Following latest fixes in std::inplace_merge and std::stable_sort you
propose Jonathan to enhance branching in the first.
Here is a proposal based on yours to do so in both algos.
libstdc++: Enhance branching in std::inplace_merge and std::stable_sort
libstdc++-v3/ChangeLog:
Hi:
This is an updated patch which implements vzeroupper as call_insn
which has a special vzeroupper ABI, also in this patch i reverted
r11-7684, r10-6451, r10-3677 which seems to fix the same issue but in
a different way.
Bootstrapped and regtested on x86_64-linux-gnux{-m32,} and
x86_64-linux-
On Linux/x86_64,
28484d00c45b7bf094a22a4fddf9ffdc7482c7e1 is the first bad commit
commit 28484d00c45b7bf094a22a4fddf9ffdc7482c7e1
Author: Uros Bizjak
Date: Wed May 26 20:44:49 2021 +0200
i386: Autovectorize 4-byte vectors
caused
FAIL: gcc.dg/vect/vect-epilogues.c -flto -ffat-lto-objects
On 5/26/21 2:04 PM, Harald Anlauf via Fortran wrote:
Dear Fortranners,
Gerhard found a case where bounds-checking for an optional,
allocatable character dummy resulted in an ICE. We'd better
not call the presence check on a non-dummy symbol, as this
will hit an assert...
Regtested on x86_64-pc
On 5/18/21 11:36 AM, Harald Anlauf via Fortran wrote:
The generation of the new runtime check picked up the wrong attributes
in the case of CLASS array arguments. There is related new code in
gfc_conv_procedure_call which served as reference for the fix.
Regtested on x86_64-pc-linux-gnu.
OK fo
Hi,
As described in PR100499, loop niters analysis for "!=" now relies on
multiple_of_p which
so far is mostly implemented for no-overflow scenarios. This patch fixes the
issue by:
1. add new parameter to multiple_of_p indicating no-wrapping behavior in top
expression.
2. pass new argument to h
On Tue, May 25, 2021 at 10:15 AM Aldy Hernandez via Gcc-patches
wrote:
>
> Adjustments per discussion.
>
> OK pending tests?
>
> Aldy
This caused:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100787
--
H.J.
On Wed, May 26, 2021 at 2:55 PM Jonathan Wakely wrote:
>
> On Wed, 26 May 2021 at 20:11, Patrick Palka via Libstdc++
> wrote:
> >
> > On Wed, 26 May 2021, Tim Song wrote:
> >
> > > I noticed that output_iterator_wrapper still has a (non-void)
> > > value_type. Perhaps we can get better coverage i
On Linux/x86_64,
fe9a499cb8775cfbcea356ab0cae5c365971cf86 is the first bad commit
commit fe9a499cb8775cfbcea356ab0cae5c365971cf86
Author: Aldy Hernandez
Date: Wed May 19 18:27:47 2021 +0200
Convert Walloca pass to get_range_query.
caused
FAIL: gcc.dg/Wstringop-overflow-55.c pr? (test
On Linux/x86_64,
28484d00c45b7bf094a22a4fddf9ffdc7482c7e1 is the first bad commit
commit 28484d00c45b7bf094a22a4fddf9ffdc7482c7e1
Author: Uros Bizjak
Date: Wed May 26 20:44:49 2021 +0200
i386: Autovectorize 4-byte vectors
caused
FAIL: gcc.target/i386/pr100637-3w.c scan-assembler pmulhrsw
This implements the wording changes of CWG 1315, which permits non-type
template arguments in a partial specialization to use template
parameters more freely. Delightfully, it seems the only change needed
is to remove a few checks from process_partial_specialization.
But that change alone reveale
On 5/26/21 3:18 PM, Martin Sebor wrote:
On 5/26/21 8:00 AM, Jason Merrill wrote:
On 5/13/21 9:07 PM, Jason Merrill wrote:
On 5/13/21 7:38 PM, Martin Sebor wrote:
On 5/13/21 1:28 PM, Jason Merrill via Gcc-patches wrote:
Ping.
On 4/28/21 9:32 AM, Jason Merrill wrote:
-Wdeprecated-copy was de
While checking objects whose addresses are passed to functions
declared to take const pointers and making sure they're initialized
the GCC 11 -Wmaybe-uninitialized enhancement assumes that the actual
argument is a pointer.
