-Original Message-
From: Jeff Law [mailto:l...@redhat.com]
Sent: Wednesday, October 17, 2012 1:02 AM
To: Bin Cheng
Cc: 'Steven Bosscher'; gcc-patches@gcc.gnu.org
Subject: Re: [PATCH RFA] Implement register pressure directed hoist pass
On 10/16/2012 01:44 AM, Bin Cheng wrote:
Hi,
In function arm_expand_compare_and_swap, oldval is converted to SImode
when its mode is QImode/HImode. After FALLTHRU to case SImode,
we should use SImode, other than mode (which is QImode/HImode).
And INSN atomic_compare_and_swapmode_1 expects the operand in
SImode.
No make check
On Fri, Oct 19, 2012 at 7:46 AM, Zhenqiang Chen
zhenqiang.c...@linaro.org wrote:
Hi,
In function arm_expand_compare_and_swap, oldval is converted to SImode
when its mode is QImode/HImode. After FALLTHRU to case SImode,
we should use SImode, other than mode (which is QImode/HImode).
And INSN
On Thu, Oct 18, 2012 at 08:40:32PM -0700, Xinliang David Li wrote:
The following is small bug fix caught during testing. The shadow value
for redzone can be negative, so the shadow type must be signed for the
check to succeed.
Ok for the branch?
Yes (with a proper ChangeLog.asan entry ;)
On Thu, Oct 18, 2012 at 4:00 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
you know richi, i did not know who i was actually talking to. i said who
is this richard beiner person and then i saw the email address.
;)
On 10/18/2012 08:58 AM, Richard Biener wrote:
On Thu, Oct 18, 2012 at
On Thu, Oct 18, 2012 at 6:21 PM, Marek Polacek pola...@redhat.com wrote:
This patch changes folding in fold_sign_changed_comparison in a way
that when we have pointer/non-pointer comparison, we give up folding
here. The issue is e.g. when we have (intptr_t) MEM[(void *)x + 4B]
== (intptr_t) y
On Thu, Oct 18, 2012 at 7:32 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
This patch replaces all instances of INT_CST_LT and INT_CST_LT_UNSIGNED
with
the new functions tree_int_cst_lts_p and tree_int_cst_ltu_p. With the new
implementation of int_cst these functions will be too big to
gfortran was ignoring the TARGET_SYSTEM_ROOT and thus searched in
/usr/include for files specified with #include/include files and for
.mod files.
The solution is to do in gcc/fortran/cpp.c the same as it is done in
gcc/c-family/c-opts.c.
However, the TARGET_SYSTEM_ROOT also has to be
On Thu, Oct 18, 2012 at 2:16 AM, Alexander Ivchenko aivch...@gmail.com wrote:
Hi,
this patch adds new intrinsics for fxsave, fxsave64, xsave, xsave64,
xsaveopt and xsaveopt64 instructions
Bootstrapped on x86-64
Is it ok for trunk?
Changelog entry:
2012-10-16 Alexander Ivchenko
The patch builds in 32-bit on my powerpc64-linux system (which showed up
Eric's problem). I can't test the patch for Dominique's problem, since it
needs a non-linux system to show up. I verified the debug problem:
Eric and Dominique could you try it on your respective systems? Thanks in
Hi,
this patch fixes off-by-one error in the testcase attached. The problem is that
dominance based test used by record_estimate to check whether the given
statement
must be executed at last iteration of the loop is wrong ignoring the side effect
of other statements that may terminate the
On 19 October 2012 14:53, Ramana Radhakrishnan
ramana@googlemail.com wrote:
On Fri, Oct 19, 2012 at 7:46 AM, Zhenqiang Chen
zhenqiang.c...@linaro.org wrote:
Hi,
In function arm_expand_compare_and_swap, oldval is converted to SImode
when its mode is QImode/HImode. After FALLTHRU to case
On Fri, 19 Oct 2012, Jan Hubicka wrote:
Hi,
this patch fixes off-by-one error in the testcase attached. The problem is
that
dominance based test used by record_estimate to check whether the given
statement
must be executed at last iteration of the loop is wrong ignoring the side
effect
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk
and branch.
Richard.
2012-10-19 Richard Biener rguent...@suse.de
PR tree-optimization/54976
* tree-vect-stmts.c (get_vectype_for_scalar_type_and_size):
Robustify against odd inner_mode inputs.
