All remarks from
https://gcc.gnu.org/ml/gcc-patches/2014-06/msg02166.html are fixed.
Updated version of liboffloadmic you can find in GIT branch
kyukhin/gomp4-offload:
https://gcc.gnu.org/git/?p=gcc.git;a=shortlog;h=refs/heads/kyukhin/gomp4-offload.
Commit: HEAD~3 ("Add liboffloadmic with offloadin
On Sat, Aug 9, 2014 at 12:00 PM, Matthew Fortune
wrote:
> Moore, Catherine writes:
>> > -Original Message-
>> > From: Steve Ellcey [mailto:sell...@mips.com]
>> > Sent: Friday, August 08, 2014 3:42 PM
>> > To: Moore, Catherine; matthew.fort...@imgtec.com; echri...@gmail.com;
>> >
>> > 2014
>>
>> -#ifdef __mips_micromips
>> +#if defined(__mips_micromips) || defined(__mips_soft_float)
>>/* DO NOTHING */
>> #else
>>
Mind adding a comment here explaining why we don't want to do anything
for soft float (and micromips)?
OK otherwise.
Thanks!
-eric
> I think the best place is below this:
>
> #if defined (__vxworks) && ! defined (__RTP__)
> #include
> #else
> #include
> #endif
>
> And just move the conditional from socket.c with comment.
Yes, that looks right to me.
Arno
This is a fix for off-by-one bug that I've recently introduced in
r213806. Bootstrapped, regtested and Asan-boostrapped on x64.
Ok to commit?
-Y
commit cc88579f072240686605783229129700f9eb78a0
Author: Yury Gribov
Date: Mon Aug 11 19:30:24 2014 +0400
2014-08-12 Yury Gribov
gc
On Tue, Aug 12, 2014 at 11:17:48AM +0400, Yury Gribov wrote:
> This is a fix for off-by-one bug that I've recently introduced in r213806.
> Bootstrapped, regtested and Asan-boostrapped on x64.
>
> Ok to commit?
Sure.
> 2014-08-12 Yury Gribov
>
> gcc/
> * internal-fn.c (init
On August 12, 2014 8:15:34 AM CEST, Roman Gareev wrote:
>> Checking for this specific AST may cause failures with future
>versions of
>> isl that choose a different schedule. Could you write a regular
>expression
>> that checks that there is no if-condition contained in a for loop? I
>think
>> thi
On 08/12/2014 11:20 AM, Jakub Jelinek wrote:
>> This is a fix for off-by-one bug that I've recently introduced in
r213806.
>> Bootstrapped, regtested and Asan-boostrapped on x64.
>>
>> Ok to commit?
>
> Sure.
Done, r213848.
-Y
We shouldn't allow friend declarations and member templates to
have virt-specifiers, as neither of those can be virtual. Tested on
Linux x86-64.
Note that the standardese can be squinted at in a way that would allow
virt-specifiers on friends, there's a DR coming for that. I do not expect
the resol
Jakub Jelinek wrote:
> On Mon, Aug 11, 2014 at 05:04:20PM +0100, Gary Benson wrote:
> > + case 's':
> > + seed = atoi (optarg);
> > + break;
> > +
> > + case 't':
> > + timeout = atoi (optarg);
> > + break;
> > +
> > + case 'm':
> > + maxcount = atoi (optarg);
> > + br
On Tue, Aug 12, 2014 at 10:02:40AM +0100, Gary Benson wrote:
> I've removed the timeout code. Users can limit the run by setting a
> maximum number of iterations. That's more consistent for testing
> anyway: 500 iterations is 500 iterations wherever you run it.
>
> How about this one?
L
Eric Christopher writes:
> On Sat, Aug 9, 2014 at 12:00 PM, Matthew Fortune
> wrote:
> > Moore, Catherine writes:
> >> > -Original Message-
> >> > From: Steve Ellcey [mailto:sell...@mips.com]
> >> > Sent: Friday, August 08, 2014 3:42 PM
> >> > To: Moore, Catherine; matthew.fort...@imgtec
Marek Polacek a écrit:
> Thise testcases use predefined identifiers, and without the
> dg-options, they would compile with -ansi -pedantic-errors and fail.
