On Wed, Sep 3, 2025 at 6:26 AM Andrew Pinski wrote:
>
> On Tue, Sep 2, 2025 at 9:22 PM Sam James wrote:
> >
> > Andi Kleen writes:
> >
> > > That said the automatic scanning of identifiers still seems like a
> > > useful option to have. I bet the kernel is not the only software
> > > with this i
On Wed, Sep 3, 2025 at 4:34 AM Andrew Pinski
wrote:
>
> Inside split_address_to_core_and_offset, this calls get_inner_reference.
>
> Take:
> ```
> _6 = t_3(D) + 12;
> _8 = &MEM[(struct s1 *)t_3(D) + 4B].t;
> _1 = _6 - _8;
> ```
>
> On the assignement of _8, get_inner_reference will return `M
On Wed, Sep 3, 2025 at 3:18 AM Li, Pan2 wrote:
>
> Thanks Richard.
>
> >> +(convert (widen_mult:c@3 (convert@4 (convert @0))
> >> So I added 2 times convert to match above sematics. Do you indicate that
> >> we only need one convert here?
> > Yes, the above extra widening looks odd to
On Tue, Sep 2, 2025 at 11:22 PM Michal Jireš wrote:
>
> On 8/28/25 7:40 PM, Andi Kleen wrote:
> >
> > Just 1:1 is slow, isn't it?
> >
>
>
> Surprisingly not by much:
>
> rm vmlinux.o; time make -j16 vmlinux
>
> # -flto-partition=1to1
>
> Exe
> -Original Message-
> From: Richard Biener
> Sent: Wednesday, September 3, 2025 8:36 AM
> To: Tamar Christina
> Cc: gcc-patches@gcc.gnu.org; nd
> Subject: RE: [PATCH v2 3/3]middle-end: Use addhn for compression instead of
> inclusive OR when reducing comparison values
>
> On Tue, 2 Sep
This patch series reworks the implementation of the ranges::_Partial,
to reuse the functors created from _Bind_back, which allows it to
benefit from it's storage optimization.
The first two patches are refactoring patches done in pepration:
* 1/3 merges _Bind_front and _Bind_back binder into one t
On Tue, Sep 2, 2025 at 11:55 PM Andrew Pinski
wrote:
>
> On Tue, Sep 2, 2025 at 5:26 AM Richard Biener
> wrote:
> >
> > On Sun, Aug 31, 2025 at 4:11 PM Andrew Pinski
> > wrote:
> > >
> > > This is a small cleanup by moving the optimization of memcmp to
> > > memcmp_eq to fab from strlen pass. Si
Dear All,
Please ignore this submission. I have crossed over PRs between the message
and the attachments. I will resubmit both.
The patch is for pr103414, which I will have to do a bit more work on
because a parameter expression of BT_BOZ fails with a double free.
Watch this space and sorry for
On Tue, Sep 2, 2025 at 9:40 PM Jonathan Wakely wrote:
> The r16-3416-g806de30f51c8b9 change to use __cpp_lib_chrono in
> preprocessor conditions broke support for for freestanding and
> the COW std::string ABI. That happened because __cpp_lib_chrono is only
> defined to the C++20 value for hoste
On Wed, Sep 3, 2025 at 3:41 AM Patrick Palka wrote:
> On Tue, 2 Sep 2025, Jonathan Wakely wrote:
>
> > For the pair(T1, T2) explicit deduction type to decay its arguments as
> > intended, we need the pair(const T1&, const T2&) constructor to not be
> > used for CTAD. Otherwise we try to instantia
On Tue, Sep 2, 2025 at 9:46 PM Jonathan Wakely wrote:
> Make the std::get overloads for rvalues use std::forward(p.first)
> not std::move(p.first), so that lvalue reference members are not
> incorrectly converted to rvalues.
>
> libstdc++-v3/ChangeLog:
>
> PR libstdc++/121745
> *
Andi Kleen writes:
> That said the automatic scanning of identifiers still seems like a
> useful option to have. I bet the kernel is not the only software
> with this issue.
Off the top of my head, there's PHP and Valgrind.. I'd have to go
looking for more (of course projects have added workarou
On Mon, 1 Sep 2025, Jakub Jelinek wrote:
> Hi!
