On February 1, 2019 12:00:26 AM GMT+01:00, Jakub Jelinek
wrote:
>Hi!
>
>As mentioned in the PR, with the introduction of ABSU_EXPR where we
>were
>previously using ABS_EXPR we've regressed various cases where VRP used
>to be
>able to determine ranges for ABS_EXPR, but as ABSU_EXPR is unknown, we
On January 31, 2019 11:56:39 PM GMT+01:00, Jakub Jelinek
wrote:
>Hi!
>
>While OpenMP parsing ensures the parallel, task, taskloop, target,
>teams
>etc. body are wrapped in ERT_MUST_NOT_THROW EH regions, for parloops we
>can
>actually end up with multiple different EH regions. If they still have
/util/pstl/parallel_utils.h:update from upstream
* testsuite/util/pstl/pstl_test_config.h:update from upstream
gg
20190131-pstl-integration.patch.bz2
Description: application/bzip
On Tue, Jan 15, 2019 at 1:10 AM Andi Kleen wrote:
>
> On Mon, Jan 14, 2019 at 04:15:20PM +0800, Bin.Cheng wrote:
> > On Mon, Jan 14, 2019 at 4:07 PM Andi Kleen wrote:
> > >
> > > Bin Cheng,
> > >
> > > I did some testing on this now. The attached patch automatically
> > > increases the iteration
On Thu, Jan 31, 2019 at 07:11:59PM -0500, Jason Merrill wrote:
> On 1/31/19 7:01 PM, Marek Polacek wrote:
> > + if (*jump_target && *jump_target == orig_jump)
>
> You shouldn't need to check that it's non-null, since the enclosing if
> established that orig_jump will be non-null. OK with that
On 1/31/19 5:51 PM, Jakub Jelinek wrote:
Hi!
THe following patch rejects invalid __attribute__ syntax where one or both
opening parens are missing, but it is accepted anyway.
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
2019-01-31 Jakub Jelinek
PR c++/87175
On 1/31/19 7:01 PM, Marek Polacek wrote:
+ if (*jump_target && *jump_target == orig_jump)
You shouldn't need to check that it's non-null, since the enclosing if
established that orig_jump will be non-null. OK with that tweak.
Jason
Here we have switch (1) with body
if (1)
{
case 1:;
return = 1;
}
else
{
default:;
}
so we're looking for "case 1:", which we found in the then branch while
evaluating the COND_EXPR, so *jump_target is cleared. But it's followed
by a jumping statement so *j
Hi!
In an if (code == OMP_FOR || code == OMP_SECTIONS) guarded block c will be
always initialized, yet clang warns anyway.
Changed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
committed to trunk.
2019-01-31 Jakub Jelinek
PR middle-end/89137
* omp-low.c (lowe
Hi!
As mentioned in the PR, with the introduction of ABSU_EXPR where we were
previously using ABS_EXPR we've regressed various cases where VRP used to be
able to determine ranges for ABS_EXPR, but as ABSU_EXPR is unknown, we
always defer to VARYING. ABSU_EXPR is actually much easier to handle tha
Hi!
I've successfully bootstrapped/regtested following patch from the PR on
x86_64-linux and i686-linux. Will you commit it?
2019-01-31 Richard Biener
PR rtl-optimization/88593
* mode-switching.c (optimize_mode_switching): Free dominators before
calling cleanup_cfg.
Hi!
While OpenMP parsing ensures the parallel, task, taskloop, target, teams
etc. body are wrapped in ERT_MUST_NOT_THROW EH regions, for parloops we can
actually end up with multiple different EH regions. If they still have some
common outer region, perhaps covering much bigger part of code than
Hi!
THe following patch rejects invalid __attribute__ syntax where one or both
opening parens are missing, but it is accepted anyway.
Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
2019-01-31 Jakub Jelinek
PR c++/87175
* parser.c (cp_parser_gnu_attribute
On 1/30/19 3:15 PM, Jason Merrill wrote:
On 1/29/19 7:15 PM, Martin Sebor wrote:
+ /* Try to convert the original SIZE to a ssizetype. */
+ if (orig_size != error_mark_node
+ && !TYPE_UNSIGNED (TREE_TYPE (orig_size)))
+ {
+ if (TREE_CODE (size) == INTEGER_CST
On Thu, Jan 31, 2019 at 10:23:17PM +0100, Thomas Koenig wrote:
> Am 27.01.19 um 14:18 schrieb Thomas Koenig:
> > Hello world,
> >
> > the attached, rather straightforward patch fixes a rejects-valid error
> > by fixing up the contiguous attribute for a class, and by using the
> > correct attribute
I was writing a test which needed to override the std::nothrow
versions of the operator new and delete to control get_temporary_buffer
behavior and noticed that it is inconsistent with
release_temporary_buffer in terms of new/delete operators.
Grepping for other std::nothrow usages I f
Am 27.01.19 um 14:18 schrieb Thomas Koenig:
Hello world,
the attached, rather straightforward patch fixes a rejects-valid error
by fixing up the contiguous attribute for a class, and by using the
correct attributes.
Regression-tested. OK for trunk?
PING ** (4./7.)?
Regards
Thomas
On Thu, Jan 31, 2019 at 03:14:39PM -0500, Jason Merrill wrote:
> OK.
Thanks, committed. Note that I plan to backport the patch to 8 after
a while.
Marek
Hello!
As exposed by gcc.dg/torture/fp-int-convert-timode-?.c testcases from
PR88931, it is expected that libgcc2 rounds calculations according to
set FP rounding mode. To achieve dynamic rounding, we have to pass
-mfp-rounding-mode=d in addition to -mieee when compiling libgcc2.
2019-01-31 Uroš
On Thu, Jan 31, 2019 at 12:32 PM Marek Polacek wrote:
> On Thu, Jan 31, 2019 at 10:17:54AM -0500, Jason Merrill wrote:
> > On 1/31/19 9:50 AM, Marek Polacek wrote:
> > > On Wed, Jan 30, 2019 at 05:37:13PM -0500, Jason Merrill wrote:
> > > > On 1/30/19 4:15 PM, Marek Polacek wrote:
> > > > > On Wed
Hi Rainer,
I'd recommend not using such a workaround:
This means getTLSRange will always return an empty range, but the GC
uses this to scan TLS memory. This means a GC collection can delete
objects which are still pointed to from TLS. This leads to hard to debug
errors, and if I remember cor
This patch allows for the use of Fortran optional arguments in the
use_device clause of a host_data directive.
I will push this into openacc-gcc-8-branch later today.
Kwok
Optional arguments should be treated as references rather than pointers
in the lowering. However, for non-present argume
Hi!
As mentioned in the PR, armv7hl-linux-gnueabi doesn't currently bootstrap
with --enable-checking=release (and profiledbootstrap is broken too),
the problem is that e.g. tree-ssa-sccvn.c is miscompiled.
The problem is that the various ldrdstrd.md peephole2s generate for
non-thumb{1,2} DImode l
Looks good to me, although ultimately Iain has to decide of course.
One nitpick: wouldn't you have to somehow mark __start/__stop _minfo as
hidden? This is important in the case where you have multiple shared
libraries and each library should have its own __start/__stop symbold to
braket the l
On Thu, Jan 31, 2019 at 10:17:54AM -0500, Jason Merrill wrote:
> On 1/31/19 9:50 AM, Marek Polacek wrote:
> > On Wed, Jan 30, 2019 at 05:37:13PM -0500, Jason Merrill wrote:
> > > On 1/30/19 4:15 PM, Marek Polacek wrote:
> > > > On Wed, Jan 30, 2019 at 04:11:11PM -0500, Marek Polacek wrote:
> > > >
The use_device_ptr patch (which needed reviewing) is needed for that
test. I bundled all the new tests into one commit, but forgot to exclude
that particular test. I will post and push that patch a little later.
Kwok
On 31/01/2019 5:20 pm, Thomas Schwinge wrote:
Hi Kwok!
On Wed, 30 Jan 2019
Hi,
This adds an extra check in omp_is_optional_argument to confirm that
Fortran is being used, to guard against it accidently matching against
some other case in another language.
This will be committed to openacc-gcc-8-branch shortly.