That's normally a safe assumption because for nonpointer arguments
the fro
> OK
Thanks. I have also installed the attached tweak to the documentation.
PR c/100653
* doc/extend.texi (scalar_storage_order): Rephrase slightly.
--
Eric Botcazoudiff --git a/gcc/doc/extend.texi b/gcc/doc/extend.texi
index 3ddeb0dee3a..bfe9c21bddf 100644
--- a/gcc/doc/exten
Hi.
Here's the new patch with the fixes requested.
Le jeudi 20 mai 2021 à 16:24 -0400, David Malcolm a écrit :
> On Mon, 2021-05-17 at 21:02 -0400, Antoni Boucher via Jit wrote:
> > Hello.
> > This patch fixes the issue with using atomic builtins in libgccjit.
> > Thanks to review it.
>
> [...sni
For some reason this never showed up on gcc-patches, trying again.
Aaron Sawdey, Ph.D. saw...@linux.ibm.com
IBM Linux on POWER Toolchain
> Begin forwarded message:
>
> From: Aaron Sawdey
> Subject: [PATCH,rs6000] Fix p10 fusion test cases for -m32
> Date: May 25, 2021 at 1:45:36 PM CDT
> To:
On Friday, 14 May 2021 18:05:03 CEST Martin Sebor wrote:
> [...]
>
> At the same time, my concern with adding another syntactic renaming
> mechanism that's specifically intended to change symbol names in
> diagnostics (and the two macros) but nowhere else is that it would
> considerably raise the
New revision which can also be compiled with GCC 4.8.
From: Matthias Kretz
Ensure dump_template_decl for function templates never prints template
parameters after the function name (it did with -fno-pretty-templates)
and skip output of irrelevant & confusing "[with T = T]" in
dump_substitution.
From: Matthias Kretz
dump_type on 'const std::string' should not print 'const string' unless
TFF_UNQUALIFIED_NAME is requested.
gcc/cp/ChangeLog:
PR c++/100763
* error.c: Call dump_scope when printing a typedef.
---
gcc/cp/error.c | 2 ++
1 file changed, 2 insertions(+)
--
Using an MMA builtin within an openmp parallel code block leads to an SSA
verification ICE on the temporaries we create while expanding the MMA builtins
at gimple time. The solution is to use create_tmp_reg_or_ssa_name(), which
knows when to create either an SSA or register temporary.
This fixes
Dear Fortranners,
Gerhard found a case where bounds-checking for an optional,
allocatable character dummy resulted in an ICE. We'd better
not call the presence check on a non-dummy symbol, as this
will hit an assert...
Regtested on x86_64-pc-linux-gnu.
OK for mainline? And backport to 11?
(Th
Hi!
On Mon, Apr 26, 2021 at 09:35:58AM -0700, Carl Love wrote:
> This patch fixes the order of the argument in the vec_rlmi and
> vec_rlnm builtins. The patch also adds a new test cases to verify
> the fix.
>
> The patch has been tested on
> powerpc64-linux instead (Power 8 BE)
> powerpc
This commit breaks the build of glibc for powerpc64le-linux-gnu. Compile
the following code with -O2 -mlong-double-128 -mabi=ibmlongdouble and I
get the error
opts-bug.c:8:1: error: '-mabi=ibmlongdouble' requires '-mlong-double-128'
8 | {
| ^
which is clearly nonsensical, since -mlon
Hi Jakub
Updated patch attached – addressing your points:
On 28.04.21 15:41, Jakub Jelinek wrote:
On Tue, Apr 27, 2021 at 03:36:38PM +0200, Tobias Burnus wrote:
[...] (a) This patch add the iterator support to the Fortran FE
and adds support for it to the depend clause.
(b) It also adds a con
On Wed, 26 May 2021 at 20:11, Patrick Palka via Libstdc++
wrote:
>
> On Wed, 26 May 2021, Tim Song wrote:
>
> > I noticed that output_iterator_wrapper still has a (non-void)
> > value_type. Perhaps we can get better coverage if it doesn't have one?
