> On 3 Dec 2021, at 03:12, Jeff Law wrote:
>
>
>
> On 11/22/2021 7:49 AM, Maxim Blinov wrote:
>> Hi all, apologies for forgetting to add the cover letter.
> No worries. I'd already assumed this was to support aarch64 trampolines on
> darwin by having them live elsewere as managed
Hi GCC community,
This patch will add the missed pattern described in bug 103514 [1] to the
match.pd. Tested on x86_64 Linux.
tree-optimization/103514 Missing XOR-EQ-AND Optimization
* match.pd (a & b) == (a ^ b) -> !(a | b): New optimization.
* match.pd (a & b) ^ (a == b) ->
I'm very sorry there seems to be encoding issue in the attachment
in my last email. Attached is the new patch.
Thanks,
Di Zhao
> -Original Message-
> From: Di Zhao OS
> Sent: Tuesday, November 16, 2021 1:24 AM
> To: 'Richard Biener'
> Cc: gcc-patches@gcc.gnu.org
> Subject: RE: [PATCH
Hi,
Jeff Law wrote:
> So what doesn't make sense here is how both stddef.h files get
> included. That's the core problem I think you need to resolve.
The libgo/sysinfo.c file includes stddef.h (for which the GCC version in
ginclude is used on my system) and stdlib.h which, on musl, causes an
Hi:
> Please also consider TARGET_INTER_UNIT_MOVES_TO_VEC and
> TARGET_INTER_UNIT_MOVES_FROM_VEC.
Here's updated patch.
Also honor TARGET_INTER_UNIT_MOVES_TO/FROM_VEC and in
preferred_{,out_}reload_class.
Bootstrapped and regtested on x86_64-pc-linux-gnu{-m32\ -march=k8,\ -march=k8}.
Ok?
Skip renaming if instruction is noop move, and it will
been removed for performance.
gcc/
* regrename.c (find_rename_reg): Return satisfied regno
if instruction is noop move.
---
gcc/regrename.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/gcc/regrename.c
Hi Segher,
Thanks for the review!
on 2021/11/30 上午12:57, Segher Boessenkool wrote:
> Hi!
>
> On Wed, Sep 01, 2021 at 02:55:51PM +0800, Kewen.Lin wrote:
>> This patch is to fix the inconsistent behaviors for non-LTO mode
>> and LTO mode. As Martin pointed out, currently the function
>>
Skip renaming if instruction is noop move, and it will
been removed for performance.
gcc/
* regrename.c (find_rename_reg): Return satisfied regno
if instruction is noop move.
---
gcc/regrename.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/gcc/regrename.c
Hi Mike,
on 2021/12/3 上午8:51, Michael Meissner wrote:
> On Mon, Nov 29, 2021 at 10:57:12AM -0600, Segher Boessenkool wrote:
>> Why are there OPTION_MASKs for separate P10 fusion types here, as well as
>> MASK_P10_FUSION?
>
> Well going back in time, before we used rs6000_isa_flags, we used the
On 11/24/2021 7:16 PM, Martin Sebor via Gcc-patches wrote:
When the optional size-index argument to attribute index is
omitted for a pointer, GCC expects the actual pointer argument
to point to an object at least as big as its size implies, or
at least one byte for void*. This is done to
On 11/22/2021 4:54 PM, Martin Sebor via Gcc-patches wrote:
In an effort to avoid false positives while still detecting
certain out-of-bounds accesses the warning code that handles
PHI nodes chooses the operand with the most space remaining
as the one representative of the PHI. That's not
On 11/13/2021 2:45 AM, Maxim Blinov wrote:
Implement the __builtin_nested_func_ptr_{created,deleted} functions
for the aarch64-linux platform. This serves to exercise the
infrastructure added in libgcc (--enable-off-stack-trampolines) and
gcc (-foff-stack-trampolines) in supporting off-stack
On 11/13/2021 2:45 AM, Maxim Blinov wrote:
Implement the __builtin_nested_func_ptr_{created,deleted} functions
for the x86_64-linux platform. This serves to exercise the
infrastructure added in libgcc (--enable-off-stack-trampolines) and
gcc (-foff-stack-trampolines) in supporting off-stack
On 11/13/2021 2:45 AM, Maxim Blinov wrote:
Add support for allocating nested function trampolines on an
executable heap rather than on the stack. This is motivated by targets
such as AArch64 Darwin, which globally prohibit executing code on the
stack.