On 10/19/2012 04:22 AM, Richard Biener wrote:
On Thu, Oct 18, 2012 at 7:32 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
This patch replaces all instances of INT_CST_LT and INT_CST_LT_UNSIGNED
with
the new functions tree_int_cst_lts_p and tree_int_cst_ltu_p. With the new
implementation
On Fri, Oct 19, 2012 at 12:59 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 04:22 AM, Richard Biener wrote:
On Thu, Oct 18, 2012 at 7:32 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
This patch replaces all instances of INT_CST_LT and INT_CST_LT_UNSIGNED
with
the
On 10/19/2012 07:13 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 12:59 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 04:22 AM, Richard Biener wrote:
On Thu, Oct 18, 2012 at 7:32 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
This patch replaces all instances of
On Fri, Oct 19, 2012 at 1:49 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 07:13 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 12:59 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 04:22 AM, Richard Biener wrote:
On Thu, Oct 18, 2012 at 7:32 PM,
On 10/19/2012 07:58 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 1:49 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 07:13 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 12:59 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 04:22 AM, Richard Biener
As a followup to:
http://gcc.gnu.org/ml/gcc/2012-10/msg00276.html
2012-10-19 Paulo Matospma...@broadcom.com
* tm.texi, tm.texi.in: Add IDENTIFIER_NODE as an alternative possibility
to possible values of decl.
Paulo Matos
doc.patch
Description: doc.patch
This fixes PR54981, loop distribution was depending on DEBUG_STMT
uses.
Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.
Richard.
2012-10-19 Richard Guenther rguent...@suse.de
PR tree-optimization/54981
* tree-loop-distribution.c
On Fri, Oct 19, 2012 at 2:12 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 07:58 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 1:49 PM, Kenneth Zadeck
zad...@naturalbridge.com wrote:
On 10/19/2012 07:13 AM, Richard Biener wrote:
On Fri, Oct 19, 2012 at 12:59 PM,
On Fri, Oct 19, 2012 at 5:48 AM, Paulo Matos pma...@broadcom.com wrote:
2012-10-19 Paulo Matospma...@broadcom.com
* tm.texi, tm.texi.in: Add IDENTIFIER_NODE as an alternative
possibility
to possible values of decl.
This is OK.
Thanks.
Ian
Eric and Dominique could you try it on your respective systems? Thanks in
advance.
It works for me, thanks.
It does not for me:
In file included from ../../work/gcc/config/rs6000/rs6000.c:1475:0:
../../work/gcc/config/rs6000/rs6000-cpus.def:54:0: error: MASK_STRICT_ALIGN
redefined
On Fri, Oct 19, 2012 at 12:26:13PM +0200, Richard Biener wrote:
+ /* If we felt back to using the mode fail if there was
+ no scalar type for it. */
+ if (scalar_type == NULL_TREE)
+ return NULL_TREE;
s/felt/fell/ ;)
Marek
On Fri, 19 Oct 2012, Jan Hubicka wrote:
Hi,
this patch fixes off-by-one error in the testcase attached. The problem is
that
dominance based test used by record_estimate to check whether the given
statement
must be executed at last iteration of the loop is wrong ignoring the side
I'd like to commit LRA patches on the Sunday. I got a lot of
feedback about LRA for a few last weeks. All patches besides RA part
were approved by Jeff Law. Steven Bosscher and I solved LRA speed
problem on famous PR54146. Now Steven does not object to LRA merge.
I got invaluable
Hi all,
this regression is haunting me a lot in real-world code. It's 'just' a
bogus warning, but it can really swamp the output for modules with
many private procedures.
I have found a fix that is rather trivial in the sense that it's a
one-liner. However, it may not be as trivial conceptually
On 10/10/12 16:03, Greta Yorsh wrote:
Generate epilogue using LDRD in Thumb mode when prefer_ldrd_strd is set in
tune_params.
ChangeLog
gcc/
2012-09-13 Sameera Deshpande sameera.deshpa...@arm.com
Greta Yorsh greta.yo...@arm.com
* config/arm/arm.c (thumb2_emit_ldrd_pop):
On 10/10/12 16:03, Greta Yorsh wrote:
As a result of adding LDRD/STRD patterns in Thumb mode, the compiler
generates LDRD/STRD instead of LDM/STM in some cases. This patch adjusts
existing tests to accept LDRD/STRD in addition to LDM/STM.
ChangeLog
gcc/testsuite
2012-09-13 Sameera Deshpande
What about the conservative variant of simply
else
delta = double_int_one;
I think it would be bad idea: it makes us to completely unroll one interation
too many that bloats code for no benefit. No optimization cancels the path in
CFG because of undefined effect and
On 2012-10-12 18:19 , Sriraman Tallam wrote:
When the front-end sees more than one decl for foo, it calls a target hook to
determine if they are versions. To prevent duplicate definition errors with
other
versions of foo, decls_match function in cp/decl.c is made to return false
when 2 decls
On 18 October 2012 14:41, Richard Earnshaw wrote:
+/* Checks whether the operands are valid for use in an LDRD/STRD
instruction.
