The following fixes PR72488, we forgot to restore SSA info when we aborted
VN due to too large SCC size. I've added a verifier that checks that
we do not share SSA_NAME_{PTR,RANGE}_INFO.
The issue is latent on the GCC 6 branch as well where I will not install
the new verifier.
Bootstrap and reg
On Wed, Jan 18, 2017 at 5:36 PM, Jeff Law wrote:
> On 01/17/2017 02:28 AM, Richard Biener wrote:
>>>
>>>
>>> This feels somewhat different, but still a hack.
>>>
>>> I don't have strong suggestions on how to approach this, but what we've
>>> got
>>> here feels like a hack and one prone to bitrot.
On Wed, Jan 18, 2017 at 5:22 PM, Jakub Jelinek wrote:
> On Wed, Jan 18, 2017 at 07:15:34PM +0300, Alexander Monakov wrote:
>> On Wed, 18 Jan 2017, Jakub Jelinek wrote:
>> > We are talking here about addressable vars, right (so if we turn it into
>> > non-addressable, in the SIMT region we just use
On Wed, Jan 18, 2017 at 10:45 PM, David Malcolm wrote:
> The jit testcase test-nested-loops.c was crashing.
>
> Root cause is that deep inside loop optimization we're now exposing
> this call within fold-const.c which wasn't being hit before:
>
> 4082 /* Compute the mask to access the bitfiel
On Thu, Jan 19, 2017 at 10:31:38AM +0100, Richard Biener wrote:
> On Wed, Jan 18, 2017 at 5:22 PM, Jakub Jelinek wrote:
> > On Wed, Jan 18, 2017 at 07:15:34PM +0300, Alexander Monakov wrote:
> >> On Wed, 18 Jan 2017, Jakub Jelinek wrote:
> >> > We are talking here about addressable vars, right (so
Hi All,
This is a slight modification of the earlier patch (Using a different constant
in the mask creation.)
< + HOST_WIDE_INT_M1 << bits));
---
> + HOST_WIDE_INT_M1U <<
> bits));
Kind Regards,
T
On Wed, Jan 18, 2017 at 4:32 PM, Bin.Cheng wrote:
> On Wed, Jan 18, 2017 at 2:54 PM, Richard Biener
> wrote:
>> On Wed, Jan 18, 2017 at 11:10 AM, Martin Liška wrote:
>>> Hello.
>>>
>>>
>>> Patch can bootstrap on ppc64le-redhat-linux and survives regression tests.
>>>
>>> Ready to be installed?
>
On Thu, 19 Jan 2017, Richard Biener wrote:
> >> What about motion in the other direction, upwards across SIMT_ENTER()?
> >
> > I think this is a question for Richard, whether it can be done in the alias
> > oracle. If yes, it supposedly can be done for both SIMT_ENTER and
> > SIMT_EXIT.
>
> Code
On Thu, Jan 19, 2017 at 10:37 AM, Jakub Jelinek wrote:
> On Thu, Jan 19, 2017 at 10:31:38AM +0100, Richard Biener wrote:
>> On Wed, Jan 18, 2017 at 5:22 PM, Jakub Jelinek wrote:
>> > On Wed, Jan 18, 2017 at 07:15:34PM +0300, Alexander Monakov wrote:
>> >> On Wed, 18 Jan 2017, Jakub Jelinek wrote:
On Thu, Jan 19, 2017 at 10:44 AM, Alexander Monakov wrote:
> On Thu, 19 Jan 2017, Richard Biener wrote:
>> >> What about motion in the other direction, upwards across SIMT_ENTER()?
>> >
>> > I think this is a question for Richard, whether it can be done in the alias
>> > oracle. If yes, it suppos
On Thu, Jan 19, 2017 at 10:45:08AM +0100, Richard Biener wrote:
> > But in the escape analysis we could consider all the specially marked
> > "omp simt private" addressable vars to escape and thus confine them into the
> > SIMT region that way, right?
>
> We could. But that doesn't prevent vars f
Hi,
this patch fixes false positive of symtab_node::equal_address_to when
comparing non-interposable alias with interposable definition.
