On Mon, Nov 28, 2011 at 11:10:56PM +0100, Eric Botcazou wrote:
Here is an attempt to make the check more complete (e.g.
the change wouldn't see overlap if addr was PLUS of two REGs,
where one of the REGs was based on internal_arg_pointer, etc.)
and less pessimistic. As tree-tailcall.c
All,
Whilst developing the Cortex-A15 integer pipeline patch it was noted
that the MD_INCLUDES variable in config/arm/t-arm has not been kept
up-to-date.
The attached patch fixes this, and rearranges the list of md files into
alphabetical order.
The list was generated using `ls -1 *.md |
Dear Tobias,
Build and regtested on x86-64-linux (trunk and trunk + Paul's patch).
OK?
Many thanks for the patch - it's OK for trunk.
Cheers
Paul
On 29/11/11 09:42, Matthew Gretton-Dann wrote:
All,
Whilst developing the Cortex-A15 integer pipeline patch it was noted
that the MD_INCLUDES variable in config/arm/t-arm has not been kept
up-to-date.
The attached patch fixes this, and rearranges the list of md files into
alphabetical
gcc/ChangeLog:
2011-11-28 Matthew Gretton-Dannmatthew.gretton-d...@arm.com
* config/arm/arm.c (arm_issue_rate): Cortex-A15 can triple
issue.
* config/arm/arm.md (mul64): Add new attribute.
(generic_sched): Cortex-A15 is not scheduled generically
For devices with 8-bit SP reading the high byte SP_H of SP will get garbage.
The patch uses CLR instead of IN SP_H to read the high part of SP.
There are two issues with this patch:
== 1 ==
I cannot really test it because for devices that small running test suite
does not give usable results.
On Mon, Nov 28, 2011 at 05:33:17PM -0800, Richard Henderson wrote:
On 11/27/2011 02:57 PM, Alan Modra wrote:
This is the mutex part. Depends on
http://gcc.gnu.org/ml/gcc-patches/2011-11/msg02454.html for MEMMODEL_*
values.
* config/linux/mutex.h: Use atomic rather than sync
Hi Jayant,
Sorry, I had not gone well through the libgcc related changes in GCC-4.7.
Hence I missed moving rest of the libgcc routines to top level libgcc
directory.
I noticed a couple of minor issues in the libgcc changes:
* LIB1ASMSRC is relative to libgcc/config, and the file is called
Since there were a number of changes requested, and some confusion
over what was happening in this code, it might be better if I post
a new patch and we start over. In this patch I'm still using the MSB
for the wait flag, and count in the low 31 bits, as that way is
slightly more efficient on
PING.
On 11/22/11 15:11, Yufeng Zhang wrote:
Hi,
This patch changes the macro CDBL in gcc.dg/compat/compat-common.h
to use complex floating-point constant rather than complex integer
constant.
CDBL is a macro to help construct complex floating-point constants in a
number of compat tests.
I
On 29/11/11 13:20, Yufeng Zhang wrote:
PING.
On 11/22/11 15:11, Yufeng Zhang wrote:
Hi,
This patch changes the macro CDBL in gcc.dg/compat/compat-common.h
to use complex floating-point constant rather than complex integer
constant.
CDBL is a macro to help construct complex
Iain, could you let me know if the attached patch fixes your problem?
The patch changes gimple_call_set_cannot_inline to update the
corresponding callgraph edge, if needed. I did not touch any of the
other calls, because sometimes we are calling this function in IPA
mode, and so we don't know
On 29 Nov 2011, at 13:44, Diego Novillo wrote:
Iain, could you let me know if the attached patch fixes your problem?
apologies for not responding to the last message -
- Richi has already resolved the Ada issue as far as it affected
x86-64/darwin.
I assumed your message was addressed
On Tue, Nov 29, 2011 at 08:50, Iain Sandoe
develo...@sandoe-acoustics.co.uk wrote:
On 29 Nov 2011, at 13:44, Diego Novillo wrote:
Iain, could you let me know if the attached patch fixes your problem?
apologies for not responding to the last message -
- Richi has already resolved the Ada
The following fails:
FAIL: g++.dg/eh/simd-3.C -std=gnu++98 execution test
FAIL: g++.dg/eh/simd-3.C -std=gnu++11 execution test
FAIL: g++.dg/eh/simd-4.C -std=gnu++98 execution test
FAIL: g++.dg/eh/simd-4.C -std=gnu++11 execution test
are the tip of an unwinding doesn't work with vectors iceberg
Way behind, sorry, and probably irrelevant anyway, but:
Aldy Hernandez al...@redhat.com writes:
+#define ITM_BARRIERS(T) \
+ extern _ITM_TYPE_##T _ITM_R##T(const _ITM_TYPE_##T *) ITM_REGPARM; \
+ extern _ITM_TYPE_##T _ITM_RaR##T(const _ITM_TYPE_##T *) ITM_REGPARM;
\
+ extern
As Rainer recently pointed out, libgcc/config/rs6000/t-darwin64
overrides the t-darwin version.
