On Wed, Dec 20, 2023 at 3:54 AM Alexandre Oliva wrote:
>
>
> Builtin expanders for memset and memcpy may involve conditionals and
> loops, but their sequences may be end up emitted in edges. Alas,
> commit_one_edge_insertion rejects sequences that end with a jump, a
> requirement that makes
On Wed, Dec 20, 2023 at 12:51 AM Alexandre Oliva wrote:
>
> On Dec 15, 2023, Richard Biener wrote:
>
> > It might be worth amending the documentation in case this
> > is unexpected to users?
>
> Oh, yes indeed, thanks!
>
> Here's a patch that brings relevant parts of the implementation comment
>
On Thu, Dec 14, 2023 at 2:23 AM wrote:
>
> From: Vladimir Mezentsev
>
> This is fixes for releases/gcc-13 for 31109 gprofng not built and installed
> in a combined binutils+gcc build
> I only cherry-picked 24552056fd5fc677c0d032f54a5cad1c4303d312 and tested my
> build.
I don't think a
On Tue, Dec 19, 2023 at 6:41 PM Jason Merrill wrote:
>
> On 12/11/23 22:00, Jason Merrill wrote:
> > OK for trunk?
>
> Ping. CCing Alex because this could plausibly be considered build
> machinery, and he's had useful feedback on my sh code before.
OK in case Alex doesn't have any comments.
On Tue, 19 Dec 2023, Andrew Pinski wrote:
> On Tue, Dec 19, 2023 at 2:40?AM Richard Sandiford
> wrote:
> >
> > Richard Biener writes:
> > > On Tue, 19 Dec 2023, juzhe.zh...@rivai.ai wrote:
> > >
> > >> Hi, Richard.
> > >>
> > >> After investigating the codes:
> > >> /* Return true if EXPR is
+ (and:VI
+ (match_operand:VI 3 "register_operand" "vr, vr, vr, vr")
+ (not:VI (match_operand:VI 4 "register_operand" "vr, vr, vr, vr")))
Swap the order:
(not:VI (match_operand:VI 4 "register_operand" "vr, vr, vr, vr")
(and:VI
(match_operand:VI 3
Patch v4: Add some RV32 vx constraint testcase.
Patch v3: Refine crypto vector api-testing cases.
Patch v2: Update march info according to the change of riscv-common.c
This patch add crypto vector api-testing cases based on
Patch v4: Add process of SEW=64 in RV32 system.
Patch v3: Moidfy constrains for crypto vector.
Patch v2: Add crypto vector ins into RATIO attr and use vr as
destination register.
This patch add the crypto machine descriptions(vector-crypto.md) and
some new iterators which are used by crypto
Patch v4:Merge crypto vector function.def into vector.
Patch v3:Define a shape for vaesz and merge vector-crypto-types.def
into riscv-vector-builtins-types.def.
Patch v2:Optimize function_shape class for crypto_vector.
This patch add the intrinsic funtions of crypto vector based on the
On Mon, 4 Dec 2023 at 14:44, Prathamesh Kulkarni
wrote:
>
> On Thu, 23 Nov 2023 at 17:06, Prathamesh Kulkarni
> wrote:
> >
> > Hi Richard,
> > For the test-case mentioned in PR111702, compiling with -O2
> > -frounding-math -fstack-protector-all results in following ICE during
> > cse2 pass:
> >
+ bool diff_seq_repeated_p = true;
+
+ for (unsigned i = 0; i < v.npatterns (); i++)
+ {
+ poly_int64 diff_0 = rtx_to_poly_int64 (builder.elt (i)) - i;
+ poly_int64 diff_1 = rtx_to_poly_int64 (
+ builder.elt
From: Pan Li
This patch would like to fix the below execution failure.
FAIL: gcc.dg/vect/pr92420.c -flto -ffat-lto-objects execution test
The will be one single step const vector like { -4, 4, -3, 5, -2, 6, -1, 7,
...}.