+ Assumes that RT, RT2, and RTN are REG. This is guaranteed by the
patterns.
+ Assumes that the address in the base register RTN is word aligned.
Pattern
+
Quoting Richard Sandiford rdsandif...@googlemail.com:
Joern Rennecke joern.renne...@embecosm.com writes:
Quoting Richard Sandiford rdsandif...@googlemail.com:
The fact that we even have shared unique ids is pretty bad -- and surely
a contradiction in terms -- but I think both ways of handling
On 19/10/12 16:20, Greta Yorsh wrote:
Removed the condition !optimize_function_for_size_p (cfun)).
The condition current_tune-prefer_ldrd_strd is needed because the
patterns
for LDRD/STRD appear before the patterns for LDM/STM that can match the same
RTL
(two register in the list). Condition
On 2012-10-19 10:31 , Vladimir Makarov wrote:
So my plan is to merge trunk into LRA branch first and, if it looks
ok, commit LRA to the trunk on Sunday to be ready solve problems (if
there are any) starting Monday morning EST. LRA merge will affect only
x86/x86-64 target. If you suspect
On 12-10-19 11:46 AM, Diego Novillo wrote:
On 2012-10-19 10:31 , Vladimir Makarov wrote:
So my plan is to merge trunk into LRA branch first and, if it looks
ok, commit LRA to the trunk on Sunday to be ready solve problems (if
there are any) starting Monday morning EST. LRA merge will
On Wed, 17 Oct 2012, Manuel L?pez-Ib??ez wrote:
This patch changes most trivial cases to use the new (Lang)EnabledBy.
Bootstrapped and regression tested on x86_64-linux-gnu. OK?
2012-10-17 Manuel L?pez-Ib??ez m...@gcc.gnu.org
PR c/53063
PR c/40989
c-family/
*
On 2012-10-19 11:51 , Vladimir Makarov wrote:
I counted 4 vectors (with about 25 lines referencing them). One vector
(point_freq_vec) is analogous to IRA.
Perfect. That shouldn't be hard at all then. Thanks.
Diego.
Hi Janus,
Janus Weil wrote:
I have found a fix that is rather trivial in the sense that it's a
one-liner.
I think it is the proper fix.
However, it may not be as trivial conceptually (in particular I'm not 100% sure
what caused this regression in the first place
I think it is the
This adds support for piecewise construction of std::pair by
scoped_allocator_adaptor. The only thing missing from
scoped_allocator_adaptor now is that my definition of OUTERMOST isn't
recursive so doesn't work for nested scoped_allocator_adaptors.
That's a suitably obscure use case that I'm not
On Wed, 17 Oct 2012, Manuel L?pez-Ib??ez wrote:
documentation but I can also implement -Wformat=0 being an alias for
-Wno-format and -Wformat=1 an alias for -Wformat and simply reject
-Wno-format=.
I think that's what's wanted; -Wno-format= should be rejected, -Wformat=
should take an
On Thu, 18 Oct 2012, Jonathan Wakely wrote:
This adds a warning switch for the existing returning address of
local variable warnings in the C and C++ FEs which are enabled by
default but have no switch controlling them. Adding a switch allows it
to be turned into an error with
I can't approve this, but some comments:
Kenneth Zadeck zad...@naturalbridge.com writes:
diff --git a/gcc/alias.c b/gcc/alias.c
index 0c6a744..9e67823 100644
--- a/gcc/alias.c
+++ b/gcc/alias.c
@@ -1490,9 +1490,9 @@ rtx_equal_for_memref_p (const_rtx x, const_rtx y)
case VALUE:
What about the conservative variant of simply
else
delta = double_int_one;
I think it would be bad idea: it makes us to completely unroll one
interation
too many that bloats code for no benefit. No optimization cancels the path
in
CFG because of undefined
-Original Message-
From: Richard Earnshaw
Sent: 19 October 2012 16:44
To: Greta Yorsh
Cc: GCC Patches; Ramana Radhakrishnan; ni...@redhat.com;
p...@codesourcery.com
Subject: Re: [PATCH, ARM][1/4] New RTL patterns for LDRD/STRD in Thumb
mode
On 19/10/12 16:20, Greta Yorsh wrote:
The table's a bit out of date
* doc/xml/manual/status_cxx2011.xml: Update.