This would make it miss the out-of-line saves.
corrected as attached,
OK for trunk?
Iain
libgcc:
* config/rs6000/t-darwin64 (LIB2ADD): Add fp and gp save routines.
Index:
The bug observes that atomic and/or operations can occasionally be
replaced with a exchange operation.
fetch_and (x, 0) - exchange (x, 0)
fetch_or (x, -1)- exchange (x, -1)
This is now trivial in the general expansion code. Others could easily
be added later.
This patch
On Mon, Nov 28, 2011 at 7:33 PM, Richard Henderson r...@redhat.com wrote:
On 11/28/2011 04:26 PM, Richard Henderson wrote:
On 11/28/2011 03:05 PM, Richard Henderson wrote:
On 11/28/2011 02:16 PM, Alan Modra wrote:
Hmm, I suppose you could argue that powerpc and others ought to not
generate
Jason Merrill ja...@redhat.com writes:
On 11/28/2011 11:54 AM, Dodji Seketeli wrote:
@@ -3028,10 +3028,12 @@ find_parameter_packs_r (tree*tp, int *walk_subtrees,
void* data)
case BOUND_TEMPLATE_TEMPLATE_PARM:
/* Check the template itself. */
- cp_walk_tree (TREE_TYPE
On 11/29/2011 07:13 AM, David Edelsohn wrote:
... actually, this one. There's no reason to differentiate between strong
and weak compare-and-swap when computing boolval.
Has anyone bootstrapped and regression-tested the patch?
Yes, I did so last night on gcc110.
r~
I've now tried to build c6x-uclinux, not only c6x-elf, and found another
problem: crtbegin.o wasn't built. I committed the following patch;
please doublecheck there wasn't a reason for how it was being done
previously.
Bernd
Index: libgcc/config.host
On 11/29/2011 04:57 PM, Bernd Schmidt wrote:
I've now tried to build c6x-uclinux, not only c6x-elf, and found another
problem: crtbegin.o wasn't built. I committed the following patch;
please doublecheck there wasn't a reason for how it was being done
previously.
It's ok, indeed we should look
On 11/29/2011 04:49 AM, Alan Modra wrote:
Since there were a number of changes requested, and some confusion
over what was happening in this code, it might be better if I post
a new patch and we start over. In this patch I'm still using the MSB
for the wait flag, and count in the low 31 bits,
Georg-Johann Lay wrote:
For devices with 8-bit SP reading the high byte SP_H of SP will get garbage.
The patch uses CLR instead of IN SP_H to read the high part of SP.
There are two issues with this patch:
== 1 ==
I cannot really test it because for devices that small running test
On Tue, Nov 29, 2011 at 10:50 AM, Richard Henderson r...@redhat.com wrote:
On 11/29/2011 07:13 AM, David Edelsohn wrote:
... actually, this one. There's no reason to differentiate between strong
and weak compare-and-swap when computing boolval.
Has anyone bootstrapped and regression-tested
On 11/26/11 12:50, Richard Sandiford wrote:
Could we use DF_REF_USES/DEFS here instead?
I suppose we could, but I can't really get excited about it...
Also:
+/* Look for sets of call-saved registers in the first block of the
+ function, and move them down into successor blocks if the
On Mon, Nov 28, 2011 at 7:02 PM, Richard Henderson r...@redhat.com wrote:
On 11/28/2011 03:54 PM, David Edelsohn wrote:
On Wed, Nov 23, 2011 at 8:57 AM, Jakub Jelinek ja...@redhat.com wrote:
On Wed, Nov 23, 2011 at 07:47:46AM -0600, Aldy Hernandez wrote:
@@ -4198,7 +4198,7 @@
On Tue, Nov 29, 2011 at 11:22:15AM -0500, David Edelsohn wrote:
A patch using DECL_ONE_ONLY works for me as well. Can someone approve
that version of the patch?
This is ok for the trunk, thanks.
* trans-mem.c (ipa_tm_create_version_alias): Mangle new_decl
if DECL_ONE_ONLY.
