There's a typo in the testcase, I've committed the patch as an obvious fix.
Fix typo in testcase.
gcc/testsuite/ChangeLog:
* gcc.target/i386/avx-pr94680.c: Fix typo in testcase.
diff --git a/gcc/testsuite/gcc.target/i386/avx-pr94680.c
On Linux/x86_64,
94de7e225c1fda079052c3f0725c926437d56c94 is the first bad commit
commit 94de7e225c1fda079052c3f0725c926437d56c94
Author: liuhongt
Date: Thu Apr 22 15:33:16 2021 +0800
Optimize __builtin_shuffle when it's used to zero the upper bits of the
dest. [PR target/94680]
caused
A logic bug in the handling of PHI arguments in compute_objsize
that are all null pointers lets an incompletely populated struct
be used in a way that triggers an assertion causing an ICE.
The attached patch corrects that by having compute_objsize fail
when the struct isn't fully populated (when
On 2021/4/30 14:32, Xionghu Luo wrote:
The vsel instruction is a bit-wise select instruction. Using an
IF_THEN_ELSE to express it in RTL is wrong and leads to wrong code
being generated in the combine pass. Per element selection is a
subset of per bit-wise selection,with the patch the
On Wed, May 12, 2021 at 10:19 PM Jakub Jelinek wrote:
>
> On Sun, Apr 25, 2021 at 02:57:08PM +0800, Hongtao Liu via Gcc-patches wrote:
> > gcc/ChangeLog:
> >
> > PR target/94680
> > * config/i386/sse.md (ssedoublevecmode): Add attribute for
> > V64QI/V32HI/V16SI/V4DI.
> >
On Wed, May 12, 2021 at 8:38 PM Uros Bizjak wrote:
>
> On Wed, May 12, 2021 at 1:42 PM Hongtao Liu wrote:
> >
> > On Wed, May 12, 2021 at 4:36 PM Uros Bizjak wrote:
> > >
> > > On Tue, Apr 27, 2021 at 1:05 PM Hongtao Liu via Gcc-patches
> > > wrote:
> > > >
> > > > Hi:
> > > > As described
On 5/12/21 8:03 PM, Marek Polacek wrote:
+ // FIXME: Add dg error when PR100339 is resolved.
+ //[[deprecated]] friend void ::foo(int, int);
This could be an xfailed dg-error.
On 5/12/21 8:03 PM, Marek Polacek wrote:
On Wed, May 12, 2021 at 12:21:26PM -0400, Jason Merrill wrote:
On 5/12/21 11:03 AM, Marek Polacek wrote:
On Wed, May 12, 2021 at 10:37:50AM -0400, Jason Merrill wrote:
On 5/11/21 10:45 PM, Marek Polacek wrote:
This patch implements
On Wed, 12 May 2021, Martin Sebor via Gcc-patches wrote:
> An erroneous VLA parameter bound in a function redeclaration causes
> an ICE as the -Wvla-parameter is trying to format a diagnostic after
> an error has already been issued. The attached change prevents
> such parameters from being
An erroneous VLA parameter bound in a function redeclaration causes
an ICE as the -Wvla-parameter is trying to format a diagnostic after
an error has already been issued. The attached change prevents
such parameters from being considered.
This patch is for trunk and GCC 11.
Martin
PR c/100550
On Wed, May 12, 2021 at 12:21:26PM -0400, Jason Merrill wrote:
> On 5/12/21 11:03 AM, Marek Polacek wrote:
> > On Wed, May 12, 2021 at 10:37:50AM -0400, Jason Merrill wrote:
> > > On 5/11/21 10:45 PM, Marek Polacek wrote:
> > > > This patch implements [dcl.attr.grammar]/5: "If an
> > > >
On Sun, Apr 25, 2021 at 08:50:18PM -0500, Bill Schmidt wrote:
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/powerpc/rop-1.c
> @@ -0,0 +1,16 @@
> +/* { dg-do compile } */
> +/* { dg-options "-O2 -mdejagnu-cpu=power10 -mrop-protect" } */
This should only run on ELFv2 currently, no?
