I have seen this a few places though the testcase from PR 95906
is an obvious place where this shows up for sure.
This convert `cmp - 1` into `-icmp` as that form is more useful
in many cases.
Bootstrapped and tested on x86_64-linux-gnu.
PR tree-optimization/110949
PR tree-optimiz
After import std, std::byteswap cannot be used.
After my observation, this function is the only one in the header file
that has not been exported.
The purpose of this patch is to export std::byteswap so that it can be used
normally after import std.
Signed-off-by: hexne <61956889
The details are exactly the same as
https://gcc.gnu.org/pipermail/gcc-patches/2025-July/689865.html, except
this is for libssp not libgcc.
It probably makes sense to just pick one of these to review first, and
then based on any feedback I can rework the other one. But for
posterity, and to illustr
I think the question is not whether this could be done somehow,
but whether it should. Why design a language feature that requires
storing tokens and parsing it outside the original context?
For an attribute this may still be acceptable, but we need the same
thing for array sizes in C. For
From: Bill Wendling
Also, this code doesn't go further than parsing. I.e., it doesn't generate the
internal gimple code that accesses the struct fields. The code is meant to show
that it *is* possible to perform a delayed parsing with no "double parsing" and
still be performant.
Minor Nomenclatu
> -Original Message-
> From: Kyrylo Tkachov
> Sent: Friday, July 18, 2025 5:48 PM
> To: Tamar Christina
> Cc: GCC Patches ; Richard Sandiford
> ; Alex Coplan ; Andrew
> Pinski
> Subject: Re: [PATCH 1/2] aarch64: NFC - Make vec_* rtx costing logic
> consiste
e implemented but not exported.
>
> The following patch exports it, and additionally appends some further
> entities to the FIXME list, those all seems to be unimplemented yet.
>
Yes, those are all C++26 additions to the header and not supported yet.
> Tested on x86_64-linux, ok fo
When we have an empty function, things can go wrong with
cfi_startproc/cfi_endproc and a few other things like exceptions. So if
the only thing the function does is a call to __builtin_unreachable,
let's replace that with a __builtin_trap instead if the target has a trap
instruction. For targets w
Lu"
Date: Wed, 9 Jul 2025 09:43:52 +0800
Subject: [PATCH] libstdc++: Update some baseline_symbols.txt (x32)
* config/abi/post/x86_64-linux-gnu/x32/baseline_symbols.txt:
Updated.
Signed-off-by: H.J. Lu
(cherry picked from commit c7baa61a583b49df63d3df8c6336f8405e24f012)
---
On 7/16/25 7:04 AM, Andreas Schwab wrote:
PR target/121121
* config/riscv/riscv.cc (riscv_allocate_and_probe_stack_space):
Use temp2 instead of temp1 for the CFA note.
---
gcc/config/riscv/riscv.cc | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/g
On 7/16/25 8:38 AM, Paul-Antoine Arras wrote:
This pattern enables the combine pass (or late-combine, depending on the case)
to merge a float_extend'ed vec_duplicate into a (possibly negated) minus-mult
RTL instruction.
Before this patch, we have six instructions, e.g.:
vse
nt via
-fdemote-register-asm I removed it from this patch series. I
compiled and used the Linux kernel and glibc successfully with it for
s390x. For x86_64, the Linux kernel compiles fine, too, except of one
small manual change. For powerpc64le, I ran into an odd case compiling
glibc which I wo
On 7/18/25 7:06 AM, Li, Pan2 wrote:
OK. Curious if you've seen this show up in practice and is using the
vaaddu and similar instructions actually profitable?
I remember avg will be presented in somewhere else but not for DImode.
I tried to add vaadd related vx combine and notice we don't ha
a few submissions to do everything we want to do.
This patch is primarily concerned with avoiding signaling that fusion can
occur in cases where it obviously should not be signaling fusion.
Hi Jeff,
With this change, we're liable to ICE whenever prev_set or curr_set are
NULL_RTX. For a fix,
e op2 of the ashiftrt but seems missed.
Thus, add it back to align the definition.
The below test suites are passed for this patch.
* The rv64gcv fully regression test.
gcc/ChangeLog:
* config/riscv/autovec.md: Add (const_int 1) as the op2 of
ashiftrt.