The target-specific routines for
On 11/22/2021 7:49 AM, Maxim Blinov wrote:
Hi all, apologies for forgetting to add the cover letter.
No worries. I'd already assumed this was to support aarch64 trampolines
on darwin by having them live elsewere as managed entities.
The motivation of this work is to provide (limited)
On 10/28/2021 10:41 AM, Richard Purdie via Gcc-patches wrote:
On Wed, 2021-10-27 at 22:56 +0200, Bernhard Reutner-Fischer wrote:
On Wed, 27 Oct 2021 21:05:03 +0100
Richard Purdie via Gcc-patches wrote:
OpenEmbedded/Yocto Project builds libgcc and the other gcc runtime libraries
separately
On 10/28/2021 10:39 AM, Richard Purdie wrote:
On Thu, 2021-10-28 at 08:51 -0600, Jeff Law wrote:
On 10/27/2021 2:05 PM, Richard Purdie via Gcc-patches wrote:
OpenEmbedded/Yocto Project builds libgcc and the other gcc runtime libraries
separately from the compiler and slightly differently to
On 10/27/2021 2:05 PM, Richard Purdie via Gcc-patches wrote:
When building in longer build paths (200+ characters), the
"echo $(PLUGIN_HEADERS)" from the install-plugins target would cause an
"argument list too long error" on some systems.
Avoid this by calling make's sort function on the
On 6/7/2021 1:30 AM, Matt Jacobson via Gcc-patches wrote:
gcc/ChangeLog:
2020-06-07 Matt Jacobson
* config.gcc: For the AVR target, populate TM_MULTILIB_CONFIG.
* config/avr/genmultilib.awk: Add ability to filter generated multilib
list.
* config/avr/t-avr:
When a basic block A has been annotated with a count and it has only one
successor (or predecessor) B, we can propagate the A's count to B.
The algorithm without this change could leave B without an annotation if B had
other unannotated predecessors (or successors). For example, in the test case I
On 10/11/2021 4:42 AM, YunQiang Su wrote:
MIPS release 6 requires the lw/ld/sw/sd can work with
unaligned address, while it can be implemented by
full hardware or trap
Since it doesn't have to be fully done by hardware, we add a
pair of options -m(no-)unaligned-access. Kernels may need them.
On 7/6/2021 5:47 AM, H.J. Lu via Gcc-patches wrote:
On Sat, Jun 19, 2021 at 7:10 PM H.J. Lu wrote:
1. On some targets, like PowerPC, reference to ifunc function resolver
must be non-local so that compiler will properly emit PLT call. Add
TARGET_IFUNC_REF_LOCAL_OK to allow binding indirect
On Mon, Nov 29, 2021 at 10:57:12AM -0600, Segher Boessenkool wrote:
> Why are there OPTION_MASKs for separate P10 fusion types here, as well as
> MASK_P10_FUSION?
Well going back in time, before we used rs6000_isa_flags, we used the default
flag word for MASK arguments. Unfortunately, the
Hi!
On Thu, Nov 11, 2021 at 04:12:08PM -0600, Peter Bergner wrote:
> This patch adds a new effective-target function that tests whether
> it is safe to emit the ROP-protect instructions and updates the
> ROP test cases to use it.
>
> Segher, as we discussed offline, this uses the double [] which
On Fri, Dec 3, 2021, at 00:07, Mark Wielaard wrote:
> Hi Eddy,
>
> On Thu, Dec 02, 2021 at 07:35:17PM +0200, Eduard-Mihai Burtescu wrote:
>> On Thu, Dec 2, 2021, at 19:17, Mark Wielaard wrote:
>> > Rust v0 symbols can have a .suffix because if compiler transformations.
>>
>> For some context, the
On Wed, Dec 01, 2021 at 11:16:27PM -0500, Jason Merrill wrote:
> On 12/1/21 10:16, Marek Polacek wrote:
> > In r11-4758, I tried to fix this problem:
> >
> >int & = 0;
> >decltype(auto) j = i; // should behave like int & = i; error
> >
> > wherein do_auto_deduction was getting confused
Hi,
On Fri, Dec 03, 2021 at 06:58:36AM +1100, Nicholas Nethercote wrote:
> On Fri, 3 Dec 2021 at 04:17, Mark Wielaard wrote:
> >
> > * rust-demangle.c (rust_demangle_callback): Ignore everything
> > after '.' char in sym for v0.