> +/* {
On Mon, Apr 26, 2021 at 02:27:45PM -0500, Bill Schmidt wrote:
> On 4/26/21 11:04 AM, will schmidt wrote:
> >Does there need to be another test to verify if -mrop-protect is on by
> >default without specifying -mrop-protect? (or is it?) Question on
> >0/4.
>
> It's off by default (see the
On Sun, Apr 25, 2021 at 08:50:17PM -0500, Bill Schmidt wrote:
> 2021-03-25 Bill Schmidt
>
> gcc/
> * config/rs6000/rs6000-c.c (rs6000_target_modify_macros): Define
> __ROP_PROTECT__ if -mrop-protect is selected.
Okay for trunk and 11. Thanks!
Segher
On Mon, Apr 26, 2021 at 11:03:22AM -0500, will schmidt wrote:
> On Sun, 2021-04-25 at 20:50 -0500, Bill Schmidt via Gcc-patches wrote:
> > @@ -602,6 +602,9 @@ rs6000_target_modify_macros (bool define_p,
> > HOST_WIDE_INT flags,
> >/* Whether pc-relative code is being generated. */
> >if
Hi!
On Sun, Apr 25, 2021 at 08:50:16PM -0500, Bill Schmidt wrote:
> Insert the hashst and hashchk instructions when -mrop-protect has been
> selected. The encrypted save slot for ROP mitigation is placed
> between the parameter save area and the alloca space (if any;
> otherwise the local
[No changes from V2]
To support multiple debug formats, we need to move away from explicit
enumeration of each individual combination of debug formats.
gcc/c-family/ChangeLog:
* c-opts.c (c_common_post_options): Adjust access to debug_type_names.
* c-pch.c (struct
[Changes from V2]
- Tested build (make all-gcc) of cross compiler for target triplets
containing c6x/mips/powerpc and darwin/cygwin.
[End of changes from V2]
This patch introduces a dwarf_debuginfo_p predicate that abstracts and
replaces complex checks on write_symbols.
[Changes across versions]
[What's new in V2]
- (Addressed Richard's comments)
- For patch 1/2 [opts: change write_symbols to support bitmasks], use
debug_set_names more uniformly. Reworded the diagnostics in c-family/c-opts.c
and c-family/c-pch.c as there can be multiple debug formats.
A recent Glibc enhancement to add attribute access to a bunch of
APIs has triggered a couple of false positives in the test suite
when using GCC 10. The problem has been fixed on trunk as part
of a bigger commit that's not suitable for backporting (it adds
new warnings). The attached change
On 5/12/21 12:31 AM, Richard Biener wrote:
Otherwise looks OK. Did you check for write_symbols uses in FEs and targets?
Richard.
Yes, I have. I must admit I have gone back and forth in my mind on this.
My initial thinking was to adjust only those checks where I expect more
than 1
On Wed, May 12, 2021 at 05:01:00PM -0400, Aldy Hernandez via Gcc-patches wrote:
>
> PR c/100521
> * gimple-range.cc (range_of_builtin_call): Skip out on
> processing __builtin_clz when varying.
> ---
> gcc/gimple-range.cc | 2 +-
> gcc/testsuite/gcc.dg/pr100521.c
The previous changes to irange::constant_p return TRUE for
VARYING, since VARYING has numerical end points like any other
constant range. The problem is that some users of constant_p
depended on constant_p excluding the full domain. The
range handler for __builtin_clz, that is shared between
On Mon, Apr 26, 2021 at 11:02:53AM -0500, will schmidt wrote:
> On Sun, 2021-04-25 at 20:50 -0500, Bill Schmidt via Gcc-patches wrote:
> > + /* If we are inserting ROP-protect instructions, disable shrink wrap. */
> > + if (rs6000_rop_protect)
> > +{
> > + flag_shrink_wrap = 0;
> > +
On Sun, Apr 25, 2021 at 08:50:15PM -0500, Bill Schmidt wrote:
> + /* If we are inserting ROP-protect instructions, disable shrink wrap. */
> + if (rs6000_rop_protect)
> +{
> + flag_shrink_wrap = 0;
> + flag_shrink_wrap_separate = 0;
> +}
Separate shrink-wrapping requires
On Linux/x86_64,
7d7d925d0d799eb9dda4a6bd6162af0a7db9c98b is the first bad commit
commit 7d7d925d0d799eb9dda4a6bd6162af0a7db9c98b
Author: Martin Liska
Date: Thu Apr 22 16:27:19 2021 +0200
Print warning diagnostics for -flto issues.