OK.
I wonder if this would have
0 for DEF_AVG_0(WT, NT).
The below test suites are passed for this patch.
* The rv64gcv fully regression test.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/avg_floor-1-i16-from-i32.c:
Leverage DEF_AVG_0_WRAP to generate the correct func name.
* gcc.target/riscv/rvv/au
Hi!
r16-442 implemented both std::extents and std::dextents (and perhaps other
stuff), but exported only std::extents.
I went through https://eel.is/c++draft/mdspan.syn and I think std::dextents
is the only one implemented but not exported.
The following patch exports it, and additionally
There were two cases where get_insns() can be fused with end_sequence().
Applied as obvious.
Johann
--
AVR: Fuse get_insns with end_sequence.
gcc/
* config/avr/avr-passes.cc (avr_optimize_casesi): Fuse
get_insns() with end_sequence().
diff --git a/gcc/config/avr/avr-passes.cc
From: 1nfocalypse <1nfocaly...@protonmail.com>
Date: Sat, 19 Jul 2025 02:08:07 +
Subject: [PATCH] [PATCH v4] libstdc++: Implement Philox Engine [PR119794]
Implemented additional changes according to feedback from Patrick Palka.
_charT, _traits -> _CharT, _Traits in stream operator tem
ack to align the definition.
The below test suites are passed for this patch.
* The rv64gcv fully regression test.
gcc/ChangeLog:
* config/riscv/autovec.md: Add (const_int 1) as the op2 of
ashiftrt.
Signed-off-by: Pan Li
---
gcc/config/riscv/autovec.md | 6 --
1 fil
> Am 18.07.2025 um 20:06 schrieb Robin Dapp :
>
>
>>
>> Can the risc-v people try to sort out this up to a point
>> where I can just s/STMT_VINFO_TYPE/SLP_TREE_TYPE there?
>
> I think for us this mainly (only?) concerns the dynamic LMUL heuristic.
> Currently we go through all vectorized in
Ping^3
在 2025/5/27 上午8:54, Lulu Cheng 写道:
Ping^2
在 2025/5/13 下午2:06, Lulu Cheng 写道:
Ping?
在 2025/5/9 上午10:14, Lulu Cheng 写道:
From: ChengLulu
PR target/99217
gcc/ChangeLog:
* config/mips/mips.cc (mips_start_function_definition):
Implements the functionality of '-fpatchable-fun
n excess of 80 chars. Since it was the result of autogen, I let it lie. If
> > this
> > was incorrect, I can adjust it.
>
>
> That's fine, we don't care about line lengths of autogenerated files.
>
> > Otherwise, thank you for the review, and please let m
d for this patch.
* The rv64gcv fully regression test.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/avg_floor-1-i16-from-i32.c:
Leverage DEF_AVG_0_WRAP to generate the correct func name.
* gcc.target/riscv/rvv/autovec/avg_floor-1-i16-from-i64.c:
This patch fixes the 2nd parameter name mismatch in
ARRAYOFCHAR.mod.
gcc/m2/ChangeLog:
PR modula2/121164
* gm2-libs/ARRAYOFCHAR.mod (Write): Rename 2nd parameter
name a to str.
(cherry picked from commit 22d8b89689769e5efefd2c4e6dda88d9f0b2a945)
Signed-off-by: Gaius
This patch fixes the 2nd parameter name mismatch in
ARRAYOFCHAR.mod.
gcc/m2/ChangeLog:
PR modula2/121164
* gm2-libs/ARRAYOFCHAR.mod (Write): Rename 2nd parameter
name a to str.
Signed-off-by: Gaius Mulley
---
gcc/m2/gm2-libs/ARRAYOFCHAR.mod | 6 +++---
1 file changed
On Fri, 18 Jul 2025, 18:51 Jonathan Wakely, wrote:
> When the C++98 std::distance and std::advance functions (and C++11
> std::next and std::prev) are used with C++20 iterators there can be
> unexpected results, ranging from compilation failure to decreased
> performance to undefined behaviour.
>
27;t use the default
> MAX_FIXED_MODE_SIZE.