> >
>
> I just applied this change to Valgrind's
Hi!
I discovered this bug while working on patches to remove the old built-ins
infrastructure. I missed a spot in converting from the rs6000_builtins enum to
the rs6000_gen_builtins enum. This fixes it. The fix is technically not right
if new_builtins_are_enabled were to be set to zero, but
On Thu, Dec 02, 2021 at 10:43:24AM -0600, Bill Schmidt wrote:
> The new built-in infrastructure is now enabled!
Congratulations, and thanks for all the work!
Segher
On Tue, Oct 26, 2021 at 2:45 AM apinski--- via Gcc-patches
wrote:
>
> From: Andrew Pinski
>
> So the problem here is that arm_md_asm_adjust would
> just create a set directly to the output memory which is wrong.
> It needs to output to a temp register first and then do a
> move.
>
> OK?
On Thu, Nov 18, 2021 at 5:55 PM apinski--- via Gcc-patches
wrote:
>
> From: Andrew Pinski
>
> The problem here is that aarch64_expand_setmem does not change the alignment
> for strict alignment case. This is a simplified patch from what I had
> previously.
> So constraining copy_limit to the
Hi Eddy,
On Thu, Dec 02, 2021 at 07:35:17PM +0200, Eduard-Mihai Burtescu wrote:
> On Thu, Dec 2, 2021, at 19:17, Mark Wielaard wrote:
> > Rust v0 symbols can have a .suffix because if compiler transformations.
>
> For some context, the suffix comes from LLVM (I believe as part of
> its LTO). If
Dear Fortranners,
specifying invalid constant array declarations (e.g. real array bounds)
could lead to an ICE because the array specs were checked for consistency.
A possible solution is to try an early simplification before doing that
check and was suggested by Steve.
However, regtesting did
On 12/2/2021 2:24 PM, Iain Sandoe wrote:
On 2 Dec 2021, at 21:08, Jeff Law via Gcc-patches
wrote:
On 11/22/2021 8:15 AM, Richard Biener via Gcc-patches wrote:
On Mon, Nov 22, 2021 at 3:40 PM Maxim Blinov wrote:
Hi Richard,
The purpose of this patch is to give more of the target code
> This hashing looks unnecessarily complex. The values we're hashing are
> vector SSA_NAMEs, so I think we should be able to hash and compare them
> as a plain pair of pointers.
>
> The type could then be std::pair and the hashing could be done using
> pair_hash from hash-traits.h.
>
Fancy..
On 9/22/2021 5:03 AM, Thomas Schwinge wrote:
Hi!
On 2021-09-19T11:35:00-0600, Jeff Law via Gcc-patches
wrote:
A couple of goacc tests do not have unique names.
Thanks for fixing this up, and sorry, largely my "fault", I suppose. ;-|
No worries. I suspect there's still a ton of these
> On 2 Dec 2021, at 21:08, Jeff Law via Gcc-patches
> wrote:
> On 11/22/2021 8:15 AM, Richard Biener via Gcc-patches wrote:
>> On Mon, Nov 22, 2021 at 3:40 PM Maxim Blinov
>> wrote:
>>> Hi Richard,
>>>
>>> The purpose of this patch is to give more of the target code access to
>>> the
On 12/2/2021 1:59 PM, Ilya Lipnitskiy wrote:
On Thu, Dec 2, 2021 at 12:48 PM Jeff Law wrote:
On 11/15/2021 10:13 PM, Ilya Lipnitskiy wrote:
musl only uses PT_GNU_STACK to set default thread stack size and has no
executable stack support[0], so there is no reason not to emit the
On 11/22/2021 8:15 AM, Richard Biener via Gcc-patches wrote:
On Mon, Nov 22, 2021 at 3:40 PM Maxim Blinov wrote:
Hi Richard,
The purpose of this patch is to give more of the target code access to
the cumulative_args_t structure in order to enable certain (otherwise
currently impossible)
On 11/11/2021 2:10 AM, Pekka Seppänen wrote:
Hi.
CPP/CPPFLAGS were changed by commit
84401ce5fb4ecab55decb472b168100e7593e01f. That commit uses CPP as a
default for CPP_FOR_BUILD. Unless CPP is defined, GNU make defaults
CPP as `$(CC) -E'. Given the context, this is now incorrect,
On Thu, Dec 2, 2021 at 12:48 PM Jeff Law wrote:
>
>
>
> On 11/15/2021 10:13 PM, Ilya Lipnitskiy wrote:
> > musl only uses PT_GNU_STACK to set default thread stack size and has no
> > executable stack support[0], so there is no reason not to emit the
> > .note.GNU-stack section on musl builds.