caused
FAIL: gcc.dg/atomic/c11-atomic-exec-2.c -O2
On Linux/x86_64,
b1f7fd8a2a5558da1e101de11bb1cdba081ce010 is the first bad commit
commit b1f7fd8a2a5558da1e101de11bb1cdba081ce010
Author: Uros Bizjak
Date: Wed May 12 08:11:18 2021 +0200
i386: Implement FP vector compares for V2SFmode [PR98218]
caused
FAIL: g++.target/i386/pr98218-1.C
Am 12.05.2021 um 19:37 schrieb Marcel Vollweiler:
-
Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201, 80634 München
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung,
Frank Thürauf
-
Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201,
On 5/7/21 4:21 AM, Richard Biener via Gcc-patches wrote:
On Fri, May 7, 2021 at 12:17 PM Richard Biener wrote:
canonicalize_constructor_val was setting TREE_ADDRESSABLE on bases
of ADDR_EXPRs but that's futile when we're dealing with CTOR values
in debug stmts. This rips out the code which
-
Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201, 80634 München
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Frank
Thürauf
commit 8982a5354d2697eeb12a77d49b6730db90053618
Author: Marcel Vollweiler
Date: Wed May 12 10:14:41 2021 -0700
On 5/12/21 12:21 PM, Marek Polacek wrote:
This warning is of questionable value when it's emitted when
instantiating a template, as in the following testcase. It could be
silenced by writing hb(i) << 1 instead of 2 * hb(i) but that's
unnecessary obfuscation.
Bootstrapped/regtested on
On 5/12/21 5:50 PM, Tobias Burnus wrote:
> Hi,
>
> On 12.05.21 16:10, Tom de Vries wrote:
>> Add nvptx option -mptx that sets the ptx ISA version. This is currently
>> hardcoded to 3.1.
>> Tested libgomp on x86_64-linux with nvptx accelerator, both with
>> default set to
>> 3.1 and 6.3.
>> Any
On 5/12/21 11:03 AM, Marek Polacek wrote:
On Wed, May 12, 2021 at 10:37:50AM -0400, Jason Merrill wrote:
On 5/11/21 10:45 PM, Marek Polacek wrote:
This patch implements [dcl.attr.grammar]/5: "If an attribute-specifier-seq
appertains to a friend declaration ([class.friend]), that declaration
This warning is of questionable value when it's emitted when
instantiating a template, as in the following testcase. It could be
silenced by writing hb(i) << 1 instead of 2 * hb(i) but that's
unnecessary obfuscation.
Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
gcc/cp/ChangeLog:
On 12/05/21 18:51 +0300, Antony Polukhin via Libstdc++ wrote:
ср, 12 мая 2021 г. в 18:38, Antony Polukhin :
ср, 12 мая 2021 г. в 17:44, Jonathan Wakely :
>
> On 12/05/21 12:58 +0300, Antony Polukhin wrote:
> >ср, 12 мая 2021 г. в 12:18, Jonathan Wakely :
> ><...>
> >> Or just leave it
ср, 12 мая 2021 г. в 18:38, Antony Polukhin :
>
> ср, 12 мая 2021 г. в 17:44, Jonathan Wakely :
> >
> > On 12/05/21 12:58 +0300, Antony Polukhin wrote:
> > >ср, 12 мая 2021 г. в 12:18, Jonathan Wakely :
> > ><...>
> > >> Or just leave it undefined, as libc++ seems to do according to your
> > >>
Hi,
On 12.05.21 16:10, Tom de Vries wrote:
Add nvptx option -mptx that sets the ptx ISA version. This is currently
hardcoded to 3.1.
Tested libgomp on x86_64-linux with nvptx accelerator, both with default set to
3.1 and 6.3.
Any comments?
