Hi maintainers,
This patch was approved in November 2011, but got missed out for commit so
reposting it.
Index: changes.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.6/changes.html,v
retrieving revision 1.137
diff -r1.137 changes
This patch to the Go frontend fixes the handling of pointer indirection
through a circular type. The gcc middle-end can not represent a
circular type, as such a construct can not exist in C (in Go you get one
by writing "type T *T"). We handle this by using void* and introducing
a type cast on th
Hi Maintainers,
This patch adds bdver2 description in changes.html.
Index: changes.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.7/changes.html,v
retrieving revision 1.90
diff -r1.90 changes.html
676a677,679
> Support for AMD
Hi Maintainers,
We want to add bdver2 description in invoke.texi for GCC 4.7. Here is the patch
that does that.
Index: gcc/doc/invoke.texi
===
--- gcc/doc/invoke.texi (revision 183891)
+++ gcc/doc/invoke.texi (working copy)
@@ -1306
This patch disables graphite loop flattening optimization due to it being
effectively unmaintained. To avoid users trying to use unsupported
optimization and reporting bugs, Sebastian Pop suggested to disable the
optimization.
Testing on x86_64 is underway. OK to apply provided no regressions
This patch to the Go frontend avoids some more crashes on invalid
input. Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu.
Committed to mainline.
Ian
2012-02-29 Ian Lance Taylor
* go-gcc.cc (class Gcc_tree): Add set_tree method.
(set_placeholder_pointer_type): W
Hi,
The attached patch is to avoid PR target/48596 which is a 4.7
regression on SH. It seems that now IRA aggressively tried to
use fp regs as the holder for memory addresses on this port.
SH requires a special fpul register to move from the fp regs to
the general regs which are legitimate for po
On 02/29/2012 08:47 PM, Jason Merrill wrote:
The substitution failure should have removed the candidate from further
consideration; sounds like something isn't propagating the failure back
up (as I was talking about in my response to your 51214 patch).
No, wait, now we're talking about the erro
On 02/29/2012 01:11 PM, Paolo Carlini wrote:
Thus I'm looking for some help about the best way to proceed. First, do
we agree that tsubst_copy_and_build should never call
unqualified_name_lookup_error unconditionally?
Yes.
Any tips about decltype32.C?
The substitution failure should have re
On 02/29/2012 06:53 AM, Paolo Carlini wrote:
Daniel noticed a couple other SFINAE issues, which I'm fixing in the
usual straightforward way. Tested x86_64-linux.
When fixing these things, please add checking for error_mark_node return
as well. For instance, in this patch the calls to decay_co
On Thu, Mar 1, 2012 at 3:20 AM, Greta Yorsh wrote:
> I'm attaching a new version of the patch. Fixed all comments and retested.
> No regression on qemu --with-cpu cortex-a9.
I assume that on the Cortex-A9 this generates a LDM instead of an
expensive LDRD. For reference, a tight load loop takes 2
The Go frontend mishandled unary ^ applied to a signed typed integer
constant. In fact it mishandled it so badly that it crashed. This
patch fixes the problem. Bootstrapped and ran Go testsuite on
x86_64-unknown-linux-gnu. Committed to mainline.
Ian
diff -r b66ddc182445 go/expressions.cc
---
Hi!
general_operand matches even CONST_INT, but none of the constraints would
allow it and we ICEd trying to reload it into a FLOAT_REGS register.
Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
approved by Uros in the PR, committed to trunk.
2012-02-29 Jakub Jelinek
This patch to the Go frontend avoids various crashes on erroneous
input. Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu.
Committed to mainline.
Ian
diff -r 3d9a01a3a527 go/expressions.cc
--- a/go/expressions.cc Wed Feb 29 14:41:03 2012 -0800
+++ b/go/expressions.cc Wed Feb 29 15:3
The Go compiler could crash when an invalid type occurred within an
invalid type, because the type verification pass was not recurring into
such types. This patch fixes the problem and actually removes more
lines than it adds. Bootstrapped and ran Go testsuite on
x86_64-unknown-linux-gnu. Commit
As discussed with DaveM, this duplicates libitm's cpu_relax into libgomp.