> Setting dg-options to "" makes the -ansi -pedantic-errors go away.
> Setting it to e.g. -std=gnu99 would work as well.
Oh, I see. Thanks.
On 05/08/14 10:31, Zhenqiang Chen wrote:
> Hi,
>
> For some large constants, ARM will split them during expanding, which
> makes impossible to hoist them out the loop or shared by different
> references (refer the test case in the patch).
>
> The patch keeps some constants in registers. If the co
Hi,
As mentioned in PR60070, there is many cases when a programmer want to ensure
that a program does not use any floating point data type. Other cases to
consider
is when a target has several floating point ABI and user want to ensure his/her
is compatible with all available floating point ABI.
Jakub Jelinek wrote:
> On Tue, Aug 12, 2014 at 10:02:40AM +0100, Gary Benson wrote:
> > I've removed the timeout code. Users can limit the run by setting a
> > maximum number of iterations. That's more consistent for testing
> > anyway: 500 iterations is 500 iterations wherever you run it
On Tue, Aug 12, 2014 at 05:25:25PM +0800, Thomas Preud'homme wrote:
> --- a/gcc/c/c-parser.c
> +++ b/gcc/c/c-parser.c
> @@ -486,6 +486,15 @@ c_token_starts_typename (c_token *token)
> case CPP_KEYWORD:
>switch (token->keyword)
> {
> + case RID_FLOAT:
> + case RID_DOUBLE:
Hi,
On 08/11/2014 08:22 PM, Jason Merrill wrote:
OK with a comment explaining how we can get there.
Thanks. I'm going to apply the below, then. I suppose the simple
parser->scope clean up (
https://gcc.gnu.org/ml/gcc-patches/2014-08/msg01000.html) is also Ok, I
would rather commit both togeth
On Tue, Aug 12, 2014 at 11:34:35AM +0200, Jakub Jelinek wrote:
> On Tue, Aug 12, 2014 at 05:25:25PM +0800, Thomas Preud'homme wrote:
> > --- a/gcc/c/c-parser.c
> > +++ b/gcc/c/c-parser.c
> > @@ -486,6 +486,15 @@ c_token_starts_typename (c_token *token)
> > case CPP_KEYWORD:
> >switch (
On Tue, Aug 12, 2014 at 05:25:25PM +0800, Thomas Preud'homme wrote:
> @@ -10078,6 +10078,10 @@ finish_declspecs (struct c_declspecs *specs)
> }
>else if (specs->complex_p)
> {
> + if (flag_no_float)
> + error_at (specs->locations[cdw_complex],
> +
On Tue, 12 Aug 2014, Thomas Preud'homme wrote:
As mentioned in PR60070, there is many cases when a programmer want to
ensure that a program does not use any floating point data type. Other
cases to consider is when a target has several floating point ABI and
user want to ensure his/her is compat
This then is the version with both issues fixed (and new test).
Bootstrapped/regtested on x86_64-linux, ok for trunk?
2014-08-12 Marek Polacek
gcc/c-family/
* c-opts.c (sanitize_cpp_opts): Make warn_long_long be set according
to warn_c90_c99_compat.
* c.opt (Wc90-c99-c
And now the version with two bugs mention elsewhere fixed as well.
Bootstrapped/regtested on x86_64-linux, ok for trunk?
2014-08-12 Marek Polacek
* doc/invoke.texi: Document -Wc99-c11-compat.
c-family/
* c.opt (Wc99-c11-compat): New option.
c/
* c-decl.c (diagnose_mism
On Tue, Aug 12, 2014 at 11:47:28AM +0200, Marc Glisse wrote:
> Are you sure you want something that strict? If there are floats in the
> code but the compiler manages to get rid of them, it won't cause any
> trouble with the ABI. Besides, if you #include a header that declares a
> function returnin
Following patch replaces the current "temporary inline assembler" implementation
of vget_high. So this patch adds a test first. We don't have any test coverage
of vget_low, either, so add that too.