:-)
ISA 3.1 = CUDA 5 (supporting sm_10 to
ср, 12 мая 2021 г. в 17:44, Jonathan Wakely :
>
> On 12/05/21 12:58 +0300, Antony Polukhin wrote:
> >ср, 12 мая 2021 г. в 12:18, Jonathan Wakely :
> ><...>
> >> Or just leave it undefined, as libc++ seems to do according to your
> >> comment in PR 89728:
> >>
> >> error: implicit instantiation of
Tested configure runs on NetBSD x86-64, Linux on x86-64 (with target s390x), and
on s390x.
Is the patch ok for master, and for gcc-11?
8<--8<--8<-
Fix a bootstrap error observed on NetBSD.
2021-05-12 Marius Hillenbrand
gcc/ChangeLog:
PR bootstrap/100552
*
On 5/12/21 2:47 AM, Richard Biener wrote:
This rejects a number of vector components that does not fit an 'int'
which is an internal limitation of RTVEC.
This regresses gcc.dg/attr-vector_size.c which checks for much larger
supported vectors. Not sure what to do about this - I'll also note
the
On Wed, May 12, 2021 at 10:37:50AM -0400, Jason Merrill wrote:
> On 5/11/21 10:45 PM, Marek Polacek wrote:
> > This patch implements [dcl.attr.grammar]/5: "If an attribute-specifier-seq
> > appertains to a friend declaration ([class.friend]), that declaration shall
> > be a definition."
> >
> >
On 12/05/2021 12:05, Prathamesh Kulkarni via Gcc-patches wrote:
> On Wed, 12 May 2021 at 16:02, Richard Earnshaw
> wrote:
>>
>>
>>
>> On 12/05/2021 08:46, Prathamesh Kulkarni via Gcc-patches wrote:
>>> On Mon, 10 May 2021 at 19:55, Richard Earnshaw
>>> wrote:
On 06/05/2021
This amends the fix for PR100053 where I failed to amend all edge
tests in dominated_by_p_w_unex.
Bootstrapped and tested on x86_64-unknown-linux-gnu, pushed to trunk
sofar.
2021-05-12 Richard Biener
PR tree-optimization/100566
* tree-ssa-sccvn.c (dominated_by_p_w_unex):
On Tue, May 11, 2021 at 11:54:58PM +0200, Tobias Burnus wrote:
> The sfield / firstprivate lookup used the wrong var decl
> for the lookup – hence it failed.
> I used an extra long diff to make it easier to follow why
> 'c' and not 'detach_clause' has the proper clause for the
> decl to be used as
On Wed, May 12, 2021 at 04:36:00PM +0200, Marcel Vollweiler wrote:
> + c_parser_error (parser, "%<#pragma omp target%> with map-type-"
> + "modifier other than % or %"
Please just write modifier rather than map-type-modifier
> + cp_parser_error (parser,
On 12/05/21 12:58 +0300, Antony Polukhin wrote:
ср, 12 мая 2021 г. в 12:18, Jonathan Wakely :
<...>
Or just leave it undefined, as libc++ seems to do according to your
comment in PR 89728:
error: implicit instantiation of undefined template
'std::__1::ctype >'
Was your aim to have a
On 4/30/21 2:32 AM, Andreas Krebbel via Gcc-patches wrote:
The problem appears to be triggered by two locations in the front-end
where non-POINTER_SIZE pointers aren't handled right now.
1. An assertion in strip_typedefs is triggered because the alignment
of the types don't match. This in turn
On 5/11/21 10:45 PM, Marek Polacek wrote:
This patch implements [dcl.attr.grammar]/5: "If an attribute-specifier-seq
appertains to a friend declaration ([class.friend]), that declaration shall
be a definition."
This restriction only applies to C++11-style attributes. There are
various forms of
Am 11.05.2021 um 17:20 schrieb Jakub Jelinek:
On Tue, May 11, 2021 at 04:27:55PM +0200, Marcel Vollweiler wrote:
The usual wording would be
"too many % modifiers"
Changed for 'always' and 'close' for C and C++.
