On Wed, Nov 19, 2025 at 6:22 PM Luc Grosheintz
wrote:
>
>
> On 11/19/25 14:54, Tomasz Kaminski wrote:
> > On Tue, Nov 18, 2025 at 3:28 PM Luc Grosheintz >
> > wrote:
> >
> >> Implement submdspan_extents as described in P3663 and adds it to the std
> >> module.
> >>
> >> There's one deviation fro
Hi!
The following testcase ICEs since r12-6080. The problem is that
lookup_fnfields can return NULL_TREE on failure, but the maybe_incomplete
handling was added before the if (!baselink) handling and assumes that
baselink is non-NULL (and BASELINK).
The following patch reorders the if (maybe_inc
Hi!
The r12-4475 change added extra code to recog_for_combine to attempt to
force some constants into the constant pool.
Unfortunately, as this (UB at runtime) testcase shows, such changes are
harmful for computed_jump_p jumps. The computed_jump_p returns false
for loads from constant pool MEMs:
Hi!
The extra handling of jumps across vacuous inits for -std=c++26
or -ftrivial-auto-var-init={zero,pattern} added for P2795R5 is undesirable
when processing_template_decl, because it creates labels without DECL_NAME
and GOTO_EXPRs to those and those can't be tsubsted.
I was afraid the pop_labels
Changes from v1:
* Corrected commit message formatting and fixed typo.
This patch removes redundant vector compare instructions and logic
from the vec_first_mismatch_or_eos_index intrinsic.
Currently, GCC generates extra vcmpneb instructions and additional
masking logic (xxland, xxlorc) to
On Wed, Nov 19, 2025 at 10:24 PM Jim Tsung-Chun Lin(林宗俊)
wrote:
> So there is not enough history on these 2. They were added to the repo as
> -O9 back in 1997 when it looks like the testsuite was merged in a whole to
> the cvs. I suspect these testcases were failing at -O3 back then. So
> changin
I believe there is no reason to use -O9 for these two testcases. So I
assumed it was a typo.
---
gcc/testsuite/g++.old-deja/g++.mike/eh49.C | 2 +-
gcc/testsuite/g++.old-deja/g++.mike/eh52.C | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/gcc/testsuite/g++.old-deja/g++.mike/
So there is not enough history on these 2. They were added to the repo as -O9
back in 1997 when it looks like the testsuite was merged in a whole to the cvs.
I suspect these testcases were failing at -O3 back then. So changing it to -O0
seems wrong and unwanted. Can you explain why you want to
On Wed, Nov 19, 2025 at 9:28 PM Jim Lin wrote:
>
> I believe there is no reason to use -O9 for these two testcases. So I
> assumed it was a typo. The number 9 is next to 0 on the keyboard.
So there is not enough history on these 2. They were added to the repo
as -O9 back in 1997 when it looks lik
I believe there is no reason to use -O9 for these two testcases. So I
assumed it was a typo. The number 9 is next to 0 on the keyboard.
---
gcc/testsuite/g++.old-deja/g++.mike/eh49.C | 2 +-
gcc/testsuite/g++.old-deja/g++.mike/eh52.C | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff -
On 11/19/25 1:57 PM, Jakub Jelinek wrote:
Hi!
The following testcase shows that range_decl in cp_hide_range_decl is
sometimes also NULL_TREE and not just error_mark_node, and the function
IMHO should treat both the same, not try to hide anything in that case
because it doesn't know what should b
Hi All,
The following patch is to fix PR122665 has been bootstrapped and regtested on
powerpc64le-linux. Is it OK for trunk?
The existing smul3_highpart and umul3_highpart patterns
incorrectly defined the high-part multiply by shifting both operands
before multiplication. This does not match the
On 19/11/25 20:47, Jakub Jelinek wrote:
External email: Use caution opening links or attachments
On Wed, Nov 19, 2025 at 08:39:48PM +0530, Dhruv Chawla wrote:
Ah, I didn't even think of that, good catch. I've attached an updated patch,
does it look okay?
