Hi,
This is a fix for pr95199.
In vectorization, vinfo->ivexpr_map is supposed to catch those "IV base and/or
step expressions". Just call cse_and_gimplify_to_preheader to handle
gathering/scattering to avoid the extra variable.
Bootstrap and tested on aarch64/x86_64 Linux platform. No new reg
This patch implements several constant folding optimizations
for __builtin_parity and friends. We canonicalize popcount(x)&1
as parity(x) in gimple, and potentially convert back again when
we expand to RTL. parity(~x) is simplified to parity(x), which
is true for all integer modes with an
Hi,
Fix codegen implementation for the builtin vec_pack_to_short_fp32.
Regtests are underway against powerpc64 (power7be,power8le,power9le).
(this is a power9 builtin, so should be a noop for most of those).
OK for trunk and backports?
Thanks
-Will
[gcc]
targ
On Fri, 2020-06-05 at 13:46 +0800, Hongtao Liu via Gcc-patches wrote:
> Hi:
>
> +/* Optimize vector MUL generation for V8QI, V16QI and V32QI
> + under TARGET_AVX512BW. i.e. for v16qi a * b, it has
> +
> + vpmovzxbw ymm2, xmm0
> + vpmovzxbw ymm3, xmm1
> + vpmullw ymm4, ymm2, ymm3
> + vp
On Sat, 2020-05-02 at 04:55 -0700, H.J. Lu wrote:
> Currently patchable area is at the wrong place. It is placed immediately
> after function label, before both .cfi_startproc and ENDBR. This patch
> adds UNSPECV_PATCHABLE_AREA for pseudo patchable area instruction and
> changes ENDBR insertion p
On Wed, 2020-06-10 at 23:41 -0300, Alexandre Oliva wrote:
> On May 26, 2020, Alexandre Oliva wrote:
>
> > On May 19, 2020, Joseph Myers wrote:
> > > Allowing a missing executable name is reasonable enough, but I was
> > > actually thinking that the messages should print "gcc" or whatever
> > >
On Thu, 2020-06-11 at 10:43 +0200, Martin Liška wrote:
> Ready for master?
>
> Thanks,
> Martin
>
> gcc/ChangeLog:
>
> * cgraphunit.c (process_symver_attribute): Wrap weakref keyword.
> * dbgcnt.c (dbg_cnt_set_limit_by_index): Do not print extra new
> line.
> * lto-wrappe
Hi Segher,
on 2020/6/12 上午6:55, Segher Boessenkool wrote:
> Hi!
>
> On Wed, Jun 10, 2020 at 08:39:19PM +0800, Kewen.Lin wrote:
>> +;; Define optab for vector access with length vectorization exploitation.
>> +(define_expand "lenload"
>> + [(match_operand:VEC_A 0 "vlogical_operand")
>> + (match
Hi,
> -Original Message-
> From: Richard Sandiford [mailto:richard.sandif...@arm.com]
> Sent: Friday, June 12, 2020 2:29 AM
> To: Yangfei (Felix)
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH PR95570] vect: ICE: Segmentation fault in
> vect_loop_versioning
>
> "Yangfei (Felix)" wr
On Tue, Jun 09, 2020 at 09:51:12AM -0600, Martin Sebor wrote:
> >I think the backend declaration is wrong, the function only takes
> >a void * argument and returns a long.
>
> Thanks. In his comment on the bug Segher (CC'd) points to
> the internals manual that documents the function:
>
> https:
On 6/11/20 10:02 PM, Nicolas Bértolo wrote:
> Hi,
>
> On 6/7/20 11:12 PM, JonY wrote:
>> Ideally, libtool is used so we get libgccjit-0.dll, unfortunately it is
>> not. So the only way to ABI version the dll would be to use Unix style
>> soname to mark when an ABI has changed.
>
> I tried generat
On Tue, Jun 09, 2020 at 02:29:13PM -0600, Jeff Law wrote:
> On Sun, 2020-05-24 at 11:22 -0500, Segher Boessenkool wrote:
> > OTOH, you don't need to name Tuple at all... It should not *have* a
> > constructor, since you declared it as class... But you can just use
> > std::tuple here?
> >
> > >
Hi!
