On Mon, Jan 11, 2021 at 10:57 PM David Malcolm via Gcc-patches
wrote:
>
> If fancy_abort is called before the diagnostic subsystem is initialized,
> internal_error will crash internally in a way that prevents a useful
> message reaching the user.
>
> This can happen with libgccjit in the case of
ChangeLog:
2021-01-12 Qian Jianhua
* MAINTAINERS (Write After Approval): Add myself
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index c0aa23df57e..e88808f9fe2 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -451,6 +451,7 @@ Daniel
On Mon, Jan 11, 2021, at 8:32 AM, Richard Earnshaw wrote:
> A general comment before we start:
>
> CLZ was added to the Arm ISA in Armv5. So all subsequent Arm versions
> (and all versions implementing thumb2) will have this instruction. So
> the only cases where you'll need a fallback are
On 1/11/21 6:19 PM, Marek Polacek wrote:
This PR wants us not to warn about missing field initializers when
the code in question takes places in decltype and similar.
Hmm, the warning seems of questionable utility with templated code like
this, but I guess this is a reasonable middle ground.
On 1/11/21 8:08 PM, Patrick Palka wrote:
On Mon, 11 Jan 2021, Jason Merrill wrote:
On 1/9/21 5:23 PM, Patrick Palka wrote:
This patch teaches find_template_parameters to visit the template
represented by a CTAD placeholder, which is normally not visited by
for_each_template_parm. This
Successfully bootstrapped & regrtested on x86_64-pc-linux-gnu.
Pushed to master as r11-6603-gab88f3607233376c3145c320e92e71943a495bb5
gcc/analyzer/ChangeLog:
PR analyzer/98628
* store.cc (binding_cluster::make_unknown_relative_to): Don't mark
dereferenced unknown pointers
On Mon, 11 Jan 2021, Jason Merrill wrote:
> On 1/9/21 5:23 PM, Patrick Palka wrote:
> > This patch teaches find_template_parameters to visit the template
> > represented by a CTAD placeholder, which is normally not visited by
> > for_each_template_parm. This template may be a template template
>
On Mon, Jan 11, 2021 at 02:52:43PM -0800, H.J. Lu wrote:
> On Mon, Jan 11, 2021 at 1:20 PM Alan Modra wrote:
> >
> > On Mon, Jan 11, 2021 at 11:53:15AM -0800, H.J. Lu via Binutils wrote:
> > > Check if AR is usable for LTO build with --enable-pgo-build=lto:
> > >
> > > checking for -plugin
On Mon, Jan 11, 2021 at 4:22 PM Alan Modra wrote:
>
> On Mon, Jan 11, 2021 at 04:07:22PM -0800, H.J. Lu wrote:
> > These are not fatal errors. Here is the updated patch to use
> > AC_MSG_WARN instead. OK for master?
>
> OK by me. Please squash the two patches.
>
I will keep 2 separate since I
On Mon, Jan 11, 2021 at 04:07:22PM -0800, H.J. Lu wrote:
> These are not fatal errors. Here is the updated patch to use
> AC_MSG_WARN instead. OK for master?
OK by me. Please squash the two patches.
--
Alan Modra
Australia Development Lab, IBM
On Mon, Jan 11, 2021 at 3:27 PM Alan Modra wrote:
>
> On Mon, Jan 11, 2021 at 08:57:05AM -0800, H.J. Lu via Binutils wrote:
> > diff --git a/config/gcc-plugin.m4 b/config/gcc-plugin.m4
> > index c5b72e9a13d..798a2054edd 100644
> > --- a/config/gcc-plugin.m4
> > +++ b/config/gcc-plugin.m4
> > @@
On Mon, Jan 11, 2021 at 08:57:05AM -0800, H.J. Lu via Binutils wrote:
> diff --git a/config/gcc-plugin.m4 b/config/gcc-plugin.m4
> index c5b72e9a13d..798a2054edd 100644
> --- a/config/gcc-plugin.m4
> +++ b/config/gcc-plugin.m4
> @@ -145,6 +145,18 @@ for plugin in $plugin_names; do
> break
>
This PR wants us not to warn about missing field initializers when
the code in question takes places in decltype and similar. Fixed
thus.
Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
gcc/cp/ChangeLog:
PR c++/98620
* typeck2.c (process_init_constructor_record):
On Mon, Jan 11, 2021, at 7:21 AM, Richard Earnshaw wrote:
> Some initial comments.
>
> On 11/01/2021 11:10, g...@danielengel.com wrote:
> > From: Daniel Engel
> >
> > These definitions facilitate subsequent patches in this series.
> >
> > gcc/libgcc/ChangeLog:
> > 2021-01-07 Daniel Engel
>
On Mon, Jan 11, 2021 at 3:05 PM Alan Modra wrote:
>
> On Mon, Jan 11, 2021 at 08:57:04AM -0800, H.J. Lu via Binutils wrote:
> > Check if AR works with --plugin and rc before passing --plugin to AR and
> > RANLIB.
>
> Thanks for looking at this, but next time please assign the bug to
> yourself.
On 11/01/2021 15:58, Daniel Engel wrote:
>
> On Mon, Jan 11, 2021, at 7:39 AM, Richard Earnshaw wrote:
>> On 11/01/2021 15:26, Richard Earnshaw wrote:
>>> On 11/01/2021 11:10, g...@danielengel.com wrote:
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
On Mon, Jan 11, 2021 at 08:57:04AM -0800, H.J. Lu via Binutils wrote:
> Check if AR works with --plugin and rc before passing --plugin to AR and
> RANLIB.
Thanks for looking at this, but next time please assign the bug to
yourself. I fixed the bug too this morning, before seeing your
email.
--
On Mon, Jan 11, 2021 at 1:20 PM Alan Modra wrote:
>
> On Mon, Jan 11, 2021 at 11:53:15AM -0800, H.J. Lu via Binutils wrote:
> > Check if AR is usable for LTO build with --enable-pgo-build=lto:
> >
> > checking for -plugin option... ar: no operation specified
> > Failed: ar --plugin
> >
I pushed this patch to update the DR table with a few new DRs. Also update
some of the older ones to reflect that they're included in C++20.
Marek
commit 6be6626d59e1c492d5e05606e5a6902feb9e5bac
Author: Marek Polacek
Date: Mon Jan 11 17:18:45 2021 -0500
C++ DRs: Add new DRs, update
On Mon, Jan 11, 2021 at 10:56 AM CHIGOT, CLEMENT
wrote:
>
> >> Hi David, Clement,
> >>
> >>> The patch is local to libstdc++ AIX support, so I believe that I can
> >>> approve it.
> >>
> >>have you considered merging the dragonfly and aix trees? I'm asking
> >>because it seems prudent to try
On 1/11/21 4:40 PM, Jason Merrill wrote:
On 1/7/21 4:06 PM, Patrick Palka wrote:
This is essentially a followup to r11-3714 -- we ICEing from another
"unguarded" call to build_concept_check, this time in do_auto_deduction,
due to the presence of templated trees when !processing_template_decl.
If fancy_abort is called before the diagnostic subsystem is initialized,
internal_error will crash internally in a way that prevents a useful
message reaching the user.
This can happen with libgccjit in the case of gcc_assert failures
that occur outside of the libgccjit mutex that guards the rest
On Sat, Dec 05, 2020 at 07:42:07PM +1030, Alan Modra wrote:
> Hi Segher,
> I've been holding off pinging these knowing you had a lot of other
> review work, but maybe that's settling down now? You already OK'd
> 1/8, 2/8 and 6/8.
Ping.
> [PATCH 3/8] [RS6000] rs6000_rtx_costs tidy AND
>
On 1/7/21 4:06 PM, Patrick Palka wrote:
This is essentially a followup to r11-3714 -- we ICEing from another
"unguarded" call to build_concept_check, this time in do_auto_deduction,
due to the presence of templated trees when !processing_template_decl.