Signed-off-by: Dhruv Chawla
在 2025/11/19 上午12:43, Rainer Orth 写道:
When building trunk on Solaris with --disable-multilib, I noticed that
the non-default multilib wasn't rejected any longer, leading to
unexpected link failures. I could trace this to DISABLE_MULTILIB no
longer being defined in multilib.h. The problem ist
On 11/19/25 15:07, Hans-Peter Nilsson wrote:
I wrote:
If I had gone with your suggestions your changes would make
sense, so I trust that you can reverse the hyphenated
vs. no-hyphenated form change without separate approval, so
"ok with that change".
...which of course does not address
On 11/20/25 1:31 AM, Marek Polacek wrote:
On Wed, Nov 19, 2025 at 09:24:26AM +0100, Jakub Jelinek wrote:
Hi!
The r16-4698 patch to use fold_convert in finish_enum_value_list instead
of copy_node + set TREE_TYPE caused UB in the compiler as detected e.g.
by valgrind on the g++.dg/parse/pr96442.C
On Wed, Nov 19, 2025 at 1:38 PM Andrew Pinski
wrote:
>
> With the addition of some `if (a) __builtin_unreachable();` in libstdc++,
> the loop ch might not be copying the full header and only copy the bb
> which contains the condition that leads to the `__builtin_unreachable()`.
> We want to contin
This patch alters the documentation to reflect the version change,
deprecations, and current implementation status of OpenMP.
Built/tested x86_64-pc-linux-gnu.
From 4fe50b2a6b9bebfb77c2dd27040529a1f1feab58 Mon Sep 17 00:00:00 2001
From: supers1ngular
Date: Mon, 17 Nov 2025 15:45:57 -0800
Subject:
This patch implements the Fortran deprecations for OpenMP 5.2, save those
either not currently available in mainline or undeprecated in OpenMP 6.0.
Deprecations use the warning introduced in patch 1/4, Wdeprecated-openmp.
It modifies the test suite accordingly.
Built/tested x86_64-pc-linux-gnu.
Fr
From: Pan Li
When the input of the scalar unsigned SAT_TRUNC is not Xmode,
the rtx need to zero extend to Xmode before the underlying
code gen. Most of other SAT_* code gen has leveraged
the API riscv_extend_to_xmode_reg but still have the ustrunc
missed. Then results in the failures mentioned
Rainer Orth writes:
> We currently have two ways to distinguish GNU and non-GNU assemblers and
> linkers:
>
> * USE_GAS and USE_GLD, defined via gcc/config/usegas.h and usegld.h
> which are included via config.gcc
>
> * HAVE_GNU_AS and HAVE_GNU_LD, determined from gcc/configure.ac
>
> This is c
On Mon, Nov 17, 2025 at 12:09 AM Sam James wrote:
>
> Lewis Hyatt writes:
>
> > Hello-
> >
> > This patch makes it easier to disable a specific language, which is a need
> > that comes up from time to time. I have tested it by trying a variety of
> > combinations of --enable-languages arguments o
On Wed, Nov 19, 2025 at 08:35:35AM -0700, Jeff Law wrote:
> Note there's still review work to do on the RISC-V bits. But it's worth
> noting your patch triggers ICEs in the RISC-V testsuite. See:
>
> https://github.com/ewlu/gcc-precommit-ci/issues/4115#issuecomment-3543758441
Ah thanks for the
On Wed, Nov 19, 2025 at 1:31 PM Robin Dapp wrote:
>
> > Yes, it's supposed to be a start.
> >
> > Btw, after Robin now implemented "type punning" for gather/stride loads
> > to handle power-of-two size groups load/store-lanes needs similar
> > support which would esp. help ARM where there's only u
On 11/16/25 4:36 PM, H.J. Lu wrote:
On Mon, Nov 17, 2025 at 12:38 AM Jeff Law wrote:
On 10/27/25 3:53 PM, H.J. Lu wrote:
As HJ indicated, I wouldn't expect significant fallout on load/store
architectures. So the scope of fallout isn't likely to be too terrible.