>
> I can't build a complete toolchain for BPF (fails building
> libgcc, PR121149), but from what I can see with that build,
> having MAX_FIXED_MODE_SIZE 64 is unintended: TImode isn't
> disabled for BPF. So, I'm not offering a patc
in extreme cases
> where the reference count somehow exceeds INT_MAX).
>
> However, in a very simple benchmark I see a significant impact in
> performance, which is not entirely surprising given that this patch adds
> a test to every increment. Maybe we could change [[__unlikely__]]
gt; > Bootstrapped/regtested on x86_64-pc-linux-gnu, ok for trunk?
> > > > > > > >
> > > > > > > > -- >8 --
> > > > > > > > This patch is an attempt to implement P2036R3 along with
> > > > > > > > P2
_64-pc-linux-gnu, ok for trunk?
> > > >
> > > > -- >8 --
> > > > This patch implements consteval blocks, as specified by P2996.
> > > > They aren't very useful without define_aggregate, but having
> > > > a reviewed implementation on
Hi Andre,
Am 18.07.25 um 21:41 schrieb Andre Vehreschild:
Hi Harald,
that looks fine to me. Ok to commit and backport. Thanks for the patch.
pushed as r16-2355-g8f9450505f8244 so far.
Thanks for the review!
Harald
- Andre
Andre Vehreschild * ve...@gmx.de
Am 18. Juli 2025 21:19:00 schrieb
On Fri, 18 Jul 2025 at 19:18, Tomasz Kaminski wrote:
>
>
>
> On Fri, Jul 18, 2025 at 8:03 PM Jonathan Wakely wrote:
>>
>> This adds a check when incrementing the shared count and weak count and
>> will trap if it overflows. This also double the effective range of the
>> counts for most 64-bit tar
Hi Harald,
that looks fine to me. Ok to commit and backport. Thanks for the patch.
- Andre
Andre Vehreschild * ve...@gmx.de
Am 18. Juli 2025 21:19:00 schrieb Harald Anlauf :
Dear all,
the attached simple and obvious patch fixes an erroneous runtime check
with -fcheck=pointer when passing a
t; alterations are needed.
Nice, thanks! A couple of more review comments added inline.
>
> Built/tested on x86_64-linux-gnu.
>
>
> Subject: [PATCH] [PATCH v3] libstdc++: implement Philox Engine [PR119794]
>
> Implemented changes according to feedb
Dear all,
the attached simple and obvious patch fixes an erroneous runtime check
with -fcheck=pointer when passing a non-associated proc-pointer to an
optional dummy.
Regtested on x86_64-pc-linux-gnu. OK for mainline / backports?
Thanks,
Harald
From 8f9450505f8244d262f8b4ff274f113f99cdc7e2
Hello,
On 7/11/25 6:47 PM, Martin Jambor wrote:
Hello,
and sorry for a rather late reaction. First and foremost, thanks for
the patch, I think it is great to have this finally implemented and
although I would like to see a couple things changed, I think the patch
is quite close to what could
gt; take
> > > a few submissions to do everything we want to do.
> > >
> > > This patch is primarily concerned with avoiding signaling that fusion can
> > > occur in cases where it obviously should not be signaling fusion.
> >
> > Hi Jeff,
> >
>
use_count to unsigned
> before returning as long.
> ---
>
> This improves safety for std::shared_ptr and std::weak_ptr, removing a
> possible source of undefined behaviour (although only in extreme cases
> where the reference count somehow exceeds INT_MAX).
>
> However,
Can the risc-v people try to sort out this up to a point
where I can just s/STMT_VINFO_TYPE/SLP_TREE_TYPE there?
I think for us this mainly (only?) concerns the dynamic LMUL heuristic.
Currently we go through all vectorized instructions of the loop's blocks,
lookup the stmt_vec_info and then get
cases
where the reference count somehow exceeds INT_MAX).
However, in a very simple benchmark I see a significant impact in
performance, which is not entirely surprising given that this patch adds
a test to every increment. Maybe we could change [[__unlikely__]] on
those branches to use __builtin
The standard requires that std::atomic::fetch_add does
not have undefined behaviour for signed overflow, instead it wraps like
unsigned integers. The compiler ensures this is true for the atomic
built-ins that std::atomic uses, but it's not currently true for the
__gnu_cxx::__exchange_and_add and _
When the C++98 std::distance and std::advance functions (and C++11
std::next and std::prev) are used with C++20 iterators there can be
unexpected results, ranging from compilation failure to decreased
performance to undefined behaviour.