> > The existing tests should have caught this ca
The standard specifies std::ranges::distance etc as function templates,
but it also requires them to not be found by ADL, and to suppress ADL
when normal unqualified lookup does find them. That means they need to
be function objects.
libstdc++-v3/ChangeLog:
PR libstdc++/100768
* i
The -Wvla-parameter checking of the bounds of VLA parameters in
function redeclarations has a few bugs: a) it scans the string
that encodes the attribute access representation of the VLA
bounds from the end, failing to consider that the string may
be followed by another for a subsequent VLA, b) it
On 5/26/21 8:00 AM, Jason Merrill wrote:
On 5/13/21 9:07 PM, Jason Merrill wrote:
On 5/13/21 7:38 PM, Martin Sebor wrote:
On 5/13/21 1:28 PM, Jason Merrill via Gcc-patches wrote:
Ping.
On 4/28/21 9:32 AM, Jason Merrill wrote:
-Wdeprecated-copy was depending only on the state of the warning
On Wed, May 26, 2021 at 1:43 PM Patrick Palka wrote:
>
> On Wed, 26 May 2021, Tim Song wrote:
> >
> > On Wed, May 26, 2021 at 12:00 PM Patrick Palka via Libstdc++
> > wrote:
> > >
> > > - else if constexpr (input_iterator<_Out>
> > > - && same_as,
> > > iter_value_
On Fri, 14 May 2021, Patrick Palka wrote:
> r11-8053 rewrote the range adaptor implementation in conformance with
> P2281, making partial application act like a SFINAE-friendly perfect
> forwarding call wrapper. Making SFINAE-friendliness coexist with
> perfect forwarding here requires adding fal
2021-05-26 Uroš Bizjak
gcc/
* config/i386/i386.c (ix86_autovectorize_vector_modes):
Add V4QImode and V16QImode for TARGET_SSE2.
* doc/sourcebuild.texi (Vector-specific attributes):
Add vect64 and vect32 description.
gcc/testsuite/
* lib/target-supports.exp (check_effective
On Wed, 26 May 2021, Tim Song wrote:
> I noticed that output_iterator_wrapper still has a (non-void)
> value_type. Perhaps we can get better coverage if it doesn't have one?
> The existing tests should have caught this case with that change, at least.
Good point, and I guess it should be fine to
On 26.05.21 14:06, Jakub Jelinek wrote:
When gcc is configured for nvptx offloading with --without-cuda-driver
and full CUDA isn't installed, many libgomp.oacc-*/* tests fail,
some of them because cuda.h header can't be found, others because
the tests can't be linked against -lcuda, -lcudart or -
On 5/26/21 7:07 AM, Andrew Pinski wrote:
On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
On Wed, May 26, 2021 at 1:43 AM Bernd Edlinger
wrote:
On 5/25/21 4:22 PM, Richard Biener via Gcc-patches wrote:
On Sun, May 23, 2021 at 12:03 PM apinski--- via Gcc-patches
wrote:
From: Andrew Pins
On Wed, 26 May 2021 at 18:00, Patrick Palka via Libstdc++
wrote:
>
> When input_iterator<_Out> isn't satisfied, we need to avoid substituting
> into iter_value_t<_Out> because the latter isn't necessarily
> well-formed in that case. To that end, this patch rewrites the
> problematic condition in
I noticed that output_iterator_wrapper still has a (non-void)
value_type. Perhaps we can get better coverage if it doesn't have one?
The existing tests should have caught this case with that change, at least.
On Wed, May 26, 2021 at 12:00 PM Patrick Palka via Libstdc++
wrote:
>
> - else if
On 5/26/21 1:07 PM, Bernd Edlinger wrote:
On 5/26/21 7:03 PM, Bernd Edlinger wrote:
On 5/26/21 2:05 PM, Richard Biener wrote:
On Wed, May 26, 2021 at 1:37 PM Andrew Pinski wrote:
On Wed, May 26, 2021 at 4:28 AM Richard Biener
wrote:
On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
On
On 5/26/21 7:03 PM, Bernd Edlinger wrote:
> On 5/26/21 2:05 PM, Richard Biener wrote:
>> On Wed, May 26, 2021 at 1:37 PM Andrew Pinski wrote:
>>>
>>> On Wed, May 26, 2021 at 4:28 AM Richard Biener
>>> wrote:
On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
>
> On Wed, May 2
On 5/26/21 2:05 PM, Richard Biener wrote:
> On Wed, May 26, 2021 at 1:37 PM Andrew Pinski wrote:
>>
>> On Wed, May 26, 2021 at 4:28 AM Richard Biener
>> wrote:
>>>
>>> On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
>
>
On 5/26/21 8:05 AM, Richard Biener wrote:
On Wed, May 26, 2021 at 1:37 PM Andrew Pinski wrote:
On Wed, May 26, 2021 at 4:28 AM Richard Biener
wrote:
On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
On Wed, May 26, 2021 at 1:43 AM Be
On 5/26/2021 8:12 AM, Eric Botcazou wrote:
Hi,
As documented in the manual, we do not support type punning that toggles the
scalar storage order, so this adds a warning for the simple case of unions.