I was going to throw
Hello Sandra.
The bpf changes LGTM.
Thanks!
> gcc/ChangeLog
> PR other/122243
> * doc/invoke.texi (Option Summary) : Fix formatting
> issues. Remove redundant entry for -mno-co-re.
> (eBPF Options): Add missing @opindex entries. Combine documentation
> for -mco-r
I wrote:
If I had gone with your suggestions your changes would make
sense, so I trust that you can reverse the hyphenated
vs. no-hyphenated form change without separate approval, so
"ok with that change".
...which of course does not address the non-hyphenated forms
not appearing in invoke.te
This removes unnecessary whitespace following blocks in the HTML
output.
libstdc++-v3/ChangeLog:
* doc/xml/faq.xml: Remove unwanted whitespace inside
programlisting elements.
* doc/xml/manual/allocator.xml: Likewise.
* doc/xml/manual/auto_ptr.xml: Likewise.
libstdc++-v3/ChangeLog:
* doc/xml/manual/abi.xml: Add versions for GCC 16.
* doc/html/manual/abi.html: Regenerate.
---
Pushed to trunk.
libstdc++-v3/doc/html/manual/abi.html | 4 ++--
libstdc++-v3/doc/xml/manual/abi.xml | 2 ++
2 files changed, 4 insertions(+), 2 deletions(-)
libstdc++-v3/ChangeLog:
PR libstdc++/122743
* doc/xml/manual/abi.xml: Add versions for GCC 15.
* doc/html/manual/abi.html: Regenerate.
---
Pushed to trunk and gcc-15.
libstdc++-v3/doc/html/manual/abi.html | 4 ++--
libstdc++-v3/doc/xml/manual/abi.xml | 2 ++
2 files ch
> From: Sandra Loosemore
> Date: Wed, 19 Nov 2025 13:07:36 -0700
> This is another patch in the series to make documentation of
> target-specific options in invoke.texi match what is in the
> corresponding .opt files.
>
> The cris.opt file is a bit strange, with many cases where negative
> forms
With the addition of some `if (a) __builtin_unreachable();` in libstdc++,
the loop ch might not be copying the full header and only copy the bb
which contains the condition that leads to the `__builtin_unreachable()`.
We want to continue if there is such a condition on to the next bb and copy
that
Hi Malcolm,
On 11/19/25 18:09, Matthew Malcomson wrote:
Hi Josef,
I've been seeing a linker error after this patch (commit 7cd91c7c42)
on a benchmark that is not open.
A `func1._omp_fn.0` gets ICF'd with a `func2._omp_fn.0`, but the
address pointer in `func1` doesn't get updated for some re
> Yes, it's supposed to be a start.
>
> Btw, after Robin now implemented "type punning" for gather/stride loads
> to handle power-of-two size groups load/store-lanes needs similar
> support which would esp. help ARM where there's only up to ld4(?).
>From what I heard so far nobody really likes seg
Hi,
For the following two cases:
A. Anonymous structure type is a top-level type:
struct {
int a;
char c[] __attribute__ ((counted_by (a)));
};
B. Anonymous structure type is the type for an unnamed field of another
structure:
struct s {
int a;
struct {
int b;
char c[] __att
Hi Sandra,
> On 19 Nov 2025, at 20:07, Sandra Loosemore wrote:
>
> The Darwin target options documentation was a bit of a mess, with several
> undocumented options, some that were listed in the option summary or mentioned
> in discussion of other options but not actually documented, and a large
On Wed, Nov 19, 2025 at 12:16 PM Sandra Loosemore
wrote:
>
> gcc/ChangeLog
> PR other/122243
> * aarch64.opt (Wexperimental-fmv-target): Mark as "Undocumented".
> * doc/invoke.texi (Option Summary) : Don't
> list "Undocumented" aarch64 options -mverbose-cost-dump or
gcc/ChangeLog
PR other/122243
* doc/invoke.texi (Option Summary) : Fix formatting
issues. Remove redundant entry for -mno-co-re.
