for mainline corresponds to the one I committed earlier on
the 4.7 branch. This sets a different version for libgo.so on mainline
going forward. Bootstrapped and ran Go testsuite on
x86_64-unknown-linux-gnu. Committed to mainline.
I actually committed this yesterday but forgot to pre
While playing around with LRA on sparc I noticed that we had some
poorly formed target memory constraints on sparc.
In particular, they were not using define_memory_constraint, so we
would not get a true return from EXTRA_MEMORY_CONSTRAINT for them.
Also, these were matching 'reg' objects for sp
On Tue, 23 Oct 2012, Vladimir Makarov wrote:
Hi, I was going to merge LRA into trunk last Sunday. It did not happen.
LRA was actively changed last 4 weeks by implementing reviewer's proposals
which resulted in a lot of new LRA regressions on GCC testsuite in comparison
with reload. Finally,
101 - 147 of 147 matches
Mail list logo