Bootstrapped and regtested on
On Mon, Jan 11, 2021 at 11:53:15AM -0800, H.J. Lu via Binutils wrote:
> Check if AR is usable for LTO build with --enable-pgo-build=lto:
>
> checking for -plugin option... ar: no operation specified
> Failed: ar --plugin
>
On Mon, 2020-12-07 at 16:31 -0800, Carl Love wrote:
> Will:
>
> I have addressed you comments with regards to the Change Log entries.
>
> The extra define vec_div was removed.
>
> Added the missing entries for DIVU_V2DI DIVS_V2DI in rs6000-call.c.
>
> The extra MULLD_V2DI case statement
On 1/9/21 5:23 PM, Patrick Palka wrote:
This patch teaches find_template_parameters to visit the template
represented by a CTAD placeholder, which is normally not visited by
for_each_template_parm. This template may be a template template
parameter (as in the first testcase), or it may
On 1/8/21 7:38 PM, Anthony Sharp wrote:
Hi Jason,
Thank you!
To start with, do you have a copyright assignment on file or in the
works already?
Good point. I incorrectly assumed it would only be a minor
contribution copyright-wise. > Mr Edelsohn gave me a template which I've
now filled out
Check if AR is usable for LTO build with --enable-pgo-build=lto:
checking for -plugin option... ar: no operation specified
Failed: ar --plugin
/usr/gcc-11.0.0-x32/libexec/gcc/x86_64-pc-linux-gnu/11.0.0/liblto_plugin.so rc
no
configure: error: AR with --plugin and rc is required for LTO build
On 1/9/21 5:43 PM, Maciej W. Rozycki wrote:
> On Mon, 21 Dec 2020, Jakub Jelinek wrote:
>
This patch adds the ~(X - Y) -> ~X + Y simplification requested
in the PR (plus also ~(X + C) -> ~X + (-C) for constants C that can
be safely negated.
>>> This regresses VAX code produced
On 1/10/21 7:45 AM, Maciej W. Rozycki wrote:
> On Fri, 8 Jan 2021, Jeff Law wrote:
>
>>> gcc/testsuite/
>>> * gcc.target/vax/cmpelim-eq-notsi.c: Use subtraction from a
>>> constant then rather than addition.
>>> * gcc.target/vax/cmpelim-le-notsi.c: Likewise.
>>> *
On 1/11/21 3:02 AM, Andreas Krebbel via Gcc-patches wrote:
> The vec-abi-varargs-1.c testcase on IBM Z currently fails.
>
> While adding an SI mode vector to a DI mode vector the first is unpacked
> using:
>
> _28 = BIT_INSERT_EXPR <{ 0, 0, 0, 0 }, _2, 0>;
> _34 = [vec_unpack_lo_expr] _28;
Hi
I had another look to this attempt to properly support alloc fancy
pointers.
I consider all your remarks appart from the big one below about all
this being a waste of time :-)
I do not see why we should use the alloc fancy pointer type in our
_Hashtable implementation
This patch adds support for both conditional and unconditional unpacked
ASRD. This meant adding a new define_insn for the unconditional form,
instead of reusing the conditional instructions. It also meant
extending the current conditional patterns to support merging with
any independent value,
This patch adds support for unpacked conditional BIC. The type suffix
could be taken from the element size or the container size, so the
patch continues to use the element size. This is consistent with
the existing support for unconditional BIC.
Tested on aarch64-linux-gnu and aarch64_be-elf.
This patch extends the SMULH and UMULH support to unpacked vectors.
The type suffix must be taken from the element size rather than the
container size.
The main use of these patterns is to support division and modulus
by a constant. The conditional forms would be hard to trigger from
non-ACLE
This patch adds support for unpacked SVE SABD and UABD.
It also rewrites the patterns so that they match as combine
patterns without the need for REG_EQUAL notes. Finally,
there was no pattern for merging with the second input,
which can be handled by reversing the operands.
The type suffix
This patch extends the ADR patterns to handle unpacked vectors.
They would work with both elements and containers, but since
the instructions only support .s and .d, we get more coverage
by using containers.
Tested on aarch64-linux-gnu and aarch64_be-elf. Pushed to trunk.