(eBPF Options): Add missing @opindex entries. Combine documentation
for -mco-re and -mno-co-re.
---
gcc/doc/invoke.texi | 30 ++
gcc/ChangeLog
PR other/122243
* config/ft32/ft32.opt (mlra): Mark obsolete option as Undocumented.
* doc/invoke.texi (Option Summary) : Remove -mlra.
(FT32 Options): Likewise. Add @opindex entries for negative
option forms.
---
gcc/config/ft32/ft32.opt | 2
The Darwin target options documentation was a bit of a mess, with several
undocumented options, some that were listed in the option summary or mentioned
in discussion of other options but not actually documented, and a large
number of options listed in darwin.opt as being obsolete.
I've undocument
Comparing the documentation in invoke.texi with hs8300.opt, I spotted
a few options in the latter that were missing documentation, and added
it.
Also -mexr and -mno-exr are listed as separate options in the .opt
file but pretty clearly are intended to be inverses of each other, so
I added RejectNe
gcc/ChangeLog
PR other/122243
* doc/invoke.texi (Option Summary) :
Correct spelling of -mthreads and add missing options.
(Cygwin and MinGW Options): Add @opindex for negative forms.
---
gcc/doc/invoke.texi | 9 +++--
1 file changed, 7 insertions(+), 2 deletions
frv.opt has a few options that have never been documented in the
manual. In the initial commit of the FRV port (prior to the adoption
of .opt files) they were marked as "Internal debug switch" so I have
explicitly made them "Undocumented", consistently with other options
similarly marked in the or
gcc/ChangeLog
PR other/122243
* doc/invoke.texi: Document -mno-android.
---
gcc/doc/invoke.texi | 4
1 file changed, 4 insertions(+)
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
index 97e60473337..07bf00219c6 100644
--- a/gcc/doc/invoke.texi
+++ b/gcc/doc/invoke.tex
gcc/ChangeLog
PR other/122243
* doc/invoke.texi (FR30 Options): Add @opindex for -mno-small-model.
---
gcc/doc/invoke.texi | 1 +
1 file changed, 1 insertion(+)
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
index 5f1fabc8b0e..4333700df4a 100644
--- a/gcc/doc/invoke.texi
+
The documentation for Blackfin options had separate entries for the
positive and negative forms of many options, both in the Option Summary
and detailed documentation. This is unnecessarily verbose and counter
to the general rule that only one form of each option is documented.
gcc/ChangeLog
gcc/ChangeLog
PR other/122243
* doc/invoke.texi (Option Summary) : Add -mdbst
and -mlong-calls.
(C6X Options): Likewise.
---
gcc/doc/invoke.texi | 13 -
1 file changed, 12 insertions(+), 1 deletion(-)
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.te
gcc/ChangeLog
PR other/122243
* doc/invoke.texi (Option Summary) : Remove
entries for "Undocumented" options -EB, -EL, -mhard-float,
-msoft-float, and nonexistent option -mcse-cc.
(C-SKY Options): Likewise. Also remove references to "Undocumented"
op
This is another patch in the series to make documentation of
target-specific options in invoke.texi match what is in the
corresponding .opt files.
The cris.opt file is a bit strange, with many cases where negative
forms are listed explicitly as separate options from the positive
forms, with both h
gcc/ChangeLog
PR other/122243
* config/alpha/alpha.opt (mgas): Mark as Undocumented.
* doc/invoke.texi (Option Summary) : Add
-mtls-kernel, -mtls-size=, -mlong-double-128, and -mlong-double-64.
(DEC Alpha Options): Likewise.
---
gcc/config/alpha/alpha.opt |
gcc/ChangeLog
PR other/122243
* aarch64.opt (Wexperimental-fmv-target): Mark as "Undocumented".
* doc/invoke.texi (Option Summary) : Don't
list "Undocumented" aarch64 options -mverbose-cost-dump or
-Wexperimental-fmv-target, or both positive and negative form
gcc/ChangeLog
PR other/122243
* config/epiphany/epiphany.opt (mlong-calls): Make it do something
useful.