On Wed, Jun 10, 2020 at 08:39:19PM +0800, Kewen.Lin wrote:
> +;; Define optab for vector access with length vectorization exploitation.
> +(define_expand "lenload"
> + [(match_operand:VEC_A 0 "vlogical_operand")
> + (match_operand:VEC_A 1 "memory_operand")
> + (match_operand:QI 2 "int_reg
On 6/11/20 5:28 PM, Marek Polacek wrote:
On Thu, Jun 11, 2020 at 03:51:29PM -0400, Jason Merrill wrote:
On 6/9/20 2:17 PM, Marek Polacek wrote:
Another part of 95369 is that we accept designated initializers with
non-aggregate types. That seems to be wrong since they're part of
aggregate initi
On 6/5/20 5:16 PM, Patrick Palka wrote:
This patch generalizes our existing functionality for deferring access
checking of typedefs when parsing a function or class template to now
defer all kinds of access checks until template instantiation time,
including member function and member object acce
Hi,
On 6/7/20 11:12 PM, JonY wrote:
> Ideally, libtool is used so we get libgccjit-0.dll, unfortunately it is
> not. So the only way to ABI version the dll would be to use Unix style
> soname to mark when an ABI has changed.
I tried generating the library as libgccjit-0.dll and naming its import
On Thu, Jun 11, 2020 at 03:51:29PM -0400, Jason Merrill wrote:
> On 6/9/20 2:17 PM, Marek Polacek wrote:
> > Another part of 95369 is that we accept designated initializers with
> > non-aggregate types. That seems to be wrong since they're part of
> > aggregate initialization. clang/icc also reje
Hi Martin,
> The compute_objsize() function started out as a thin wrapper around
> compute_builtin_object_size(), but over time developed its own
> features to compensate for the other function's limitations (such
> as its inability to work with ranges). The interaction of these
> features and th
On 6/5/20 3:41 PM, Martin Sebor wrote:
+ location_t loc
+= EXPR_HAS_LOCATION (param) ? EXPR_LOCATION (param) : pctx->loc;
This could be EXPR_LOC_OR_LOC (param, pctx->loc)
+ location_t loc = (EXPR_HAS_LOCATION (ptr)
+ ? EXPR_LOCATION (ptr) : EXPR_LOCATION
Hi,
I am currently looking at making libgfortran use the
TOUPPER et al functions from ctype-safe.h. Adding this
in a gfortran header results in lots of
Gcc/trunk-bin/x86_64-pc-linux-gnu/./libgfortran/.libs/libgfortran.so:
undefined reference to `_sch_istable'
errors.
So, can we use this, or
Hi
We had omitted the copying of function attributes (including the
'used' status). Mark the outlined functions as artificial, since
they are; some diagnostic processing tests this.
tested on Linux and Darwin,
OK for master?
10.2?
thanks
Iain
gcc/cp/ChangeLog:
PR c++/95518
* co
On 6/9/20 2:17 PM, Marek Polacek wrote:
Another part of 95369 is that we accept designated initializers with
non-aggregate types. That seems to be wrong since they're part of
aggregate initialization. clang/icc also reject it.
(Un)fortunately there are multiple contexts where we can use design
Hi,
The reason that the code fails is that (by a somewhat complex
implied route), when a user adds a 'get-return-on-alloc-fail’
to their coroutine promise, this implies the use of some content
from ; we should not ICE because the user forgot that tho.
Jonathan and I were discussing whether there’
On 6/9/20 2:18 PM, Patrick Palka wrote:
This fixes two issues in our handling of constrained class template
friend declarations.
The first issue is that we fail to set the constraints on the injected
class template declaration during tsubst_friend_class.