>
> The following patch implements
> https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3577.txt
> No big deal on the GCC side, for uimaxabs we just won't
> recognize it as builtin and I don't see it worth preserving
> __builtin_uimaxabs, I doubt anyt
Andrew Pinski writes:
> On Tue, Sep 2, 2025 at 9:22 PM Sam James wrote:
>>
>> Andi Kleen writes:
>>
>> > That said the automatic scanning of identifiers still seems like a
>> > useful option to have. I bet the kernel is not the only software
>> > with this issue.
>>
>> Off the top of my head, t
On Tue, Sep 2, 2025 at 9:22 PM Sam James wrote:
>
> Andi Kleen writes:
>
> > That said the automatic scanning of identifiers still seems like a
> > useful option to have. I bet the kernel is not the only software
> > with this issue.
>
> Off the top of my head, there's PHP and Valgrind.. I'd have
On 2025-08-27 17:07, Torbjörn SVENSSON wrote:
Reduce fp16-aapcs testcases to return value testing since parameter
passing are already tested in aapcs/vfp*.c
gcc/testsuite/ChangeLog:
* gcc.target/arm/fp16-aapcs.c: New test.
* gcc.target/arm/fp16-aapcs-1.c: Removed.
* gc
On 2025-09-01 16:59, Christophe Lyon wrote:
On Wed, 27 Aug 2025 at 13:25, Torbjorn SVENSSON
wrote:
On 2025-08-18 19:24, Christophe Lyon wrote:
Like we do in other effective-targets, add
"-mcpu=unset -march=armv8-a"
directly when setting et_arm_v8_neon_flags in arm_v8_neon_ok_nocache,
to
This patch add basic support for the following XAndes ISA extensions:
XANDESPERF
XANDESBFHCVT
XANDESVBFHCVT
XANDESVSINTLOAD
XANDESVPACKFPH
XANDESVDOT
gcc/ChangeLog:
* config/riscv/riscv-ext.def: Include riscv-ext-andes.def.
* config/riscv/riscv-ext.opt (riscv_xandes_subext): New
On arm, overriding -march can lead to warnings if the testsuite
options try to pass -mcpu. Avoid these by ensuring the -mcpu is unset
before adding the architecture.
Also, improve the compatibility of asm-hard-reg-error-3.c for
hard-float environment by allowing FP instructions in the
architectur
Argh... I realized I sent an old version of the patch, sorry. Will update
tomorrow.
--
Regards
Robin
From: Pan Li
This patch would like to try to match the the unsigned
SAT_MUL form 5, aka below:
#define DEF_SAT_U_MUL_FMT_5(NT, WT) \
NT __attribute__((noinline))\
sat_u_mul_##NT##_from_##WT##_fmt_5 (NT a, NT b) \
{
Given a sequence such as
int foo ()
{
#pragma GCC unroll 4
for (int i = 0; i < N; i++)
if (a[i] == 124)
return 1;
return 0;
}
where a[i] is long long, we will unroll the loop and use an OR reduction for
early break on Adv. SIMD. Afterwards the sequence is followed by a compression
Hi,
This patch changes is_vlmax_len_p to handle VLS modes properly.
Before we would check if len == GET_MODE_NUNITS (mode). This works vor
VLA modes but not necessarily for VLS modes. We regularly have e.g.
small VLS modes where LEN equals their number of units but which do not
span a full vec
> The last sucessful build was 20250605, which has this in the build
> directory:
>
> $ find -name s-vai.ads
> ./gcc/ada/rts_32/s-vai.ads
> ./gcc/ada/rts/s-vai.ads
>
> Both files are now missing, and reverting this commit brings them back.
The intent was to remove only the former and
The extension subset check logic in riscv_ext_is_subset was incorrectly
inverted, causing functions with more extensions to be incorrectly
rejected from being inlined into functions with fewer extensions.
This patch fixes the logic to correctly check if the callee's required
extensions are a subse
This is a partial implementation of P2781R9. It adds std::cw and
std::constant_wrapper, but doesn't modify __integral_constant_like for
span/mdspan.
libstdc++-v3/ChangeLog:
* include/bits/version.def (constant_wrapper): Add.