Richard
gcc/
This patch adds support for conditional binary ADD, SUB, MUL, SMAX,
UMAX, SMIN, UMIN, LSL, LSR, ASR, AND, ORR and EOR. It's not really
possible to split it up further given how the patterns are written.
Min, max and right-shift need the element size rather than the container
size. The others
This patch makes the SVE_INT_BINARY_IMM patterns support
unpacked arithmetic, covering MUL, SMAX, SMIN, UMAX and UMIN.
For min and max, the type suffix must be taken from the element
size rather than the container size.
The XFAILs are due to PR98602.
Tested on aarch64-linux-gnu and
This patch adds support for unpacked SVE LSL, ASR and LSR.
For right shifts, the type suffix needs to be taken from the
element size rather than the container size.
Tested on aarch64-linux-gnu and aarch64_be-elf. Pushed to trunk.
Richard
gcc/
* config/aarch64/aarch64-sve.md (3)
On 1/11/21 5:55 PM, Jan Hubicka wrote:
The patch is about not using delete for a memory that
is allocated by GGC.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
gcc/ChangeLog:
PR jit/98615
* symtab-clones.h
On 1/11/21 9:30 AM, Matthias Klose wrote:
On 1/10/21 10:18 PM, Martin Sebor wrote:
On 1/10/21 3:29 AM, Matthias Klose wrote:
is the newline intended? It's followed by a debug_rtx call.
To avoid the warning there shouldn't be any trailing punctuation
or whitespace in the message. The GCC
Could this commit be backported to the gcc-10 branch? We would like to
switch the default name mangling scheme in Rust but are currently
blocked on upstream tool support.
commit 84096498a7bd788599d4a7ca63543fc7c297645e
libiberty: Support the new ("v0") mangling scheme in rust-demangle
AR from older binutils doesn't work with --plugin and rc:
[hjl@gnu-cfl-2 bin]$ touch foo.c
[hjl@gnu-cfl-2 bin]$ ar --plugin
/usr/libexec/gcc/x86_64-redhat-linux/10/liblto_plugin.so rc libfoo.a foo.c
[hjl@gnu-cfl-2 bin]$ ./ar --plugin
/usr/libexec/gcc/x86_64-redhat-linux/10/liblto_plugin.so rc
Check if AR works with --plugin and rc before passing --plugin to AR and
RANLIB.
bfd/
PR ld/27173
binutils/
PR ld/27173
* configure: Regenerated.
gas/
PR ld/27173
* configure: Regenerated.
gprof/
PR ld/27173
* configure: Regenerated.
AR from older binutils doesn't work with --plugin and rc:
[hjl@gnu-cfl-2 bin]$ touch foo.c
[hjl@gnu-cfl-2 bin]$ ar --plugin
/usr/libexec/gcc/x86_64-redhat-linux/10/liblto_plugin.so rc libfoo.a foo.c
[hjl@gnu-cfl-2 bin]$ ./ar --plugin
/usr/libexec/gcc/x86_64-redhat-linux/10/liblto_plugin.so rc
> The patch is about not using delete for a memory that
> is allocated by GGC.
>
> Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
>
> Ready to be installed?
> Thanks,
> Martin
>
> gcc/ChangeLog:
>
> PR jit/98615
> * symtab-clones.h (clone_info::release):
The patch is about not using delete for a memory that
is allocated by GGC.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
gcc/ChangeLog:
PR jit/98615
* symtab-clones.h (clone_info::release): Release
> From: Luis Machado via Gcc-patches
> Date: Mon, 11 Jan 2021 15:58:43 +0100
> This seems to have broken the builds on AArch64-Linux Ubuntu 18.04.