Honza
Index: ChangeLog
===
--- ChangeLog (revision 244611)
+++ ChangeLog (working copy)
@@ -
On Thu, Jan 19, 2017 at 9:42 AM, Richard Biener
wrote:
> On Wed, Jan 18, 2017 at 4:32 PM, Bin.Cheng wrote:
>> On Wed, Jan 18, 2017 at 2:54 PM, Richard Biener
>> wrote:
>>> On Wed, Jan 18, 2017 at 11:10 AM, Martin Liška wrote:
Hello.
Patch can bootstrap on ppc64le-redhat-linu
On Thu, 19 Jan 2017, Jakub Jelinek wrote:
> On Thu, Jan 19, 2017 at 10:45:08AM +0100, Richard Biener wrote:
> > > But in the escape analysis we could consider all the specially marked
> > > "omp simt private" addressable vars to escape and thus confine them into
> > > the
> > > SIMT region that wa
On 18/01/17 21:07, Jeff Law wrote:
> On 01/18/2017 11:08 AM, Richard Earnshaw (lists) wrote:
>> PR 79121 is a silent wrong code regression where, when generating a
>> shift from an extended value moving from one to two machine registers,
>> the type of the right shift is for the most significant wo
Hi Andrew,
On 18 Jan 15:45, Andrew Senkevich wrote:
> 2017-01-17 16:51 GMT+03:00 Jakub Jelinek :
> > On Tue, Jan 17, 2017 at 04:03:08PM +0300, Andrew Senkevich wrote:
> >> > I've played a bit w/ SDE. And looks like operands are not early clobber:
> >> > TID0: INS 0x004003ee AVX5
This innocent looking patch shaves 1749 warnings in stage 1
when building with clang (such as on newer versions of FreeBSD).
Richi, at one point you indicated such changes would be fine
(though usually they go from struct to class)?
Gerald
2016-09-28 Gerald Pfeifer
* coretypes.h (cla
Hi!
Apparently the Windows SEH has problems with the [xyz]mm16 and later
registers that were added with AVX512F (usable in 64-bit code only),
if any of the zmm16 to zmm31 registers are used in some function compiled
with avx512f or later, GCC emits .seh_savexmm %xmm16, ... directive or
similar and
On Thu, Jan 19, 2017 at 11:25 AM, Bin.Cheng wrote:
> On Thu, Jan 19, 2017 at 9:42 AM, Richard Biener
> wrote:
>> On Wed, Jan 18, 2017 at 4:32 PM, Bin.Cheng wrote:
>>> On Wed, Jan 18, 2017 at 2:54 PM, Richard Biener
>>> wrote:
On Wed, Jan 18, 2017 at 11:10 AM, Martin Liška wrote:
> Hel
On Thu, 19 Jan 2017, Gerald Pfeifer wrote:
> This innocent looking patch shaves 1749 warnings in stage 1
> when building with clang (such as on newer versions of FreeBSD).
>
> Richi, at one point you indicated such changes would be fine
> (though usually they go from struct to class)?
In this ca
On Thu, Jan 19, 2017 at 12:23:34PM +0100, Richard Biener wrote:
> On Thu, 19 Jan 2017, Gerald Pfeifer wrote:
>
> > This innocent looking patch shaves 1749 warnings in stage 1
> > when building with clang (such as on newer versions of FreeBSD).
> >
> > Richi, at one point you indicated such change
On Thu, Jan 19, 2017 at 11:00 AM, Jakub Jelinek wrote:
> On Thu, Jan 19, 2017 at 10:45:08AM +0100, Richard Biener wrote:
>> > But in the escape analysis we could consider all the specially marked
>> > "omp simt private" addressable vars to escape and thus confine them into
>> > the
>> > SIMT regi
On Thu, 19 Jan 2017, Jakub Jelinek wrote:
> On Thu, Jan 19, 2017 at 12:23:34PM +0100, Richard Biener wrote:
> > On Thu, 19 Jan 2017, Gerald Pfeifer wrote:
> >
> > > This innocent looking patch shaves 1749 warnings in stage 1
> > > when building with clang (such as on newer versions of FreeBSD).
>
This patch removes spurious dimensionality errors in programs that are compiled
with front-end inlining (as in SPARK mode). The original model of dimensionality
checking analyzes only nodes coming from source because the correctness of the
dimensioned program only depends on source user code. Howev
This patch modifies the processing of subprograms to properly frag a function
as returning by reference when the return type is a limited view and the full
view of the type requires the secondary stack.