* include/bits/version.h: Regenerate.
* include/
On Tue, Sep 02, 2025 at 03:01:27PM +0530, Ayappan wrote:
> > Wow. So this shortcoming harks back to the origin of time.
> >
> > Why was it not discovered before, do you think? No one ever tried it?
> >
> > Should we have a testcase for it? _Do_ we have a testcase for it?
>
> I came to know abou
Hi Iain,
>> What about Solaris and *BSD. Are there any outstanding PRs for druntime
>> hanging
>> on those ports too? Or is specific to Darwin (or possibly emutls?).
>
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944
>
> says darwin and i686-linux (so my exclusion of linux is not right), b
Hi!
The following patch implements
https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3577.txt
No big deal on the GCC side, for uimaxabs we just won't
recognize it as builtin and I don't see it worth preserving
__builtin_uimaxabs, I doubt anything but gcc testsuite used
that.
But on the glibc side
On Tue, 26 Aug 2025 at 18:31, Nathan Myers wrote:
>
> Changes in v4:
> * Move global symbols to correct list.
> * Fix up commit annotations: Changelog, title.
>
> Changes in v3:
> * Add to global symbols list.
>
> Changes in v2:
> * tai_clock and gps_clock remain "is_steady = false".
> * Note
On Fri, Aug 29, 2025 at 5:23 PM Tomasz Kaminski wrote:
>
> On Fri, Aug 29, 2025 at 5:21 PM Tomasz Kaminski
> wrote:
>
>>
>>
>> On Fri, Aug 29, 2025 at 5:16 PM Patrick Palka wrote:
>>
>>> On Fri, 29 Aug 2025, Tomasz Kamiński wrote:
>>>
>>> > The changes the _Variadic_union implementation, in a w
@@ -3047,7 +3047,7 @@ can_find_related_mode_p (machine_mode vector_mode,
scalar_mode element_mode,
GET_MODE_SIZE (element_mode), nunits))
return true;
if (riscv_v_ext_vls_mode_p (vector_mode)
- && multiple_p (TARGET_MIN_VLEN * TARGET_MAX_LMUL,
+ && multip
On Tue, Sep 2, 2025 at 1:11 PM Tomasz Kaminski wrote:
>
>
> On Tue, Sep 2, 2025 at 11:44 AM Jonathan Wakely
> wrote:
>
>> On Tue, 2 Sept 2025 at 09:01, Luc Grosheintz wrote:
>> >
>> > As preparation for implementing std::constant_wrapper that's part of the
>> > C++26 version of the header, the
On Fri, Aug 29, 2025 at 2:20 PM Matteo Nicoli
wrote:
>
> Here’s the patch with the modified comment before the rule in match.pd
OK. Do you have git access? Btw, the ChangeLog is currently wrong:
Subject: [PATCH] tree-optimization/121595 - the optimization fabs(a+0.0) ->
fabs(a) should apply o
From: Ayappan Perumal
---
stack-protector is not supported in GCC on AIX. This patch is to fail the
compilation if -fstack-protector option is passed.
gcc/config/rs6000/aix.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/gcc/config/rs6000/aix.h b/gcc/config/rs6000/aix.h
This implements the new vector optabs vec_addh_narrow
adding support for in-vectorizer use for early break.
Bootstrapped Regtested on aarch64-none-linux-gnu,
arm-none-linux-gnueabihf, x86_64-pc-linux-gnu
-m32, -m64 and no issues.
Ok for master?
Thanks,
Tamar
gcc/ChangeLog:
* config/aar
Hi!
In r16-3414 libstdc++ changed ABI for (still experimental C++20) and uses
unordered value -128 instead of 2. Generally the change improved code
generation on all targets tested, see
https://gcc.gnu.org/pipermail/gcc-patches/2025-August/693534.html
for details.
In r16-3474 I've adjusted the mi
On 9/1/25 2:47 AM, Jiawei wrote:
This patch update RISC-V Zba extension 'sext' instructions generation.
Supplemented the instruction generation detection of 'sext.h' and
'sext.b'.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/zbb-sext.c: New test.
I went ahead and pushed this as well.