Passing on aarch64-none-elf and aarch64_be-none-elf.
diff --git a/gcc/testsuite/gcc.target/aarch6
This patch replaces the current inline assembler for the vget_high intrinsics in
arm_neon.h with a sequence of other calls, in a similar fashion to vget_low.
Unlike the assembler, these are all transparent to the front-end, so should
enable better optimization through the mid-end.
Tested check
The following patch forces you to write convert?, convert1? or
convert2? to get a conditional convert.
Committed.
Richard.
2014-08-12 Richard Biener
* genmatch.c (enum tree_code): Add CONVERT0.
(main): Likewise.
(parse_operation): Handle conditional converts speciall
> From: Marc Glisse [mailto:marc.gli...@inria.fr]
> Sent: Tuesday, August 12, 2014 5:47 PM
>
>
> Are you sure you want something that strict? If there are floats in the
> code but the compiler manages to get rid of them, it won't cause any
> trouble with the ABI. Besides, if you #include a header
Hello,
Attached patch allows wider mask types.
Is it ok for trunk?
Bootstrapped.
gcc/
* config/i386/i386.c (print_reg): Сorrectly print 64-bit mask
registers.
(inline_secondary_memory_needed): Allow 64 bit wide mask registers.
(ix86_hard_regno_mode_ok): Allow 32/6
On Tue, Aug 12, 2014 at 12:15:20PM +0200, Richard Biener wrote:
> @@ -2422,6 +2441,7 @@ main(int argc, char **argv)
>add_operator (SYM, # SYM, # TYPE, NARGS);
> #define END_OF_BASE_TREE_CODES
> #include "tree.def"
> +add_operator (CONVERT1, "CONVERT0", "tcc_unary", 1);
Pasto? Shouldn't that
Hello,
Patch allows wider vec int modes.
Bootstrapped.
Is it ok for trunk?
gcc/
* config/i386/i386.c (ix86_preferred_simd_mode): Allow V64QImode and
V32HImode.
--
Thanks, K
diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 5a3b67a..183b7be 100644
--- a/gcc/conf
On Tue, 12 Aug 2014, Jakub Jelinek wrote:
> On Tue, Aug 12, 2014 at 12:15:20PM +0200, Richard Biener wrote:
> > @@ -2422,6 +2441,7 @@ main(int argc, char **argv)
> >add_operator (SYM, # SYM, # TYPE, NARGS);
> > #define END_OF_BASE_TREE_CODES
> > #include "tree.def"
> > +add_operator (CONVERT
On Tue, Aug 12, 2014 at 12:58 AM, Prathamesh Kulkarni
wrote:
> Add pattern: abs (abs (x)) -> abs (x)
>
> * match.pd: Add new pattern.
Thanks - applied.
Richard.
> [gcc/testsuite/gcc.dg/tree-ssa]
> * match.c: New test-case.
>
> Thanks,
> Prathamesh
Hello,
This patch extends support of masking instructions.
Bootstrapped.
Is it ok for trunk?
gcc/
* config/i386/i386.md
(define_attr "isa"): Add avx512bw,noavx512bw.
(define_attr "enabled"): Ditto.
(define_split): Add 32/64-bit mask logic.
(define_insn "*k
On Tue, Aug 12, 2014 at 4:51 AM, Kugan
wrote:
>
> On 11/08/14 18:03, Richard Biener wrote:
>> On Sat, Aug 9, 2014 at 2:33 PM, Kugan
>> wrote:
>>> Hi,
>>>
>>> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52904
>>>
>>> Tescase was generating warning: assuming signed overflow does not occur
>>> whe
While looking at another failure, I noticed that COMDAT group support on
Solaris with Sun as has been accidentally disabled for quite some time
by this patch:
https://gcc.gnu.org/ml/gcc-patches/2012-08/msg01847.html
Unfortunately, the fix was anything but obvious: by setting
gcc_cv_as_com
Once COMDAT group support on Solaris 11+ with Sun as got re-enabled
https://gcc.gnu.org/ml/gcc-patches/2014-08/msg01120.html
g++.dg/debug/dwarf2/nested-4.C is FAILing in that configuration:
FAIL: g++.dg/debug/dwarf2/nested-4.C -std=gnu++98 scan-assembler-times
debug_types 2
FAIL: g++.