-- Source --
-- pack_1.ads
with Pack_2;
with Ada.Finalization; us
This patch inproves the error report produced on a selected component Nam.Comp
which appears in an object declaration, when Nam has seeveral interpretations
as a function, and there is non-visible package with the same name.
Compiling pqi.adb must yield:
pqi.adb:15:07: no legal interpretation
Hello.
Following patch fixes asan bootstrap, as mentioned in the PR.
Ready to be installed?
Thanks,
Martin
>From 6a3d1b85e124751fdb804ae86596d30ea98b54af Mon Sep 17 00:00:00 2001
From: marxin
Date: Thu, 19 Jan 2017 10:25:55 +0100
Subject: [PATCH] Fix false positive for -Walloc-size-larger-than (
This patch modifies the generation of the invariant procedure body as follows:
* The body of the "partial" invariant procedure is still generated at the
end of the visible declarations.
* The body of the "full" invariant procedure is generated when the related
type is frozen or at the
On 18/01/2017 19:24, DJ Delorie wrote:
> Joe Seymour writes:
>>> the msp430 -mlarge multilib failing to build with...
configure: error: Unknown underlying type for size_t
make[1]: *** [configure-target-libstdc++-v3] Error 1
>>
>> This is still reproducible.
>
> FYI the underlying type i
On Thu, Jan 19, 2017 at 11:22 AM, Richard Biener
wrote:
> On Thu, Jan 19, 2017 at 11:25 AM, Bin.Cheng wrote:
>> On Thu, Jan 19, 2017 at 9:42 AM, Richard Biener
>> wrote:
>>> On Wed, Jan 18, 2017 at 4:32 PM, Bin.Cheng wrote:
On Wed, Jan 18, 2017 at 2:54 PM, Richard Biener
wrote:
>
Hi all,
unfortunately triggered this patch a regression in the opencoarray's testsuite,
which also occurs outside of opencoarray, when a caf-function is used in a
block in the main-program. This patch fixes the error and adds a testcase.
Bootstrapped and regtested ok on x86_64-linux/f25. Ok for t
Am 2017-01-12 um 21:16 schrieb Jakub Jelinek:
libmpx/
* configure.ac: Add GCC_BASE_VER.
* Makefile.am (gcc_version): Use @get_gcc_base_ver@ instead of cat to
get version from BASE-VER file.
* configure: Regenerated.
Hi,
it seems libmpx/configure.ac is missing th
glibc compiled with current gcc-7 fails one test due to strcmp and
strncmp appearing in the PLT. This is because the inline expansion of
those functions falls back to a function call, but not using the asm
name for the call. Fixed by retrieving the asm name from the builtin
decl. I used the buil
The inline expansions are non-trivial, so aren't really appropriate
for -Os.
Bootstrapped and regression tested powerpc64le-linux. OK to apply?
* config/rs6000/rs6000.md (cmpstrnsi, cmpstrsi): Fail if
optimizing for size.
diff --git a/gcc/config/rs6000/rs6000.md b/gcc/config/rs6
In the attached testcase, we have a clearly bounded case of alloca which
is being incorrectly reported:
void g (int *p, int *q)
{
size_t n = (size_t)(p - q);
if (n < 10)
f (__builtin_alloca (n));
}
The problem is that VRP gives us an anti-range for `n' which may be out
of range:
Hi all,
The patch that renamed all uses of aarch64_nopcrelative_literal_loads into
aarch64_pcrelative_literal_loads missed out
its extern declaration in aarch64-protos.h and a couple of its uses in
aarch64.md.
The aarch64_nopcrelative_literal_loads doesn't get initialised anywhere (since
it's
On Thu, Jan 19, 2017 at 01:13:14PM +0100, Franz Sirl wrote:
> Am 2017-01-12 um 21:16 schrieb Jakub Jelinek:
> > libmpx/
> > * configure.ac: Add GCC_BASE_VER.
> > * Makefile.am (gcc_version): Use @get_gcc_base_ver@ instead of cat to
> > get version from BASE-VER file.