For such const vector generation with single step, we will generate vid
+
While trying to fix bugs of PR113097, notice this following situation we
generate redundant vsetvli
_255 = SELECT_VL (3, POLY_INT_CST [4, 4]);
COND_LEN (..., _255)
Before this patch:
vsetivli a5, 3...
...
vadd.vv (use a5)
After this patch:
...
vadd.vv (use AVL = 3)
The reason we can do this
This patch fixes bugs in the fusion of this following case:
li a5,-1
vmv.s.x v0,a5 -> demand any non-zero AVL
vsetvli a5, ...
Incorrect fusion after VSETVL PASS:
li a5,-1
vsetvli a5...
vmv.s.x v0, a5 --> a5 is modified as incorrect value.
We disallow this incorrect fusion above.
Full coverage
On Dec 20, 2023, Jeff Law wrote:
> So the strub tests in c-c++-common are problematical. They get run
> twice, once for C, once for C++. Yet the name of the test is the same
> in both runs. (by the name, I mean the name emitted into the dejagnu
> summary and log files).
> Thus if you have a
Here is a new version of this patch.
The previous one had some flaws that were unnoticed by testsuite tests,
only the performance tests were spotting it. So I'm adding checks on the
consistency of the unordered containers in this patch.
I also forget to signal that after this patch gnu
On 12/19/23 02:53, Sergei Lewis wrote:
gcc/ChangeLog
* config/riscv/riscv-protos.h (riscv_vector::expand_vec_setmem): New
function
declaration.
* config/riscv/riscv-string.cc (riscv_vector::expand_vec_setmem): New
function: this generates an inline vectorised memory
On 12/19/23 02:53, Sergei Lewis wrote:
gcc/ChangeLog
* config/riscv/riscv.md (movmem): Use
riscv_vector::expand_block_move,
if and only if we know the entire operation can be performed using one
vector
load followed by one vector store
gcc/testsuite/ChangeLog
PR
On 12/19/23 10:21, Siddhesh Poyarekar wrote:
Narrow down scope of the unknowns bitmap so that it is only accessible
within the reexamination process. This also removes any role of unknown
propagation from object_sizes_set, thus simplifying that code path a
bit.
gcc/ChangeLog:
*
On 12/11/23 20:00, Jason Merrill wrote:
OK for trunk?
-- 8< --
I thought it could be easier to use check_GNU_style.py. With this alias,
'git gcc-style' will take a git revision as argument instead of a file, or
check HEAD if no argument is given.
contrib/ChangeLog:
*
On 12/11/23 21:44, Alexandre Oliva wrote:
untyped calls: use wrapper class type for implicit plus_one
Instead of get and set macros to apply a delta, use a single macro
that resorts to a temporary wrapper class to apply it.
To be combined (or not) with the previous patch.
I'd be OK with
So the strub tests in c-c++-common are problematical. They get run
twice, once for C, once for C++. Yet the name of the test is the same
in both runs. (by the name, I mean the name emitted into the dejagnu
summary and log files).
Thus if you have a test in there which passes in one
Dear all,
In the spirit of c/c++ using the .i/.ii extensions for intermediates,
use the .fi/.fii intermediate extensions for gfortran fixed/free form
sources when -save-temps is invoked to avoid various issues.
The documentation part will be submitted separately, because it
involves adding a
Oh, I see. Thanks Jeff for suggestion, will refine the commit log in V2.
Pan
-Original Message-
From: Jeff Law
Sent: Wednesday, December 20, 2023 12:03 PM
To: juzhe.zh...@rivai.ai; Li, Pan2 ; gcc-patches
Cc: Wang, Yanzhang ; kito.cheng
Subject: Re: [PATCH v1] RISC-V: Bugfix for the
Pushed to the trunk.
jeffcommit e56dc0003729ea6f7d26594dae34d218543edb49
Author: Jeff Law
Date: Tue Dec 19 21:28:03 2023 -0700
Document that bitfields are signed on mcore now.
diff --git a/htdocs/gcc-14/changes.html b/htdocs/gcc-14/changes.html
index eb14e09d..11c7ca7e 100644
---
The GCC manual has a whole section on signedness of bitfields with the
ultimate conclusion that the property really isn't an ABI issue, but
instead a C dialect issue (agreed). Furthermore it concludes that all
targets should behave the same by default.