>>> Did you verify the testcase fails before the revision that fixed it?
>>> Esp. the placement of the dg-bogus looks bogus to me.
>>
>> I tried it on Linaro 4.9 (It should be the same in fsf gcc 4.9 branch)
>> and the test cases is failing there. Passes on trunk.
>
> Well, it probably fails becau
Hello,
This patch extends substs/subst_attrs to be used
with new patterns.
Bootstrapped.
Is it ok for trunk?
gcc/
* config/i386/sse.md: Allow V64QI, V32QI, V32HI, V4HI modes.
* config/i386/subst.md
(define_mode_iterator SUBST_V): Update.
(define_mode_iterator SUBS
On Tue, Aug 12, 2014 at 2:02 AM, Gary Benson wrote:
> +#include
Include demangle.h with "".
> +int
> +main (int argc, char *argv[])
> +{
> + char symbol[2 + MAXLEN + 1] = "_Z";
> + int seed = -1, seed_set = 0;
> + int count = 0, maxcount = -1;
I think that by default the program should sto
This didn't get any comments earlier. Is it OK to comment?
Somewhere along the way, the Ada run-time Makefile's quit
honoring CFLAGS_FOR_TARGET. This just adds it back.
2014-08-12 Joel Sherrill
* Makefile.in: Add CFLAGS_FOR_TARGET to GNATLIBCFLAGS_FOR_C.
---
libada/Makefile.in | 2 +-
Hopefully this addresses the comments.
OK to comment?
2014-08-12 Joel Sherrill
* socket.c: For RTEMS, use correct prototype of gethostbyname_r().
* gsocket.h Add include of on RTEMS.
---
gcc/ada/gsocket.h | 5 +
gcc/ada/socket.c | 2 +-
2 files changed, 6 insertions(+),
Hi,
This fixes the issue in PR target/62098 where the pattern is clearly
bogus and it's not clear what I was thinking when I ok'd it.
The vcvt instruction in this form writes to it's source register and
this should have been modelled correctly in the backend in this form
rather than the bog
(It is no more expensive.)
gcc/ChangeLog:
* config/aarch64/aarch64.md (subdi3, adddi3_aarch64): Don't penalize
SIMD reg variant.diff --git a/gcc/config/aarch64/aarch64.md b/gcc/config/aarch64/aarch64.md
index f8eb305140e7b0aed006b33f1724a90939e48316..0a8ca4bcc7941f912c8d42200b332
This patch adds SIMD register variants for and, ior, xor and not - similarly to
add/sub, the H/W supports it, and it'll be more efficient if the values are
there already, e.g. if passed as [u]int64x1_t parameters.
gcc/ChangeLog:
* config/aarch64/aarch64.md (3, one_cmpl2):
Add S
Following the change to make arm_neon.h's (u?)int64x1_t types into vectors,
these types are now passed in the SIMD registers rather than general purpose
registers, which often results in poor quality code in functions taking or
returning these types. Often values are moved from vector registers
[When I wrote that xor was broken on GPRs and this fixes it, I meant
xor_one_cmpl rather than xor, sorry!]
The pattern for xor_one_cmpl never matched, due to the action of
combine_simplify_rtx; hence, separate this pattern out from that for ORN/BIC.
ORN/BIC have equivalent SIMD-reg variants,
On Tue, 12 Aug 2014, Thomas Preud'homme wrote:
> 3) __builtin_sqrt
>
> True, I shall try if it works with builtins. Thanks for the advice.
In addition to built-in functions, you have built-in typedefs. I don't
think your code would catch uses of __fp16, __float128 etc. (such types
are defined
...patch attached...
Alan Lawrence wrote:
[When I wrote that xor was broken on GPRs and this fixes it, I meant
xor_one_cmpl rather than xor, sorry!]
The pattern for xor_one_cmpl never matched, due to the action of
combine_simplify_rtx; hence, separate this pattern out from that for ORN/BIC.