(may-round-for-trunc): Make this undocumented option with a weird
name an alias for -mmay-round-for-trunc.
(mfp-iarith): Fix doc string.
*
This patch undocuments ARM target-specific options that have never
been implemented, are already marked as "Undocumented" in arm.opt
file, and/or are presently documented as obsolete or only useful for
back end debugging. I've also cleaned up the option summary to list
only one of the positive or
The ARC front end presently has a large number of options that are
explicitly deprecated, either by the "Warn" option or by being
documented as such in the GCC manual. The manual text has documented
a long list of obsolete options with a warning that they will be removed
completely in a future rel
These two options are defined in the negative form and don't have the
RejectNegative property so they end up having positive forms beginning
with "-mno-no-", which is confusing. I've inverted the sense of the
option instead (so that the positive forms are -mbrcc and
-mdpfp-lrsr).
I'm not set up t
On Wed, Nov 19, 2025 at 09:24:26AM +0100, Jakub Jelinek wrote:
> Hi!
>
> The r16-4698 patch to use fold_convert in finish_enum_value_list instead
> of copy_node + set TREE_TYPE caused UB in the compiler as detected e.g.
> by valgrind on the g++.dg/parse/pr96442.C g++.dg/cpp0x/auto9.C testcases.
>
Per PR target/122288, gcn.opt contained some invalid syntax that was
quietly accepted by the options processor. This patch fixes that,
marks some useless options as "Undocumented", and brings the
documentation into sync with the options file.
I tested the .opt file changes on both a standalone gc
Continuing with PR122243 fixes, this set of patches covers about the
first third of the target-specific options sections in invoke.texi.
The goal I've been working towards with these patches is that all
options that aren't marked as "Undocumented" in the .opt file ought to
have documentation in the
On Wed, Nov 19, 2025 at 05:03:05PM +0100, Daniele Sahebi wrote:
> Fixed the issues from the previous version.
> Bootstrapped and regtested on x86_64-linux-unknown-gnu, OK?
> ---
>
> Currently, build_over_call calls build_cplus_new in template decls, generating
> a TARGET_EXPR that it then passes t
On 11/5/25 23:38, Jørgen Kvalsvik wrote:
Hi,
These patches are a mix of features, bugfixes, and quality of life
improvements.
* fixes a wrong line count when multiple blocks touch the same line
* -fpath-coverage and -fcondition-coverage implies -ftest-coverage
(.gcno, a prerequisite for run
Now that SCEV is calling ranger for values, we need to make sure ranger
doesnt call back into SCEV again or we get infinite loops like this PR.
fold_using_range::range_of_ssa_name_with_loop_info tries to use SCEV
to see if it can refinea PHI result based on loop information. This
patch adds
Hello,
On Thu, Nov 13 2025, Richard Biener wrote:
> We currently remove stmts inside of a FOR_EACH_IMM_USE_STMT iteration
> which can be problematical. The following adjusts purge_all_uses
> to gather all stmts to remove and remove them in reverse order
> afterwards which also better deals with d
On 2025-11-19 12:49 p.m., Rainer Orth wrote:
> Hi Dave,
>
>> On 2025-11-19 3:29 a.m., Rainer Orth wrote:
>>> diff --git a/gcc/config/ia64/hpux.h b/gcc/config/ia64/hpux.h
>>> --- a/gcc/config/ia64/hpux.h
>>> +++ b/gcc/config/ia64/hpux.h
>>> @@ -62,10 +62,6 @@ do {
On Wed, 19 Nov 2025, Rainer Orth wrote:
> This patch fixes this by providing a Solaris definition instead. At the
> same time, it introduces __PID_TYPE__ etc. macros matching e.g. wint_t
> as well as two testcases.