Hi Yuao,
Yuao Ma wrote:
Hi Tobias,
I have some updates about this patch.
First, some good news:
1. The patch has been bootstrapped and tested with no regressions.
This was achieved by limiting the type with only one 'sorry' case.
2. The frontend parsing now considers outer parentheses.
:-)
Hi,
In a two-source gather we unconditionally overwrite target with the
first gather's result already. If op1 == target this clobbers the
source operand for the second gather. This patch uses a temporary in
that case.
Posting a v2 for the CI with the proper PR# and reg_overlap_mentioned_p.
Re
Hi Kito,
On 02/09/2025 05:17, Kito Cheng wrote:
Hi Paul-Antoine:
diff --git gcc/config/riscv/autovec-opt.md gcc/config/riscv/autovec-opt.md
index d56fb5f237c..5cb34cef418 100644
--- gcc/config/riscv/autovec-opt.md
+++ gcc/config/riscv/autovec-opt.md
@@ -2163,3 +2163,41 @@ (define_insn_and_spli
On Tue, Sep 02, 2025 at 03:28:46PM +0200, Richard Biener wrote:
> On Tue, 2 Sep 2025, Jakub Jelinek wrote:
>
> > On Tue, Sep 02, 2025 at 11:34:59AM +0200, Richard Biener wrote:
> > >
> > > Oops, now also to the list.
> >
> > If so, it needs to punt on anything that uses something more than once.
gcc/ChangeLog:
* config/aarch64/aarch64-c.cc (aarch64_update_cpp_builtins):
Conditonally define
`__ARM_FEATURE_SME_LUTv2" macro.
* config/aarch64/aarch64-option-extensions.def
(AARCH64_OPT_EXTENSION("sme-lutv2")): New
optional architecture extension.
* con
On 8/31/25 9:21 AM, Nathaniel Shead wrote:
On Fri, Aug 29, 2025 at 12:33:23PM +0200, Jason Merrill wrote:
On 8/24/25 9:04 AM, Nathaniel Shead wrote:
Bootstrapped and regtested on x86_64-pc-linux-gnu, OK for trunk?
-- >8 --
On looking again at [basic.lookup.argdep] p4, I believe GCC hasn't ful
Hi!
The following patch implements the
https://www.open-std.org/jtc1/sc22/wg14/www/docs/n3457.htm
paper without the first 3 lines in Recommended practice.
Seems GCC behavior already matches the expected behavior except for
diagnostics of more than 2147483648 __COUNTER__ expansions, so the
patch ad
From: Claudiu Zissulescu
Add a new target instruction. Hardware-assisted sanitizers on
architectures providing instructions to tag/untag memory can then
make use of this new instruction pattern. For example, the
memtag-stack sanitizer uses these instructions to tag and untag a
memory granule.
gc
On 9/2/25 1:29 PM, Harald Anlauf wrote:
Dear all,
here's an attempt to fix a technical regression for TRANSFER with
an unlimited polymorphic SOURCE argument. As this is something
where others are more familiar with (Paul?), I might have missed
something, so be extra careful when reviewing.
Oth
On 8/29/25 9:51 AM, Jakub Jelinek wrote:
On Thu, Aug 28, 2025 at 02:24:33PM -0700, Andi Kleen wrote:
Jakub Jelinek writes:
On Wed, Aug 27, 2025 at 03:52:11PM +0200, Michal Jires wrote:
This new pass heuristically detects symbols referenced by toplevel
assembly to prevent their optimization.
Inside split_address_to_core_and_offset, this calls get_inner_reference.
Take:
```
_6 = t_3(D) + 12;
_8 = &MEM[(struct s1 *)t_3(D) + 4B].t;
_1 = _6 - _8;
```
On the assignement of _8, get_inner_reference will return `MEM[(struct s1
*)t_3(D) + 4B]`
and an offset but that does not match up w
On Mon, 1 Sep 2025, Aleksandar Rakic wrote:
> The compressed MIPS ISAs (microMIPS and MIPS16) require the LSB of
> an address to indicate which ISA to execute. The non-conformant
> patterns used in these tests cannot set the ISA mode bit and may
> attempt to directly call the variable which trigg
On Mon, 1 Sep 2025, Tomasz Kaminski wrote:
>
>
> On Fri, Aug 29, 2025 at 5:23 PM Tomasz Kaminski wrote:
>
> On Fri, Aug 29, 2025 at 5:21 PM Tomasz Kaminski
> wrote:
>
>
> On Fri, Aug 29, 2025 at 5:16 PM Patrick Palka wrote:
> On Fri, 29 Aug 2025, Tomasz Kamiński wrote:
>
>
Thanks Richard.