On Tue, 12 Aug 2014, Thomas Preud'homme wrote:
> However it's true that it prevents including math.h for instance.
And stdlib.h (strtod) and time.h (difftime) and wchar.h (wcstod). Is that
your intent? Note that glibc stdlib.h has an atof inline definition, it's
not just a matter of prototype
Hi,
The ACLE
http://infocenter.arm.com/help/topic/com.arm.doc.ihi0053c/IHI0053C_acle_2_0.pdf
expects this macro to be defined in terms of byte sizes but we return
bit sizes *and* additionally do not handle -fshort-wchar.
Applied to trunk.
regards
Ramana
2014-08-12 Ramana Radhakrishnan
On Mon, Aug 11, 2014 at 01:25:37PM -0400, Jason Merrill wrote:
> >+convert_like_real (location_t loc, conversion *convs, tree expr, tree fn,
> >+ int argnum, int inner, bool issue_conversion_warnings,
> >bool c_cast_p, tsubst_flags_t complain)
> > {
> >tree totype
We're missing these overloads required by C++11.
Tested x86_64-linux, committed to trunk.
I think this should go on the branches too, any objections?
commit c904682994e3c1d77abe16a4a7366e191b6ef6bc
Author: Jonathan Wakely
Date: Tue Aug 12 15:24:35 2014 +0100
* include/bits/basic_string
Ping. Rebased patch to current gcc master.
gcc/ChangeLog
PR debug/59051
* dwarf2out.h (modified_type_die): Handle TYPE_QUAL_RESTRICT.
gcc/testsuite/ChangeLog
PR debug/59051
* gcc.dg/guality/restrict.c: New test.
---
gcc/ChangeLog | 5 +
Ping. Rebased patch to current gcc master.
modified_type_die and add_type_attribute take two separate arguments
for whether the type should be const and/or volatile. To help add
more type modifiers pass the requested modifiers as one cv_quals argument
to these functions. And introduce helper funct
Hi,
this patch adds warning on mismatching virtual tables. It is update/rewrite of
my earlier version https://gcc.gnu.org/ml/gcc-patches/2014-02/msg00741.html
The main difference is that we no longer stream all variable initializers, so
the new code uses ipa-reference instead of walk of the DECL_IN
On Tue, 2014-08-12 at 00:07 -0700, Eric Christopher wrote:
> >>
> >> -#ifdef __mips_micromips
> >> +#if defined(__mips_micromips) || defined(__mips_soft_float)
> >>/* DO NOTHING */
> >> #else
> >>
>
> Mind adding a comment here explaining why we don't want to do anything
> for soft float (and
Ping. This patch is based on the proposed patches to support the
restrict type. But isn't meant to be integrated at this time. DWARFv5
is still only a collection of proposals and hasn't seen any draft
yet. But feedback is highly appreciated. Also if there are tips on how
(and where) to maintain an
Ping. As with the patch for DW_TAG_atomic_type this is only experimental
at this time, but feedback appreciated.
There is a corresponding patch for GDB:
https://sourceware.org/ml/gdb-patches/2014-07/msg00199.html
gcc/ChangeLog
* dwarf2out.c (decl_user_align): New function.
(type_
On 12/08/14 16:20 +0100, Jonathan Wakely wrote:
We're missing these overloads required by C++11.
Tested x86_64-linux, committed to trunk.
I missed an 'inline' on one of the new overloads, fixed by this patch,
which I'll commit once testing finishes.
commit 64240d493d1d72f78538d8bb06445d3ccb7a
> On Aug 12, 2014, at 7:40 AM, Alan Lawrence wrote:
>
> (It is no more expensive.)
Yes on some processors it could be.
Thanks,
Andrew
>
> gcc/ChangeLog:
>
>* config/aarch64/aarch64.md (subdi3, adddi3_aarch64): Don't penalize
>SIMD reg variant.
> diff --git a/gcc/config/aarch64/aa
Ping.
-Maxim
Original Message
Subject:[PATCH] Add patch for debugging compiler ICEs.