The second issue is that the template pa
On Tue, 9 Jun 2020, Jeff Law via Gcc-patches wrote:
> On Tue, 2020-06-09 at 17:26 +0200, Jakub Jelinek wrote:
> > On Tue, Jun 09, 2020 at 09:18:25AM -0600, Jeff Law wrote:
> > > On Tue, 2020-06-09 at 16:59 +0200, Jakub Jelinek wrote:
> > > > On Tue, Jun 09, 2020 at 08:54:47AM -0600, Jeff Law via G
On 6/11/20 3:32 PM, Jason Merrill wrote:
On 6/10/20 5:11 PM, Marek Polacek wrote:
Another indication that perhaps this warning is emitted too early. We
crash because same_type_p gets a null type: we have an enumerator
without a fixed underlying type and finish_enum_value_list hasn't yet
run. S
On 6/11/20 3:34 PM, Jason Merrill wrote:
On 6/11/20 3:32 PM, Jason Merrill wrote:
On 6/10/20 5:11 PM, Marek Polacek wrote:
Another indication that perhaps this warning is emitted too early. We
crash because same_type_p gets a null type: we have an enumerator
without a fixed underlying type and
On 6/10/20 5:11 PM, Marek Polacek wrote:
Another indication that perhaps this warning is emitted too early. We
crash because same_type_p gets a null type: we have an enumerator
without a fixed underlying type and finish_enum_value_list hasn't yet
run. So check if the type is null before calling
My apologies in advance for a middle-end patch without a test case. The
patch below
implements a simple/safe missing transformation in the RTL optimizers, that
transforms
(A&C)^(B&C) into the equivalent (A^B)&C, when C doesn't side-effect, such as
a constant.
I originally identified this
Found by Steve. Committed as obvious.
Thanks,
Harald
PR fortran/95611 - ICE in access_attr_decl, at fortran/decl.c:9075
When reporting a duplicate access specification of an operator, refer to
the proper symbol.
2020-06-11 Harald Anlauf
gcc/fortran/
PR fortran/95611
* decl.
Hi!
On Thu, Jun 11, 2020 at 01:49:57PM +0200, Martin Liška wrote:
> Since stmt_vec_info is not generated for debug info statements, these needs
> to
> be skipped in rs6000_density_test.
> PR target/95627
> * config/rs6000/rs6000.c (rs6000_density_test): Skip debug
> statements.
Hi Thomas,
> one remark: Instead of
>
> + && !(strcmp(gfc_current_intrinsic, "associated") == 0
> + || strcmp(gfc_current_intrinsic, "null") == 0
> + || strcmp(gfc_current_intrinsic, "present") == 0))
>
> could you maybe test sym->id ?
>
> OK with that change.
done.
On 6/10/20 5:11 PM, Marek Polacek wrote:
Since r10-7096 convert_like, when called in a template, creates an
IMPLICIT_CONV_EXPR when we're converting to/from array type.
In this test, we have e[f], and we're converting f (of type class A) to
int, so convert_like in build_new_op_1 created the IMPL
"Yangfei (Felix)" writes:
> From 30a0196b0afd45bae9291cfab3fee4ad6b90cbcb Mon Sep 17 00:00:00 2001
> From: Fei Yang
> Date: Thu, 11 Jun 2020 19:33:22 +0800
> Subject: [PATCH] vect: Fix an ICE in vect_loop_versioning [PR95570]
>
> In the test case for PR95570, the only data reference in the loop i
On 6/10/20 4:43 PM, Iain Sandoe wrote:
Hi Jason,
Jason Merrill wrote:
On Tue, Jun 9, 2020 at 5:04 AM Iain Sandoe wrote:
/* Don't bother reversing an operator with two identical parameters.
*/
- else if (args->length () == 2 && (flags & LOOKUP_REVERSED))
+ else if (ar
The current code assumes that if the next character in the stream is
equal to the delimiter then we stopped because we saw that delimiter,
and so discards it. But in the testcase for the PR we stop because we
reached the maximum number of characters, and it's coincidence that the
next character eq
On 6/11/20 2:43 AM, Martin Liška wrote:
Ready for master?
I've been meaning to do some of this to resolve pr94982. Thanks
a lot for taking the lead on it! I'll try to get to it soon!
Martin
Thanks,
Martin
gcc/ChangeLog:
* cgraphunit.c (process_symver_attribute): Wrap weakref keywor
Hi Harald,
one remark: Instead of
+ && !(strcmp(gfc_current_intrinsic, "associated") == 0
+ || strcmp(gfc_current_intrinsic, "null") == 0
+ || strcmp(gfc_current_intrinsic, "present") == 0))
could you maybe test sym->id ?
OK with that change.
Best regards,
On Wed, Jun 10, 2020 at 12:11:00PM -0400, David Malcolm wrote:
> Thanks for the patch; sorry about the delay in reviewing it.