One extra thing, sorry, forgot to mention, for the translators it might be
On Wed, May 12, 2021 at 3:06 PM Jakub Jelinek wrote:
>
> On Fri, Apr 23, 2021 at 03:46:59PM +0200, Jakub Jelinek via Gcc-patches wrote:
> > On Tue, Feb 09, 2021 at 12:12:24PM +0100, Jakub Jelinek via Gcc-patches
> > wrote:
> > > As mentioned in the PR, we don't support arithmetic right V2DImode
On Sun, Apr 25, 2021 at 02:57:08PM +0800, Hongtao Liu via Gcc-patches wrote:
> gcc/ChangeLog:
>
> PR target/94680
> * config/i386/sse.md (ssedoublevecmode): Add attribute for
> V64QI/V32HI/V16SI/V4DI.
> (ssehalfvecmode): Add attribute for V2DI/V2DF.
>
Hi,
Add nvptx option -mptx that sets the ptx ISA version. This is currently
hardcoded to 3.1.
Tested libgomp on x86_64-linux with nvptx accelerator, both with default set to
3.1 and 6.3.
Any comments?
Thanks,
- Tom
[nvptx] Add -mptx=3.1/6.3
gcc/ChangeLog:
2021-05-12 Tom de Vries
libstdc++-v3/ChangeLog:
* testsuite/25_algorithms/pstl/alg_nonmodifying/find_end.cc:
Increase dg-timeout-factor to 4. Fix -Wunused-parameter
warnings. Replace bitwise AND with logical AND in loop
condition.
*
On Fri, Apr 23, 2021 at 03:46:59PM +0200, Jakub Jelinek via Gcc-patches wrote:
> On Tue, Feb 09, 2021 at 12:12:24PM +0100, Jakub Jelinek via Gcc-patches wrote:
> > As mentioned in the PR, we don't support arithmetic right V2DImode or
> > V4DImode on x86 without -mavx512vl or -mxop. The ISAs
On Wed, May 12, 2021 at 1:42 PM Hongtao Liu wrote:
>
> On Wed, May 12, 2021 at 4:36 PM Uros Bizjak wrote:
> >
> > On Tue, Apr 27, 2021 at 1:05 PM Hongtao Liu via Gcc-patches
> > wrote:
> > >
> > > Hi:
> > > As described in the subject line, this patch is about to do the
> > > below
ср, 12 мая 2021 г. в 12:58, Antony Polukhin :
<...>
> Should we care about those users?
Looks like the answer is "yes". There are tests for that in
22_locale/ctype/requirements/explicit_instantiation.cc and
22_locale/money_get/cons/3.cc
Attaching a patch with review fixes (with ctype
On Wed, May 12, 2021 at 2:06 PM H.J. Lu wrote:
>
> On Wed, May 12, 2021 at 2:00 AM Richard Biener
> wrote:
> >
> > On Wed, May 12, 2021 at 1:35 AM H.J. Lu wrote:
> > >
> > > It is only defined for i386 and everyone uses the default:
> > >
> > > #define MAX_BITSIZE_MODE_ANY_INT
On Wed, May 12, 2021 at 2:06 PM Martin Liška wrote:
>
> Silents the following warning:
> lto-wrapper: warning: using serial compilation of 2 LTRANS jobs
>
> Ready to be installed?
OK.
> Thanks,
> Martin
>
> gcc/testsuite/ChangeLog:
>
> * lib/lto.exp: When running tests without
Silents the following warning:
lto-wrapper: warning: using serial compilation of 2 LTRANS jobs
Ready to be installed?
Thanks,
Martin
gcc/testsuite/ChangeLog:
* lib/lto.exp: When running tests without jobserver, one can see
the following warning for tests that use 1to1
On Wed, May 12, 2021 at 2:00 AM Richard Biener
wrote:
>
> On Wed, May 12, 2021 at 1:35 AM H.J. Lu wrote:
> >
> > It is only defined for i386 and everyone uses the default:
> >
> > #define MAX_BITSIZE_MODE_ANY_INT (64*BITS_PER_UNIT)
>
> The default is determined by genmodes.c which looks at the
On Wed, May 12, 2021 at 4:36 PM Uros Bizjak wrote:
>
> On Tue, Apr 27, 2021 at 1:05 PM Hongtao Liu via Gcc-patches
> wrote:
> >
> > Hi:
> > As described in the subject line, this patch is about to do the
> > below transformation.