>> +(convert (widen_mult:c@3 (convert@4 (convert @0))
>> So I added 2 times convert to match above sematics. Do you indicate that we
>> only need one convert here?
> Yes, the above extra widening looks odd to me.
Remove the inner convert of (convert @0)) will make the
On Tue, 2 Sep 2025, Jonathan Wakely wrote:
> For the pair(T1, T2) explicit deduction type to decay its arguments as
> intended, we need the pair(const T1&, const T2&) constructor to not be
> used for CTAD. Otherwise we try to instantiate pair without
> decaying, which is ill-formed for function lv
On Tue, 2025-09-02 at 22:59 +0800, Xi Ruoyao wrote:
> On Tue, 2025-09-02 at 16:51 +0200, Jakub Jelinek wrote:
> > On Tue, Aug 19, 2025 at 07:37:29PM +0800, Yang Yujie wrote:
> > > This patch fixes regressions of the gcc.dg/torture/bitint-* tests
> > > caused by r16-3036-ga76a032354ee48 with --enabl
On Tue, Sep 2, 2025 at 5:26 AM Richard Biener
wrote:
>
> On Sun, Aug 31, 2025 at 4:11 PM Andrew Pinski
> wrote:
> >
> > This is a small cleanup by moving the optimization of memcmp to
> > memcmp_eq to fab from strlen pass. Since the copy of the other
> > part of the memcmp strlen optimization to
Excerpts from Iain Sandoe's message of September 1, 2025 6:20 pm:
> Tested on a cross to powerpc-darwin9 and on powerpc64le-linux,
> OK for trunk?
> thanks
> Iain
>
OK, and please backport to gcc-15 too.
I've raised upstream pull request here:
https://github.com/dlang/dmd/pull/21803
Regards,
I
On Tue, Sep 2, 2025 at 5:21 AM Richard Biener
wrote:
>
> On Sun, Aug 31, 2025 at 12:29 AM Andrew Pinski
> wrote:
> >
> > Inside split_address_to_core_and_offset, this calls get_inner_reference.
> >
> > Take:
> > ```
> > _6 = t_3(D) + 12;
> > _8 = &MEM[(struct s1 *)t_3(D) + 4B].t;
> > _1 = _
For the pair(T1, T2) explicit deduction type to decay its arguments as
intended, we need the pair(const T1&, const T2&) constructor to not be
used for CTAD. Otherwise we try to instantiate pair without
decaying, which is ill-formed for function lvalues.
Use std::type_identity_t to make the constru
On Tue, Sep 2, 2025 at 5:32 AM Richard Biener
wrote:
>
> On Tue, Sep 2, 2025 at 6:29 AM Andrew Pinski
> wrote:
> >
> > I noticed that when looking into g++.dg/tree-ssa/vector-compare-1.C
> > failure on arm, the wrong alignment was being used for the load.
> > There needs to be an unaligned type h
On 8/28/25 7:40 PM, Andi Kleen wrote:
Just 1:1 is slow, isn't it?
Surprisingly not by much:
rm vmlinux.o; time make -j16 vmlinux
# -flto-partition=1to1
Executed in 235.38 secsfish external
usr time 37.97 mins0
Hi All,
This is another straightforward patch that detects variables with
non-integer type being used as PDT parameters. Note that much of the time,
when being parsed, variable expressions have type BT_PROCEDURE and so the
type of the symbol must be checked. The parameter fed to the PDT
in pr95090
Bootstrapped and regtested on x86_64-cp-linux-gnu, does this look
OK for trunk and eventual backports?