On 11/11/11 17:22, Bernd Schmidt wrote:
On 11/11/11 16:30, Joey Ye wrote:
-fstrict-volatile-bitfields doesn't work incorrectly in some cases
when storing into a volatile bit-field.
Bernd provided a fix here about 1 year ago:
http://gcc.gnu.org/ml/gcc-patches/2010-12/msg00217.html.
But it
On 11/29/2011 10:41 AM, Dodji Seketeli wrote:
+ cp_walk_tree (DECL_ORIGINAL_TYPE (TYPE_NAME (t)),
+ find_parameter_packs_r,
+ ppd, ppd-visited);
I still don't think we want to walk the underlying type, since it isn't
part of the syntactic form of the type
So, I still think this patch is the best way to go forward, and it
does
fix incorrect code generation. Would appreciate an OK.
Ping.
If you don't hear any objections within a week, please proceed.
Thank you,
Mark
On 11/29/2011 04:27 AM, Alan Modra wrote:
Did you see my comment about the OpenMP spec requirement that critical,
ordered and parallel regions have a flush on entry and on exit? So
should I commit a version with
No, I missed that. Too many branches to this thread. ;-)
/* FIXME: This
Ian Lance Taylor wrote:
Georg-Johann Lay a...@gjlay.de writes:
So if a frontend can define address spaces and it is a generic feature, the
question is how to get the name of an address space in a generic, language
independent way.
We could decide that all frontends that use address spaces
I've observed failures on SH of the glibc test csu/tst-atomic where the
problem is that
#define __arch_compare_and_exchange_val_32_acq(mem, newval, oldval) \
({ __typeof (*(mem)) __result; \
__asm __volatile (\
.align 2\n\
mova 1f,r0\n\
nop\n\
mov r15,r1\n\
On Tue, Nov 29, 2011 at 5:44 AM, Diego Novillo dnovi...@google.com wrote:
Iain, could you let me know if the attached patch fixes your problem?
The patch changes gimple_call_set_cannot_inline to update the
corresponding callgraph edge, if needed. I did not touch any of the
other calls,
On 11/29/2011 08:09 AM, Richard Henderson wrote:
+static inline bool
+likely_compare_exchange (int *ptr, int *expected, int val, bool weak,
+ enum memmodel succ, enum memmodel fail)
{
+ return __builtin_expect (__atomic_compare_exchange_n (ptr, expected, val,
+
Hi Janne,
Regression-tested on x86_64-unknown-linux-gnu. OK for trunk?
Nice! Ok for trunk.
Sende ChangeLog
Sende module.c
Übertrage Daten ..
Revision 181810 übertragen.
Thanks for the review!
Thomas
On Tue, Nov 29, 2011 at 12:49, H.J. Lu hjl.to...@gmail.com wrote:
This caused:
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51346
Thanks. I'm on it.
Diego.
Hello!
We don't need scratch memory to convert value from memory.
2011-11-29 Uros Bizjak ubiz...@gmail.com
* config/i386/i386.md (*floathimode2_i387_with_temp): Do not
allocate scratch memory for alternative 0.
Tested on x86_64-pc-linux-gnu {,-m32} and committed to mainline
Richard Henderson wrote:
On 11/21/2011 11:31 AM, Georg-Johann Lay wrote:
;; The caveat is that if there are insns for some mode, there must also be a
;; respective move insn that describes reloads. Therefore, this
;; implementation uses an accumulator-based model with two hard-coded,
;;
Hello!
Attached patch splits movdi_via_fpu to loaddi_via_fpu and
storedi_via_fpu. As a consequence, independent insns can be scheduled
between load and store, and insn length is now calculated correctly.
2011-11-29 Uros Bizjak ubiz...@gmail.com
* config/i386/sync.md (UNSPEC_LDA,
On 11/29/2011 09:10 AM, Joseph S. Myers wrote:
So in my
view the best fix is to add a new constraint that means any
general-purpose register other than r15 and to use it in glibc when the
compiler version is new enough. That is what these GCC and glibc patches
do. Comments? OK?
I
On 11/29/2011 10:11 AM, Georg-Johann Lay wrote:
The only question that remains is what the -m64 option should be like?
[ ] Omit it altogether
[ ] Leave it as is (off per default)
[ ] Set it on per default
As soon as the direction is clear, I'll post a follow-up patch to add the
missing
Dear all,
gfortran 4.6 and 4.7 have a too tight check whether a variable can be
modified if the actual argument is INTENT(IN). This patch relaxes/fixes
the checking.
Build and regtested on x86-64-linux.
OK for the trunk - and for 4.6 (as it is a regression)?