> >
> > - vpcmpeqd%ymm3, %ymm3, %ymm3
> > -
This splits can_associate_p into checks for SSA defs and checks
for the type so it can be called from is_reassociable_op to
catch cases not catched by the earlier fix.
Bootstrapped and tested on x86_64-unknown-linux-gnu, pushed.
2021-05-11 Richard Biener
PR tree-optimization/100519
On Wed, 12 May 2021 at 16:02, Richard Earnshaw
wrote:
>
>
>
> On 12/05/2021 08:46, Prathamesh Kulkarni via Gcc-patches wrote:
> > On Mon, 10 May 2021 at 19:55, Richard Earnshaw
> > wrote:
> >>
> >>
> >>
> >> On 06/05/2021 01:14, Prathamesh Kulkarni via Gcc-patches wrote:
> >>> Hi,
> >>> The
On 12/05/2021 09:51, Christophe Lyon via Gcc-patches wrote:
arm.h has had this error message since 1997, but it is no longer
needed since option parsing has been improved: -mXXX-endian is handled
via arm.opt and updates the BIG_END mask. So, the last
instance of -mXXX-endian on the command
On 12/05/2021 08:46, Prathamesh Kulkarni via Gcc-patches wrote:
On Mon, 10 May 2021 at 19:55, Richard Earnshaw
wrote:
On 06/05/2021 01:14, Prathamesh Kulkarni via Gcc-patches wrote:
Hi,
The attached patch replaces __builtin_neon_vtst* (a, b) with (a & b) != 0.
Bootstrapped and tested on
ср, 12 мая 2021 г. в 12:18, Jonathan Wakely :
<...>
> Or just leave it undefined, as libc++ seems to do according to your
> comment in PR 89728:
>
> error: implicit instantiation of undefined template
> 'std::__1::ctype >'
>
> Was your aim to have a static_assert that gives a more descriptive
>
Hi,
This is a backport to GCC-10 branch, this patch got applied cleanly on the
branch.
This patch removes several duplicated intrinsic definitions from
arm_mve.h mentioned in PR100419 and also fixes the wrong arguments
in few of intrinsics polymorphic variants.
Ok for GCC-10 branch?
Hi,
This is a backport to GCC-11 branch, this patch got applied cleanly on the
branch.
This patch removes several duplicated intrinsic definitions from
arm_mve.h mentioned in PR100419 and also fixes the wrong arguments
in few of intrinsics polymorphic variants.
Ok for GCC-11 branch?
Hi all,
The various sqdmlal2 and sqdmlsl2 expanders perform almost identical functions
and can be
merged using code iterators and attributes to reduce the code in the MD file.
No behavioural change is expected.
Bootstrapped and tested on aarch64-non-linux-gnu.
Pushing to trunk.
Thanks,
Kyrill
On Wed, May 12, 2021 at 11:10 AM Martin Liška wrote:
>
> May I please ping this Richi?
OK.
Thanks,
Richard.
> Thanks,
> Martin
>
> On 4/22/21 4:30 PM, Martin Liška wrote:
> > On 4/22/21 2:47 PM, Richard Biener wrote:
> >> On Thu, Apr 22, 2021 at 2:21 PM Martin Liška wrote:
> >>>
> >>> On
On 11/05/21 21:00 +0100, Jonathan Wakely wrote:
Alternatively, would it be even simpler to just define a partial
specialization of ctype?
template
class ctype >
{
#if __cplusplus >= 201103L
static_assert(something dependent,
"std::basic_string used as a character
On Wed, May 12, 2021 at 11:08 AM Martin Liška wrote:
>
> On 4/29/21 2:22 PM, Richard Biener wrote:
> > On Wed, Apr 21, 2021 at 11:12 AM Martin Liška wrote:
> >>
> >> Now living in the 21st century, we don't longer need using the following
> >> tuple:
> >> cl_decoded_option **decoded_options,
>
May I please ping this Richi?