So it was a mistake for the mcore
On 12/19/23 19:50, juzhe.zh...@rivai.ai wrote:
+ if (known_eq (ele_0 - 0, ele_n - v.npatterns ()))
->
for (i = 0; i < v.npatterns (); )
check each nelt of npatterns is equal to vid.
Pan -- please indicate what testing was performed. The standard is to
test with and without the
The patches are look ok and no problem in spec2017 correctness check.
On 2023/12/19 15:26, chenglulu wrote:
We will read and test these patches as soon as possible.
Thanks!
在 2023/12/19 下午2:59, Xi Ruoyao 写道:
Superseds
https://gcc.gnu.org/pipermail/gcc-patches/2023-December/640871.html.
Per
On 12/19/23 19:32, Alexandre Oliva wrote:
Hi, Jeff,
On Dec 18, 2023, Jeff Law wrote:
These are timing sporadically on the embedded platforms. Given they
include a test that has a timeout factor, it seems to me you should
duplicate the timeout factor in the new tests.
Remember when you
On 12/6/23 22:11, Ken Matsui wrote:
This patch series optimizes type traits compilation performance by
implementing built-in type traits and using them in libstdc++.
I'm finding that all the new g++.dg/ext/is_*.C testcases added by this patch
series are failing due
to the "#include ". If
Builtin expanders for memset and memcpy may involve conditionals and
loops, but their sequences may be end up emitted in edges. Alas,
commit_one_edge_insertion rejects sequences that end with a jump, a
requirement that makes sense for insertions after expand, but not so
much during expand.
+ if (known_eq (ele_0 - 0, ele_n - v.npatterns ()))
->
for (i = 0; i < v.npatterns (); )
check each nelt of npatterns is equal to vid.
juzhe.zh...@rivai.ai
From: pan2.li
Date: 2023-12-20 10:39
To: gcc-patches
CC: juzhe.zhong; pan2.li; yanzhang.wang; kito.cheng
Subject: [PATCH v1]
From: Pan Li
For generating the const vector with single step, we have code
gen similar as below. We have npatterns = 4.
v1= {3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8... }
v2 (diff) = {3 - 0, 2 - 1, 1 - 2, 0 - 3, 7 - 4, 6 - 5, 5 - 6, 4 - 7...}
= {3, 1, -1, 3, 3, 1, -1, 3 ...}
v1 = vd +
Hi, Jeff,
On Dec 18, 2023, Jeff Law wrote:
> These are timing sporadically on the embedded platforms. Given they
> include a test that has a timeout factor, it seems to me you should
> duplicate the timeout factor in the new tests.
> Remember when you include another file, the dg- directives
Thanks Andrew.
Ok. I think I can apply suggestions from both you and Richards.
That is, fix it in const_binop but with creating new helper function
simplify_const_binop (suggestion from Richard).
But I don't know how to write codes in simplify_const_ binop... Could you give
me some hints ?
On Fri, Dec 15, 2023 at 10:34 AM Haochen Jiang wrote:
>
> Hi all,
>
> There is a recent change in AVX10 documentation which allows 64 bit mask
> register instructions in AVX10-256, the documentation comes following:
>
> Intel Advanced Vector Extensions 10 (Intel AVX10) Architecture Specification
On Tue, Dec 19, 2023 at 2:40 AM Richard Sandiford
wrote:
>
> Richard Biener writes:
> > On Tue, 19 Dec 2023, juzhe.zh...@rivai.ai wrote:
> >
> >> Hi, Richard.