Tested on SPARC64/Linux, applied on the mainline.
2012-02-29 Eric Botcazou
* config/linux/sparc/futex.h (cpu_relax): Read from CC register.
--
Eric Botcazou
Index: config/linux/sparc/futex.h
This patch to libgo fills out the syscall package for GNU/Linux to match
all the functions in the syscall package in the master Go library.
There is a test case for this patch at
http://code.google.com/p/go/issues/detail?id=3071 . Bootstrapped and
ran Go testsuite on x86_64-unknown-linux-gnu. Com
Hello,
This fixes two files that still have the GPLv2 copyright notice.
OK for trunk?
Ciao!
Steven
* config/mn10300/mn10300-modes.def: Fix copyright notice.
* config/v850/v850-modes.def: Fix copyright notice.
Index: config/mn10300/mn10300-modes.def
===
Hi,
today Jon filed this PR about 'reporting routines re-entered' with
-fdump-tree-gimple and first I tried to figure where we are trying to
produce an error from inside the diagnostic code itself.
Turns out that tsubst_copy_and_build, case CALL_EXPR, calls
unqualified_name_lookup_error unco
On 29/02/12 14:48, Andrew Stubbs wrote:
> Hi all,
>
> This patch adds support for the DImode not operation in NEON.
>
> Currently the compiler must move the value to core registers, invert it
> there, and move it back again. This is bonkers because the VMVN
> instruction will do the job perfect
Hi!
This fixes misaligned store expansion if we aren't overwriting all bits
of the misaligned memory chunk to perform a RMW cycle. Preapproved
by richi, bootstrapped/regtested on x86_64-linux and i686-linux,
committed to trunk.
For C++11/C11 memory model we probably will need some further improve
On Mon, Feb 27, 2012 at 7:21 AM, Jonathan Wakely wrote:
> On 27 February 2012 15:18, Jakub Jelinek wrote:
>> Hi!
>>
>> Apparently after I've updated baseline_symbols.txt
>> files for a couple of targets Jonathan added some extra
>> exported symbols. This patch adds them.
>> Ok for trunk?
>
> OK -
On Wed, Feb 29, 2012 at 5:07 PM, Steve Kargl
wrote:
> On Wed, Feb 29, 2012 at 08:27:42AM +0100, Paul Richard Thomas wrote:
>> Regression fix okayed by Tobias Burnus on #gfortran and committed as
>> revision 184651.
>>
>> Cheers
>>
>> Paul
>>
>
> Thanks!
>
> Just a reminder. Did you get approval f
On Wed, Feb 29, 2012 at 08:27:42AM +0100, Paul Richard Thomas wrote:
> Regression fix okayed by Tobias Burnus on #gfortran and committed as
> revision 184651.
>
> Cheers
>
> Paul
>
Thanks!
Just a reminder. Did you get approval from the
Release Mananger? The last GCC 4.7.0 status
reported ind
On Wed, Feb 29, 2012 at 4:31 PM, Jakub Jelinek wrote:
> Hi!
>
> Since debug stmts can contain LABEL_DECLs we can ICE in parloops on them,
> fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux. Ok
> for trunk?
Ok.
Thanks,
Richard.
> 2012-02-29 Jakub Jelinek
>
> PR tree
On Wed, 29 Feb 2012, Michael Matz wrote:
> Hi,
>
> On Wed, 29 Feb 2012, Richard Guenther wrote:
>
> > the whole base object - instead I mark the base MEM_REF of such accesses
> > with TREE_NO_WARNING (ugh) and special-case for them in dr_may_alias_p.
>
> At least use a flag that isn't curren
Hi,
On Wed, 29 Feb 2012, Richard Guenther wrote:
> the whole base object - instead I mark the base MEM_REF of such accesses
> with TREE_NO_WARNING (ugh) and special-case for them in dr_may_alias_p.