Tobias
2011-11-29 Tobias Burnus
gfortran had an ICE when trying to insert a check whether the character
length between actual and dummy argument matches. This check is
pointless for deferred-length character arguments - thus, no bounds
check should be generated.
Build and regtested on x86-64-linux.
OK for the trunk?
Tobias
On 11/29/2011 07:08 AM, Andrew MacLeod wrote:
PR target/50123
* optabs.c (maybe_optimize_fetch_op): New. Look for more optimal
instructions for a FECTH_OP or OP_FECTH sequence.
(expand_atomic_fetch_op): Call maybe_optimize_fetch_op.
*
On 11/28/2011 12:28 PM, DJ Delorie wrote:
Ping? Anything else for this?
http://gcc.gnu.org/ml/gcc-patches/2011-11/msg02178.html
http://gcc.gnu.org/ml/gcc-patches/2011-11/msg01467.html
http://gcc.gnu.org/ml/gcc-patches/2011-11/msg01356.html
Not that I know of.
As far as I'm concerned you
Excellent. Thanks!
On 11/28/2011 08:49 PM, Hans-Peter Nilsson wrote:
On Sat, 26 Nov 2011, Richard Henderson wrote:
The m68k-linux failure for the various omp atomic tests
is due to the fact that BIGGEST_ALIGNMENT is 16 bits on
that platform. I think it's pretty reasonable to assume
that if something is aligned
Bernd Schmidt ber...@codesourcery.com writes:
The reason I'm suddenly reviewing the code now is that it
doesn't prevent shrink-wrapping, because nothing adds register 2
to the liveness info of the affected blocks. The temporary prologue
value of register 2 is then moved into register 15.
Rainer Orth r...@cebitec.uni-bielefeld.de writes:
Ian Lance Taylor i...@google.com writes:
This patch updates the implementations of locks and notes used in libgo
to use the current version from the master Go library. This now uses
futexes when running on GNU/Linux, while still using
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 11/24/11 08:11, Jakub Jelinek wrote:
Hi!
Since Richard's build_int_cst changes to make it effectively
build_int_cst_type when extract_range_from_assert wants to
subtract or add 1 to min or max of a signed 1-bit precision type,
On 28.11.11 20:45, Andreas Tobler wrote:
All,
I'd like to commit the attached patch to trunk and gcc-4.6.
It is a follow up from this one:
http://gcc.gnu.org/ml/gcc-cvs/2011-11/msg00886.html
I missed this one because in my setup I didn't build any shared
libraries from classpath. The
Richard Sandiford rdsandif...@googlemail.com writes:
Bernd Schmidt ber...@codesourcery.com writes:
The reason I'm suddenly reviewing the code now is that it
doesn't prevent shrink-wrapping, because nothing adds register 2
to the liveness info of the affected blocks. The temporary prologue
Hello,
the attached patch addresses PR 51337.
Tested against rev 181782 with:
make -k check RUNTESTFLAGS=--target_board=sh-sim
\{-m2/-ml,-m2/-mb,-m2a-single/-mb,-m4-single/-ml,-m4-single/-mb,-m4a-single/-ml,-m4a-single/-mb}
...with no new failures and 24 unexpected failure fixes for
-Original Message-
From: Richard Henderson
Sent: Tuesday, November 29, 2011 11:30 AM
To: Georg-Johann Lay
Cc: gcc-patches@gcc.gnu.org; Denis Chertykov; Weddington, Eric;
Anatoly
Sokolov
Subject: Re: [Patch,AVR] Light-weight DImode implementation.
On 11/29/2011 10:11 AM,
On Tue, 29 Nov 2011, Richard Henderson wrote:
On 11/28/2011 08:49 PM, Hans-Peter Nilsson wrote:
On Sat, 26 Nov 2011, Richard Henderson wrote:
The m68k-linux failure for the various omp atomic tests
is due to the fact that BIGGEST_ALIGNMENT is 16 bits on
that platform. I think it's pretty
Rainer Orth r...@cebitec.uni-bielefeld.de writes:
Rainer Orth r...@cebitec.uni-bielefeld.de writes:
This broke bootstrap on Linux/x86_64 (CentOS 5.5), which lacks
O_CLOEXEC.