Committed to trunk.
diff --git a/libstdc++-v3/doc/xml/manual/status_cxx2011.xml
b/libstdc++-v3/doc/xml/manual/status_cxx2011.xml
index 226eef9..6b1ad51 100644
--- a/libstdc++-v3/doc/xml/manual/status_cxx2011.xml
gfortran's INTENT(IN) check was too strict for do variables. While a
variable in the normal do-stmt and in an io-implied-do is in the scope
and, hence, the variable may not be modified for a nonpointer intent(in)
variable.
However, ac-implied-do and data-implied-do live in their own scope and
On 19/10/12 17:51, Greta Yorsh wrote:
-Original Message-
From: Richard Earnshaw
Sent: 19 October 2012 16:44
To: Greta Yorsh
Cc: GCC Patches; Ramana Radhakrishnan; ni...@redhat.com;
p...@codesourcery.com
Subject: Re: [PATCH, ARM][1/4] New RTL patterns for LDRD/STRD in Thumb
mode
On
Unwind/debug info on x86-64/Windows:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg00013.html
PR bootstrap/54820 (stage #1 bootstrap failure)
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg01093.html
--
Eric Botcazou
I tried it, and this version works for me.
Your probably do not need to add the m4 subdir under libasan. The
required m4 files are either in .. or ../config dir. See how
libmudflap does it.
Other than that, if there are no other comments, the change is good to
check into the branch. Remaining
Hi Tobias,
I have found a fix that is rather trivial in the sense that it's a
one-liner.
I think it is the proper fix.
good :)
However, it may not be as trivial conceptually (in particular I'm not 100%
sure what caused this regression in the first place
I think it is the combination
On 2012-10-19 13:16 , Xinliang David Li wrote:
Other than that, if there are no other comments, the change is good to
check into the branch. Remaining bugs can always be found and fixed
later.
Agreed. David, could you commit the patch after Wei has addressed your
feedback? I have just
David, I put the m4 subdir under libasan because once I use the .m4
files (libtool.m4 lt~obsolete.m4 ltoptions.m4 ltsugar.m4
ltversion.m4) and ltmain.sh under $topsrcdir, the problem that a bad
libtool was generated under
$topbuilddir/x86_64-unknown-linux-gnu/libasan you met yesterday
appeared.
Jim,
How do you want to move forward with the VLE patch? Can you localize
more of the changes?
Thanks, David
2012-10-18 Jakub Jelinek ja...@redhat.com
PR debug/54970
* cfgexpand.c (expand_debug_expr): Expand MEM_REF[var, n]
as DEBUG_IMPLICIT_PTR + n if var expands to DEBUG_IMPLICIT_PTR.
* tree-sra.c (create_access_replacement): Allow also MEM_REFs
with
I checked in your patch along with my patches into the trunk.
[gcc]
2012-10-19 Michael Meissner meiss...@linux.vnet.ibm.com
Dominique Dhumieres domi...@lps.ens.fr
* config/rs6000/rs6000-cpus.def (POWERPC_MASKS): Don't assume
OPTION_MASK_STRICT_ALIGN is defined.
Quoting Richard Sandiford rdsandif...@googlemail.com:
Joern Rennecke joern.renne...@embecosm.com writes:
When the condition is not fulfilled, we want to keep the length from the
previous iteration.
Right, that's what I mean. So we need to make sure that the difference
between the address
The library builds fine with the following diff. The file acinclude.m4
is cloned from libmudflap.
David
index 485d169..3e847f1 100644
--- a/libasan/aclocal.m4
+++ b/libasan/aclocal.m4
@@ -1037,8 +1037,8 @@ AC_SUBST([am__tar])
AC_SUBST([am__untar])
]) # _AM_PROG_TAR
On Fri, Oct 19, 2012 at 3:52 PM, Xinliang David Li davi...@google.com wrote:
The library builds fine with the following diff. The file acinclude.m4
is cloned from libmudflap.
This is OK with the corresponding ChangeLog entry (in ChangeLog.asan).
Diego.
From: Jonathan Wakely jwakely@gmail.com
Date: Fri, 19 Oct 2012 18:16:51 +0200
This adds support for piecewise construction of std::pair by
scoped_allocator_adaptor. The only thing missing from
scoped_allocator_adaptor now is that my definition of OUTERMOST isn't
recursive so doesn't
(Adding Joseph in Cc: because we'll want the same behavior in C
afterwards. Conversation starts here:
http://gcc.gnu.org/ml/gcc-patches/2012-10/msg01665.html )
On Thu, 18 Oct 2012, Marc Glisse wrote:
Clang's webpage says they support ?: for vector types, but my version (3.1)
rejects it, I'll
On Fri, 19 Oct 2012, Hans-Peter Nilsson wrote:
Looks like _U is one of those identifiers that should be
avoided:
[...]