> >>
> >> After investigating the codes:
> >> /* Return true if EXPR is the integer constant zero or a complex constant
> >>of zero,
On Thu, 14 Dec 2023, David Malcolm wrote:
> Are these OK for trunk, assuming I followup with adding CI for this?
> (that said, I disappear for the rest of 2023 at the end of this week, so
> I'd work on the CI in early January)
Patches 2 (updated at the time of commit to reflect the latest
On Thu, 14 Dec 2023, David Malcolm wrote:
> diff --git a/gcc/doc/sourcebuild.texi b/gcc/doc/sourcebuild.texi
> index 26a7e9c35070..9a394b3e2c77 100644
> --- a/gcc/doc/sourcebuild.texi
> +++ b/gcc/doc/sourcebuild.texi
> @@ -813,6 +813,10 @@ options supported by this target (@pxref{Run-time
>
On Dec 15, 2023, Richard Biener wrote:
> I think __noipa__ is more complete and will make the libgcc functions appear
> as black boxes to callers.
I was hesitant to use __noipa__ because I thought it might disable
relevant optimizations even when not using LTO, but it is likely safer
in the
On Dec 15, 2023, Richard Biener wrote:
> It might be worth amending the documentation in case this
> is unexpected to users?
Oh, yes indeed, thanks!
Here's a patch that brings relevant parts of the implementation comment
to the user-facing documentation, so that it reflects the change in
On 2023-12-19 17:57, Jakub Jelinek wrote:
On Mon, Dec 18, 2023 at 11:42:52AM -0500, Siddhesh Poyarekar wrote:
It is always safe to set the computed bit for dynamic object sizes at
the end of collect_object_sizes_for because even in case of a dependency
loop encountered in nested calls, we have
Hello-
May I please ping this one? Thanks...
https://gcc.gnu.org/pipermail/gcc-patches/2023-November/638692.html
-Lewis
On Wed, Nov 29, 2023 at 7:05 PM Lewis Hyatt wrote:
>
> On Thu, Nov 09, 2023 at 04:16:10PM -0500, Lewis Hyatt wrote:
> > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=111918
>
On Mon, Dec 18, 2023 at 11:42:52AM -0500, Siddhesh Poyarekar wrote:
> It is always safe to set the computed bit for dynamic object sizes at
> the end of collect_object_sizes_for because even in case of a dependency
> loop encountered in nested calls, we have an SSA temporary to actually
> finish
On Tue, Dec 19, 2023 at 05:19:35PM -0500, Marek Polacek wrote:
> Should be obvious, but... tested on x86_64, ok for trunk?
>
> -- >8 --
>
> PR tree-optimization/113069
>
> gcc/ChangeLog:
>
> * gimple-ssa-sccopy.cc (scc_discovery): Remove unused member.
Ok, thanks.
> diff --git
On Sun, 17 Dec 2023, Martin Uecker wrote:
> + tree f = build_decl (input_location, FIELD_DECL, DECL_NAME (a),
> +composite_type_internal (ta, tb, cache));
> +
> + DECL_PACKED (f) = DECL_PACKED (a);
> + SET_DECL_ALIGN (f, DECL_ALIGN
Should be obvious, but... tested on x86_64, ok for trunk?
-- >8 --
PR tree-optimization/113069
gcc/ChangeLog:
* gimple-ssa-sccopy.cc (scc_discovery): Remove unused member.
---
gcc/gimple-ssa-sccopy.cc | 1 -
1 file changed, 1 deletion(-)
diff --git a/gcc/gimple-ssa-sccopy.cc
Tested x86_64-pc-linux-gnu, OK for trunk?
-- 8< --
-Werror=foo implying -Wfoo wasn't working for -Wdeprecated-copy-dtor,
because it is specified as the value 2 of warn_deprecated_copy, which shows
up as CLVC_EQUAL, which is not one of the three var_types handled by
control_warning_option. It
On Sun, 17 Dec 2023, Martin Uecker wrote:
> +/* While tese tests check that incompatible definitions
> + * of enums can alias. */
s/tese/these/
Patch 3 is OK with that fix.