>
> make[2]: Entering directory 'binutils-gdb-master-bionic/libiberty'
> rm -f ./libiberty.a pic/./libiberty.a noasan/./libiberty.a
> ar --plugin
Excerpts from Iain Buclaw's message of January 7, 2021 6:48 pm:
> Excerpts from Rainer Orth's message of January 7, 2021 5:17 pm:
>> Hi Iain,
>>
The Solaris assemblers don't support UTF-8 identifiers. Unless gdc can
encode them in some way for toolchains like this (no idea if this is
On Mon, 11 Jan 2021 at 17:18, Daniel Engel wrote:
>
> On Mon, Jan 11, 2021, at 8:07 AM, Christophe Lyon wrote:
> > On Sat, 9 Jan 2021 at 14:09, Christophe Lyon
> > wrote:
> > >
> > > On Sat, 9 Jan 2021 at 13:27, Daniel Engel wrote:
> > > >
> > > > On Thu, Jan 7, 2021, at 4:56 AM, Richard
On 11/01/2021 11:10, g...@danielengel.com wrote:
> From: Daniel Engel
>
> This version combines __ctzdi2() with __ctzsi2() into a single object with
> an efficient tail call. The former implementation of __ctzdi2() was in C.
>
> On architectures without a clz instruction, this version merges
A general comment before we start:
CLZ was added to the Arm ISA in Armv5. So all subsequent Arm versions
(and all versions implementing thumb2) will have this instruction. So
the only cases where you'll need a fallback are armv6m (and derivatives)
and pre-armv5 (Arm or thumb1). So there's no
On 1/10/21 10:18 PM, Martin Sebor wrote:
> On 1/10/21 3:29 AM, Matthias Klose wrote:
>> is the newline intended? It's followed by a debug_rtx call.
>
> To avoid the warning there shouldn't be any trailing punctuation
> or whitespace in the message. The GCC quoting directives should
> be
On Mon, Jan 11, 2021, at 8:07 AM, Christophe Lyon wrote:
> On Sat, 9 Jan 2021 at 14:09, Christophe Lyon
> wrote:
> >
> > On Sat, 9 Jan 2021 at 13:27, Daniel Engel wrote:
> > >
> > > On Thu, Jan 7, 2021, at 4:56 AM, Richard Earnshaw wrote:
> > > > On 07/01/2021 00:59, Daniel Engel wrote:
> > > >
On Sat, 9 Jan 2021 at 14:09, Christophe Lyon wrote:
>
> On Sat, 9 Jan 2021 at 13:27, Daniel Engel wrote:
> >
> > On Thu, Jan 7, 2021, at 4:56 AM, Richard Earnshaw wrote:
> > > On 07/01/2021 00:59, Daniel Engel wrote:
> > > > --snip--
> > > >
> > > > On Wed, Jan 6, 2021, at 9:05 AM, Richard
On Mon, Jan 11, 2021, at 7:39 AM, Richard Earnshaw wrote:
> On 11/01/2021 15:26, Richard Earnshaw wrote:
> > On 11/01/2021 11:10, g...@danielengel.com wrote:
> >> From: Daniel Engel
> >>
> >> gcc/libgcc/ChangeLog:
> >> 2021-01-07 Daniel Engel
> >>
> >>* config/arm/lib1funcs.S: Move
>> Hi David, Clement,
>>
>>> The patch is local to libstdc++ AIX support, so I believe that I can
>>> approve it.
>>
>>have you considered merging the dragonfly and aix trees? I'm asking
>>because it seems prudent to try and avoid creating more and more
>>almost-but-not-quite-similar
On 11/01/21 16:35 +0100, Rainer Orth wrote:
Hi David, Clement,
The patch is local to libstdc++ AIX support, so I believe that I can approve it.
have you considered merging the dragonfly and aix trees? I'm asking
because it seems prudent to try and avoid creating more and more
On 11/01/2021 15:26, Richard Earnshaw wrote:
> On 11/01/2021 11:10, g...@danielengel.com wrote:
>> From: Daniel Engel
>>
>> gcc/libgcc/ChangeLog:
>> 2021-01-07 Daniel Engel
>>
>> * config/arm/lib1funcs.S: Move __clzsi2() and __clzdi2() to
>> * config/arm/bits/clz2.S: New file.