What's the preferred replacement?
std::pair seems to use _U1 and _U2 rather than _U and _V.
--
Marc Glisse
2012/10/19 Marc Glisse marc.gli...@inria.fr:
On Fri, 19 Oct 2012, Hans-Peter Nilsson wrote:
Looks like _U is one of those identifiers that should be
avoided:
[...]
What's the preferred replacement?
std::pair seems to use _U1 and _U2 rather than _U and _V.
When not numbered, I have seen
As I beging work on extensions for a potential future machine, I am running
into the overload table. In the power7 timeframe, I just gritted my teeth, and
added all of the appropriate entries, but I would like to visit making the
table somewhat friendlier to edit and add new overloaded functions.
On Thu, 4 Oct 2012, Iyer, Balaji V wrote:
Here is a link to the latest spec. This should clear several of the
questions you are seeking.
(http://software.intel.com/sites/default/files/m/4/e/7/3/1/40297-
Intel_Cilk_plus_lang_spec_2.htm#array)
This specification is much improved, especially as
On 10/18/2012 01:33 PM, Bernhard Reutner-Fischer wrote:
On 18 October 2012 17:30:20 Meador Inge mead...@codesourcery.com wrote:
Ping ^ 2
Been a while but wasn't --with-build-sysroot for exactly this?
AFAICT, no. --with-build-sysroot seems to be used for setting a different
sysroot to use
CC'ing the LTO maintainers.
On 10/18/2012 10:30 AM, Meador Inge wrote:
Ping ^ 2.
On 10/09/2012 09:44 PM, Meador Inge wrote:
Ping.
On 10/04/2012 03:45 PM, Meador Inge wrote:
Hi All,
Currently the gcc-{ar,nm,ranlib} utilities assume that binutils is in
path when invoking the wrapped
Hi Joseph,
Thank you very much for your response. I will look into this and get
back to you soon!
-Balaji V. Iyer.
-Original Message-
From: Joseph Myers [mailto:jos...@codesourcery.com]
Sent: Friday, October 19, 2012 5:38 PM
To: Iyer, Balaji V
Cc: Richard Guenther;
const struct altivec_builtin_types altivec_overloaded_builtins[] = {
[big fat table constructed using many macros]
Now, it bothers me that every time we do the overloaded builtins, we
essentially have to scan the table in a linear fashion. I'm
thinking that we
may want to move these tables
On Thu, Oct 18, 2012 at 3:36 PM, Easwaran Raman era...@google.com wrote:
Hi,
During expression reassociation, statements are conservatively moved
downwards to ensure that dependences are correctly satisfied after
reassocation. This could lead to lengthening of live ranges. This
patch moves
On Fri, Oct 19, 2012 at 5:13 PM, Xinliang David Li davi...@google.com wrote:
On Thu, Oct 18, 2012 at 3:36 PM, Easwaran Raman era...@google.com wrote:
Hi,
During expression reassociation, statements are conservatively moved
downwards to ensure that dependences are correctly satisfied after
As suggested in http://gcc.gnu.org/ml/gcc/2012-10/msg00285.html, I
have updated the attached patch to rename 'dump_enabled_phase' to
'dump_enabled_phase_p'. The 'dump_enabled_p ()' doesn't take any
argument and can be used as a predicate for the dump calls.
Once this patch gets in, the plan is to
On 19 October 2012 21:28, Hans-Peter Nilsson wrote:
Looks like _U is one of those identifiers that should be
avoided: Grep yields:
src/newlib/libc/include/ctype.h:#define _U 01
Argh! my bad, sorry - fix on the way ...
On 20 October 2012 02:35, Jonathan Wakely wrote:
On 19 October 2012 21:28, Hans-Peter Nilsson wrote:
Looks like _U is one of those identifiers that should be
avoided: Grep yields:
src/newlib/libc/include/ctype.h:#define _U 01
Argh! my bad, sorry - fix on the way ...
Sorry about that,
Thanks.
On Oct 19, 2012, at 2:27 PM, Michael Meissner meiss...@linux.vnet.ibm.com
wrote:
As I beging work on extensions for a potential future machine, I am running
into the overload table.
I still like the built-in mechanism Kenny and I have Not perfect, but
vastly better... If you wanted to
77 matches
Mail list logo