-- >8 --
Here although the local templated variables x and y have the same
reduced constant value, only x's initializer is well-formed as written
since A::m has private access. We correctly rej
Dear all,
here's an attempt to fix a technical regression for TRANSFER with
an unlimited polymorphic SOURCE argument. As this is something
where others are more familiar with (Paul?), I might have missed
something, so be extra careful when reviewing.
Otherwise regtests fine on x86_64-pc-linux-g
On Sun, 10 Aug 2025, Prathamesh Kulkarni wrote:
> Hi Matthew,
> Thanks for the suggestions! In the attached patch, I have modified
> libatomic/configure.ac to use
> __libatomic_save_CFLAGS__ instead of save_CFLAGS, so it isn't (accidentally)
> modified by ACX_PROG_CC_WARNING_OPTS.
>
> The patch
The r16-3416-g806de30f51c8b9 change to use __cpp_lib_chrono in
preprocessor conditions broke support for for freestanding and
the COW std::string ABI. That happened because __cpp_lib_chrono is only
defined to the C++20 value for hosted and for the new ABI, because the
full set of C++20 features ar
On Fri, 22 Aug 2025, David Faust wrote:
> +{
> + /* Treat btf_type_tag applied to a function type as applying to the
> + return type instead. Otherwise with GNU syntax there is no way to
> + apply type_tag to the return type; the parser always associates it
> + to the func
On Wed, 28 May 2025, Trevor Gross wrote:
> Documentation for `__cmpsf2` and similar functions currently indicate a
> return type of `int`. This is not correct however; the `libgcc`
> functions return `CMPtype`, the size of which is determined by the
> `libgcc_cmp_return` mode.
>
> Update document
Hi Jerry,
For some reason (fingers!) pdt_41.f03 had disappeared from my tree. Now
that I have pulled afresh, there it is. Consider that the numbers of the
tests in today's submissions are bumped by one.
Thanks for spotting that and for the reviews. I'll push first thing
tomorrow.
Regards
Paul
On Tue, Sep 02, 2025 at 12:12:43PM +0200, Jakub Jelinek wrote:
> On Tue, Sep 02, 2025 at 05:01:37AM -0500, Segher Boessenkool wrote:
> > On Tue, Sep 02, 2025 at 07:56:58AM +0200, Jakub Jelinek wrote:
> > > On Tue, Sep 02, 2025 at 09:32:15AM +0530, Surya Kumari Jangala wrote:
> > > > Ping.
> > > >
On 9/2/25 6:06 AM, Paul Richard Thomas wrote:
Hi All,
This is another relatively straight forward PDT patch that enables
typebound procedures with a parameterized kind specification for the interface.
Regtests with FC42/x86_64. OK for mainline
Paul
PS Patches for PRs 84432, 103414 & 114815
On 9/2/25 4:43 AM, Paul Richard Thomas wrote:
Hi All,
The attached ChangeLogs explain the chunk in expr.cc. The chunk
in is_extension_of was something of a surprise to me since I thought that we had
stopped using the hash to identify derived types and to restrict their use to
intrinsic types
On 02/09/2025 16:16, Jeff Law wrote:
On 9/1/25 5:07 AM, Paul-Antoine Arras wrote:
This pattern enables the combine pass (or late-combine, depending on
the case)
to merge a vec_duplicate into an smax RTL instruction.
[ ... ]
diff --git gcc/config/riscv/autovec-opt.md gcc/config/riscv/autove
This patch adds support for the new LUTv2 features as described in the ACLE.
It adds the `+sme-lutv2` target flag, feature test macro and intrinsics.
Making use of the new instructions without the intrinsics will be done in a
follow up patch.
ChangeLog:
* V1: Initial series.
Karl Meakin (2):
AA
From: Claudiu Zissulescu
Hi,
Please find series of patches for adding memtag-stack sanitizer using
AArch64 MTE instructions.
Most of the patches are ACKed, however, I have included them for
reference. The remaining ones to be reviews are:
* target-insns.def: (compose_tag) New pattern.
* aa
On 9/2/25 13:45, Tomasz Kaminski wrote:
On Tue, Sep 2, 2025 at 10:07 AM Luc Grosheintz
wrote:
This is a partial implementation of P2781R9. It adds std::cw and
std::constant_wrapper, but doesn't modify __integral_constant_like for
span/mdspan.
libstdc++-v3/ChangeLog:
* include/bit
Yeah, it's not a strict bound, so the function needs to return -1 aka
UNKNOWN.