Tested on x86-64/Linux, OK for the mainline?
2021-05-26 Eric Botcazou
PR c/100
When input_iterator<_Out> isn't satisfied, we need to avoid substituting
into iter_value_t<_Out> because the latter isn't necessarily
well-formed in that case. To that end, this patch rewrites the
problematic condition in ranges::unique_copy into a nested requirement
which has the correct short-ci
On 5/26/21 10:31 AM, Patrick Palka wrote:
On Tue, 25 May 2021, Jason Merrill wrote:
On 5/25/21 10:50 AM, Patrick Palka wrote:
On Mon, 24 May 2021, Jason Merrill wrote:
On 5/21/21 4:35 PM, Patrick Palka wrote:
Here, during ahead of time access checking for the private member
EnumeratorRange:
> But with --enable-linux-futex it used to work reliably.
> Isn't the problem we are talking about here introduced with
> r12-1044-gfd97aeb494cdcffe0d21e7f15ab4593662e065bdAA and
> worked around with r12-1064-ga0d371a25148d113f7c3657f36fbf7dcb5e1f63d ?
Yes, dangling references apparently, henc
On Wed, May 26, 2021 at 03:45:06PM +0100, Iain Sandoe via Gcc-patches wrote:
> on several platforms ...
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88707
But with --enable-linux-futex it used to work reliably.
Isn't the problem we are talking about here introduced with
r12-1044-gfd97aeb494cdcff
Andrew MacLeod via Gcc-patches wrote:
On 5/25/21 11:03 PM, sunil.k.pandey wrote:
On Linux/x86_64,
a6e94287d31525b3ad0963ad22a92e9f3dbcd3cf is the first bad commit
commit a6e94287d31525b3ad0963ad22a92e9f3dbcd3cf
Author: Andrew MacLeod
Date: Tue May 25 14:59:54 2021 -0400
Remove the log
On Wed, 26 May 2021 at 11:32, Richard Sandiford
wrote:
>
> Christophe Lyon writes:
> > This patch adds support for the reduc_plus_scal optab with MVE, which
> > maps to the vaddv instruction.
> >
> > It moves the reduc_plus_scal_ expander from neon.md to
> > vec-common.md and adds support for MVE
On Tue, 25 May 2021, Jason Merrill wrote:
> On 5/25/21 10:50 AM, Patrick Palka wrote:
> > On Mon, 24 May 2021, Jason Merrill wrote:
> >
> > > On 5/21/21 4:35 PM, Patrick Palka wrote:
> > > > Here, during ahead of time access checking for the private member
> > > > EnumeratorRange::end_reached_ in
On 5/25/21 11:03 PM, sunil.k.pandey wrote:
On Linux/x86_64,
a6e94287d31525b3ad0963ad22a92e9f3dbcd3cf is the first bad commit
commit a6e94287d31525b3ad0963ad22a92e9f3dbcd3cf
Author: Andrew MacLeod
Date: Tue May 25 14:59:54 2021 -0400
Remove the logical stmt cache for now.
caused
FAIL:
Hi,
As documented in the manual, we do not support type punning that toggles the
scalar storage order, so this adds a warning for the simple case of unions.
Tested on x86-64/Linux, OK for the mainline?
2021-05-26 Eric Botcazou
PR c/100653
* c-decl.c (finish_struct): Warn for
On 5/13/21 9:07 PM, Jason Merrill wrote:
On 5/13/21 7:38 PM, Martin Sebor wrote:
On 5/13/21 1:28 PM, Jason Merrill via Gcc-patches wrote:
Ping.