... and also Solaris 8 and 9 bootstrap which lack sem_timedwait:
On 11/29/11 20:02, Richard Sandiford wrote:
Bernd Schmidt ber...@codesourcery.com writes:
The reason I'm suddenly reviewing the code now is that it
doesn't prevent shrink-wrapping, because nothing adds register 2
to the liveness info of the affected blocks. The temporary prologue
value of
Joseph S. Myers jos...@codesourcery.com wrote:
This asm doesn't tell the compiler that r15 is temporarily saved and
restored, and marking it as clobbering the stack pointer is unlikely to
work. It might be possible to avoid problems by marking all the inputs
that are used while the stack
Oleg Endo oleg.e...@t-online.de wrote:
2011-11-29 Oleg Endo oleg.e...@t-online.de
PR target/51337
* config/sh/sh.c (sh_secondary_reload): Add case when FPUL
register is being loaded from a pseudo in memory.
OK. Thakns for the patch!
I've applied your patch as revision
Richard Henderson r...@redhat.com wrote:
It does raise the question of whether it's worth while to move these
sequences inline in the compiler? We'd need to keep the symbols from
linux-atomic.S of course, but there's no reason we couldn't satisfy them with
now-trivial C code.
The same
This patch copies the runtime_panicstring function from the master Go
library, to simplify future merges. The runtime_panicstring function
replaces the __go_panic_msg function. Bootstrapped and ran Go testsuite
on x86_64-unknown-linux-gnu. Committed to mainline.
Ian
diff -r 3992ad49b04e
Hi Diego,
This is a backport for two upstream patches into our 4.6 branch.
I submitted the first patch by Julian a while ago for backport but
Richard Earnshaw pointed out a problem with the first patch. The second
patch from Joey fixes that problem. This was tested on x86 and ARM.
-Doug
Hi,
Martin, do you plan to have this pushed in for GCC 4.7?
well, there were two patches. I have managed to update and push
trough one of them in time
(http://gcc.gnu.org/ml/gcc-patches/2011-11/msg00086.html) but
unfortunately I have not managed to do the same with the second one.
It's
On Tue, Nov 29, 2011 at 08:09:31AM -0800, Richard Henderson wrote:
Tight loops like this should use the weak compare-exchange so that we don't
get two nested loops without need.
Done.
I do think we need more commentary here so that next month I can still follow
it.
Added this in
On Tue, Nov 29, 2011 at 08:37:07AM -0800, Richard Henderson wrote:
How about MEMMODEL_ACQ_REL
Right. Committed with that change revision 181832.
--
Alan Modra
Australia Development Lab, IBM
Sorry. Wrong patch. New patch attached.
I am getting the following error in the new patch though.
../../gimple-front-end/gcc/gimple/parser.c: In function ‘gp_parse_var_decl’:
../../gimple-front-end/gcc/gimple/parser.c:927:3: error: implicit
declaration of function
On Tue, Nov 29, 2011 at 04:15:36PM +1030, Alan Modra wrote:
On Mon, Nov 28, 2011 at 05:42:15PM -0800, Richard Henderson wrote:
On 11/28/2011 06:02 AM, Alan Modra wrote:
- unsigned int ret = bar-generation ~3;
- /* Do we need any barrier here or is __sync_add_and_fetch acting
-
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
On 11/29/11 15:03, Vladimir Makarov wrote:
This patch contains a part of my several last weeks work on LRA.
1. The major change is to add pseudo live range split
transformation to the existing inheritance transformations. The
idea is to create
2011/11/29 Georg-Johann Lay a...@gjlay.de:
For devices with 8-bit SP reading the high byte SP_H of SP will get garbage.
The patch uses CLR instead of IN SP_H to read the high part of SP.
There are two issues with this patch:
== 1 ==
I cannot really test it because for devices that small
2011/11/30 Weddington, Eric eric.wedding...@atmel.com:
-Original Message-
From: Richard Henderson
Sent: Tuesday, November 29, 2011 11:30 AM
To: Georg-Johann Lay
Cc: gcc-patches@gcc.gnu.org; Denis Chertykov; Weddington, Eric;
Anatoly
Sokolov
Subject: Re: [Patch,AVR] Light-weight
2011/11/29 Georg-Johann Lay a...@gjlay.de:
Ian Lance Taylor wrote:
Georg-Johann Lay a...@gjlay.de writes:
So if a frontend can define address spaces and it is a generic feature, the
question is how to get the name of an address space in a generic, language
independent way.
We could decide
On 11/29/2011 03:02 PM, Iain Sandoe wrote:
The following fails:
FAIL: g++.dg/eh/simd-3.C -std=gnu++98 execution test
FAIL: g++.dg/eh/simd-3.C -std=gnu++11 execution test
FAIL: g++.dg/eh/simd-4.C -std=gnu++98 execution test
FAIL: g++.dg/eh/simd-4.C -std=gnu++11 execution test
are the tip of an
75 matches
Mail list logo