--
Joseph S. Myers
jos...@codesourcery.com
On Sun, 17 Dec 2023, Martin Uecker wrote:
> diff --git a/gcc/testsuite/gcc.dg/c23-tag-enum-1.c
> b/gcc/testsuite/gcc.dg/c23-tag-enum-1.c
> new file mode 100644
> index 000..a81a5afc456
> --- /dev/null
> +++ b/gcc/testsuite/gcc.dg/c23-tag-enum-1.c
> +void test2(void)
> +{
> + enum ee
On Sun, 17 Dec 2023, Martin Uecker wrote:
> Here is the revised series. The first three patches only
> have changes in the tests as well as the return value
> changes. The fourth patch was now also revised,
> with changes and tests to make sure that the composite
> type works correctly for
PING.
If the patch is approved, can anyone apply this patch.
I don't have permissions for `git push`.
Thank you,
-Vladimir
Forwarded Message
Subject:[PATCH] gprofng: a new GNU profiler
Date: Wed, 13 Dec 2023 17:23:01 -0800
From: vladimir.mezent...@oracle.com
To:
Hi Julian & Thomas,
the patch LGTM - and seemingly also Thomas is somewhat fine with it -
and it includes the stand-alone testcase.
* * *
I guess, you don't know the answer to Thomas question, i.e. whether
that's a bug in CUDA or in our use of the CUDA API?
CUDA's spec itself,
On 12/19/23 14:08, Patrick Palka wrote:
On Mon, 1 May 2023, Jason Merrill wrote:
Tested x86_64-pc-linux-gnu, applying to trunk.
Patrick, can you verify that this resolves 109506 and add whatever testcase(s)
seem appropriate from that PR?
-- 8< --
Here it turns out I also needed to adjust
On Mon, 18 Dec 2023, Jakub Jelinek wrote:
> Hi!
>
> The following patch changes -Walloc-size warning to no longer warn
> about int *p = calloc (1, sizeof (int));, because as discussed earlier,
> the size is IMNSHO sufficient in that case, for alloc_size with 2
> arguments warns if the product of
On Mon, 1 May 2023, Jason Merrill wrote:
> Tested x86_64-pc-linux-gnu, applying to trunk.
>
> Patrick, can you verify that this resolves 109506 and add whatever testcase(s)
> seem appropriate from that PR?
>
> -- 8< --
>
> Here it turns out I also needed to adjust cfun when stepping out of the
Am Dienstag, dem 19.12.2023 um 12:20 -0500 schrieb Jason Merrill:
> On 12/19/23 03:47, Jakub Jelinek wrote:
> > On Tue, Dec 19, 2023 at 08:11:11AM +0100, Martin Uecker wrote:
> > > Am Montag, dem 18.12.2023 um 20:14 +0100 schrieb Jakub Jelinek:
> > > > Hi!
> > > >
> > > > The following patch
On 12/11/23 22:00, Jason Merrill wrote:
OK for trunk?
Ping. CCing Alex because this could plausibly be considered build
machinery, and he's had useful feedback on my sh code before.
-- 8< --
I thought it could be easier to use check_GNU_style.py. With this alias,
'git gcc-style' will
On 12/12/23 14:35, David Faust wrote:
The BTF specification does not formally define a representation for
forward-declared enum types such as:
enum Foo;
Forward-declarations for struct and union types are represented by
BTF_KIND_FWD, which has a 1-bit flag distinguishing the two.
The
Narrow down scope of the unknowns bitmap so that it is only accessible
within the reexamination process. This also removes any role of unknown
propagation from object_sizes_set, thus simplifying that code path a
bit.
gcc/ChangeLog:
* tree-object-size.cc (object_size_info): Remove
On 12/19/23 03:47, Jakub Jelinek wrote:
On Tue, Dec 19, 2023 at 08:11:11AM +0100, Martin Uecker wrote:
Am Montag, dem 18.12.2023 um 20:14 +0100 schrieb Jakub Jelinek:
Hi!