Kwok
gcc/
* omp-general.c (omp_is_opti
On Thu, Jan 31, 2019 at 09:13:24AM -0500, David Malcolm wrote:
> In the meantime, is the original patch OK?
Yes, thanks.
Jakub
Hi Kwok!
On Wed, 30 Jan 2019 22:24:48 +, Kwok Cheung Yeung
wrote:
> diff --git
> a/libgomp/testsuite/libgomp.oacc-fortran/optional-host_data.f90
> b/libgomp/testsuite/libgomp.oacc-fortran/optional-host_data.f90
> new file mode 100644
I'm seeing this one fail at run time: "libgomp: use_dev
On 1/31/19 7:13 AM, David Malcolm wrote:
On Thu, 2019-01-31 at 10:35 +0100, Richard Biener wrote:
On Thu, Jan 31, 2019 at 12:09 AM David Malcolm
wrote:
PR c/89122 reports that we emit a bogus fix-it hint for the case
where
the code uses FLT_MAX, but has included rather than
:
x.c:3:11: erro
Hi Kwok!
On Wed, 30 Jan 2019 22:18:43 +, Kwok Cheung Yeung
wrote:
> A non-present passed-by-reference Fortran optional argument is
> represented by a null pointer. When passed to an update directive, it
> should be ignored as variable mappings are not created for null
> pointers. This s
Hi!
On Thu, 21 Jan 2016 22:54:26 +0100, I wrote:
> Committed to gomp-4_0-branch in r232709:
>
> commit 41a76d233e714fd7b79dc1f40823f607c38306ba
> Author: tschwinge
> Date: Thu Jan 21 21:52:50 2016 +
>
> Un-parallelized OpenACC kernels constructs with nvptx offloading: "avoid
> offloa
Hi Bernd,
>> those configure files have been generated with some modified version of
>> autoconf: --runstatedir isn't present in vanilla autoconf 2.69, but only
>> in git autoconf. Since there are no uses of it I could find, please do
>> regenerate with autoconf 2.69 (preferably self-built since
On 1/31/19 1:21 PM, Rainer Orth wrote:
> Hi Bernd,
>
>> I have fixed the installation with the attached patch, but when I regenerate
>> the automake
>> files using automake-1.15.1 and autoconf-2.69, I have an issue that
>> apparently
>> the configure.ac must be out of sync, and the the generated
On Thu, Jan 31, 2019 at 04:05:51PM +0100, Richard Biener wrote:
> > On Thu, Jan 31, 2019 at 03:56:12PM +0100, Richard Biener wrote:
> > > Don't you alreday have
> > >
> > > @@ -4200,10 +4202,34 @@ pass_sprintf_length::execute (function *fun)
> > >init_target_to_host_charmap ();
> > >
> > >c
On 1/30/19 8:07 PM, Alexandre Oliva wrote:
It's too risk to reuse the type field for USING_DECL_SCOPE.
Language-independent parts of the compiler, such as location and
non-lvalue wrappers, happily take the TREE_TYPE of a USING_DECL as if
it was a type rather than an unrelated scope.
For better o
On 1/31/19 9:50 AM, Marek Polacek wrote:
On Wed, Jan 30, 2019 at 05:37:13PM -0500, Jason Merrill wrote:
On 1/30/19 4:15 PM, Marek Polacek wrote:
On Wed, Jan 30, 2019 at 04:11:11PM -0500, Marek Polacek wrote:
On Tue, Jan 29, 2019 at 09:40:18PM -0500, Jason Merrill wrote:
On Tue, Jan 29, 2019 a
On Thu, Jan 31, 2019 at 4:02 PM Jakub Jelinek wrote:
>
> On Thu, Jan 31, 2019 at 03:56:12PM +0100, Richard Biener wrote:
> > Don't you alreday have
> >
> > @@ -4200,10 +4202,34 @@ pass_sprintf_length::execute (function *fun)
> >init_target_to_host_charmap ();
> >
> >calculate_dominance_inf
In this testcase, we look for an instantiation of the outer lambda from
within the inner lambda. enclosing_instantiation_of didn't handle this
properly, as it assumed that any references would be from the same lambda
nesting depth. Fixed thus.
Tested x86_64-pc-linux-gnu, applying to trunk.