> >
On Thu, Dec 02, 2021 at 03:32:58PM -0500, Jason Merrill wrote:
> > So IMHO we need the patch I've posted (with the testcases slightly adjusted
> > not to do those extra copies afterwards for now),
> > and try to deal with the lvalue-to-rvalue conversion of indeterminate later,
> > and once done,
On 11/15/2021 10:13 PM, Ilya Lipnitskiy wrote:
musl only uses PT_GNU_STACK to set default thread stack size and has no
executable stack support[0], so there is no reason not to emit the
.note.GNU-stack section on musl builds.
[0]:
On 12/1/21 02:57, Richard Biener wrote:
On Tue, 30 Nov 2021, Jason Merrill wrote:
On 11/29/21 10:03, Richard Biener via Gcc-patches wrote:
This cleans up unreachable code diagnosed by -Wunreachable-code-ctrl.
It largely follows the previous series but discovers a few extra
cases, namely dead
On 12/1/21 04:46, Jakub Jelinek wrote:
On Tue, Nov 30, 2021 at 03:19:19PM -0500, Jason Merrill wrote:
On 11/30/21 07:17, Jakub Jelinek wrote:
On Mon, Nov 29, 2021 at 10:25:58PM -0500, Jason Merrill wrote:
It's a DR. Really, it was intended to be part of C++20; at the Cologne
meeting in 2019
On Fri, 3 Dec 2021 at 04:17, Mark Wielaard wrote:
>
> * rust-demangle.c (rust_demangle_callback): Ignore everything
> after '.' char in sym for v0.
>
I just applied this change to Valgrind's copy of rust-demangle.c and I can
confirm that it works -- the symbols that were failing
We need to revise the PCH memory allocation scheme to enable
support for PIE on aarch64. The rewrite uses a similar scheme
to the one used on Linux.
We attempt to identify VM segments for each arch/OS version that
are always available to the compiler (note this is not general,
it only needs to
---
htdocs/projects/cfg.html | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/htdocs/projects/cfg.html b/htdocs/projects/cfg.html
index dac75c68..eaa71687 100644
--- a/htdocs/projects/cfg.html
+++ b/htdocs/projects/cfg.html
@@ -472,7 +472,7 @@ Compilation Tools for Alpha;
FreeBSD 1 and FreeBSD 2, both still a.out, have been end of life for
over two decades and GCC has not been supporting them for ages, too,
so simply remove references.
gcc:
* doc/install.texi (*-*-freebsd*): Remove references to
FreeBSD 1 and FreeBSD 2.
---
gcc/doc/install.texi |
Successfully regrtested on x86_64-pc-linux-gnu.
Pushed to trunk as r12-5757-g38a0ee2649ef236ea2763bb9cfc42dc917c7d3fd.
gcc/testsuite/ChangeLog:
PR analyzer/103526
* gcc.dg/analyzer/pr103526.c: New test.
Signed-off-by: David Malcolm
---
gcc/testsuite/gcc.dg/analyzer/pr103526.c |
We've been referring to the project as GCC since GCC 2.95.
---
htdocs/git.html | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/htdocs/git.html b/htdocs/git.html
index 493b5734..5fbd98bf 100644
--- a/htdocs/git.html
+++ b/htdocs/git.html
@@ -327,7 +327,7 @@ in Git.
by
On 11/18/2021 1:04 AM, Jakub Jelinek wrote:
On Mon, Nov 08, 2021 at 08:48:07PM +0100, Jakub Jelinek via Gcc-patches wrote:
On Mon, Nov 08, 2021 at 12:46:04PM +0100, Jakub Jelinek via Gcc-patches wrote:
So, if we want to make PCH work for PIEs, I'd say we can:
1) add a new GTY option, say
On 11/18/2021 11:23 PM, Jojo R wrote:
— Jojo
在 2021年11月19日 +0800 AM12:13，Jeff Law ，写道：
On 11/16/2021 7:20 PM, Jojo R via Gcc-patches wrote:
— Jojo
在 2021年11月16日 +0800 PM8:12，Richard Biener
，写道：
On Tue, Nov 16, 2021 at 12:45 PM Jojo R via
[Review for patch 1]
Joel Hutton writes:
> From e7b3017b7b5879204e9d61760a85cc84beeb4fe0 Mon Sep 17 00:00:00 2001
> From: Joel Hutton
> Date: Wed, 25 Aug 2021 14:31:15 +0100
> Subject: [PATCH 1/3] [vect-patterns] Refactor to allow internal_fn's
>
> Hi all,
>
> This refactor allows widening
On Sat, 27 Nov 2021, Petter Tomner wrote:
> Ye it is supposed to compile cleanly for 32bit too.