Thanks,
Martin
On 4/22/21 4:30 PM, Martin Liška wrote:
On 4/22/21 2:47 PM, Richard Biener wrote:
On Thu, Apr 22, 2021 at 2:21 PM Martin Liška wrote:
On 4/22/21 1:19 PM, Richard Biener wrote:
On Thu, Apr 22, 2021 at 11:02 AM Martin Liška wrote:
On 4/22/21
On 4/29/21 2:22 PM, Richard Biener wrote:
On Wed, Apr 21, 2021 at 11:12 AM Martin Liška wrote:
Now living in the 21st century, we don't longer need using the following tuple:
cl_decoded_option **decoded_options,
unsigned int *decoded_options_count)
but we can rather use a standard (our)
On Wed, May 12, 2021 at 3:44 AM Alexandre Oliva wrote:
>
>
> default_zero_call_used_regs currently requires all potentially zeroed
> registers to offer a move opcode that accepts zero as an operand.
>
> This is not the case e.g. for ARM's r12/ip in Thumb mode, and it was
> not the case of FP
On Wed, May 12, 2021 at 1:35 AM H.J. Lu wrote:
>
> It is only defined for i386 and everyone uses the default:
>
> #define MAX_BITSIZE_MODE_ANY_INT (64*BITS_PER_UNIT)
The default is determined by genmodes.c which looks at the maximum
size of any [partial] integer mode. Since we have
INT_MODE
Pushed as obvious.
Martin
gcc/ChangeLog:
PR bootstrap/100560
* Makefile.in: Remove version.h from linker command line.
---
gcc/Makefile.in | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/gcc/Makefile.in b/gcc/Makefile.in
index 487db220d8c..1b5d3f4696c
arm.h has had this error message since 1997, but it is no longer
needed since option parsing has been improved: -mXXX-endian is handled
via arm.opt and updates the BIG_END mask. So, the last
instance of -mXXX-endian on the command line wins.
Tested on many arm* configurations, with no impact on
This rejects a number of vector components that does not fit an 'int'
which is an internal limitation of RTVEC.
This regresses gcc.dg/attr-vector_size.c which checks for much larger
supported vectors. Not sure what to do about this - I'll also note
the RTVEC limitation is a host specific
This makes the rtvec_alloc argument size_t catching overflow and
truncated arguments (from "invalid" testcases), verifying the
argument against INT_MAX which is the limit set by the int
typed rtvec_def.num_elem member.
Bootstrapped and tested on x86_64-unknown-linux-gnu, pushed.
2021-05-12
Hi.
The patch fixes Dially bump strict.
Pushed to master.
Martin
contrib/ChangeLog:
* gcc-changelog/git_commit.py: Simplify needle lookup.
* gcc-changelog/git_update_version.py: Pass ref_name to
parse_git_revisions.
---
contrib/gcc-changelog/git_commit.py | 2
On Tue, Apr 27, 2021 at 1:05 PM Hongtao Liu via Gcc-patches
wrote:
>
> Hi:
> As described in the subject line, this patch is about to do the
> below transformation.
>
> - vpcmpeqd%ymm3, %ymm3, %ymm3
> - vpandn %ymm3, %ymm2, %ymm2
> - vpblendvb %ymm2, %ymm1,
Ping
On 4/30/21 8:32 AM, Andreas Krebbel via Gcc-patches wrote:
> The problem appears to be triggered by two locations in the front-end
> where non-POINTER_SIZE pointers aren't handled right now.
>
> 1. An assertion in strip_typedefs is triggered because the alignment
> of the types don't match.
> It's about more porting to Python3.
>
> Ready for master?
OK, thanks.
> gcc/ada/ChangeLog:
>
> * doc/share/conf.py: Do not use binary mode.
> Do not use u' literals as Python3 uses unicode by default.
On Wed, 12 May 2021, Jakub Jelinek wrote:
> Hi!
>
> The inliner doesn't remap DEBUG_EXPR_DECLs, so the same decls can appear
> in multiple functions.
> Furthermore, expansion reuses corresponding DEBUG_EXPRs too, so they again
> can be reused in multiple functions.
> Neither of that is a major
On Wed, 12 May 2021, Bernd Edlinger wrote:
> Hi,
>
> this fixes another regression from my previous patch.
>
>
> Bootstrapped and reg-tested on x86_64-pc-linux-gnu.