> > * configure
On Thu, Jan 19, 2017 at 1:17 PM, Aldy Hernandez wrote:
> In the attached testcase, we have a clearly bounded case of alloca which is
> being incorrectly reported:
>
> void g (int *p, int *q)
> {
>size_t n = (size_t)(p - q);
>
>if (n < 10)
> f (__builtin_alloca (n));
> }
>
> The proble
Hi,
On 18 January 2017 at 22:45, Louis Krupp wrote:
> Fixed in revision 244601.
>
I've noticed a new failure on arm/aarch64:
compiler driver --help=fortran option(s): "^ +-.*[^:.]$" absent from
output: " -ftest-forall-temp Force creation of temporary to
test infrequently-executed fo
On 01/19/2017 07:45 AM, Richard Biener wrote:
On Thu, Jan 19, 2017 at 1:17 PM, Aldy Hernandez wrote:
In the attached testcase, we have a clearly bounded case of alloca which is
being incorrectly reported:
void g (int *p, int *q)
{
size_t n = (size_t)(p - q);
if (n < 10)
f (__builti
> 2017-01-19 Jakub Jelinek
>
> PR target/79127
> * acinclude.m4 (LIBGFOR_CHECK_AVX512F): Ensure the test clobbers
> some zmm16+ registers to verify they are handled by unwind info
> properly if needed.
> * configure: Regenerated.
OK to commit, with an additional c
On Wed, 18 Jan 2017, Jakub Jelinek wrote:
> > Inlining needs to do just like omp-low; if we take the current framework, it
> > would need to collect addressable locals into one struct, replace
> > references to
> > those locals by field references in the inlined body. Then it needs to
> > appropr
On Thu, Jan 19, 2017 at 2:02 PM, Aldy Hernandez wrote:
> On 01/19/2017 07:45 AM, Richard Biener wrote:
>>
>> On Thu, Jan 19, 2017 at 1:17 PM, Aldy Hernandez wrote:
>>>
>>> In the attached testcase, we have a clearly bounded case of alloca which
>>> is
>>> being incorrectly reported:
>>>
>>> void
On 17/01/17 15:11, Jiong Wang wrote:
>
>
> On 17/01/17 13:57, Richard Earnshaw (lists) wrote:
>> On 16/01/17 14:29, Jiong Wang wrote:
>>>
I can see the reason for doing this is if you want to seperate the
interpretion
of GCC CFA reg-note and the final DWARF CFA operation. My
On 18/01/17 17:07, Jiong Wang wrote:
> On 12/01/17 18:10, Jiong Wang wrote:
>> On 06/01/17 11:47, Jiong Wang wrote:
>>> This is the update on libgcc unwinder support according to new DWARF
>>> proposal.
>>>
>>> As Joseph commented, duplication of unwind-dw2.c is not encouraged in
>>> libgcc,
>>> Bu
On Thu, Jan 19, 2017 at 04:36:25PM +0300, Alexander Monakov wrote:
> On Wed, 18 Jan 2017, Jakub Jelinek wrote:
> > > Inlining needs to do just like omp-low; if we take the current framework,
> > > it
> > > would need to collect addressable locals into one struct, replace
> > > references to
> > >
Hi Joseph & Jeff,
Thanks for the feedback!
> > Generally, I don't see tests added that these new functions are correct
> > for float, double and long double, which would detect such issues if run
> > for a target with IBM long double.
>
> Specifically, I think gcc.dg/tg-tests.h should have tests
Jakub pointed out that parenthesized decomposition of an array wasn't
properly using direct-initialization. Rather than pass the flags down
into build_vec_init at this point in GCC 7 development, let's turn the
initializer into something that build_vec_init recognizes as
direct-initialization.
Te
On 01/19/2017 03:37 AM, Richard Earnshaw (lists) wrote:
On 18/01/17 21:07, Jeff Law wrote:
On 01/18/2017 11:08 AM, Richard Earnshaw (lists) wrote:
PR 79121 is a silent wrong code regression where, when generating a
shift from an extended value moving from one to two machine registers,
the type
On Thu, 19 Jan 2017, Tamar Christina wrote:
> > Also, I don't think the call to perform_ibm_extended_fixups in
> > is_subnormal is correct. Subnormal for IBM long double is *not* the same
> > as subnormal double high part. Likewise it's incorrect in is_normal as
> > well.
>
> The calls to is_ze
Thanks for the review.