An iterator which satisfies std::input_iterator but does not
On 7/18/25 11:03 AM, Artemiy Volkov wrote:
On Wed, Jul 02, 2025 at 08:15:40PM -0600, Jeff Law wrote:
A number of folks have had their fingers in this code and it's going to take
a few submissions to do everything we want to do.
This patch is primarily concerned with avoiding signaling
costs set ins_gp to COSTS_N_INSNS (1) to
>> preserve the current codegen, though I'd be happy to increase it for generic
>> tuning.
>>
>> For -Os we don't add any extra cost so the shorter INS-WZR form is still
>> generated always.
>>
>> Bootstrap
On Wed, Jul 02, 2025 at 08:15:40PM -0600, Jeff Law wrote:
> A number of folks have had their fingers in this code and it's going to take
> a few submissions to do everything we want to do.
>
> This patch is primarily concerned with avoiding signaling that fusion can
> occ
On Fri, 18 Jul 2025, Jeff Law wrote:
> On 7/18/25 8:37 AM, Georg-Johann Lay wrote:
>> Ok to apply removal of backends no more present in v13?
> Yes.
And feel free to simply push such changes (under the obvious rule).
Gerald
Hi Tamar,
> On 18 Jul 2025, at 18:25, Tamar Christina wrote:
>
> Hi Kyrill,
>
>> -Original Message-
>> From: Kyrylo Tkachov
>> Sent: Friday, July 18, 2025 10:40 AM
>> To: GCC Patches
>> Cc: Tamar Christina ; Richard Sandiford
>> ; Alex C
Hi Kyrill,
> -Original Message-
> From: Kyrylo Tkachov
> Sent: Friday, July 18, 2025 10:40 AM
> To: GCC Patches
> Cc: Tamar Christina ; Richard Sandiford
> ; Alex Coplan ; Andrew
> Pinski
> Subject: [PATCH 1/2] aarch64: NFC - Make vec_* rtx costing logi
Hi Andre,
After a false start, the patch was applied to mainline, did what it was
supposed to do and regtests fine.
Ok for mainline.
Thanks
Paul
On Fri, 18 Jul 2025 at 13:42, Andre Vehreschild wrote:
> Hi all, hi Harald,
>
> attached patch fixes a runtime out-of-bounds erro
On 7/18/25 8:37 AM, Georg-Johann Lay wrote:
Ok to apply removal of backends no more present in v13?
Yes.
jeff
Hi,
Sorry for the very long delay (more than one year O.o).
I finally took time to apply suggestions and to update the patch to be
up-to-date.
Joined to this email.
Cordially.
Le mer. 26 juin 2024 à 17:10, David Malcolm a écrit :
>
> On Mon, 2024-04-22 at 19:56 +0200, Guillaume Gomez
Ok to apply removal of backends no more present in v13?
Johann
--
backends.html (cr16, tilegx, tilepro): Remove entries.
The cr16, tilegx and tilepro backends have been removed in v13,
which is the oldest supported GCC release.
diff --git a/htdocs/backends.html b/htdocs/backends.html
index 52
For master, let's change the default; see other recent patch.
Committed to gcc-15 branch.
Pietro, thanks again for the patch that led up to this. This is
in a different place and with different contents, but you made
the first move in this direction so IMHO attribution is in order.
s building
libgcc, PR121149), but from what I can see with that build,
having MAX_FIXED_MODE_SIZE 64 is unintended: TImode isn't
disabled for BPF. So, I'm not offering a patch to keep
MAX_FIXED_MODE_SIZE 64 for BPF; with this patch, it moves to 2 *
BITS_PER_WORD == 128 as all other 64-bitte
>are !ok().
> >
> >This patch lifts debug checks from each conversion function into
> _M_check_ok,
> >that in case of !ok() values return a string_view containing the kind of
> >calendar data, to be included after "is not a valid" string. The localized
>
On 7/18/25 05:39, Kyrylo Tkachov wrote:
> External email: Use caution opening links or attachments
>
>
> Hi all,
>
> For inserting zero into a vector lane we usually use an instruction like:
> ins v0.h[2], wzr
>
> This, however, has not-so-great performance on some CPUs.