On 4/28/21 9:32 AM, Jason Merrill wrote:
-Wdeprecated-copy was depending only on the state of the warning
at the
point where we call the function,
Excerpts from Iain Buclaw's message of May 19, 2021 5:41 pm:
> Hi,
>
> This is a belated patch which covers some of the more notable changes
> that have gone into the GCC 11 release of the D front-end.
>
> Ran this through the W3 validator and no new warnings are generated.
>
> I will go through
On Wed, May 26, 2021 at 03:24:00PM +0200, Jan-Benedict Glaw wrote:
> thanks, committed with above changes.
This patch removes useless register keywords from several backends and one
spot in the Ada FE.
Committed to trunk as obvious.
2021-05-26 Jakub Jelinek
gcc/
* config/epiphany/epi
Ping.
On 5/17/21 1:58 PM, Jason Merrill wrote:
On 5/17/21 3:56 AM, Richard Biener wrote:
On Fri, May 14, 2021 at 2:23 AM Martin Sebor via Gcc-patches
wrote:
On 5/13/21 1:26 PM, Jason Merrill via Gcc-patches wrote:
Ping.
On 5/1/21 12:29 PM, Jason Merrill wrote:
Like my recent patch to add o
On 5/25/21 12:16 PM, Aldy Hernandez wrote:
The interface is now an inline function for get_range_query() and an
external function for get_global_range_query), as discussed.
There are no magic cfun uses behind the scenes.
The annoying downcast is gone.
Passes can now decide if they want to exp
Hi Jukub & Richard,
On Wed, 2021-05-26 14:47:20 +0200, Jakub Jelinek wrote:
> On Wed, May 26, 2021 at 02:42:47PM +0200, Jan-Benedict Glaw wrote:
> > The "register" keyword was removed in C++17, remove them to get the
> > backen
>
> s/backend/&d/
>
> > to build again.
> >
> > g
On Wed, May 26, 2021 at 11:50:07AM +0200, Richard Biener wrote:
> > We can split the loop into two loops:
> >
> > while (++k > n)
> > a[k] = b[k] + 1;
> > while (l++ < n)
> > a[k] = b[k] + 1;
> >
> > then for the second loop, it could be optimized.
>
> Btw, I think even the first l
On Wed, May 26, 2021 at 3:12 PM Jan-Benedict Glaw wrote:
>
> Hi!
>
> I'd like to commit the following patch.
>
> Okay for master?
OK.
Richard.
> Thanks,
> Jan-Benedict
>
>
>
>
> arc: Remove useless register keyword
>
> The "register" keyword was removed in C++17, remove them to get th
On Wed, May 26, 2021 at 02:42:47PM +0200, Jan-Benedict Glaw wrote:
> The "register" keyword was removed in C++17, remove them to get the backen
s/backend/&d/
> to build again.
>
> gcc/
> * config/arc/arc.c (arc_address_cost, arc_print_operand_address,
> ar
Hi!
I'd like to commit the following patch.
Okay for master?
Thanks,
Jan-Benedict
arc: Remove useless register keyword
The "register" keyword was removed in C++17, remove them to get the backen
to build again.
gcc/
* config/arc/arc.c (arc_address_cost,
On Wed, May 26, 2021 at 7:06 AM Hongtao Liu wrote:
>
> On Tue, May 25, 2021 at 6:24 PM Richard Biener
> wrote:
> >
> > On Mon, May 24, 2021 at 11:52 AM Hongtao Liu wrote:
> > >
> > > Hi:
> > > Details described in PR.
> > > Bootstrapped and regtest on
> > > x86_64-linux-gnu{-m32,}/x86_64-lin
On Wed, May 26, 2021 at 10:44 AM Uros Bizjak wrote:
>
> On Tue, May 25, 2021 at 4:29 PM Richard Biener
> wrote:
> >
> > On Fri, May 21, 2021 at 5:00 PM Uros Bizjak via Gcc-patches
> > wrote:
> > >
> > > Here it is, the patch that enables auto-vectorization for 32bit modes.
> > >
> > > Sent as RF
Hi Richard,
> Are we actually planning to do any linker relaxations here, or is this
> purely theoretical? If doing relaxations is a realistic possiblity then
> I agree that would be a good/legitimate reason to use a single define_insn
> for both instructions. In that case though, there should b
Hi!