On 19/01/17 14:18, Richard Earnshaw (lists) wrote:
diff --git a/libgcc/unwind-dw2.c b/libgcc/unwind-dw2.c
index
8085a42ace15d53f4cb0c6681717012d906a6d47..cf640135275deb76b820f8209fa51eacfd64c4a2
100644
--- a/libgcc/unwind-dw2.c
+++ b/libgcc/unwind-dw2.c
@@ -136,6 +13
On Thu, Jan 19, 2017 at 01:07:50PM +0100, Andre Vehreschild wrote:
> Hi all,
>
> unfortunately triggered this patch a regression in the opencoarray's
> testsuite,
> which also occurs outside of opencoarray, when a caf-function is used in a
> block in the main-program. This patch fixes the error a
On 01/18/2017 11:18 PM, Jan Hubicka wrote:
>>
>> 2016-12-19 Martin Liska
>>
>> * cgraphclones.c (cgraph_node::create_virtual_clone):
>> Create either IPA_REF_LOAD of IPA_REF_READ depending on
>> whether new_tree is a VAR_DECL or an ADDR_EXPR.
>> * ipa-cp.c (create_specialized
On Wed, 18 Jan 2017, Jakub Jelinek wrote:
> On Wed, Jan 18, 2017 at 10:52:32PM +0300, Alexander Monakov wrote:
> > Sorry for not noticing this earlier, but ...
> >
> > > +#ifdef __LP64__
> > > +typedef unsigned long long CUdeviceptr;
> > > +#else
> > > +typedef unsigned CUdeviceptr;
> > > +#endif
> >> 2016-12-19 Martin Liska
> >>
> >>* cgraphclones.c (cgraph_node::create_virtual_clone):
> >>Create either IPA_REF_LOAD of IPA_REF_READ depending on
> >>whether new_tree is a VAR_DECL or an ADDR_EXPR.
> >>* ipa-cp.c (create_specialized_node): Add reference just for
> >>ADD
On 01/18/2017 06:22 AM, Richard Biener wrote:
> On Wed, Jan 18, 2017 at 3:11 PM, Alexander Monakov wrote:
>> On Wed, 18 Jan 2017, Richard Biener wrote:
After OpenMP lowering, inlining might break this by inlining functions with
address-taken locals into SIMD regions. For now, such inlin
> > The calls to is_zero and is_subnormal were incorrect indeed. I've
> > corrected them by not calling the fixup code and to instead make sure it
> > falls through into the old fp based code which did normal floating point
> > operations on the number. This is the same code as was before in
> > f
Hi!
This is on top of the PR78835 patch I've posted yesterday.
We ICE on the following testcase reduced from libstdc++ ios_failure.cc.
If there are nested comdat types, as B (outer type) and A (inner type) in the
testcase, we first move A to a type unit and add there copies of its
children, and pr
> honza,
> this is the fix for the partitioned WPA bug I was tracking down.
>
> We have base and complete dtors sharing a comdat group (one's an alias for
> the other). The linker tells us the complete dtor is PREVAILING_DEF, as
> it's referenced from some other library. The base dtor is UNKNOWN
Hi All,
It seems the entry in config/aarch64/aarch64-builtins.c isn't needed, as such
I've simplified the patch
and the changelog.
Ok for trunk?
Tamar
gcc/
2017-01-19 Tamar Christina
* config/aarch64/aarch64.c (aarch64_simd_gen_const_vector_dup):
Changed int to HOST_WIDE_I
Hi Ian,
> On Mon, Jan 16, 2017 at 7:21 AM, Rainer Orth
> wrote:
>>
>> I'm getting further on Solaris now, but the build still fails:
>
> I committed this patch to fix the remaining build problems on Solaris.
> Bootstrapped and ran some of the Go tests on i386-sun-solaris11 and
> x86_64-pc-linux-g
Hi Steve,
thanks for the review. Committed as r244637.
Regards,
Andre
On Thu, 19 Jan 2017 06:51:19 -0800
Steve Kargl wrote:
> On Thu, Jan 19, 2017 at 01:07:50PM +0100, Andre Vehreschild wrote:
> > Hi all,
> >
> > unfortunately triggered this patch a regression in the opencoarray's
> >
On Thu, Jan 19, 2017 at 06:09:35PM +0300, Alexander Monakov wrote:
> > -#ifdef __LP64__
> > +#if defined(__LP64__) || defined(_WIN64)
> >
> > (is that the right define for 64-bit MingW, right?).