The point of those macros is for use in the implementation of
(because that's a
Hi,
On Sat, Oct 18 2025, Josef Melcr wrote:
> Due to the if statement in ipa_compute_jump_functions_for_bb, callback
> edges were never constructed for builtin functions unless LTO was
> enabled. This patch corrects this behavior, allowing GCC to optimize
> callbacks more broadly. It also extend
Hello,
On Sat, Oct 18 2025, Josef Melcr wrote:
> The inclusion of this early return statement has been discussed before,
> it was ultimately left out of the original patch, but it turns out to be
> necessary.
>
> When a callback edge is being created, it is first created by
> symbol_table::create_
Hi Dave,
> On 2025-11-19 3:29 a.m., Rainer Orth wrote:
>> diff --git a/gcc/config/ia64/hpux.h b/gcc/config/ia64/hpux.h
>> --- a/gcc/config/ia64/hpux.h
>> +++ b/gcc/config/ia64/hpux.h
>> @@ -62,10 +62,6 @@ do {
>> \
>> #undef ASM_EXTRA_SPEC
>>
On 11/19/25 14:54, Tomasz Kaminski wrote:
On Tue, Nov 18, 2025 at 3:28 PM Luc Grosheintz
wrote:
Implement submdspan_extents as described in P3663 and adds it to the std
module.
There's one deviation from the standard. Doesn't (under all
circumstances) require:
0 <= begin[k] <= end[k] <
Hi Josef,
I've been seeing a linker error after this patch (commit 7cd91c7c42) on
a benchmark that is not open.
A `func1._omp_fn.0` gets ICF'd with a `func2._omp_fn.0`, but the address
pointer in `func1` doesn't get updated for some reason and the linker
ends up complaining that there is no
> I know I was pushing for it to be enabled more widely as it's painfully hard
> to forward from a narrow store to a wider load. But based on earlier
> discussions I've backed off that position.
FWIW I would expect any slightly better OOO core aimed at general
purpose code to have some form of ha
Hi Alfie,
> This is a fix for the regression I caused with my preserve_none patch.
>
> Bootstrapped and reg tested on aarch64-linux-gnu.
LGTM.
Cheers,
Wilco
When laying out arguments of size 0, previously would return the next argument
passing register, without checking that there was a next
Hi Alfie,
> -Original Message-
> From: Alfie Richards
> Sent: 14 November 2025 13:45
> To: [email protected]
> Cc: Richard Earnshaw ; Tamar Christina
> ; [email protected]; Alice Carlotti
> ; Alex Coplan ; Wilco
> Dijkstra ; [email protected]; Alfie
> Richards
> Subj
On 2025-11-19 3:29 a.m., Rainer Orth wrote:
> diff --git a/gcc/config/ia64/hpux.h b/gcc/config/ia64/hpux.h
> --- a/gcc/config/ia64/hpux.h
> +++ b/gcc/config/ia64/hpux.h
> @@ -62,10 +62,6 @@ do {
> \
> #undef ASM_EXTRA_SPEC
> #define ASM_EXTR
On Fri, 14 Nov 2025, Tomasz Kamiński wrote:
> The iterators for transform views (views::transform, views::zip_transform,
> and views::adjacent_transform) now store a __detail::__func_handle instead
> of a pointer to the view object (_M_parent).
>
> The behavior of the __func_handle specialization
Hi all,
This is a fix for the regression I caused with my preserve_none patch.
Bootstrapped and reg tested on aarch64-linux-gnu.
Kind Regards,
Alfie
-- >8 --
When laying out arguments of size 0, previously would return the next argument
passing register, without checking that there was a next
Martin Liška writes:
> On 11/17/25 2:41 PM, Sam James wrote:
>> Martin Liška writes:
>>
>>> Hello.
>>>
So this might break there, I don't know if this test would work with
the hp linker though.
(and yes hppa64-hp-hpux11 was working on the trunk as of a little over
a month ag
Fixed the issues from the previous version.
Bootstrapped and regtested on x86_64-linux-unknown-gnu, OK?