But how this -1 should be interpreted differs on context.
But for sure -1 cannot be interpreted as actual bound. For frequency scaling
I'd use the same logic as for costing - use estimated_poly_value / 2
In the att
Hi Iain
> On 2 Sep 2025, at 15:25, Rainer Orth wrote:
>>> What about Solaris and *BSD. Are there any outstanding PRs for druntime
>>> hanging
>>> on those ports too? Or is specific to Darwin (or possibly emutls?).
>>
>> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=103944
>>
>> says darwin a
From: Claudiu Zissulescu
MEMTAG sanitizer, which is based on the HWASAN sanitizer, will invoke
the target-specific hooks to create a random tag, add tag to memory
address, and finally tag and untag memory.
Implement the target hooks to emit MTE instructions if MEMTAG sanitizer
is in effect. Con
On Tue, Aug 19, 2025 at 07:37:29PM +0800, Yang Yujie wrote:
> This patch fixes regressions of the gcc.dg/torture/bitint-* tests
> caused by r16-3036-ga76a032354ee48 with --enable-checking=all.
>
> The errors are similar to the following:
>
> ../../gcc/testsuite/gcc.dg/torture/bitint-14.c:54:1: er
On Tue, Sep 02, 2025 at 10:03:20AM +0200, Jakub Jelinek wrote:
> Hi!
>
> In r16-3414 libstdc++ changed ABI for (still experimental C++20) and uses
> unordered value -128 instead of 2. Generally the change improved code
> generation on all targets tested, see
> https://gcc.gnu.org/pipermail/gcc-pa
On 8/30/25 12:45 PM, Iain Sandoe wrote:
A small amount of preparation for upstreaming the C++26 implementation.
We need to add (and share) some APIs and then, at some point, remove the
c++2a edition which was not adopted. Keeping the related material
together is helpful at this stage.
Tested on
LGTM and I just have a wording nit that was already there before your patch.
When reading it I really wondered why we need to "pollute" the flags.
Wouldn't a function that performs a "backup" make sense, analogous to
..._restore? And then rename the pollute function to something like
enable_
On Tue, 26 Aug 2025 at 18:31, Nathan Myers wrote:
> It also changes #if guards to mention the actual __cpp_lib_*
> feature gated, not just the language version, for clarity.
To expand on my previous mail about this part ...
> --- a/libstdc++-v3/include/std/chrono
> +++ b/libstdc++-v3/include/st
On 8/13/25 11:54 AM, H.J. Lu wrote:
Set a tentative TLS model in grokvardecl and update DECL_TLS_MODEL with
the default TLS access model after a TLS variable has been fully processed
if the default TLS access model is stronger.
If the non-template problem is needing to recalculate the default a
gcc/ChangeLog:
* config/aarch64/aarch64-sme.md (@aarch64_sme_write_zt):
New insn.
(aarch64_sme_lut_zt): Likewise.
* config/aarch64/aarch64-sve-builtins-shapes.cc (parse_type): New type
format "%T".
(struct luti_lane_zt_base): New function shape.
(SHAPE): L
> -Original Message-
> From: Richard Biener
> Sent: Tuesday, September 2, 2025 3:45 PM
> To: Tamar Christina
> Cc: gcc-patches@gcc.gnu.org; nd
> Subject: RE: [PATCH v2 3/3]middle-end: Use addhn for compression instead of
> inclusive OR when reducing comparison values
>
> On Tue, 2 Sep 2
On Tue, Sep 2, 2025 at 3:35 PM Li, Pan2 wrote:
>
> Thanks Richard for comments.
>
> > Are the inner (convert ..) around @0 and @1 necessary? The whole
> > thing as written seems to match (long)(short)x * (long)(unsigned int)y
> > for 'char' x and y?
>
> The gimple after wide-mul has 2 times conve
Thanks Richard for comments.
> Are the inner (convert ..) around @0 and @1 necessary? The whole
> thing as written seems to match (long)(short)x * (long)(unsigned int)y
> for 'char' x and y?