>
> Yes, _WIN64; libsanitizer has a similar test. Alternatively, I guess,
>
> #if __SIZEOF_POINTER
It seems I can drop even more:
gcc/
2017-01-19 Tamar Christina
* config/aarch64/aarch64.c (aarch64_simd_gen_const_vector_dup):
Changed int to HOST_WIDE_INT.
* config/aarch64/aarch64-protos.h
(aarch64_simd_gen_const_vector_dup): Likewise.
* config/aarch6
On Thu, Jan 19, 2017 at 9:43 AM, Jason Merrill wrote:
> Jakub pointed out that parenthesized decomposition of an array wasn't
> properly using direct-initialization. Rather than pass the flags down
> into build_vec_init at this point in GCC 7 development, let's turn the
> initializer into somethi
On Thu, 2017-01-19 at 10:36 +0100, Richard Biener wrote:
> On Wed, Jan 18, 2017 at 10:45 PM, David Malcolm
> wrote:
> > The jit testcase test-nested-loops.c was crashing.
> >
> > Root cause is that deep inside loop optimization we're now exposing
> > this call within fold-const.c which wasn't bei
On Thu, 19 Jan 2017, Jakub Jelinek wrote:
> On Thu, Jan 19, 2017 at 04:36:25PM +0300, Alexander Monakov wrote:
> > > One of the problems with that is that it means that you can't easily turn
> > > addressable private variables into non-addressable ones once you force
> > > them
> > > into such str
Hi Doug,
I've committed this on your behalf to get the testcases in and also
add the description of when this feature is required. Thanks for the
patch. Committed code inline below.
r244640
gcc/
PR target/78176
* config.gcc (supported_defaults): Add lxc1-sxc1.
(with_lx
Hi Paul,
Your latest version of the patch is now committed. I have been doing some
work on the recursive build failure but the issue is complex and involves
LRA so I went ahead with committing your change independently. It also
turns out that (at least when targeting loongson3a) there are stage2
On Thu, 19 Jan 2017, Tamar Christina wrote:
> > The old code set orig_arg before converting IBM long double to double.
> > Your code sets it after the conversion. The old code set min_exp based on
> > a string set from REAL_MODE_FORMAT (orig_mode)->emin - 1; your code uses
> > the adjusted mode.
On Wed, Jan 18, 2017 at 04:34:48PM +0100, Martin Liška wrote:
> Hello.
>
> During bootstrap, I came to following test-case:
>
> struct A
> {
> int regno;
> };
> struct
> {
> A base;
> } typedef *df_ref;
> int *a;
> void
> fn1 (int N)
> {
> for (int i = 0; i < N; i++)
> {
> df_ref
On Thu, Jan 19, 2017 at 12:17:40PM +, Kyrill Tkachov wrote:
> Hi all,
>
> The patch that renamed all uses of aarch64_nopcrelative_literal_loads into
> aarch64_pcrelative_literal_loads missed out its extern declaration in
> aarch64-protos.h and a couple of its uses in aarch64.md.
> The aarch64_
On Mon, 2017-01-16 at 14:42 -0700, Jeff Law wrote:
> On 01/09/2017 07:38 PM, David Malcolm wrote:
> > gcc/ChangeLog:
> > * passes.c: Include "insn-addr.h".