---
Currently, build_over_call calls build_cplus_new in template decls, generating
a TARGET_EXPR that it then passes to fold_non_dependent_expr, which ends up
calling tsubst_expr, and since tsub
On 11/17/25 1:12 PM, Kees Cook wrote:
Implement RISC-V-specific KCFI backend. Nothing is conceptually rv64
specific, but using an alternative set of instructions for rv32 would be
needed, and at present the only user of KCFI on riscv is the rv64 build
of the Linux kernel.
- Scratch register a
On 11/17/25 2:41 PM, Sam James wrote:
> Martin Liška writes:
>
>> Hello.
>>
>>> So this might break there, I don't know if this test would work with
>>> the hp linker though.
>>> (and yes hppa64-hp-hpux11 was working on the trunk as of a little over
>>> a month ago:
>>> https://gcc.gnu.org/pi
On Wed, Nov 19, 2025 at 08:39:48PM +0530, Dhruv Chawla wrote:
> Ah, I didn't even think of that, good catch. I've attached an updated patch,
> does it look okay?
> Signed-off-by: Dhruv Chawla
>
> PR tree-optimization/122733
>
> gcc/ChangeLog:
>
> * match.pd: Remove patterns. Also c
On 19/11/25 19:45, Jakub Jelinek wrote:
External email: Use caution opening links or attachments
On Wed, Nov 19, 2025 at 03:07:52PM +0100, Richard Biener wrote:
--- a/gcc/match.pd
+++ b/gcc/match.pd
@@ -1339,37 +1339,19 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
(if (INTEGRAL_TYPE_P (type))
On Wed, Nov 19, 2025 at 03:49:28PM +0100, Daniele Sahebi wrote:
> > --- /dev/null
> > +++ b/gcc/testsuite/g++.dg/pr122658.C
> > @@ -0,0 +1,15 @@
> > +/* PR c++/53055 */
> Well I just noticed this is wrong.
> I was looking around the other tests to see what they were doing and
> that must have slipp
Hello,
On Mon, Nov 10 2025, David Malcolm wrote:
> On Sat, 2025-11-08 at 00:12 +0100, Martin Jambor wrote:
>
> [...snip...]
>
>> diff --git a/gcc/testsuite/gcc.dg/tree-ssa/vrp-from-cst-agg-3.c
>> b/gcc/testsuite/gcc.dg/tree-ssa/vrp-from-cst-agg-3.c
>> new file mode 100644
>> index 000..d4
On 11/17/25 7:59 AM, Richard Braun wrote:
From: Richard Braun
Instructions have two time-related units associated with them: the
number of delay slots, and the functional unit latency. But some
floating-point multiplication instructions have a functional unit
latency that actually varies dep
On Wed, Nov 19, 2025 at 09:28:42AM +0100, Daniele Sahebi wrote:
> Fixed what was asked.
> Bootstrapped and regtested on x86_64-linux-unknown-gnu, please tell me
> if there are any other issues.
>
> ---
> Currently, build_over_call calls build_cplus_new in template decls, generating
> a TARGET_EXPR
On Wed, Nov 19, 2025 at 10:32 AM Richard Biener wrote:
> On Tue, 18 Nov 2025, Konstantinos Eleftheriou wrote:
>
> > On Mon, Nov 17, 2025 at 2:51 PM Richard Biener
> wrote:
> >
> > > On Mon, 17 Nov 2025, Konstantinos Eleftheriou wrote:
> > >
> > > > Hi Richard, thanks for the feedback.
> > > >
>
On 11/19/25 1:32 AM, Richard Biener wrote:
Well. I think it's OK to enable on a subset of targets by default,
but it should be target maintainers say whether a target is among the
crowd or not. There's the target specific optimization_table as
mechanism to achieve such defaults, like
gcc/
Sometimes, `store_bit_field` copies the destination register into a new one,
which leads to the old register being used in the instructions that follow
the ones generated by `store_bit_field`, while the bit field insertion is
performed on the new register.
This patch copies back the new destinatio
From: kelefth
This patch enables the avoid-store-forwarding patch by default at O2 or
higher on AArch64.