Date: Mon, 04 Aug 2014 21:03:22 +0400
From: Maxim Ostapenko
To: GCC Patches
CC: Jeff Law , Jakub Jelinek ,
tsaund...@mozilla.com, Yury Gribov , Slava
Garbuzov
On 08/04/2
On 08/12/2014 11:09 AM, Marek Polacek wrote:
One possibility (and I think you mentioned this at Cauldron) would be to
create a new tree code, called e.g. LOCATION_EXPR.
I don't think it needs a new code: I think we can use VIEW_CONVERT_EXPR
for decls and NOP_EXPR for constants. There will pro
Hi,
On 08/12/2014 05:09 PM, Marek Polacek wrote:
Furthermore, the arguments should have their own locations in EXPR_LOCATION;
I don't think we need to track them separately. If their locations are
wrong that's something to fix directly, rather than adding a special case
for function arguments.
On Mon, Aug 11, 2014 at 1:44 PM, Yi Yang wrote:
> I ported this to trunk.
>
> Shall I commit this patch to the Google 4_8/4_9 branches first?
Ok with me.
Teresa
>
> On Mon, Aug 11, 2014 at 12:46 PM, Teresa Johnson wrote:
>> Ok, thanks. This seems reasonable. Can you send the patch to trunk as w
On 12 August 2014 11:47, Ville Voutilainen wrote:
> We shouldn't allow friend declarations and member templates to
> have virt-specifiers, as neither of those can be virtual. Tested on
> Linux x86-64.
> Note that the standardese can be squinted at in a way that would allow
> virt-specifiers on fri
Hi Richard,
I find that you missed the testcase with when committing the patch.
GCC 4.8 & 4.9 branch also has the same issue, may be we need to
fix for them too?
Cheers,
Felix
On Mon, Aug 11, 2014 at 7:24 PM, Richard Biener
wrote:
> On Sat, Aug 9, 2014 at 6:28 AM, Felix Yang wrote:
>>
On Tue, 12 Aug 2014, Jason Merrill wrote:
> On 08/12/2014 11:09 AM, Marek Polacek wrote:
> > One possibility (and I think you mentioned this at Cauldron) would be to
> > create a new tree code, called e.g. LOCATION_EXPR.
>
> I don't think it needs a new code: I think we can use VIEW_CONVERT_EXPR
On 08/12/2014 12:48 PM, Joseph S. Myers wrote:
For C, it is definitely expected within the front end that anything that
is an integer constant expression is an INTEGER_CST and that a NOP_EXPR
wrapping an INTEGER_CST is something folded to an integer constant that is
neither an integer constant ex
On 08/12/2014 12:22 PM, Paolo Carlini wrote:
why wrapping ("somebody" also suggested a
NOP_EXPR at some point ;) instead of simply changing integer_cst & co to
natively have a location?!?
That also seems like a possibility, though I don't know to what extent
different parts of the compiler rel
Applied, thanks.
Jason
> On Tue, 12 Aug 2014, Jason Merrill wrote:
>
>> On 08/12/2014 11:09 AM, Marek Polacek wrote:
>> > One possibility (and I think you mentioned this at Cauldron) would be to
>> > create a new tree code, called e.g. LOCATION_EXPR.
>>
>> I don't think it needs a new code: I think we can use VIEW_CONVER
Ian Lance Taylor wrote:
> On Tue, Aug 12, 2014 at 2:02 AM, Gary Benson wrote:
> > +#include
>
> Include demangle.h with "".
Ok.
> > +int
> > +main (int argc, char *argv[])
> > +{
> > + char symbol[2 + MAXLEN + 1] = "_Z";
> > + int seed = -1, seed_set = 0;
> > + int count = 0, maxcount = -1;
On Tue, Aug 12, 2014 at 10:11 AM, Gary Benson wrote:
> Ian Lance Taylor wrote:
>>
>> I think that by default the program should stop. That will make it
>> possible to eventually run as part of "make check". Give it some
>> number of iterations that stops it in a second or so. You can still
>> h
Hello
The patch fixes some typos found by Lintian, the Debian static analyzer.
(it is mainly the opportunity to test my commit permissions).