>
> No,
Hi Clement,
> Would "Skip if target = aix" be ok, especially for the few that I have
> absolutely no idea why they are failing ?
you should try to avoid such skips if at all possible. Should the tests
start to PASS in the future, either due to AIX changes or changes in
libstdc++, this would go
Hi David, Clement,
> The patch is local to libstdc++ AIX support, so I believe that I can approve
> it.
have you considered merging the dragonfly and aix trees? I'm asking
because it seems prudent to try and avoid creating more and more
almost-but-not-quite-similar configurations (Solaris
On 11/01/2021 11:10, g...@danielengel.com wrote:
> From: Daniel Engel
>
> gcc/libgcc/ChangeLog:
> 2021-01-07 Daniel Engel
>
> * config/arm/lib1funcs.S: Move __ashldi3(), __ashrdi3(), __lshldi3() to
> * config/arm/bits/shift.S: New file.
Again, drop the bits/
R.
> ---
>
On 11/01/2021 11:10, g...@danielengel.com wrote:
> From: Daniel Engel
>
> gcc/libgcc/ChangeLog:
> 2021-01-07 Daniel Engel
>
> * config/arm/lib1funcs.S: Move __ctzsi2() to
> * config/arm/bits/ctz2.S: New file.
Similar to the previous patch, just use ctz.S without the bits subdir.
On 11/01/2021 11:10, g...@danielengel.com wrote:
> From: Daniel Engel
>
> gcc/libgcc/ChangeLog:
> 2021-01-07 Daniel Engel
>
> * config/arm/lib1funcs.S: Move __clzsi2() and __clzdi2() to
> * config/arm/bits/clz2.S: New file.
No, please don't push these down into a subdirectory.
Hi David,
Alright, I'll see what I can do.
Would "Skip if target = aix" be ok, especially for the few that I have
absolutely no idea why they are failing ?
Clément
From: David Edelsohn
Sent: Monday, January 11, 2021 4:03 PM
To: CHIGOT, CLEMENT
Cc:
Some initial comments.
On 11/01/2021 11:10, g...@danielengel.com wrote:
> From: Daniel Engel
>
> These definitions facilitate subsequent patches in this series.
>
> gcc/libgcc/ChangeLog:
> 2021-01-07 Daniel Engel
>
> * config/arm/t-elf: Organize functions into logical groups.
> *
Hi, Clement
The patch is local to libstdc++ AIX support, so I believe that I can approve it.
libstdc++ loads the gcc testsuite target-supports.exp extensions. The
patch needs to update the libstdc++ testcases to require wchar32 or
utf-8 as appropriate to avoid the failures.
Thanks, David
On
Hello
Thanks for the review. Due to the Christmas holidays I have not finished
addressing all these issues yet, but I expect to be done by the end of this
week. Can this patch still make it for GCC 10, as I believe stage 4 is starting
soon?
Thanks
Kwok
On 10/12/2020 2:38 pm, Jakub Jelinek
This seems to have broken the builds on AArch64-Linux Ubuntu 18.04.
make[2]: Entering directory 'binutils-gdb-master-bionic/libiberty'
rm -f ./libiberty.a pic/./libiberty.a noasan/./libiberty.a
ar --plugin /usr/lib/gcc/aarch64-linux-gnu/7/liblto_plugin.so rc
./libiberty.a \
./regex.o
On 1/9/21 7:52 PM, Matthias Klose wrote:
> The attached patch makes the link targets a little bit more verbose. Ok to
> commit?
approved by Jakub on IRC, checked in.
> It shows that --enable-link-serialization=1 doesn't work:
>
> $ grep ^Linking ../log
> Linking gnat1 |==-- |
On 09/01/2021 18:04, Daniel Engel wrote:
> On Sat, Jan 9, 2021, at 5:09 AM, Christophe Lyon wrote:
>> On Sat, 9 Jan 2021 at 13:27, Daniel Engel wrote:
>>>
>>> On Thu, Jan 7, 2021, at 4:56 AM, Richard Earnshaw wrote:
On 07/01/2021 00:59, Daniel Engel wrote:
> --snip--
>
> On Wed,
On Mon, 11 Jan 2021, Tamar Christina wrote:
> Hi Richi,
>
> Attached is the updated patch.
>
> Note that testcases for all of these will be committed with the patch but I'm
> Finishing up the 32-bit Arm changes to mirror the changes the AArch64
> maintainer
> wanted and then have to do
I'm suggesting postponing this to GCC 12 as I'm planning a bigger
target/optimize attribute (pragma) overhaul.