On Thu, Jan 31, 2019 at 03:56:12PM +0100, Richard Biener wrote:
> Don't you alreday have
>
> @@ -4200,10 +4202,34 @@ pass_sprintf_length::execute (function *fun)
>init_target_to_host_charmap ();
>
>calculate_dominance_info (CDI_DOMINATORS);
> + bool optimizing_late = optimize > 0 && fold
On Thu, Jan 31, 2019 at 3:45 PM Jakub Jelinek wrote:
>
> On Thu, Jan 31, 2019 at 09:30:13AM -0500, Aldy Hernandez wrote:
> > > > (LOOPS_HAVE_PREHEADERS). This avoids unnecessary changes to the CFG.
> > > > But really, I'm just guessing here. LOOPS_NORMAL would also work,
> > > > albeit creating
On 23/01/2019 10:22, Sam Tebbs wrote:
> On 14/01/2019 10:43, Kyrill Tkachov wrote:
>
>> On 08/01/19 11:38, Sam Tebbs wrote:
>>
>>
>>
>> On 1/7/19 6:28 PM, James Greenhalgh wrote:
>>
>> > On Fri, Dec 21, 2018 at 09:00:10AM -0600, Sam Tebbs wrote:
>>
>> >> On 11/9/18 11:04 AM, Sam Tebbs wrote:
>>
>>
On Wed, Jan 30, 2019 at 05:37:13PM -0500, Jason Merrill wrote:
> On 1/30/19 4:15 PM, Marek Polacek wrote:
> > On Wed, Jan 30, 2019 at 04:11:11PM -0500, Marek Polacek wrote:
> > > On Tue, Jan 29, 2019 at 09:40:18PM -0500, Jason Merrill wrote:
> > > > On Tue, Jan 29, 2019 at 6:53 PM Marek Polacek
>
Hi,
PR89008 observes that SLSR gets confused when provided with an expression of
the form X * 0.
The bug is currently latent on trunk, gcc 8, and gcc 7. This patch avoids
placing such
expressions in SLSR's candidate table, since they aren't of any use anyway.
Bootstrapped and tested on powerpc
On Thu, Jan 31, 2019 at 09:30:13AM -0500, Aldy Hernandez wrote:
> > > (LOOPS_HAVE_PREHEADERS). This avoids unnecessary changes to the CFG.
> > > But really, I'm just guessing here. LOOPS_NORMAL would also work,
> > > albeit creating forwarder blocks.
> > >
> > > Tested on x86-64 Linux.
> > >
>
On 1/31/19 4:52 AM, Richard Biener wrote:
On Thu, Jan 31, 2019 at 8:40 AM Aldy Hernandez wrote:
Hi folks.
The problem here is that Wprintf* uses the evrp_range_analyzer engine,
and doesn't understand that the x_5 != 256 conditional below should make
the RANGE [0, 255], not [0, 256].
On Thu, 2019-01-31 at 10:35 +0100, Richard Biener wrote:
> On Thu, Jan 31, 2019 at 12:09 AM David Malcolm
> wrote:
> >
> > PR c/89122 reports that we emit a bogus fix-it hint for the case
> > where
> > the code uses FLT_MAX, but has included rather than
> > :
> >
> > x.c:3:11: error: 'FLT_MAX'
On Thu, 31 Jan 2019 at 13:30, Iain Sandoe wrote:
>
>
> > On 29 Jan 2019, at 10:22, Christophe Lyon
> > wrote:
> >
> > On Tue, 29 Jan 2019 at 11:12, Iain Sandoe wrote:
> >>
>
>
> >>> Did it work for you?
> >>
> >> So far it’s looking nice… (anything that makes the dg output more stable
> >> is
> On 29 Jan 2019, at 10:22, Christophe Lyon wrote:
>
> On Tue, 29 Jan 2019 at 11:12, Iain Sandoe wrote:
>>
>>> Did it work for you?