>
> Some high-level review points
>
> - I like the patch overall
>
> - This will deserve an item in the release notes
>
> - I don't like adding "global_tabstop" (I don't
> I am not very good at reasoning about reverse storage order stuff but
> can it ever happen that this reverse is not the same as racc->reverse?
>
> In order for that to happen, we'd have to have an assignment (folded
> memcpy?) between two aggregates in the original code that, at the same
> offse
> Gesendet: Montag, 08. Juni 2020 um 22:25 Uhr
> Von: "Harald Anlauf"
> An: "fortran" , "gcc-patches"
> Betreff: [PATCH] PR fortran/95544 - ICE in gfc_can_put_var_on_stack, at
> fortran/trans-decl.c:494
OK, now with a brown bag over my head, here comes the patch instead of
just the testcase.
T
Hi Mike,
On Tue, Jun 09 2020, Mike Stump wrote:
> I think I'd prefer the fix on the other side, if reasonable. I'd give
> them some time to see about a fix there before selecting this patch.
given Alexandre's email, are you OK with the patch? It essentially
manually keeps the input name "rootna
Hi,
On Thu, Jun 11 2020, Eric Botcazou wrote:
> Hi,
>
> this fixes an issue with reverse storage order in SRA, which is caught by the
> built-in verifier:
>
> ===GNAT BUG DETECTED==+
> | 11.0.0 20200610 (experimental) (x86_64-suse-linux) GCC err
Hi Harald,
The following patch fixes an almost obvious ICE in invalid.
Regtested on x86_64-pc-linux-gnu.
OK for master, and backports to 9/10?
OK. Thanks for the patch!
Regards
Thomas
Hi Jose,
Proposed patch to PR95331 - Unlimited polymorphic arrays have wrong bounds.
Patch tested only on x86_64-pc-linux-gnu.
reviewed, ChangeLog reformatted, committed as
r11-1235-g2ee70f5d161edd99a7af97d166b251bcf83cd91b .
Thanks a lot for the patch!
Do you have interest in getting write
On Thu, Jun 11, 2020 at 01:49:57PM +0200, Martin Liška wrote:
> diff --git a/gcc/config/rs6000/rs6000.c b/gcc/config/rs6000/rs6000.c
> index 0bbd06ad1de..00daf979856 100644
> --- a/gcc/config/rs6000/rs6000.c
> +++ b/gcc/config/rs6000/rs6000.c
> @@ -4987,6 +4987,9 @@ rs6000_density_test (rs6000_cost
On June 11, 2020 1:49:57 PM GMT+02:00, "Martin Liška" wrote:
>Hi.
>
>Since stmt_vec_info is not generated for debug info statements, these
>needs to
>be skipped in rs6000_density_test.
>
>Patch can bootstrap on ppc64le-linux-gnu and survives regression tests.
>
>Ready to be installed?
OK.
Richa
Hi Jose,
Proposed patch to PRs 52351, 85868 Wrong array section bounds when
passing to an intent-in pointer dummy.
First, thanks for working on this and for this patch.
Regarding the patch, there are a few style issues which I fixed
for the commit. If you could try to adhere to a few more of
Hi Alexandre,
> On Jun 9, 2020, Rainer Orth wrote:
>
>> this is wrong unfortunately: braces are the Tcl equivalent of single
>> quotes so you're setting skip_lto to the string inside.
>
> Aah, thanks. So when $skip_lto is expanded in the ifs, the whole thing
> is evaluated, and that's why it wo
Hi,
This is a experimental fix for pr94274.
For if/else structure, when the expressions is binary operation and have a
common subexpression, and the opcode is the same, we can
fold the merging phi node in tree_ssa_phiopt_worker (ssa-phiopt). It can be
optimized to do csel first and then do binar
Hi Jose,
Proposed patch to Bug 94022 - Array slices of assumed-size arrays.
Patch tested only on x86_64-pc-linux-gnu.
Reviewed, regression-tested and commited as
r11-1228-g6a07010b774cb5a0b1790b857e69d3d8534eebd2 .
Thanks for the patch!
Regards
Thomas
On Thu, Jun 11, 2020 at 01:47:30PM +0200, Martin Liška wrote:
> The argument can be useful to update arbitrary branch, the changes
> are added to git index and user is supposed to make a commit.