> On Grace, f
: gcc-patches@gcc.gnu.org
Cc: kito.ch...@gmail.com; richard.guent...@gmail.com; tamar.christ...@arm.com;
juzhe.zh...@rivai.ai; Li, Pan2 ; jeffreya...@gmail.com;
rdapp@gmail.com; panciyan
Subject: [PATCH] RISC-V: Add testcase for unsigned scalar SAT_ADD form 8 and
form 9
From: panciyan
This patch add
On 18/07/25 14:01 +0200, Tomasz Kamiński wrote:
Previously for localized output, if _M_debug option was send, the _M_check_ok
completed succesfully and _M_locale_fmt was called for months/weekdays that
are !ok().
This patch lifts debug checks from each conversion function into _M_check_ok,
that
cture and conditionalize the appropriate
> code.
I see, thanks Jeff.
Pan
-Original Message-
From: Jeff Law
Sent: Friday, July 18, 2025 1:23 AM
To: Li, Pan2 ; gcc-patches@gcc.gnu.org
Cc: juzhe.zh...@rivai.ai; kito.ch...@gmail.com; rdapp@gmail.com; Chen, Ken
; Liu, Hongtao
Subject:
Hi all, hi Harald,
attached patch fixes a runtime out-of-bounds error when an iterator of an
implied do-loop was not substituted as expected. The resulting code would still
use the uninitialized iterator variable which then lead to out-of-bounds
accesses.
Harald, I did not look at the patch you
I am at a point where I want to store additional information from
analysis (from loads and stores) to re-use them at transform stage
without repeating the analysis. I do not want to add to
stmt_vec_info at this point, so this starts adding kind specific
sub-structures by moving the STMT_VINFO_TYPE
Previously for localized output, if _M_debug option was send, the _M_check_ok
completed succesfully and _M_locale_fmt was called for months/weekdays that
are !ok().
This patch lifts debug checks from each conversion function into _M_check_ok,
that in case of !ok() values return a string_view
Previously for localized output, if _M_debug option was send, the _M_check_ok
completed succesfully and _M_locale_fmt was called for months/weekdays that
are !ok().
This patch lifts debug checks from each conversion function into _M_check_ok,
that in case of !ok() values return a string_view
esult
(and with the patch verify that does not change in unexpected
ways). The 2nd step would elide the repeated calls, instead
access the available info directly.
Bootstrapped and tested on x86_64-unknown-linux-gnu.
Note this is for the CI to catch non-x86 (aka IFN) issues, also
due to low covera
On Fri, 18 Jul 2025 at 11:04, Tomasz Kamiński wrote:
>
> Previously for localized output, if _M_debug option was send, the _M_check_ok
> completed succesfully and _M_locale_fmt was called for months/weekdays that
> are !ok().
>
> This patch lifts debug checks from each conver
Previously for localized output, if _M_debug option was send, the _M_check_ok
completed succesfully and _M_locale_fmt was called for months/weekdays that
are !ok().
This patch lifts debug checks from each conversion function into _M_check_ok,
that in case of !ok() values return a string_view
Hi all,
For inserting zero into a vector lane we usually use an instruction like:
ins v0.h[2], wzr
This, however, has not-so-great performance on some CPUs.
On Grace, for example it has a latency of 5 and throughput 1.
The alternative sequence:
moviv31.8b, #0
ins
size cost only.
This patch makes the logic consistent by using the extra_cost fields to
increment the cost when speed is set. This requires reducing the extra_cost
values
of the movi, dup and extract fields by COSTS_N_INSNS (1), as every insn being
costed
has a cost of COSTS_N_INSNS (1) at the
This removes paths gated by !slp_node and propagates out ncopies == 1,
thereby reducing the vectorizable_live_operation_1 API.
Bootstrapped and tested on x86_64-unknown-linux-gnu, pushed.
* tree-vect-loop.cc (vectorizable_live_operation_1):
Remove stmt_info and ncopies parameters.
This patch introduces the procedure GetFile into the supplementary
ISO style library IOChanUtils.
gcc/m2/ChangeLog:
PR modula2/120912
* gm2-libs-iso/IOChanUtils.def (GetFile): New procedure function.