>>
>> So far it’s looking nice… (anything that makes the dg output more stable is
>> Good in my book)
>>
>> … I am not sure (not checked thoroughly) but I t
Hi Bernd,
> I have fixed the installation with the attached patch, but when I regenerate
> the automake
> files using automake-1.15.1 and autoconf-2.69, I have an issue that apparently
> the configure.ac must be out of sync, and the the generated files are missing
> the option --runstatedir no ma
On Thu, Jan 31, 2019 at 12:40 PM Tom de Vries wrote:
>
> Hi,
>
> The 'debugaltlink_name_len =+ 1' bug reported in PR89136 exposes the fact that
> the build-id is not verified for the .gnu_debugaltlink.
>
> Fix both problems.
>
> OK for trunk?
OK.
Richard.
> Thanks,
> - Tom
>
> [libbacktrace] Fi
Hi,
The 'debugaltlink_name_len =+ 1' bug reported in PR89136 exposes the fact that
the build-id is not verified for the .gnu_debugaltlink.
Fix both problems.
OK for trunk?
Thanks,
- Tom
[libbacktrace] Fix .gnu_debugaltlink build-id check
2019-01-31 Tom de Vries
PR libbacktrace/891
Hi,
I have an issue with the installation of gcc when configured with
--enable-languages=all
on an arm-target where install-sh is used, and make install fails at libphobos
as follows:
if test -f $file; then \
/home/ed/gnu/gcc-9-20190127-0/install-sh -c -m 644 -D $file
/home/ed/gnu/arm-li
Hi,
This patch adds support for array slices on dereferenced struct
members, e.g.:
#pragma acc parallel copy(mystruct->a[0:n])
This works by making a new mapping pair for each struct pointer used in
the directive ("alloc(mystruct) firstprivate_pointer(mystruct)").
The C/C++ parsers permit chain
Hello!
Attached patch (partially) avoids emitting XOR dependency breaking
insn by removing SSE reg dependency in the AVX instructions
themselves.
2019-01-31 Uroš Bizjak
PR target/89071
* config/i386/i386.md (*extendsfdf2): Split out reg->reg
alternative to avoid partial SSE regist
On Thu, 31 Jan 2019, 10:09 Tamar Christina Hi Gerard,
>
> Thanks I'll make the suggested changes.
>
> About the duplication, we can avoid it if we have a "general" section for
> all Arm ports first and then subsections for Arm and AArch64 specific
> things.
>
> I'm not sure how the maintainers fee
Hi Gerard,
Thanks I'll make the suggested changes.
About the duplication, we can avoid it if we have a "general" section for all
Arm ports first and then subsections for Arm and AArch64 specific things.
I'm not sure how the maintainers feel about such a re-organization though.
Any opinions guy
On Thu, Jan 31, 2019 at 8:40 AM Aldy Hernandez wrote:
>
> Hi folks.
>
> The problem here is that Wprintf* uses the evrp_range_analyzer engine,
> and doesn't understand that the x_5 != 256 conditional below should make
> the RANGE [0, 255], not [0, 256].
>
> [local count: 1063004407]:
># RA
On Thu, Jan 31, 2019 at 12:09 AM David Malcolm wrote:
>
> PR c/89122 reports that we emit a bogus fix-it hint for the case where
> the code uses FLT_MAX, but has included rather than :
>
> x.c:3:11: error: 'FLT_MAX' undeclared here (not in a function); did you
> mean 'INT_MAX'?
> 3 | float
On Wed, Jan 30, 2019 at 3:42 PM Alexander Monakov wrote:
>
> On Mon, 21 Jan 2019, Alexander Monakov wrote:
>
> > Ah, I see now. I agree transparent_union ought to work, but today both GCC
> > and Clang will reject such attempt; I've filed PR 88955 for the GCC issue.
> >
> > So unfortunately such
I am testing the following patch to fix PR89135.
Bootstrap / regtest running on x86_64-unknown-linux-gnu.
Richard.
2019-01-31 Richard Biener
PR tree-optimization/89135
* tree-ssa-phiprop.c (pass_phiprop::execute): Skip blocks
with abnormal preds.
* gcc.dg/t
On Thu, 31 Jan 2019, Jakub Jelinek wrote:
> Hi!
>
> As mentioned in the PR, we refuse to inline with -fsanitize=address
> no_sanitize_address functions into functions without that attribute,
> which is good and has been requested in PR59600.
> We also refuse to inline functions without that attri
62 matches
Mail list logo