>
> I'm going to install it if there are no objections.
>
> Martin
>
> contrib/ChangeLog:
>
>
Hi,
gimplify_modify_expr_rhs has an optimization whereby the assignment to an
aggregate variable from a read-only object with a DECL_INITIAL is optimized
into the direct assignment of the DECL_INITIAL, provided that no temporary is
created in the process.
The optimization is blocked if the rea
Hi.
Since stmt_vec_info is not generated for debug info statements, these needs to
be skipped in rs6000_density_test.
Patch can bootstrap on ppc64le-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
gcc/ChangeLog:
PR target/95627
* config/rs6000/rs
The argument can be useful to update arbitrary branch, the changes
are added to git index and user is supposed to make a commit.
I'm going to install it if there are no objections.
Martin
contrib/ChangeLog:
* gcc-changelog/git_update_version.py: Add --curent argument.
---
contrib/gcc-
Hi,
> -Original Message-
> From: Richard Sandiford [mailto:richard.sandif...@arm.com]
> Sent: Thursday, June 11, 2020 12:23 AM
> To: Yangfei (Felix)
> Cc: gcc-patches@gcc.gnu.org
> Subject: Re: [PATCH PR95570] vect: ICE: Segmentation fault in
> vect_loop_versioning
>
> "Yangfei (Felix)"
Hi,
this fixes an issue with reverse storage order in SRA, which is caught by the
built-in verifier:
===GNAT BUG DETECTED==+
| 11.0.0 20200610 (experimental) (x86_64-suse-linux) GCC error:|
| in verify_sra_access_forest, at tree-sra
The following patch enables vector permutations optimization by using another
vector element size when applicable.
It allows usage of simpler instructions in applicable cases.
example:
#define vector __attribute__((vector_size(16) ))
vector float f(vector float a, vector float b)
{
return __bu
"Kewen.Lin" writes:
> @@ -9195,12 +9222,13 @@ optimize_mask_stores (class loop *loop)
> }
>
> /* Decide whether it is possible to use a zero-based induction variable
> - when vectorizing LOOP_VINFO with a fully-masked loop. If it is,
> - return the value that the induction variable must b
Work around bug in Put_Image of types in Remote_Types packages. Use the
switch -gnatd_z to control enabling of Put_Image. Put_Image is still
disabled by default for all types.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Bob Duff
gcc/ada/
* exp_put_image.adb (Build_
The Safe_Unchecked_Type_Conversion predicate was invoking the
Has_Discriminant predicate without checking that the Etype really
Is_Type, which is not the case for Standard_Void_Type "returned" by
entry calls.
No functional changes.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Er
This adjusts the new Resolve_Implicit_Dereference procedure to the cases
where nodes in an instance do not have the proper view of a type that
was declared as private (a well-known limitation of the current
implementation of generic instantiations for types that are only
implicitly referenced in th
Routine Get_Value is only called with its Compon parameter equal to
entity ids of components. This is now reflected in the type of this
parameter. Code cleanup only; semantics is unaffected.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Piotr Trojanek
gcc/ada/
* sem_a
Previously, the four aspects Async_Readers, Async_Writers,
Effective_Reads, and Effective_Writes could only be specified for
volatile variables and for state abstractions. Allow specifying these
aspects for volatile types.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Steve Baird
SPARK RM 7.1.3(8) has been deleted, but there were actually plenty of
mistakes in references to rules 7.1.3(X). This patch fixes them in both
comments and error messages.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Piotr Trojanek
gcc/ada/
* sem_ch4.adb, sem_ch6.adb,
Large bit-packed arrays, i.e. whose size is greater than 64 bits,
are implemented under the hood by means of arrays of storage units,
the front-end generating the required mask-and-shifts operations to
go back and forth between the two representations.
These operations depend on the endianness of
Work around the fact that Put_Image doesn't work for private types whose
full type is real. Make Put_Image_Unknown print out the name of the
type. Put_Image is still disabled by default for all types.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Bob Duff
gcc/ada/
*
This consolidates the handling of almost all the implicit dereferences
allowed in Ada 95 into the semantic analysis phase of the compiler, and
more precisely in the Resolve routine of the front-end.