> Is it OK for trunk?
OK.
Richard.
>
> Thanks
> Bernd.
>
Hi!
The inliner doesn't remap DEBUG_EXPR_DECLs, so the same decls can appear
in multiple functions.
Furthermore, expansion reuses corresponding DEBUG_EXPRs too, so they again
can be reused in multiple functions.
Neither of that is a major problem, DEBUG_EXPRs are just magic value holders
and what
Hi,
this fixes another regression from my previous patch.
Bootstrapped and reg-tested on x86_64-pc-linux-gnu.
Is it OK for trunk?
Thanks
Bernd.
From 62da66525c4b7ac1cfd5cad5b8e690ce928802e5 Mon Sep 17 00:00:00 2001
From: Bernd Edlinger
Date: Tue, 11 May 2021 17:55:18 +0200
Subject: [PATCH]
On Mon, 10 May 2021 at 19:55, Richard Earnshaw
wrote:
>
>
>
> On 06/05/2021 01:14, Prathamesh Kulkarni via Gcc-patches wrote:
> > Hi,
> > The attached patch replaces __builtin_neon_vtst* (a, b) with (a & b) != 0.
> > Bootstrapped and tested on arm-linux-gnueabihf and cross-tested on arm*-*-*.
> >
On Tue, May 11, 2021 at 4:45 PM Indu Bhagat wrote:
>
> On 5/10/21 6:11 AM, Richard Biener wrote:
> > On Thu, May 6, 2021 at 2:31 AM Indu Bhagat via Gcc-patches
> > wrote:
> >>
> >> To support multiple debug formats, we need to move away from explicit
> >> enumeration of each individual
ping.
On Sun, Apr 25, 2021 at 2:57 PM Hongtao Liu wrote:
>
> On Fri, Apr 23, 2021 at 5:13 PM Jakub Jelinek wrote:
> >
> > On Fri, Apr 23, 2021 at 12:53:58PM +0800, Hongtao Liu via Gcc-patches wrote:
> > > + if (!CONST_INT_P (er))
> > > + return 0;
> > > + ei = INTVAL (er);
> > > +
ping
On Tue, Apr 27, 2021 at 5:58 PM Hongtao Liu wrote:
>
> Hi:
> As described in the subject line, this patch is about to do the
> below transformation.
>
> - vpcmpeqd%ymm3, %ymm3, %ymm3
> - vpandn %ymm3, %ymm2, %ymm2
> - vpblendvb %ymm2, %ymm1, %ymm0, %ymm0
>
ping
On Fri, Apr 30, 2021 at 12:42 PM Hongtao Liu wrote:
>
> Hi:
> This patch is to fix ice which was introduced by my
> r11-5696-g35c4c67e6c534ef3d6ba7a7752ab7e0fbc91755b.
> Bootstrapped and regtested on x86_64-linux-gnu{-m32,}.
> Ok for trunk and backport to GCC11?
>
> gcc/ChangeLog
>
ping
On Fri, Apr 30, 2021 at 12:49 PM Hongtao Liu wrote:
>
> Hi:
> For v{,p}expand* When mask is 0, -1, or has all all one bits in its
> lower part, it can be optimized to simple mov or mask mov.
> Bootstrapped and regtested on x86_64-linux-gnu{-m32,} and
> x86_64-linux-gnu{m32\
On 2021-05-12, Rainer Orth wrote:
Hi Fangrui,
Hi Rainer,
for -fPIE. Now that we have H.J. Lu's GOTPCRELX for years which mostly
nullify the benefit of HAVE_LD_PIE_COPYRELOC, HAVE_LD_PIE_COPYRELOC
should retire now.
Solaris/x86 ld doesn't support this, so HAVE_LD_PIE_COPYRELOC needs to
Implement FP vector compares for V2SFmode for TARGET_MMX_WITH_SSE.
2021-05-12 Uroš Bizjak
gcc/
PR target/98218
* config/i386/i386-expand.c (ix86_expand_sse_movcc): Handle V2SF mode.
* config/i386/mmx.md (MMXMODE124): New mode iterator.
(V2FI): Ditto.
(mmxintvecmode): New
96 matches
Mail list logo