The following patch changes -Walloc-size warning to no longer warn
about int *p = calloc (1, sizeof (int));, because as
On Tue, 12 Dec 2023 04:08:09 PST (-0800), jia...@iscas.ac.cn wrote:
Supports RISC-V profiles[1] in -march option.
Default input set the profile is before other formal extensions.
V2: Fixes some format errors and adds code comments for parse function
Thanks for Jeff Law's review and comments.
Hi Julian,
On 16.12.23 14:25, Julian Brown wrote:
OpenMP/OpenACC: Reorganise OMP map clause handling in gimplify.cc
This patch has been separated out from the C++ "declare mapper"
support patch. It contains just the gimplify.cc rearrangement
work, mostly moving
Marc Poulhiès writes:
>> Perhaps this script could also deal directly with Sphinx-generated
>> HTML?
>
> I investigated a bit... The Ada part doesn't handle the html target, so
> it's expected you don't have anything to parse. The online docs are
> generated using a different script, not using
From: Eric Botcazou
The problem is that the predicated subtype does not inherit all the required
attributes of a string subtype with a static predicate.
gcc/ada/
* sem_ch3.adb (Analyze_Subtype_Declaration): Remove a short-circuit
for subtypes without aspects when it comes to
From: Gary Dismukes
The compiler fails to reject a container aggregate written using positional
notation when the container type specifies an Add_Named operation in its
Aggregate aspect. Container aggregates for such types must be written using
named associations. The compiler ignores the
From: Sheri Bernstein
Remove GNATcheck violations by refactoring code and also using
pragma Annotate to exempt them.
gcc/ada/
* libgnat/a-comlin.adb (Argument_Count): Rewrite code so there is
only one return, to remove Improper_Returns violation.
(Command_Name): Add
From: Patrick Bernardi
It was previously assumed that configurable runtimes could not return exit
statuses, however this assumption no longer holds. Instead, only import
the required symbols from Ada.Command_Line's support packages if
Ada.Command_Line is in the closure of the partition when a
From: Eric Botcazou
The flag is set on the constructed subtype of an object with unconstrained
nominal subtype that is aliased and is used by the code generator to adjust
the layout of the object.
But it is actually only used for array subtypes, where it determines whether
the object is
From: Eric Botcazou
Ancestor_Type is overloaded with Aggregate_Bounds on N_Aggregate nodes
so its access needs to be guarded in Copy_Generic_Node.
gcc/ada/
* sem_ch12.adb (Copy_Generic_Node): Add guard for Ancestor_Type.
Tested on x86_64-pc-linux-gnu, committed on master.
---
From: Eric Botcazou
This expands on the reason for properly guarding the transformation.
gcc/ada/
* exp_ch6.adb (Expand_Ctrl_Function_Call): Rework last comment.
Tested on x86_64-pc-linux-gnu, committed on master.
---
gcc/ada/exp_ch6.adb | 7 ++-
1 file changed, 6 insertions(+),
From: Viljar Indus
gcc/ada/
* sem_ch13.adb (Valid_Default_Iterator): Check all interfaces for
valid iterator type. Also improve error reporting.
Tested on x86_64-pc-linux-gnu, committed on master.
---
gcc/ada/sem_ch13.adb | 103 +--
1
From: Vasiliy Fofanov
gcc/ada/
* libgnat/i-cstrin.adb (Value): Optimize.
Tested on x86_64-pc-linux-gnu, committed on master.