At least use a flag that isn't currently used for MEM_REF. E.g.
addressable_flag, and make a
Hi!
Since debug stmts can contain LABEL_DECLs we can ICE in parloops on them,
fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux. Ok
for trunk?
2012-02-29 Jakub Jelinek
PR tree-optimization/52429
* tree-parloops.c (separate_decls_in_region_debug): Return earl
On Tue, 28 Feb 2012, Richard Guenther wrote:
>
> I am testing the following patch to fix PR52406. We cannot simply
> feed DR_BASE_OBJECT to the alias-oracle as it does not reflect
> a real memory access. Only if we query two 'structurally compatible'
> references we may do this. So the followi
Hi ,
This patch turns on hot cold partitioning for the ARM target. Given that
NOTE_INSN_SWITCH_SECTIONS gets created once - we should see
that this works.
Passed testing with cross on qemu - Simple handwritten tests for this
appear to work.
I'll commit this for 4.8 and would like some feedback f
On 02/28/2012 05:06 PM, Fabien Chêne wrote:
I agree, this is not efficient but I didn't find a better place.
perhaps in cp_parser_enumerator_list, that would require adding an
additional parameter to keep track of all the enum DECLs. Is it what
you have in mind ?
I was thinking of finish_enum_
Hi all,
This patch adds support for the DImode not operation in NEON.
Currently the compiler must move the value to core registers, invert it
there, and move it back again. This is bonkers because the VMVN
instruction will do the job perfectly.
The patch adds a pattern to support VMVN in DIm
On 24/02/12 01:03 , Sandeep Soni wrote:
+ name_token = gl_consume_expected_token (parser->lexer, CPP_NAME);
+ name = gl_token_as_text (name_token);
+
+ e = ggc_alloc_cleared_gimple_symtab_entry_def ();
+ e->id = get_identifier(name);
+ slot = htab_find_slot (gimple_symtab, e, NO_INSERT);
+
On 29 February 2012 14:20, Greta Yorsh wrote:
> I'm attaching a new version of the patch. Fixed all comments and retested.
> No regression on qemu --with-cpu cortex-a9.
OK by me but please give 24 hours for an RM to comment / object.
cheers
Ramana
I'm attaching a new version of the patch. Fixed all comments and retested.
No regression on qemu --with-cpu cortex-a9.
Thank you,
Greta
gcc/ChangeLog
2012-02-29 Greta Yorsh
* config/arm/arm-ldmstm.ml (write_ldm_commutative_peephole):
Improved
conditions of peepholes generati
On 28/02/12 17:45, Andrew Stubbs wrote:
This patch adds a DImode negate pattern for NEON.
Oops, that patch completely failed to allow for the fallback to
instructions that work in core registers. It also forgot to mention that
the CC register was clobbered.
This patch is the same except tha
On 29 February 2012 13:25, Ulrich Weigand wrote:
> Ramana Radhakrishnan wrote:
>> > ("SAT", "SATrev"): New code iterators.
>> > ("SATlo", "SAThi"): New code iterator attributes.
>> > ("*satsi_"): New pattern.
>> > ("*satsi__shift"): Likewise.
>>
>> Iterators must go into iterators.md.
Ramana Radhakrishnan wrote:
> >("SAT", "SATrev"): New code iterators.
> >("SATlo", "SAThi"): New code iterator attributes.
> >("*satsi_"): New pattern.
> >("*satsi__shift"): Likewise.
>
> Iterators must go into iterators.md.
Hmm, I left them in arm.md deliberately, because they ap
> ("SAT", "SATrev"): New code iterators.
> ("SATlo", "SAThi"): New code iterator attributes.
> ("*satsi_"): New pattern.
> ("*satsi__shift"): Likewise.
Iterators must go into iterators.md.
Ok with that change for 4.8 if no regressions.
Ramana
On 29 February 2012 13:06, Ulrich Weigand wrote:
> Mike Stump wrote:
>> On Feb 28, 2012, at 4:08 PM, Ramana Radhakrishnan wrote:
>> > Certainly no objections to it - in fact I'd suggest it for 4.6 as well.