>
> I pushed a patch for it as a "free for all". With %zu specifiers.
Thank you, Petter. I just updated the lang/gcc12-devel port in FreeBSD
to Sunday's snapshot that has those changes, so we shall
On 12/2/21 10:27, Marek Polacek wrote:
On Wed, Dec 01, 2021 at 11:24:58PM -0500, Jason Merrill wrote:
On 12/1/21 10:16, Marek Polacek wrote:
In C++23, auto(x) is valid, so decltype(auto(x)) should also be valid,
so
void f(decltype(auto(0)));
should be just as
void f(int);
but
On 2021-12-02 12:21, Vladimir Makarov via Gcc-patches wrote:
On 2021-12-02 12:06, Vladimir Makarov wrote:
So simple problem and so many details :)
This will require that long long is at least twice as large as int
everywhere, I thought you wanted to do that only when
Prathamesh Kulkarni writes:
> Hi Richard,
> I have attached a WIP untested patch for PR96463.
> IIUC, the PR suggests to transform
> lhs = svld1rq ({-1, -1, ...}, [0])
> into:
> lhs = vec_perm_expr
> if v is vector of 4 elements, and each element is 32 bits on little
> endian target ?
>
> I am
On Thu, Dec 2, 2021, at 19:17, Mark Wielaard wrote:
> Rust v0 symbols can have a .suffix because if compiler transformations.
For some context, the suffix comes from LLVM (I believe as part of its LTO).
If this were a semantic part of a Rust symbol, it would have an encoding within
v0 (as we
On 2021-12-02 12:06, Vladimir Makarov wrote:
On 2021-12-02 11:13, Jakub Jelinek wrote:
On Thu, Dec 02, 2021 at 11:03:46AM -0500, Vladimir Makarov wrote:
--- a/gcc/ira-color.c
+++ b/gcc/ira-color.c
@@ -2797,6 +2797,7 @@ static void
setup_allocno_priorities (ira_allocno_t
Rust v0 symbols can have a .suffix because if compiler transformations.
These can be ignored it the demangled name. Which is what this patch
implements). But an alternative implementation could be to follow what
C++ does and represent these as [clone .suffix] tagged onto the
demangled name. But
Pushed to wwwdocs.
---
htdocs/gcc-12/changes.html | 5 +
1 file changed, 5 insertions(+)
diff --git a/htdocs/gcc-12/changes.html b/htdocs/gcc-12/changes.html
index 000501fb..c2b87a53 100644
--- a/htdocs/gcc-12/changes.html
+++ b/htdocs/gcc-12/changes.html
@@ -41,6 +41,11 @@ a
On 2021-12-02 11:13, Jakub Jelinek wrote:
On Thu, Dec 02, 2021 at 11:03:46AM -0500, Vladimir Makarov wrote:
--- a/gcc/ira-color.c
+++ b/gcc/ira-color.c
@@ -2797,6 +2797,7 @@ static void
setup_allocno_priorities (ira_allocno_t *consideration_allocnos, int n)
{
int i, length, nrefs,
I'd like to push this to trunk to fix PR64135.
It think this fixes the use of a non-reserved name, while preserving the
ABI as far as class layout goes. It will break any programs which rely
on std::allocator having a __gnu_cxx::new_allocator base class, e.g. so
that
Apart from "don't bother changing the COW string", does anybody see a
reason we shouldn't do this? This passes all tests for normal COW
strings and fully-dynamic COW strings.
When non-const references, pointers or iterators are obtained to the
contents of a COW std::basic_string, the
Tested powerpc64le-linux using --enable-fully-dynamic-string and
x86_64-linux using the default config. Pushed to trunk.