When gcc is configured for nvptx offloading with --without-cuda-driver
and full CUDA isn't installed, many libgomp.oacc-*/* tests fail,
some of them because cuda.h header can't be found, others because
the tests can't be linked against -lcuda, -lcudart or -lcublas.
I usually only have akmod-nv
On Wed, May 26, 2021 at 1:37 PM Andrew Pinski wrote:
>
> On Wed, May 26, 2021 at 4:28 AM Richard Biener
> wrote:
> >
> > On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
> > >
> > > On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
> > > >
> > > > On Wed, May 26, 2021 at 1:43 AM Bernd Ed
On Wed, May 26, 2021 at 4:28 AM Richard Biener
wrote:
>
> On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
> >
> > On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
> > >
> > > On Wed, May 26, 2021 at 1:43 AM Bernd Edlinger
> > > wrote:
> > > >
> > > > On 5/25/21 4:22 PM, Richard Biener
On Wed, May 26, 2021 at 1:07 PM Andrew Pinski wrote:
>
> On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
> >
> > On Wed, May 26, 2021 at 1:43 AM Bernd Edlinger
> > wrote:
> > >
> > > On 5/25/21 4:22 PM, Richard Biener via Gcc-patches wrote:
> > > > On Sun, May 23, 2021 at 12:03 PM apinski--
On Wed, 12 May 2021, Qing Zhao wrote:
> Hi,
>
> This is the 3rd version of the patch for the new security feature for GCC.
>
> Please take look and let me know your comments and suggestions.
>
> thanks.
>
> Qing
>
> **Compare with the 2nd version, the following are the major changes:
>
On Wed, May 26, 2021 at 2:01 AM Andrew Pinski wrote:
>
> On Wed, May 26, 2021 at 1:43 AM Bernd Edlinger
> wrote:
> >
> > On 5/25/21 4:22 PM, Richard Biener via Gcc-patches wrote:
> > > On Sun, May 23, 2021 at 12:03 PM apinski--- via Gcc-patches
> > > wrote:
> > >>
> > >> From: Andrew Pinski
> >
> -Original Message-
> From: Tamar Christina
> Sent: 25 May 2021 16:02
> To: gcc-patches@gcc.gnu.org
> Cc: Richard Earnshaw ; nd ;
> Ramana Radhakrishnan ; Kyrylo
> Tkachov
> Subject: RE: [PATCH 3/4][AArch32]: Add support for sign differing dot-
> product usdot for NEON.
>
> Forgot to
On Mon, 17 May 2021, Jiufu Guo wrote:
> When there is the possibility that overflow/wrap may happen on the loop index,
> a few optimizations would not happen. For example code:
>
> foo (int *a, int *b, unsigned k, unsigned n)
> {
> while (++k != n)
> a[k] = b[k] + 1;
> }
>
> For this code
Christophe Lyon writes:
> This patch adds support for the reduc_plus_scal optab with MVE, which
> maps to the vaddv instruction.
>
> It moves the reduc_plus_scal_ expander from neon.md to
> vec-common.md and adds support for MVE to it.
>
> Since vaddv uses a 32-bits accumulator, we have to truncat
Hi!
OpenMP Nesting of Regions restrictions say:
- If a target update, target data, target enter data, or target exit data
construct is encountered during execution of a target region, the behavior is
unspecified.
- If a target construct is encountered during execution of a target region and
a de
merged.
On 4/30/21 9:02 PM, Geng Qi wrote:
gcc/ChangeLog:
* config/csky/csky.c (csky_can_change_mode_class): Delete.
For csky, HF/SF mode use the low bits of VREGS.
---
gcc/config/csky/csky.c | 16
1 file changed, 16 deletions(-)
diff --git a/gcc/config/csk
Wilco Dijkstra writes:
> @@ -23746,6 +23767,24 @@ aarch_macro_fusion_pair_p (rtx_insn *prev, rtx_insn
> *curr)
> }
> }
>
> + /* Always treat GOT accesses as a pair to ensure they can be easily
> + identified and optimized in linkers. */
Sorry, I guess I'd not given enough wei
On Wed, May 26, 2021 at 1:43 AM Bernd Edlinger
wrote:
>
> On 5/25/21 4:22 PM, Richard Biener via Gcc-patches wrote:
> > On Sun, May 23, 2021 at 12:03 PM apinski--- via Gcc-patches
> > wrote:
> >>
> >> From: Andrew Pinski
> >>
> >> Instead of some of the more manual optimizations inside phi-opt,
On Tue, 25 May 2021, Tamar Christina wrote:
> Hi Richi,
>
> Here's a respun version of the patch.