Thanks,
Sylvestre
libstdc++-v3/ChangeLog:
2014-08-12 Sylvestre Ledru
* include/profile/impl/profiler_hash_func.h: Fix a typo
gcc/fortran/Cha
> Hopefully this addresses the comments.
>
> OK to comment?
>
> 2014-08-12 Joel Sherrill
>
> * socket.c: For RTEMS, use correct prototype of gethostbyname_r().
> * gsocket.h Add include of on RTEMS.
OK
> This didn't get any comments earlier. Is it OK to comment?
This one is a bit tricky so requires more thoughts. On which targets did
you test it?
> Somewhere along the way, the Ada run-time Makefile's quit
> honoring CFLAGS_FOR_TARGET. This just adds it back.
>
> 2014-08-12 Joel Sherrill
>
>
> gcc/ada/ChangeLog:
>
> 2014-08-12 Sylvestre Ledru
>
> * 9drpc.adb: Fix a typo
> * s-interr.ads: Likewise
> * s-taskin.ads: Likewise
> * s-traces.ads: Likewise
> * sysdep.c: Likewise
The Ada part is OK.
On Tue, 12 Aug 2014, Sylvestre Ledru wrote:
The patch fixes some typos found by Lintian, the Debian static analyzer.
Did you check them yourself? We end up with "these information" which
doesn't seem correct to me.
--
Marc Glisse
After talking to Jeff Law at the GCC Cauldron I have updated my switch
shortcut plugin pass to try and address this optimization in the hopes of
getting it added to GCC as a static pass. I fixed the code to build with
the various C++ changes that have been happening in GCC but the current
version
On Sun, 10 Aug 2014, Marek Polacek wrote:
> We weren't properly diagnosing neither the __func__ (introduced in C99),
> nor the __FUNCTION__/__PRETTY_FUNCTION__ (GNU extension) predefined
> identifiers. I believe we should; the compiler ought to have a
> compile-time switch for turning off extensi
On Mon, 11 Aug 2014, Sylvestre Ledru wrote:
> > The test Wmissing-return2.c only has one of the two warnings. But as per
> > "-Wreturn-type => Run both", and for backwards compatibility with the
> > existing definition of -Wreturn-type, both warnings should appear for this
> > test.
> Make s
On 08/12/2014 01:10 PM, Manuel López-Ibáñez wrote:
Wouldn't the folding that already occurs in the FEs prevent this anyway?
Currently, probably. But we're planning to delay that folding for other
reasons.
In any case, we don't need to fix location info for constants/decls in
order to fix l
On 12/08/2014 19:48, Joseph S. Myers wrote:
> On Mon, 11 Aug 2014, Sylvestre Ledru wrote:
>
>>> The test Wmissing-return2.c only has one of the two warnings. But as per
>>> "-Wreturn-type => Run both", and for backwards compatibility with the
>>> existing definition of -Wreturn-type, both warnin
On Tue, Aug 12, 2014 at 12:45 PM, Kirill Yukhin wrote:
> Hello,
> Attached patch allows wider mask types.
>
> Is it ok for trunk?
>
> Bootstrapped.
>
> gcc/
> * config/i386/i386.c (print_reg): Сorrectly print 64-bit mask
> registers.
> (inline_secondary_memory_needed): Allo
On Tue, Aug 12, 2014 at 12:49 PM, Kirill Yukhin wrote:
> Hello,
> Patch allows wider vec int modes.
>
> Bootstrapped.
>
> Is it ok for trunk?
>
> gcc/
> * config/i386/i386.c (ix86_preferred_simd_mode): Allow V64QImode and
> V32HImode.
OK.
Thanks,
Uros.
On Tue, Aug 12, 2014 at 10:46:46AM -0700, Steve Ellcey wrote:
> --- /dev/null
> +++ b/gcc/tree-switch-shortcut.c
> +/* This file implements an optimization where, when a variable is set
> + to a constant value and there is a path that leads from this definition
> + to a switch statement that us
Thanks!