* gm2-libs-iso/IOChanUtils.mod (GetFile): New procedure function
rther
alterations are needed.
Built/tested on x86_64-linux-gnu.From a175a89dd87426d1724812040bf8bb33680a9cb8 Mon Sep 17 00:00:00 2001
From: 1nfocalypse <1nfocaly...@protonmail.com>
Date: Fri, 18 Jul 2025 01:16:48 +
Subject: [PATCH] [PATCH v3] libstdc++: implement Philox Engine [PR119794]
could
kindly commit this v3 patch(link given below) on my behalf.
https://gcc.gnu.org/pipermail/gcc-patches/2025-July/689920.html
Do you want backports? Okay for those as well (if they are wanted).
In addition to master branch, I would like this change to be backported to
GCC 15 and GCC 14
On Fri, 18 Jul 2025 at 07:59, Tomasz Kaminski wrote:
>
>
>
> On Thu, Jul 17, 2025 at 7:02 PM Jonathan Wakely wrote:
>>
>> This implements the missing functions in _Utf_iterator to support
>> reverse iteration. All existing tests pass when the view is reversed, so
>> that the same code units are s
> Hi,
>
> Thanks for the review! No, there is no reason for it being CamelCase. I had
> written
> the code in CamelCase originally then 'sed s/CamelCase/snake_case/g'd the
> functions.
> I forgot to update this one :)
>
> Looks like I also forgot to update the Colors class as well.
I'd actuall
ch:
+return c_match
+return re.match(self.ignored_cxx, line)
+
+
+
+# Main routine
+####
+
+
+def main() -> None:
+a
Added changelog entries.
-- >8 --
If a check-function-bodies test is compiled using -fstack-protector*,
-fhardened, -fstack-check*, or -fstack-clash-protection, but the test is
not asking explicitly for those via dg-options or
dg-additional-options, then mark the test as unsupported. Since these
at I prepare separate
> backport
> patches for these branches, or will the backports be handled by maintainers?
Like I said above, all backports are approved. Please send emails here
(i.e. gcc-patches@) when you do one, maybe with the patch if it needs
modification, etc.
Segher
Hi Segher,
On 7/17/2025 10:20 PM, Segher Boessenkool wrote:
CAUTION: This email comes from a non Wind River email account!
Do not click links or open attachments unless you recognize the sender and know
the content is safe.
Hi!
On Thu, Jul 17, 2025 at 05:58:47AM -0700, harish.sadin...@windriv
From: Harish Sadineni
Avoid encoding full build paths into file headers, use only the
basename of the source file. This improves build reproducibility,
particularly in environments where build paths vary and source
files are saved for debugging.
For example, in addkf3-sw.c the file header previo
and future glibcs, starting
with 2.42, that would pass hwcap, hwcap2,3,4.
I should probably add tests to the patch, will do in v2. It is easy to miss the
check for size at runtime.
Thanks,
Yury
C_PATTERN
> +self.xgpp = XGPP_PATTERN
> +self.gfortran = GFORTRAN_PATTERN
> +
> +def match_c_pattern(self, line: str):
> +cc_match = re.match(self.cc, line)
> +if cc_match:
> +return cc_match
> +return re.match(self.xgcc
On Fri, Jul 18, 2025 at 09:07:35AM +0200, Richard Biener wrote:
> The PR shows that the uninit analysis limits are set too low in
> cases we lower switches to ifs as happens on s390x for a linux
> kernel TU. This causes false positive uninit diagnostics as we
> abort the attempt to prove that a va
The PR shows that the uninit analysis limits are set too low in
cases we lower switches to ifs as happens on s390x for a linux
kernel TU. This causes false positive uninit diagnostics as we
abort the attempt to prove that a value is initialized on all
paths. The new testcase only would require up
This patch reformats three comments in the GNU GCC style.
gcc/m2/ChangeLog:
* gm2-compiler/M2GCCDeclare.mod (StartDeclareModuleScopeSeparate):
Reformat statement comments.
(StartDeclareModuleScopeWholeProgram): Ditto.
(cherry picked from commit
R_INT_TYPE_MODE ()) and TREE_INT_CST_LOW (TYPE_SIZE_UNIT
> ()) for the BLKmode ones.