> > (should_skip_pass_p): Add logging. Update logic for running
> > "expand" to be compatible with both __GIMPLE and __RTL. Guard
2017-01-19 13:39 GMT+03:00 Kirill Yukhin :
> Hi Andrew,
> On 18 Jan 15:45, Andrew Senkevich wrote:
>> 2017-01-17 16:51 GMT+03:00 Jakub Jelinek :
>> > On Tue, Jan 17, 2017 at 04:03:08PM +0300, Andrew Senkevich wrote:
>> >> > I've played a bit w/ SDE. And looks like operands are not early clobber:
>>
On Mon, Jan 09, 2017 at 02:39:30PM +0100, Marek Polacek wrote:
> On Mon, Jan 09, 2017 at 12:18:01PM +0100, Jakub Jelinek wrote:
> > On Mon, Jan 09, 2017 at 10:21:47AM +0100, Marek Polacek wrote:
> > > +/* Callback function to determine whether an expression TP or one of its
> > > + subexpressions
On 01/19/2017 05:45 AM, Richard Biener wrote:
On Thu, Jan 19, 2017 at 1:17 PM, Aldy Hernandez wrote:
In the attached testcase, we have a clearly bounded case of alloca which is
being incorrectly reported:
void g (int *p, int *q)
{
size_t n = (size_t)(p - q);
if (n < 10)
f (__builti
On Thu, Jan 19, 2017 at 05:52:14PM +0100, Marek Polacek wrote:
> > I agree that not warning for
> > if (foo)
> > return NULL;
> > else
> > return NULL;
> > is bad. But how can I compare those expressions side-by-side? I'm not
> > finding
> > anything. :(
>
> Seems like ENOTIME to a
On Thu, Jan 19, 2017 at 03:55:36PM +, Tamar Christina wrote:
>
> It seems I can drop even more:
The AArch64 parts of this look fine to me, and based on benchmarking of the
patch they are low risk for high reward (and other targets have had a
vectorized copysign for a while without issue).
Ho
On Wed, Jan 18, 2017 at 11:08 PM, Uros Bizjak wrote:
> On Wed, Jan 18, 2017 at 10:48 PM, Uros Bizjak wrote:
>> Hello!
>>
>>> This fix follows the same approach that glibc uses to disable TSX on
>>> processors on which it is broken. TSX can also be disabled through a
>>> microcode update on these
On January 19, 2017 5:37:09 PM GMT+01:00, David Malcolm
wrote:
>On Thu, 2017-01-19 at 10:36 +0100, Richard Biener wrote:
>> On Wed, Jan 18, 2017 at 10:45 PM, David Malcolm
>> wrote:
>> > The jit testcase test-nested-loops.c was crashing.
>> >
>> > Root cause is that deep inside loop optimizatio
Hi,
I've rewritten/simplified this patch as it provides far too much control
to end users who will undoubtedly shoot themselves in the foot so to
speak. The option I intend to support is simply --with-madd4 --without-madd4
and -mmadd4 -mno-madd4. This is a simple enable/disable on top of
architect
Hi Jakub and Richard,
Attached is an updated BRIG FE patch which adds the HSAIL related
builtins only internally in the BRIG FE. I didn't add LTO support as I
believe it's not
useful for BRIG FE due to it always inputting fully linked BRIGs and
not mixing with
other frontends.
BR,
Pekka
On Mon
As described in the PR, gcc.dg/attr-alloc_size-4.c was FAILing on 32-bit
x86 targets. Fixed as follows by removing the mismatch between #if
conditions and target selectors.
Tested with the appropriate runtest invocations on i386-pc-solaris2.12
and amd64-pc-solaris2.12, installed on mainline.
On 19 Jan 19:42, Andrew Senkevich wrote:
> 2017-01-19 13:39 GMT+03:00 Kirill Yukhin :
> > Hi Andrew,
> > On 18 Jan 15:45, Andrew Senkevich wrote:
> >> 2017-01-17 16:51 GMT+03:00 Jakub Jelinek :
> >> > On Tue, Jan 17, 2017 at 04:03:08PM +0300, Andrew Senkevich wrote:
> >> >> > I've played a bit w/ S
Hi Joseph,
I made the requested changes and did a quick pass over the rest
of the fp cases.
Regards,
Tamar
From: Joseph Myers
Sent: Thursday, January 19, 2017 4:33:13 PM
To: Tamar Christina
Cc: Jeff Law; GCC Patches; Wilco Dijkstra; rguent...@suse.de; Mi
Hi James,
I have corrected the testsuite changes and attached is the new file and
changelog.
Ok for trunk?
Tamar
Hi All,
This patch vectorizes the copysign builtin for Aarch64
similar to how it is done for Arm.
AArch64 now generates:
...
.L4:
ldr q1, [x6, x3]
add w4,
On Wed, Jan 18, 2017 at 5:29 PM, David Malcolm wrote:
> Here's a version of the patch which simply tweaks
> cp_parser_primary_expression's call to finish_id_expression so that
> it passes the location of the id_expression, rather than that of
> id_expr_token.
>
> The id_expression in question came
On Thu, Jan 19, 2017 at 06:05:52PM +, Tamar Christina wrote:
> Hi James,
>
> I have corrected the testsuite changes and attached is the new file and
> changelog.