-eric
On Tue, Aug 12, 2014 at 8:33 AM, Steve Ellcey wrote:
> On Tue, 2014-08-12 at 00:07 -0700, Eric Christopher wrote:
>> >>
>> >> -#ifdef __mips_micromips
>> >> +#if defined(__mips_micromips) || defined(__mips_soft_float)
>> >>/* DO NOTHING */
>> >> #else
>> >>
>>
>> Mind adding a
On Tue, Aug 12, 2014 at 1:29 PM, Kirill Yukhin wrote:
> This patch extends support of masking instructions.
>
> Bootstrapped.
>
> Is it ok for trunk?
>
> gcc/
> * config/i386/i386.md
> (define_attr "isa"): Add avx512bw,noavx512bw.
> (define_attr "enabled"): Ditto.
>
On Tue, Aug 12, 2014 at 2:58 PM, Kirill Yukhin wrote:
> Hello,
> This patch extends substs/subst_attrs to be used
> with new patterns.
>
> Bootstrapped.
>
> Is it ok for trunk?
In a couple of places, checks are changed to GET_MODE_SIZE
(GET_MODE (operands[0])) and in a similar way, mode checks w
On 08/12/14 11:46, Steve Ellcey wrote:
After talking to Jeff Law at the GCC Cauldron I have updated my switch
shortcut plugin pass to try and address this optimization in the hopes of
getting it added to GCC as a static pass. I fixed the code to build with
the various C++ changes that have been
There is a syntax error in r213009 causing a FTBFS in trunk with
--target=i686-pc-cygwin. Patch attached, with which trunk now builds
for said target.
--
Yaakov Selkowitz
Associate Software Engineer, ARM
Red Hat, Inc.
2014-08-12 Yaakov Selkowitz
* config/i386/cygming-crtbegin.c (de
2014-08-12 20:34 GMT+02:00 Yaakov Selkowitz :
> There is a syntax error in r213009 causing a FTBFS in trunk with
> --target=i686-pc-cygwin. Patch attached, with which trunk now builds for
> said target.
>
> --
> Yaakov Selkowitz
> Associate Software Engineer, ARM
> Red Hat, Inc.
Patch is ok. Plea
The attached patch enabled libatomic for Cygwin targets.
check-target-libatomic for i686-pc-cygwin:
Running gcc/libatomic/testsuite/libatomic.c/c.exp ...
=== libatomic Summary ===
# of expected passes44
# of unsupported tests 5
check-target-libatomic results
2014-08-12 20:40 GMT+02:00 Yaakov Selkowitz :
> The attached patch enabled libatomic for Cygwin targets.
>
> check-target-libatomic for i686-pc-cygwin:
>
> Running gcc/libatomic/testsuite/libatomic.c/c.exp ...
> === libatomic Summary ===
> # of expected passes44
> # of u
On Tue, Aug 12, 2014 at 05:46:09PM +, Joseph S. Myers wrote:
> On Sun, 10 Aug 2014, Marek Polacek wrote:
>
> > We weren't properly diagnosing neither the __func__ (introduced in C99),
> > nor the __FUNCTION__/__PRETTY_FUNCTION__ (GNU extension) predefined
> > identifiers. I believe we should;
On 8/12/2014 12:33 PM, Arnaud Charlet wrote:
>> This didn't get any comments earlier. Is it OK to comment?
> This one is a bit tricky so requires more thoughts. On which targets did
> you test it?
native CentOS and sparc-rtems4.11. I don't have access to much else
except *-rtems* which is basical
> It was in the rts Makefile back when it was under gcc. I suspect it
> disappeared when this
> all moved to libada. Could check the history if we need to.
Well, libada has been introduced many many years ago. Are you saying you
haven't built Ada since then, or were using local patches?
Arno
On 05/08/2014 22:23, Paolo Carlini wrote:
Hi,
On 08/05/2014 10:10 PM, Jonathan Wakely wrote:
It doesn't have to be noexcept, but IMHO there is no point changing
the containers to avoid allocation unless that allows us to mark it
noexcept. If it can throw anyway, we might as well allocate the
i
OK.
Jason
1 - 100 of 166 matches
Mail list logo