This both means that the generic code handling them in the expander
is removed, and that various co
When resolving a record aggregate with a box as the value of one of its
component that is itself of a discriminated record type, this box is
replaced with an inner record aggregate. However, while a constrained
itype is created for the outer record aggregate (as described in the
comment of Resolve_
A local variable that is used as a counter (which is clear from both its
comment and its used) will only be assigned with natural numbers. This
is now reflected in its type. Code cleanup only; semantics is
unaffected.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Piotr Trojanek
When resolving record aggregates, frontend was creating "others => <>"
association to represent nested components, for example:
type T1 (D1 : Boolean := False) is record
C1 : Boolean := True;
end record;
type T2 is record
C2 : T1;
end record;
X2 : T2 := (others => <>);
Routine Build_Constrained_Itype was created as an exact duplicate
Build_Subtype with a comment suggesting that their code should be
exported from Sem_Util and reused. Unfortunately, since then both
routines diverged and now are subtly different, so reusing is not
straightforward. However, it is sti
Routine Next is implemented both as a procedure and as a function. The
procedure variant seems meant to be used when iterating, e.g.:
Next (Decl);
because it is more readable than the corresponding functions:
Decl := Next (Decl);
(and it is inlined anyway, so there is no performance penal
This patch refines the handling of warnings on overlapping actuals of
composite types when only one of them is writable. Formals of a generic
type are excluded, given that the warning will be given on any instance.
Uniform treatment of formals and actuals.
Tested on x86_64-pc-linux-gnu, committed
When looking for references to discriminants within constraint
expressions we now stop once the first such a reference is found. This
is just a tiny performance improvement; semantics is unaffected.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Piotr Trojanek
gcc/ada/
This patch corrects an issue whereby the compiler would incorrectly
calculate accessibility levels of objects within type conversions -
leading to potentially missing static and dynamic errors.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Justin Squirek
gcc/ada/
* sem
When iterating over pairs of formal parameters we now finish as soon as
we find a single problematic pair; previously we continued iteration.
This is just a simplification and a compiler performance improvement.
Semantics is not affected.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-
Internally generated static dispatch tables are preferably put in ROM,
so they are declared as constant where possible. However, the Object
Specific Dispatch tables were declared as variables, even though they
are initialized with static aggregates (with only literal integers) and
are never modifie
Routines Sem_Aggr.Build_Constrained_Itype and Sem_Ch3.Build_Subtype that
create discriminated itypes were originally identical, but now they are
subtly different. This patch removes one of their two subtle
differences, namely a call to Set_Size_Known_At_Compile_Time that was
meant as a very narrow
Tables with case-choices are sorted by Sort_Case_Table with an insertion
sort. Contrary to what comments for the Case_Table_Type says, this
routine doesn't use the table element at index 0 as a placeholder.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Piotr Trojanek
gcc/ada/
The call to Set_Size_Known_At_Compile_Time in
Backend_Processing_Possible happened just after querying
Size_Known_At_Compile_Time several lines before. Both calls operated on
the same type entity, so the call to "Set" routine had no effect and now
is removed.
Tested on x86_64-pc-linux-gnu, committ
This Ada 202x AI clarifies that Root_Storage_Pool_With_Subpools and
Root_Subpool should have pragma Preelaborable_Initialization.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Arnaud Charlet
gcc/ada/
* libgnat/s-stposu.ads (Root_Storage_Pool_With_Subpools,
Root
This patch fixes a bug where if you have a Predicate_Failure aspect on a
nested type, but no Predicate, Static_Predicate, or Dynamic_Predicate,
the compiler crashes when compiled with assertions enabled.
Tested on x86_64-pc-linux-gnu, committed on trunk
2020-06-11 Bob Duff
gcc/ada/
*
When an assignment of an allocator the type of which has predicate
checks is made inside of a record, GNAT generates a call to the
predicate function. However, before this commit, GNAT wouldn't check if
the subtype mark had a predicate, which would result in the predicate
check function not being c
On 6/11/20 11:47 AM, Jonathan Wakely wrote:
Is the attached patch what you're asking for?