>
> Ok for trunk?
>
> Tamar
>
> Hi All,
>
> This patch vectorizes the copysign builtin for Aarch64
> similar to how it is done
On Thu, 19 Jan 2017, Tamar Christina wrote:
> Hi Joseph,
>
> I made the requested changes and did a quick pass over the rest
> of the fp cases.
I've no further comments, but watch out for any related test failures
being reported.
--
Joseph S. Myers
jos...@codesourcery.com
This turns lots and lots of copies into moves, which can make a huge
difference to performance (consider a std::function which has to
allocate on every copy).
PR libstdc++/67085
* include/bits/stl_heap.h (push_heap, __adjust_heap, __pop_heap)
(pop_heap, __make_heap, make_h
On 19/01/17 18:28 +, Jonathan Wakely wrote:
@@ -397,7 +401,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
while (__last - __first > 1)
{
--__last;
- std::__pop_heap(__first, __last, __last, __comp);
+ std::__pop_heap(__first, __last, __last, _GLIBCXX_MOVE(__comp
On 19/01/17 14:46, Jiong Wang wrote:
> Thanks for the review.
>
> On 19/01/17 14:18, Richard Earnshaw (lists) wrote:
>>
>>>
>>>
>>> diff --git a/libgcc/unwind-dw2.c b/libgcc/unwind-dw2.c
>>> index
>>> 8085a42ace15d53f4cb0c6681717012d906a6d47..cf640135275deb76b820f8209fa51eacfd64c4a2
>>> 100644
>>>
On 19/01/17 18:50 +, Jonathan Wakely wrote:
On 19/01/17 18:28 +, Jonathan Wakely wrote:
@@ -397,7 +401,7 @@ _GLIBCXX_BEGIN_NAMESPACE_VERSION
while (__last - __first > 1)
{
--__last;
- std::__pop_heap(__first, __last, __last, __comp);
+ std::__pop_he
Hello!
Attached patch avoids bootstrap error when building libgfortran for
soft-float x86 targets. Configure detects when _SOFT_FLOAT is defined
and uses fpu-generic.h instead of fpu-387.h header.
The patch also imports ax_check_define.m4 from autoconf archive [1] -
the macro is really handy, an
On Thu, 2017-01-19 at 13:15 -0500, Jason Merrill wrote:
> On Wed, Jan 18, 2017 at 5:29 PM, David Malcolm
> wrote:
> > Here's a version of the patch which simply tweaks
> > cp_parser_primary_expression's call to finish_id_expression so that
> > it passes the location of the id_expression, rather th
On Thu, Jan 19, 2017 at 10:43 PM, Uros Bizjak wrote:
> Hello!
>
> Attached patch avoids bootstrap error when building libgfortran for
> soft-float x86 targets. Configure detects when _SOFT_FLOAT is defined
> and uses fpu-generic.h instead of fpu-387.h header.
>
> The patch also imports ax_check_d
On 10/01/2017 13:39, Jonathan Wakely wrote:
I've committed the attached patch, which passes the tests for the
default configuration and the versioned namespace configuration.
I added another helper function, strip_versioned_namespace, which is
more expressive than doing typename.replace(vers_nsp
On 19/01/17 22:01 +0100, François Dumont wrote:
On 10/01/2017 13:39, Jonathan Wakely wrote:
I've committed the attached patch, which passes the tests for the
default configuration and the versioned namespace configuration.
I added another helper function, strip_versioned_namespace, which is
mor
Hello!
This patch (partially) reverts my change from 2013-11-05. Apparently,
LONG_DOUBLE_TYPE_SIZE interferes with soft-float handling.
2017-01-19 Uros Bizjak
PR target/78478
Revert:
2013-11-05 Uros Bizjak
* config/i386/rtemself.h (LONG_DOUBLE_TYPE_SIZE): New define.
Test
Hello!
This file is now the same as config/i386/rtemself.h. Remove one copy.
2017-01-19 Uros Bizjak
* config.gcc (x86_64-*-rtems*): Use i386/rtemself.h
instead of i386/rtems-64.h.
* config/i386/rtems-64.h: Remove.
Committed as obvious.
Uros.
Index: config/i386/rtems-64.h
===
1 - 100 of 125 matches
Mail list logo