The assembly patterns in `bitfield-bitint-abi-align16.c` and
`bitfield-bitint-abi-align8.c` have been updated to account for
the asf transformations.
gcc/ChangeLog:
* common/config/
The avoid-store-forwarding pass is disabled by default and therefore
in the risk of bit-rotting. This patch addresses this by enabling
the pass at O2 or higher on AArch64.
Bootstrapped/regtested on AArch64 and x86_64.
Changes in v4:
- Enable asf on AArch64 only.
- Remove `avoid_store_forwardin
On Wed, Nov 19, 2025 at 09:27:27AM +0100, Jakub Jelinek wrote:
> Hi!
>
> The following testcase shows that range_decl in cp_hide_range_decl is
> sometimes also NULL_TREE and not just error_mark_node, and the function
> IMHO should treat both the same, not try to hide anything in that case
> becaus
On Wed, Nov 19, 2025 at 03:07:52PM +0100, Richard Biener wrote:
> > --- a/gcc/match.pd
> > +++ b/gcc/match.pd
> > @@ -1339,37 +1339,19 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> > (if (INTEGRAL_TYPE_P (type))
> >(rshift (op @0 @2) @1
> >
> > -/* (y << x) == x -> 0 when y != 0. */
On Wed, 19 Nov 2025, [email protected] wrote:
> From: Dhruv Chawla
>
> These patterns should not be in match.pd as they require range
> information checks that ideally belong in VRP. They were also causing
> breakages as the checks weren't tight enough.
>
> Bootstrapped and regtested on aarch64
From: Dhruv Chawla
These patterns should not be in match.pd as they require range
information checks that ideally belong in VRP. They were also causing
breakages as the checks weren't tight enough.
Bootstrapped and regtested on aarch64-linux-gnu.
Signed-off-by: Dhruv Chawla
PR tree-op
On Tue, Nov 18, 2025 at 3:28 PM Luc Grosheintz
wrote:
> Implement submdspan_extents as described in P3663 and adds it to the std
> module.
>
> There's one deviation from the standard. Doesn't (under all
> circumstances) require:
>
> 0 <= begin[k] <= end[k] <= exts.extent(k)
>
> where the k-th s
On Tue, 18 Nov 2025 at 21:38, François Dumont wrote:
>
> Ok, here is the new patch, only fixing the std::erase_if behavior for
> the _GLIBCXX_DEBUG.
Is that right? I thought this change is to benefit direct uses of
__gnu_debug::vector not std::vector with _GLIBCXX_DEBUG? The first new
test alread
Trivial bootstrap fix for the RISC-V port where a recent change
introduced a signed vs unsigned warning.
Pushing to the trunk.
jeffcommit a3e97daf1f7452d060d2e5e4eb2fea7717343f18
Author: Jeff Law
Date: Wed Nov 19 06:34:33 2025 -0700
[RISC-V] Fix trivial bootstrap failure on RISC-V
On Wed, Nov 19, 2025 at 02:13:28PM +0100, Richard Biener wrote:
> > > Makes me wonder if we couldn't set DECL_NONADDRESSABLE_P on private
> > > FIELD_DECLs in classes where none of the methods take address of it and
> > > either it doesn't have friends, or none of the friends take address of it
> >
On Wed, Nov 19, 2025 at 2:06 PM Jakub Jelinek wrote:
> On Wed, Nov 19, 2025 at 01:49:39PM +0100, Richard Biener wrote:
> > Then, there is DECL_NONADDRESSABLE_P on a FIELD_DECL which for,
> > say
> >
> > struct{
> > int len;
> > char *data;
> > };
> >
> > could say that 'len' cannot ha
On Wed, Nov 19, 2025 at 2:03 PM Jakub Jelinek wrote:
>
> On Wed, Nov 19, 2025 at 02:00:03PM +0100, Jakub Jelinek wrote:
> > On Wed, Nov 19, 2025 at 01:49:39PM +0100, Richard Biener wrote:
> > > Then, there is DECL_NONADDRESSABLE_P on a FIELD_DECL which for,
> > > say
> > >
> > > struct{
> > >
1 - 100 of 155 matches
Mail list logo