---
gcc/ada/libgnat/i-cstrin.adb | 17 ++---
1 file changed, 10 insertions(+), 7 deletions(-)
diff --git a/gcc/ada/libgnat/i-cstrin.adb
From: Gary Dismukes
The compiler hangs (or may crash, if assertions are enabled) when compiling
an iterated association of a container aggregate that has a key expression
given by a function call. The resolution of the call leads to a blowup in
Build_Call_Marker, because the temporary copy of
From: Eric Botcazou
This removes the setting of the Is_Ignored_Transient flag on the temporaries
needing finalization created by Expand_Ctrl_Function_Call when invoked from
within the dependent expressions of conditional expressions.
This flag tells the general finalization machinery to
From: Piotr Trojanek
Code cleanup.
gcc/ada/
* exp_ch9.adb, sem_ch10.adb, sem_util.adb: Fix comments.
Tested on x86_64-pc-linux-gnu, committed on master.
---
gcc/ada/exp_ch9.adb | 4 ++--
gcc/ada/sem_ch10.adb | 2 +-
gcc/ada/sem_util.adb | 2 +-
3 files changed, 4 insertions(+), 4
From: Piotr Trojanek
Routine Get_Logical_Line_Number_Img was introduced for splitting of
Pre/Post contracts, but subsequent patch for that feature removed its
only use. It was then used by GNATprove, but that use is now removed
as well.
gcc/ada/
* sinput.adb, sinput.ads
From: Ronan Desplanques
Before this patch, the compiler would fail to examine the corresponding
record types of concurrent types when building aggregate components.
This patch fixes this, and adds a precondition and additional documentation
on the subprogram that triggered the crash, as it never
Add the ada.html and ada.install-html targets so that we can build the
html with `make html`.
gcc/ada/
* gcc-interface/Make-lang.in (ada.html, ada.install-html): Add.
Tested on x86_64-pc-linux-gnu, committed on master.
---
gcc/ada/gcc-interface/Make-lang.in | 35
From: Piotr Trojanek
The current wording of SPARK RM 6.1.5(5) about the inputs for the
Depends contract doesn't mention "a record with at least one
unconstrained component".
gcc/ada/
* sem_prag.adb (Is_Unconstrained_Or_Tagged_Item): Update comment
and body.
Tested on
From: Johannes Kliemann
Some of our restricted runtimes support dynamic priorities. The binder
needs to generate code for a restricted runtime even if the restriction
No_Dynamic_Priorities is not in place.
gcc/ada/
* libgnat/s-rident.ads: Remove No_Dynamic_Priorities from
From: Eric Botcazou
This occurs for an aggregate of a derived tagged type in the body of the
instance, because the full view of the parent type, which was visible in
the generic construct (otherwise the aggregate would have been illegal),
is not restored in the body of the instance.
From: Yannick Moy
GNATprove supports container aggregates, except for indexed aggregates.
It needs all expressions to have suitable target types and Do_Range_Check
flags, which are added by the special expansion for GNATprove.
There is no impact on code generation.
gcc/ada/
*
From: Steve Baird
Sem_Util.Enclosing_Declaration can return a non-empty result which is not
a declaration; clients may need to compensate for the case where an
N_Subprogram_Specification node is returned. One such client is the function
Is_Actual_Subp_Of_Inst.
gcc/ada/
* sem_ch8.adb
From: Patrick Bernardi
The behaviour of the binder when handling command line arguments and exit
codes is simplified so that references to the corresponding runtime symbols
are always generated when the runtime is configured with command line
argument and exit code support. This allows
From: Eric Botcazou
The path in Expand_N_If_Expression implementing the special optimization for
an unidimensional array type and dependent expressions with static bounds
fails to call Process_Transients_In_Expression on their list of actions.
gcc/ada/
* exp_ch4.adb
From: Ronan Desplanques
This patch relaxes the requirement that discriminants values should be
known at compile time for a particular optimization to be applied. That
optimization is the one that treats an unconstrained object as constrained
when the object is of a limited type, in order to
From: Yannick Moy
Move one SPARK legality check from GNAT to GNATprove, and cleanup
other uses of SPARK_Mode for legality checking.
gcc/ada/
* sem_ch4.adb (Analyze_Selected_Component): Check correct mode
variable for GNATprove.