The move constructor for the fully-dynamic std::basic_string was not
noexcept until recently, so the std::logic_error and std::runtime_error
move constructors were defined to
On Wed, 24 Nov 2021 at 13:25, Jonathan Wakely via Libstdc++
wrote:
>
> I plan to commit this Real Soon. Please yell if you need these
> alternative std::allocator back-ends to stay (and explain how you're
> using them when they've been broken for years, and start sending test
> results to the
On Wed, 1 Dec 2021 at 18:24, Jonathan Wakely wrote:
>
> On Wed, 1 Dec 2021 at 18:16, Florian Weimer wrote:
> >
> > * Jonathan Wakely via Libstdc:
> >
> > > diff --git a/libstdc++-v3/include/bits/cow_string.h
> > > b/libstdc++-v3/include/bits/cow_string.h
> > > index ced395b80b8..4fae1d02981
On Thu, Dec 02, 2021 at 04:31:36PM +, Andrew Stubbs wrote:
> On 02/12/2021 16:05, Andrew Stubbs wrote:
> > On 02/12/2021 12:58, Jakub Jelinek wrote:
> > > > I've tried modifying offload_handle_link_vars but that spot
> > > > doesn't catch
> > > > the omp_data_sizes variables emitted by
> > > >
Hi!
On 12/1/21 5:00 PM, Segher Boessenkool wrote:
> On Thu, Nov 18, 2021 at 10:36:52AM -0600, Bill Schmidt wrote:
>> Hi! This is the last patch broken out of the previous test suite patch
>> for the new builtins support.
> Whew :-)
>
>> One advantage of the new builtins support is uniform error
On 11/26/2021 9:57 AM, soeren--- via Gcc-patches wrote:
From: Sören Tempel
The stddef.h header checks/sets various hardcoded toolchain/os specific
macro guards to prevent redefining types such as ptrdiff_t, wchar_t, or
size_t. However, without this patch, the file does not check/set the
On 02/12/2021 16:05, Andrew Stubbs wrote:
On 02/12/2021 12:58, Jakub Jelinek wrote:
I've tried modifying offload_handle_link_vars but that spot doesn't
catch
the omp_data_sizes variables emitted by
libgomp.c-c++-common/target_42.c,
which was one of the motivating examples.
Why doesn't catch
This package is a leftover after the change for using the base
compiler's runtime during bootstrap. It is not used anymore.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* gcc-interface/a-assert.ads, gcc-interface/a-assert.adb: Remove.diff --git
This prevents gigi from back-annotating a maximum size for the Esize of
limited record and concurrent types, in keeping with the implementation
of Analyze_Object_Declaration for objects of these types.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* gcc-interface/decl.c
This restores the packing for a record type that contains a component of
an array type whose component type is a record type with a subcomponent
that is of a discriminated record type with variable size.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* gcc-interface/gigi.h
This replicates in a simpler setting the proofs performed for generic
unit System.Arith_Double.
Note that it makes a difference to declare functions Big as expression
functions here instead of renamings, as some checks are not proved with
renamings, so expression functions are used instead.
This patch is the first part of a correction for issues in the compiler
whereby finalization of a heap-allocated class-wide type may cause an
invalid memory read at runtime when the type in question contains a
component whose type has alignment specified.
Tested on x86_64-pc-linux-gnu, committed
Filter out nodes not coming from source before emitting the warning.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* sem_aggr.adb (Resolve_Array_Aggregate): Filter out nodes not
coming from source before emitting the warning.diff --git a/gcc/ada/sem_aggr.adb
This resets the internal flags associated with the -gnatD and -gnatG
switches once the generated code is printed, so that it does not end
up being printed twice in case something goes wrong after this point.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* sprint.adb
The proof of System.Arith_Double contained 11 occurrences of pragma
Annotate to justify unproved checks, 10 of which for calls to
Raise_Error denoting a case where the input leads to a division by zero
or an overflow, and one for a loop invariant regarding an assumption on
Single_Size. That should
This replicates the proofs performed for 'Width on modular integers to
the units that support 'Width on signed integers. Also add a minimal
postcondition to the System.Width_U.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* libgnat/s-widint.ads: Mark in SPARK.
*
CodePeer issues a spurious message on the analysis of the loop variant
in s-widthi.adb. Similarly to the same for s-widthu.adb, add a pragma
Annotate to justify this message.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* libgnat/s-widthi.adb: Add pragma Annotate.diff
Current implementation of Is_Suspension_Object is a leftover from an old
code of Is_Descendant_Of_Suspension_Object, which used RTE_Available and
indeed couldn't be called from GNATprove.