The gimple after wide-mul has 2 times convert similar as below.
21[local count: 1073741824]:
On Tue, Sep 2, 2025 at 2:31 PM Robin Dapp wrote:
>
> >> So even though our strided loads do support signed strides, we cannot go
> >> via
> >> the
> >>recognize gather
> >> -> recognize strided offset
> >> -> strided load
> >> route because the initial signed-offset gather will be unsupported
cc-ing Rainer
> On 2 Sep 2025, at 14:47, Iain Buclaw wrote:
>
> Excerpts from Iain Sandoe's message of September 1, 2025 6:23 pm:
>> In use for 2 years or so on Darwin branches, OK for trunk?
>> thanks
>> Iain
>>
>> --- 8< ---
>>
>> The collect fork version has two issues on Darwin, first that
On 8/31/25 10:04 PM, Jiawei wrote:
This patch update RISC-V Zba extension 'shNadd.uw' instruction generation.
Supplemented the instruction generation detection of 'sh1add.uw' and
'sh3add.uw'.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/zba-shadd.c: New test functions.
I went ahead a
On Sep 02 2025, Eric Botcazou wrote:
> What does the s-atopri.ads file point to in the 2 directories?
They both point to s-atopri__32.ads.
--
Andreas Schwab, sch...@linux-m68k.org
GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510 2552 DF73 E780 A9DA AEC1
"And now for something completely differen
Excerpts from Iain Sandoe's message of September 1, 2025 6:23 pm:
> In use for 2 years or so on Darwin branches, OK for trunk?
> thanks
> Iain
>
> --- 8< ---
>
> The collect fork version has two issues on Darwin, first that it seems
> to hang quite frequently and second that ___fork() is not avai
On Tue, 2 Sep 2025, Tamar Christina wrote:
> > -Original Message-
> > From: Richard Biener
> > Sent: Tuesday, September 2, 2025 3:08 PM
> > To: Tamar Christina
> > Cc: gcc-patches@gcc.gnu.org; nd
> > Subject: RE: [PATCH v2 3/3]middle-end: Use addhn for compression instead of
> > inclusi
On 9/2/25 5:49 AM, Robin Dapp wrote:
Hi,
This patch changes is_vlmax_len_p to handle VLS modes properly.
Before we would check if len == GET_MODE_NUNITS (mode). This works vor
VLA modes but not necessarily for VLS modes. We regularly have e.g.
small VLS modes where LEN equals their number
On 9/2/25 5:48 AM, Robin Dapp wrote:
Hi,
In a two-source gather we unconditionally overwrite target with the
first gather's result already. If op1 == target this clobbers the
source operand for the second gather. This patch uses a temporary in
that case.
Posting a v2 for the CI with the pr
On 9/2/25 6:17 AM, Anton Blanchard wrote:
Did you run Kito's pipeline checker script? I suspect it'll tell you
you need a few more cases in your model. Essentially we have checking
asserts that require every insn to have a mapping to a reservation. So
someone could ask for code gen for
On Tue, 2 Sep 2025, Jakub Jelinek wrote:
> On Tue, Sep 02, 2025 at 11:34:59AM +0200, Richard Biener wrote:
> >
> > Oops, now also to the list.
>
> If so, it needs to punt on anything that uses something more than once.
> So, alg_add_factor and alg_sub_factor are always bad (so return NULL),
> be
On Tue, 2 Sep 2025, Tamar Christina wrote:
> > -Original Message-
> > From: Richard Biener
> > Sent: Tuesday, September 2, 2025 1:44 PM
> > To: Tamar Christina
> > Cc: gcc-patches@gcc.gnu.org; nd
> > Subject: Re: [PATCH 1/3]middle-end: clear the user unroll flag if the cost
> > model h
On Tue, Sep 2, 2025 at 11:36 AM Andrew Stubbs wrote:
>
> Ping.
OK.
> On 31/07/2025 12:15, Andrew Stubbs wrote:
> > The NoOffload flag was introduced recently (commit "Don't pass vector params
> > through to offload targets").
> >
> > gcc/ChangeLog:
> >
> > * doc/options.texi: Document NoOf
1 - 100 of 167 matches
Mail list logo