* sem_prag.adb (Refined_State): Call
From: Steve Baird
If G1 is a generic package and G1.G2 is a child unit (also a generic package)
then it would be illegal if some third generic unit (declared outside of G1)
takes a formal instance of G1.G2, as in "with package I2 is new G1.G2;".
This construct was incorrectly accepted in some
From: Eric Botcazou
This streamlines the submachinery that makes it so that the finalization of
temporaries created for EWAs and conditional expressions is deferred to the
enclosing context.
The original implementation was using a deep tree traversal for EWAs, which
was later restricted to
On Tue, 19 Dec 2023, Tamar Christina wrote:
> > Do you mean for ARM SVE, these tests need to be specified as only ARM SVE ?
>
> I think that would be the right thing to do. I think these tests are
> checking if we support VLA SLP.
> changing it to a PASS unconditionally means that if someone
On Tue, 19 Dec 2023, Tamar Christina wrote:
> > > > > + /* Save destination as we go, BB are visited in order and the
> > > > > last one
> > > > > + is where statements should be moved to. */
> > > > > + if (!dest_bb)
> > > > > + dest_bb = gimple_bb (c);
> > > > > + else
On 19/12/2023 13:38, Richard Sandiford wrote:
> Alex Coplan writes:
> > On 19/12/2023 10:15, Richard Sandiford wrote:
> >> Alex Coplan writes:
> >> > We were missing validation of the candidate register operands in the
> >> > ldp/stp pass. I was relying on recog rejecting such cases when we
>
Alex Coplan writes:
> On 19/12/2023 10:15, Richard Sandiford wrote:
>> Alex Coplan writes:
>> > We were missing validation of the candidate register operands in the
>> > ldp/stp pass. I was relying on recog rejecting such cases when we
>> > formed the final pair insn, but the testcase shows
> Do you mean for ARM SVE, these tests need to be specified as only ARM SVE ?
I think that would be the right thing to do. I think these tests are checking
if we support VLA SLP.
changing it to a PASS unconditionally means that if someone runs the testsuite
in SVE only mode they’ll fail.
>
Do you mean for ARM SVE, these tests need to be specified as only ARM SVE ?
Actually, for RVV, is same situation as ARM. We are using VLS modes
(fixed-length vectors) to vectorize these cases so that they are XPASS.
The difference between RVV and ARM is that: variable-length and fixed-length
在 2023/12/19 下午8:37, Xi Ruoyao 写道:
On Tue, 2023-12-19 at 19:04 +0800, Lulu Cheng wrote:
+(define_insn "@add_tls_le_relax"
+ [(set (match_operand:P 0 "register_operand" "=r")
+ (unspec:P [(match_operand:P 1 "register_operand" "r")
+ (match_operand:P 2 "register_operand"
On Tue, Dec 19, 2023 at 6:39 AM liuhongt wrote:
>
> Similar for A < B ? B : A to MAX_EXPR.
> There're codes in the frontend to optimize such pattern but failed to
> handle testcase in the PR since it's exposed at gimple level when
> folding backend builtins.
>
> pr95906 now can be optimized to
On Tue, 19 Dec 2023, Tamar Christina wrote:
> Hi All,
>
> While we don't support SLP for early break vectorization, we
> can land in the situation where the induction was vectorized
> through hybrid SLP. This means when vectorizing the early
> break live operation we need to get the results of
On Tue, 2023-12-19 at 19:04 +0800, Lulu Cheng wrote:
> +(define_insn "@add_tls_le_relax"
> + [(set (match_operand:P 0 "register_operand" "=r")
> + (unspec:P [(match_operand:P 1 "register_operand" "r")
> + (match_operand:P 2 "register_operand" "r")
> +
1 - 100 of 142 matches
Mail list logo