Martin
On 12/7/20 12:03 PM, Martin Liška wrote:
PING^2
On 11/26/20 2:56 PM, Martin Liška wrote:
PING^1
On 11/9/20 11:35 AM, Martin Liška wrote:
On 11/3/20 2:34 PM, Jakub Jelinek
On 1/11/21 5:38 AM, Jakub Jelinek wrote:
On Mon, Jan 11, 2021 at 11:26:42AM +0100, Martin Liška wrote:
Problem here was that GCC-related options are not applied in stage2 (and later
stages).
It's caused by fact that CXX is xg++ in stage2 (and later stages).
Fixed with the following patch.
The vectorizer, for large permuted grouped loads, generates
inefficient intermediate code (cleaned up only later) that runs
into complexity issues in SCEV analysis and elsewhere. For the
non-single-element interleaving case we already put a hard limit
in place, this applies the same limit to the
"Qian, Jianhua" writes:
> Hi Richard
>
>> -Original Message-
>> From: Richard Sandiford
>> Sent: Friday, January 8, 2021 7:04 PM
>> To: Qian, Jianhua/钱 建华
>> Cc: gcc-patches@gcc.gnu.org
>> Subject: Re: [PATCH v2] aarch64: Add cpu cost tables for A64FX
>>
>> Qian Jianhua writes:
>> >
Hi Bernd,
> On 1/8/21 8:27 PM, David Edelsohn wrote:
>> Hi, Bernd
>>
>> Thanks for investigating this and creating a revised version of the
>> patch. With the second patch, the gcc.misc-test/outputs.exp results
>> are clean on AIX.
>>
>
> Many thanks for confirming that the patch works.
>
> Is
Hi Alexandre,
> On Jan 7, 2021, Bernd Edlinger wrote:
>
>> I don't know why that is there in the first place, as there
>> are no C++ test cases, these files should not be created at all.
>
> collect2, on platforms that use it, create .cdtor files even for C.
> David Edelsohn told me so back
On Mon, 11 Jan 2021, Tamar Christina wrote:
> Hi Richi,
>
> This adds support for complex multiply and complex multiply and accumulate to
> the vect pattern detector.
>
> Bootstrapped Regtested on aarch64-none-linux-gnu, x86_64-pc-linux-gnu
> and no issues.
>
> Ok for master? (I replied to a
On Mon, Jan 11, 2021 at 11:28 AM Hongtao Liu wrote:
>
> Hi:
> d->flag is always 0 for builtins located in
> BDESC_FIRST (comi,COMI,...)
> ...
> BDESC_END (COMI, PCMPESTR)
> So the condition can be deleted, so did BUILTIN_DESC_SWAP_OPERANDS.
>
> Regtested and bootstrapped on
This fixes a double-counting in the reduction cost when vectorizing
the reduction through the regular vectorizable_* functions.
Bootstrapped / tested on x86_64-unknown-linux-gnu, pushed.
2021-01-11 Richard Biener
PR tree-optimization/98526
* tree-vect-loop.c
Excerpts from Andreas Schwab's message of January 10, 2021 9:08 am:
> ../../gcc/d/dmd/parse.c: In member function 'Dsymbols*
> Parser::parseDeclDefs(int, Dsymbol**, PrefixAttributes*)':
> ../../gcc/d/dmd/parse.c:647:29: error: unused variable 'e'
> [-Werror=unused-variable]
> 647 |
Hi,
This patch merges the D front-end implementation with upstream dmd
2d3d13748, removing the visibility and lookup deprecation.
The deprecation phase for access checks is finished.
The `-ftransition=import` and `-ftransition=checkimports` switches no
longer have an effect and are now removed.
On 1/8/21 8:27 PM, David Edelsohn wrote:
> Hi, Bernd
>
> Thanks for investigating this and creating a revised version of the
> patch. With the second patch, the gcc.misc-test/outputs.exp results
> are clean on AIX.
>
Many thanks for confirming that the patch works.
Is it OK to push?