>>
>> The backport is fine, if someone wants to do the work. Just let it bake a =
>> little
Mike Stump wrote:
> On Feb 28, 2012, at 4:08 PM, Ramana Radhakrishnan wrote:
> > Certainly no objections to it - in fact I'd suggest it for 4.6 as well.
>
> The backport is fine, if someone wants to do the work. Just let it bake a =
> little in the top of the tree as usual first.
I've already ch
Hello,
this patch adds support for generating usat/ssat instructions to match code
along the lines of:
if (a < amin) return amin;
else if (a > amax) return amax;
else return a;
for appropriate values of amin/amax. This type code actually occurs in
real-life code (e.g. c
On Wed, 29 Feb 2012, Eric Botcazou wrote:
> -feliminate-unused-debug-types will be enabled for Ada as well in GCC 4.7 and
> we have implemented a counter-measure to be able to retain apparently unused
> types in a few specific cases. But it was inadvertently broken some time ago.
>
> Fixed thu
Hi,
Daniel noticed a couple other SFINAE issues, which I'm fixing in the
usual straightforward way. Tested x86_64-linux.
Thanks,
Paolo.
PS: while working on the patch, I noticed that a few functions, which in
the meanwhile acquired a tsubst_flags_t parameter, aren't fully
propagating it, ar
2012/2/29 Georg-Johann Lay :
> This patch moves the bulk of TARGET_* macros to the end of avr.c and just
> prior
> to their usage by TARGET_INITIALIZER to initialize targetm.
>
> It's just more convenient to have that bulk of code at the end of the file...
>
> Ok to commit?
>
> Johann
>
>
>
2012/2/29 Georg-Johann Lay :
> This is more of code clean-up:
> avr_output_bld is unused and can be removed.
> To print a bit it is more convenient to use %T than patching a string.
>
> Ok for trunk?
>
> Johann
>
> * config/avr/avr-protos.h (avr_output_bld): Remove unused prototype.
>
-feliminate-unused-debug-types will be enabled for Ada as well in GCC 4.7 and
we have implemented a counter-measure to be able to retain apparently unused
types in a few specific cases. But it was inadvertently broken some time ago.
Fixed thusly, tested on i586-suse-linux, applied on the mainli
This is more of code clean-up:
avr_output_bld is unused and can be removed.
To print a bit it is more convenient to use %T than patching a string.
Ok for trunk?
Johann
* config/avr/avr-protos.h (avr_output_bld): Remove unused prototype.
* config/avr/avr.c (avr_output_bld): Remove
On Wed, 29 Feb 2012, Rainer Orth wrote:
> Richard Guenther writes:
>
> >> Moving them to libitm/testsuite is an option, but that might be more
> >> difficult for the LTO case. I thought about adding
> >> dg-require-effective-target tm_runtime (to check for working
> >> -litm/libitm.spec) and dg
On Wed, Feb 29, 2012 at 11:26:54AM +0100, Eric Botcazou wrote:
> > It's still used by the working trans-mem-3_[01].c testcase.
>
> That test is a little puzzling:
>
> /* Test that we can build one object file with -fgnu-tm
>(trans-mem-3_1.c), but do the final link of all objects without
>
> It's still used by the working trans-mem-3_[01].c testcase.
That test is a little puzzling:
/* Test that we can build one object file with -fgnu-tm
(trans-mem-3_1.c), but do the final link of all objects without
-fgnu-tm. */
So, in the end, we cannot do that without the trans-mem.h klud
Richard Guenther writes:
>> Moving them to libitm/testsuite is an option, but that might be more
>> difficult for the LTO case. I thought about adding
>> dg-require-effective-target tm_runtime (to check for working
>> -litm/libitm.spec) and dg-add-options tm_runtime to add the necessary
>> -B/-L
On Wed, 29 Feb 2012, Rainer Orth wrote:
> Richard Guenther writes:
>
> > This fixes the LTO TM testcases by removing them. TM testcases
> > belong in libitm testsuite (but dg-additional-sources does not
> > work there currently for some weird reasons, so I gave up
> > trying).