> In this case, it wants bits rather than bytes, so I've used
> GET_MODE_BITSIZE like before and TYPE_SIZE otherwise.
>
> Furthermore, the patch only computes encoding_size for big endian
> tar
when we eventually cannot
> scalarize, such as when a scalar access has children accesses) but the
> actual bug is that the detection if all data in an aggregate is indeed
> covered by replacements just assumes that is always the case if total
> scalarization triggers which however may no
On Thu, Jul 17, 2025 at 7:02 PM Jonathan Wakely wrote:
> This implements the missing functions in _Utf_iterator to support
> reverse iteration. All existing tests pass when the view is reversed, so
> that the same code units are seen when iterating forwards or backwards.
>
> libstdc++-v3/ChangeLo
TAG_GNU_annotation and DW_AT_GNU_annotation are used to represent
> >> these user annotations in DWARF.
> >>
> >> This patch introduces the new DWARF extension DIE and attribute, and
> >> generates them as necessary to represent user annotations from
>
From: Jonathan Wakely
Implement std::inplace_vector as specified in P0843R14, without follow
up papers, in particular P3074R7 (trivial unions). In consequence
inplace_vector can be used inside constant evaluations only
if T is trivial or N is equal to zero.
We provide a separate specialization f
The following makes sure we analyze live LC PHIs not part of
a double reduction.
Bootstrap and regtest running on x86_64-unknown-linux-gnu.
PR tree-optimization/121126
* tree-vect-stmts.cc (vect_analyze_stmt): Analyze the
live lane extract for LC PHIs that are vect_interna
On Thu, Jul 17, 2025 at 10:39 PM Jonathan Wakely wrote:
> On 17/07/25 15:23 -0400, Patrick Palka wrote:
> >On Thu, 17 Jul 2025, Tomasz Kamiński wrote:
> >> + constexpr inplace_vector&
> >> + operator=(const inplace_vector& __other)
> >> + noexcept(is_nothrow_copy_assignable_v<_Tp>
On 18/07/25 2:09 am, Segher Boessenkool wrote:
> So, how do you resolve it in this backport, simply leaving out those
> hunks?
>
> If so: okay for 14. You don't want to backport further, right? More
> stuff becomes missing :-)
Yes, We need to ignore those patterns in this backport!
Thanks,
Kis
l email: Use caution opening links or attachments
>
>
> writes:
>> From: Soumya AR
>>
>> Hi,
>>
>> This RFC and subsequent patch series introduces support for printing and
>> parsing
>> of aarch64 tuning parameters in the form of JSON.
>
>
> On 7 May 2025, at 6:18 PM, David Malcolm wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Tue, 2025-05-06 at 14:00 +0530, soum...@nvidia.com wrote:
>> From: Soumya AR
>>
>> Hi,
>>
>> This RFC and subsequent pat
From: Soumya AR
This patch adds a get_map () method to the JSON object class to provide access
to the underlying hash map that stores the JSON key-value pairs.
It also reorganizes the private and public sections of the class to expose the
map_t typedef, which is the return type of get_map
From: Soumya AR
To allow runtime updates to tuning parameters, the const keyword is removed from
the members of the tune_params structure and the members of its nested
structures.
Since this patch also touches tuning structures in the arm backend, it was
bootstrapped on aarch64-linux-gnu as
From: Soumya AR
This patch adds support for loading custom CPU tuning parameters from a JSON
file for AArch64 targets. The '-muser-provided-CPU=' flag accepts a user
provided JSON file and overrides the internal tuning parameters at GCC runtime.
This patch was bootstrapped and re
From: Soumya AR
This patch adds functionality to dump AArch64 CPU tuning parameters to a JSON
file. The new '-fdump-tuning-model=' flag allows users to export the current
tuning model configuration to a JSON file.
This patch was bootstrapped and regtested on aarch64-linux-gnu, no
From: Soumya AR
This commit introduces a Python maintenance script that generates C++ code
for parsing and serializing AArch64 JSON tuning parameters based on the
schema defined in aarch64-json-schema.h.
The script generates two include files:
- aarch64-json-tunings-parser-generated.inc
- aa
1 - 100 of 16943 matches
Mail list logo