Thanks
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
* config/arm/lib1funcs.S: Move __ashldi3(), __ashrdi3(), __lshldi3() to
* config/arm/bits/shift.S: New file.
---
libgcc/config/arm/bits/shift.S | 102
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
* config/arm/bpabi.c: Deleted unused file.
* config/arm/eabi/ldiv.S: Replaced the __aeabi_ldivmod() and
__aeabi_uldivmod() wrapper functions with a full implementation.
* config/arm/t-bpabi:
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-08 Daniel Engel
* config/arm/eabi/fdiv.S: New file for __divsf3().
* config/arm/lib1funcs.S: #include eabi/fdiv.S (v6m only).
* config/arm/t-elf: Add _divsf3 and _fp_divloopf3 to LIB1ASMFUNCS.
---
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-08 Daniel Engel
* config/arm/bpabi-lib.h: Remove obsolete RENAME_LIBRARY directives.
* config/arm/eabi/ffixed.S: New file for __float[un]sisf/disf().
* config/arm/lib1funcs.S: #include eabi/ffixed.S (v6m only).
*
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-08 Daniel Engel
* config/arm/eabi/fmul.S: New file for __mulsf3().
* config/arm/lib1funcs.S: #include eabi/fmul.S (v6m only).
* config/arm/t-elf: Move _mulsf3 to global scope in LIB1ASMFUNCS
(this object was
From: Daniel Engel
Since this is the first phase of the floating point functions, some
common parsing and formatting routines and also included. These common
routines will be referenced by other functions in subsequent commits.
However, even if the size penalty is accounted entirely to
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-08 Daniel Engel
* config/arm/bpabi-lib.h: Remove obsolete RENAME_LIBRARY directives.
* config/arm/eabi/ffloat.S: New file for __float[un]sisf/disf().
* config/arm/lib1funcs.S: #include eabi/ffloat.S (v6m only).
*
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-08 Daniel Engel
* config/arm/eabi/fcast.S: New file for __aeabi_f2d/__extendsfdf2()
__aeabi_d2f(), __truncdfsf2().
* config/arm/lib1funcs.S: #include eabi/fcast.S (v6m only).
* config/arm/t-elf: Add _arm_d2f and
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-09 Daniel Engel
* config/arm/eabi/fcast.S: Add __aeabi_f2h*() and __aeabi_h2f*().
* config/arm/fp16: Disable duplicate C routines (v6m only).
* config/arm/t-bpabi: Add _arm_f2h* and _arm_h2f* objects to
LIB1ASMFUNCS.
---
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
* config/arm/eabi/lmul.S: New file for __muldi3(), __mulsidi3(), and
__umulsidi3().
* config/arm/lib1funcs.S: #eabi/lmul.S (v6m only).
* config/arm/t-elf: Add the new objects to LIB1ASMFUNCS.
---
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
* config/arm/bpabi-v6m.S: Moved __aeabi_frsub() to
* config/arm/eabi/fadd.S: New file.
* config/arm/lib1funcs.S: #include eabi/fadd.S (v6m only).
---
libgcc/config/arm/bpabi-v6m.S | 16
From: Daniel Engel
These functions are significantly smaller and faster than the wrapper
functions and soft-float implementation they replace. Using the first
comparison operator (e.g. '<=') costs about 70 bytes initially, but every
additional operator incrementally adds just 4 bytes.
NOTE: It
From: Daniel Engel
gcc/libgcc/ChangeLog:
2021-01-07 Daniel Engel
* config/arm/bpabi-v6m.S: Moved __aeabi_fcmp*() functions to
* config/arm/eabi/fcmp.S: New file.
* config/arm/lib1funcs.S: #include eabi/fcmp.S (v6m only).
---
libgcc/config/arm/bpabi-v6m.S | 63
From: Daniel Engel
In Thumb mode, the new functions are each 1-2 instructions smaller
and faster, and branchless when the IT instruction is available.
The ARM versions were converted to the "xxl/xxh" big-endian register
convention, but are otherwise unchanged.
gcc/libgcc/ChangeLog:
2021-01-07
1 - 100 of 133 matches
Mail list logo