>
> Moving them
On Wed, 29 Feb 2012, Eric Botcazou wrote:
> > 2012-02-29 Richard Guenther
> >
> > PR testsuite/52297
> > * gcc.dg/lto/trans-mem-1_0.c: Remove.
> > * gcc.dg/lto/trans-mem-1_1.c: Likewise.
> > * gcc.dg/lto/trans-mem-2_0.c: Likewise.
> > * gcc.dg/lto/trans-mem-2_1.c: Likewise.
This patch moves the bulk of TARGET_* macros to the end of avr.c and just prior
to their usage by TARGET_INITIALIZER to initialize targetm.
It's just more convenient to have that bulk of code at the end of the file...
Ok to commit?
Johann
* config/avr/avr.c: Move definition of TARGET m
On 02/28/2012 05:10 PM, Joseph S. Myers wrote:
> On Tue, 28 Feb 2012, Richard Earnshaw wrote:
>
>> Here's an updated patch and a suggested web-page patch:
>
> uClinux targets do not have -gnu in their target triplets.
>
> You missed the WinCE port from the list in the webpage patch of ports
> b
2012/2/29 Georg-Johann Lay :
> The condition/branch in decrement-and-branch peepholes is ne/brcc.
>
> This can be generalized to cover eq/brcs.
>
> Tested without regressions.
>
> Because of PR52417, I had to run the tests with -fno-dse. There is one
> UNSUPPORTED because of that (missing dse dump)
Richard Guenther writes:
> This fixes the LTO TM testcases by removing them. TM testcases
> belong in libitm testsuite (but dg-additional-sources does not
> work there currently for some weird reasons, so I gave up
> trying).
Moving them to libitm/testsuite is an option, but that might be more
> 2012-02-29 Richard Guenther
>
> PR testsuite/52297
> * gcc.dg/lto/trans-mem-1_0.c: Remove.
> * gcc.dg/lto/trans-mem-1_1.c: Likewise.
> * gcc.dg/lto/trans-mem-2_0.c: Likewise.
> * gcc.dg/lto/trans-mem-2_1.c: Likewise.
> * gcc.dg/lto/trans-mem-4_0.c: Likewise.
This fixes the LTO TM testcases by removing them. TM testcases
belong in libitm testsuite (but dg-additional-sources does not
work there currently for some weird reasons, so I gave up
trying).
Committed.
Richard.
2012-02-29 Richard Guenther
PR testsuite/52297
* gcc.dg/lto/t
The condition/branch in decrement-and-branch peepholes is ne/brcc.
This can be generalized to cover eq/brcs.
Tested without regressions.
Because of PR52417, I had to run the tests with -fno-dse. There is one
UNSUPPORTED because of that (missing dse dump).
Ok for the trunk?
Johann
* co
On 28/02/12 23:42, John Tytgat wrote:
> In message <4f4d0e64.1020...@arm.com>
> Richard Earnshaw wrote:
>
>> On 28/02/12 17:10, Joseph S. Myers wrote:
>>> On Tue, 28 Feb 2012, Richard Earnshaw wrote:
>>>
Here's an updated patch and a suggested web-page patch:
>>>
>>> uClinux target
On Wed, Feb 29, 2012 at 4:12 AM, William J. Schmidt
wrote:
> This patch fixes PR52424 by pushing a missing marker onto
> const_and_copies_stack before calling dom_thread_across_edge. This bug
> has resulted in lost copy propagation opportunities in the presence of
> edge threading.
>
> Bootstrapp
On Tue, Feb 28, 2012 at 9:11 PM, Aldy Hernandez wrote:
> On 02/28/12 13:12, Richard Henderson wrote:
>>
>> On 02/28/12 09:44, Aldy Hernandez wrote:
>>>
>>> PR middle-end/51752
>>> * gimple.h (gimple_in_transaction): New.
>>> (gimple_set_in_transaction): New.
>>> (struct
64 matches
Mail list logo