>
> Bootstrapped Regtested on aarch64-none-linux-gnu and no issues.
>
> Ok for master?
index
7e3aae5f9c28a49feedc7cc66e8ac0d476b9f28a..13e405edd765dde704c64348d2d0b3cd88f0af7c
100644
--- a/gcc/
Tamar Christina writes:
> Hi All,
>
> The current RTL for the vectorizer patterns for dot-product are incorrect.
> Operand3 isn't an output parameter so we can't write to it.
>
> This fixes this issue and reduces the number of RTL.
>
> Bootstrapped Regtested on aarch64-none-linux-gnu and no issues
Tamar Christina writes:
>> -Original Message-
>> From: Richard Sandiford
>> Sent: Monday, May 10, 2021 5:49 PM
>> To: Tamar Christina
>> Cc: gcc-patches@gcc.gnu.org; nd ; Richard Earnshaw
>> ; Marcus Shawcroft
>> ; Kyrylo Tkachov
>> Subject: Re: [PATCH 2/4]AArch64: Add support for sign
Think list got dropped on my last reply.
Forwarding to archive the OK.
From: Richard Biener
Sent: Wednesday, May 26, 2021 9:40 AM
To: Tamar Christina
Cc: nd
Subject: RE: [PATCH 4/4]middle-end: Add tests middle end generic tests for sign
differing dotproduct.
O
Richard Biener writes:
> On Mon, May 17, 2021 at 3:18 PM Joern Wolfgang Rennecke
> wrote:
>>
>> Attached is the updated version of the patch.
>> Bootstrapped and regtested on x86_64-pc-linux-gnu.
>>
>> OK to apply?
>
> + machine_mode m = mode_for_size ((prec + 1) / 2, MODE_INT, 1).require
On Tue, May 25, 2021 at 4:29 PM Richard Biener
wrote:
>
> On Fri, May 21, 2021 at 5:00 PM Uros Bizjak via Gcc-patches
> wrote:
> >
> > Here it is, the patch that enables auto-vectorization for 32bit modes.
> >
> > Sent as RFC, because the patch fails some vectorizer scans, as it
> > obviously ena
On Wed, 26 May 2021, Prathamesh Kulkarni via Gcc-patches wrote:
The attached patch removes calls to builtins in vmul_n* (a, b) with __a * __b.
I am not familiar with neon, but are __a and __b unsigned here? Otherwise,
is vmul_n already undefined in case of overflow?
--
Marc Glisse
On 5/25/21 4:22 PM, Richard Biener via Gcc-patches wrote:
> On Sun, May 23, 2021 at 12:03 PM apinski--- via Gcc-patches
> wrote:
>>
>> From: Andrew Pinski
>>
>> Instead of some of the more manual optimizations inside phi-opt,
>> it would be good idea to do a lot of the heavy lifting inside match
On Tue, May 25, 2021 at 5:12 PM H.J. Lu wrote:
>
> On Tue, May 25, 2021 at 7:34 AM Richard Biener
> wrote:
> >
> > On Thu, May 20, 2021 at 10:50 PM H.J. Lu wrote:
> > >
> > > On Wed, May 19, 2021 at 5:55 AM H.J. Lu wrote:
> > > >
> > > > On Wed, May 19, 2021 at 2:25 AM Richard Biener
> > > > w
> * gimplify.c (gimplify_decl_expr): Clear TREE_READONLY on the DECL
> when really creating an initialization statement for it.
Quite an interesting can of worms I have opened it seems. :-( So the change
apparently causes OMP lowering to create dangling references, I have applied
t
Hi,
The attached patch removes calls to builtins in vmul_n* (a, b) with __a * __b.
For float variants, it gates multiplication on __FAST_MATH__.
Since we are not removing all calls to builtins, I am not sure if we
should remove
entry for vmul_n from arm_neon_builtins.def ?
Testing the patch showed
87 matches
Mail list logo