Now Is_Descendant_Of_Suspension_Object can work with Is_RTE, which can
be safely called from GNATprove.
Replace calls to Ensure_Freeze_Node & Append_Freeze_Actions (plural)
with a simple call to Append_Freeze_Action (singular), which calls
Ensure_Freeze_Node itself.
Cleanup related to expansion of dispatching primitives for GNATprove;
semantics is unaffected.
Tested on x86_64-pc-linux-gnu,
I'd like to ping the following patch.
The test case in the previously OK'd fix for PR101324 depends on this change,
so I've had to hold off committing it until this is in. Thanks.
https://gcc.gnu.org/pipermail/gcc-patches/2021-November/584208.html
Peter
Message-ID:
This patch adds a
Make bracket syntax array and container aggregates available with
-gnat2022. Parenthesis syntax is not accepted anymore for container
aggregate but can still be used for array aggregates. The latter is
considered obsolete and a warning is emitted with -gnatwj.
The warning is also temporarily
Assorted cleanups related to expansion of dispatching primitives on
derived types for GNATprove; semantics of the compiler is unaffected.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* doc/gnat_rm/standard_and_implementation_defined_restrictions.rst
Fix style issues in GNAT after spotting a similar problem in one of the
tests for GNATprove. These are easy to find with grep, but still require
a manual inspection, because we want to preserve layout in declaration
lists like:
A := 1;
B := 10;
C := 100;
Tested on x86_64-pc-linux-gnu,
This makes it possible for the freezing code to let the back-end
establish a proper order of elaboration of package and subprogram
instantiations in more cases, in particular with circularities, by
placing freeze nodes for them later in the expanded code in these cases.
Tested on
Code cleanup; all routines in Ada.Task_Identification had pragma Inline
except Activation_Is_Complete (and Image, which should rather stay like
that).
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* libgnarl/a-taside.ads (Activation_Is_Complete): Add pragma
For GNAT the dispatching equality is generated in expansion of freezing
nodes. For GNATprove this expansion needs to be specifically enabled and
must occur in a carefully setup context.
Previously type freezing for GNATprove only involved building of the DIC
procedure, which didn't require much
For GNAT we create dispatching equality together with other dispatching
routines (e.g. for stream reading and writing). For GNATprove we need to
create only the dispatching equality. This patch separates it, so that
it can be created from a SPARK-specific expansion.
Tested on x86_64-pc-linux-gnu,
Due to a latent bug with freezing, the expression function with
Subprogram_Variant which was recently added to the System.Val_Util unit
triggers a crash in CodePeer.
Ordinary compilation was not affected by this bug, because of the
Assertion_Policy (Ghost => Ignore) applied to this unit.
Tested
For warning about unassigned out parameter we had a loop over all formal
parameters which contained another loop over all formal parameters. This
was inefficient.
Cleanup related to spurious warnings about 'Initialized.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
*
When we add an entry to the implicit with chain of a unit while adding
an RTS unit, we must not mark it as a ignored ghost statement because it
points to the next with in the chain.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* rtsfind.adb (Maybe_Add_With): Ensure that
Iterate over record components with First_Component/Next_Component and
not with First_Entity/Next_Entity.
Change in Sem_Warn unit is related to spurious warning on 'Initialized;
other occurrences of the same pattern were found with grep.
Tested on x86_64-pc-linux-gnu, committed on trunk
Code for warnings about read-before-write of a variable had one
condition in the outer IF statement and then the very same condition in
an inner IF statement.
Cleanup related to spurious warning on 'Initialized.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* sem_warn.adb
Vx7r2cert ppc/ppc64 targets use the gcc toolchain, which requires a
slightly different forumulation of libraries for the
rts-light-tasking-rtp runtime.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* vxworks7-cert-rtp-link__ppcXX.spec: New file.
* Makefile.rtl: Use
This proves the functionality of Interfaces.C with GNATprove. Ada RM
specifications are added as comments in the spec, next to the
formalization of specifications as contracts.
As a side-effect of this proof, fix two errors in the 4 procedure
versions of To_Ada, which may raise a range check
Description of Activation_Is_Complete was amended in AI12-0231-1. This
routine raises a Program_Error when called with Null_Task_Id. Add an
explicit contract to make GNATprove aware of the restriction.
Tested on x86_64-pc-linux-gnu, committed on trunk
gcc/ada/
* libgnarl/a-taside.ads
1 - 100 of 148 matches
Mail list logo