Yes ;)
Martin
On 11/06/20 10:54 +0200, Martin Liška wrote:
On 6/11/20 10:48 AM, Jonathan Wakely wrote:
On 11/06/20 09:54 +0200, Martin Liška wrote:
On 6/10/20 2:20 PM, Jonathan Wakely wrote:
Oops, this line was left in while I was testing it by amending
existing commits!
Here's an updated patch without tha
On 6/11/20 10:50 AM, Jakub Jelinek wrote:
On Thu, Jun 11, 2020 at 10:12:14AM +0200, Martin Liška wrote:
gcc/ChangeLog:
PR sanitizer/95634
* asan.c (asan_emit_stack_protection): Fix emission for ilp32
by using Pmode instead of ptr_mode.
---
gcc/asan.c | 4 ++--
1 file
On 6/11/20 10:48 AM, Jonathan Wakely wrote:
On 11/06/20 09:54 +0200, Martin Liška wrote:
On 6/10/20 2:20 PM, Jonathan Wakely wrote:
Oops, this line was left in while I was testing it by amending
existing commits!
Here's an updated patch without that line.
I generally like the suggested idea
On 6/9/20 3:42 PM, Richard Biener wrote:
I think we need to fix that before merging.
There's updated version of the patch that should handle the EH properly.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
>From fc5a59e8c8887c102bff
On Thu, Jun 11, 2020 at 10:12:14AM +0200, Martin Liška wrote:
> gcc/ChangeLog:
>
> PR sanitizer/95634
> * asan.c (asan_emit_stack_protection): Fix emission for ilp32
> by using Pmode instead of ptr_mode.
> ---
> gcc/asan.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions
On 11/06/20 09:54 +0200, Martin Liška wrote:
On 6/10/20 2:20 PM, Jonathan Wakely wrote:
Oops, this line was left in while I was testing it by amending
existing commits!
Here's an updated patch without that line.
I generally like the suggested idea and I have suggestion to usage of the
GCC_FO
Ready for master?
Thanks,
Martin
gcc/ChangeLog:
* cgraphunit.c (process_symver_attribute): Wrap weakref keyword.
* dbgcnt.c (dbg_cnt_set_limit_by_index): Do not print extra new
line.
* lto-wrapper.c (merge_and_complain): Wrap option names.
---
gcc/cgraphunit.c
Hello.
There's a patch for ilp32 where we should use Pmode instead of ptr_mode.
Patch can bootstrap on x86_64-linux-gnu and survives regression tests.
Ready to be installed?
Thanks,
Martin
gcc/ChangeLog:
PR sanitizer/95634
* asan.c (asan_emit_stack_protection): Fix emission fo
On 6/11/20 9:45 AM, Martin Liška wrote:
On 6/10/20 2:15 PM, Jonathan Wakely wrote:
And this is another little patch that just avoids running 'git diff'
twice, by using tee(1) to write to $GCC_GIT_DIFF_FILE as a side effect
of generating the diff to pipe to mklog.py.
Thanks, please install the
On 6/10/20 2:20 PM, Jonathan Wakely wrote:
Oops, this line was left in while I was testing it by amending
existing commits!
Here's an updated patch without that line.
I generally like the suggested idea and I have suggestion to usage of the
GCC_FORCE_MKLOG.
Right now, we use the env variable
On 6/10/20 2:15 PM, Jonathan Wakely wrote:
And this is another little patch that just avoids running 'git diff'
twice, by using tee(1) to write to $GCC_GIT_DIFF_FILE as a side effect
of generating the diff to pipe to mklog.py.
Thanks, please install the improvement.
Martin
on 2020/6/11 上午12:58, Richard Sandiford wrote:
> "Kewen.Lin" writes:
>> diff --git a/gcc/tree-vect-loop-manip.c b/gcc/tree-vect-loop-manip.c
>> index ca68d04a919..1fac5898525 100644
>> --- a/gcc/tree-vect-loop-manip.c
>> +++ b/gcc/tree-vect-loop-manip.c
>> @@ -420,8 +420,8 @@ vect_set_loop_control
The patch removes bunch of warnings:
options-save.c:12004:29: warning: unquoted identifier or keyword
‘global_options’ in format [-Wformat-diag]
12004 | internal_error ("Error: global_options are modified in local
context\n");
gcc/ChangeLog:
* optc-save-gen.awk: Quote error string
98 matches
Mail list logo
