Hi,
Many thanks for all comments on previous versions.
Comparing with the previous version, I dropped the checks
on costly niter and invalid stmt scanning. As Richard's
suggestion, keep this as simple as possible, refine it
if finding any cases which matter later.
Bootstrapped and regression t
Hi Steve,
That's good to go - thanks
Paul
On Wed, 12 Jun 2019 at 23:44, Steve Kargl
wrote:
>
> The attach patch has been sitting in my tree for a year.
> It has been tested and updated as others have changed
> the gfortran code. The patch has been compiled and
> regression tested on x86_64-*-f
Hi Steve,
As the original author of the confusing code, I want to thank you for
de-confusing it!
OK
Paul
On Wed, 12 Jun 2019 at 20:01, Steve Kargl
wrote:
>
> This PR flags a section of confusing but correct code.
> The attach patch rewrites the code to hopefully provide
> some clarity. It has
Hi Steve,
OK - thanks
Paul
On Wed, 12 Jun 2019 at 19:42, Steve Kargl
wrote:
>
> The attach patch has lived in my tree for 4 months.
> It's time to submit it. Passed regression testing
> for a long time.
>
> An INTENT(in) entity that has CLASS(*) dummy argument
> should not use SELECT TYPE to t
Hi Kyrill,
Thanks for the comments. Committed as you suggested.
Thanks,
Kugan
On Wed, 12 Jun 2019 at 18:07, Kyrill Tkachov
wrote:
>
> Hi Kugan,
>
> On 6/12/19 4:59 AM, Kugan Vivekanandarajah wrote:
> > AArch64 comment for ADDSUB iterator is a typo or copy-and-paste error.
> > Attached patch fix
This one was fixed by r259717, but that commit didn't include any tests. So
it's reasonable to add something that exercises that code path.
Tested on x86_64-linux, applying to trunk.
2019-06-12 Marek Polacek
PR c++/87410
* g++.dg/cpp1y/pr87410.C: New test.
diff --git gcc/tes
Hello,
This patch adds libgcc configuration option to disable TM clone
registry. This option helps to reduce code size for embedded targets
which do not need transactional memory support.
Tested on x86_64 and riscv64-unknown-elf-gcc, the last is with the
option enabled.
If the change is Ok
The attach patch has been sitting in my tree for a year.
It has been tested and updated as others have changed
the gfortran code. The patch has been compiled and
regression tested on x86_64-*-freebsd. OK to commit?
Either testcase should provide sufficient information
about the problem that thi
Revision 262909 remvoed the gfc_free_dt_list function, but
failed to remove the prototype in gfortran.h. I have committed
the following patch.
Index: gcc/fortran/ChangeLog
===
--- gcc/fortran/ChangeLog (revision 272219)
+++ gcc/
On 6/12/19 5:36 PM, Paolo Carlini wrote:
Hi,
On 12/06/19 22:24, Jason Merrill wrote:
On 6/6/19 9:00 AM, Paolo Carlini wrote:
Hi,
only minor functional changes here - more precise locations for two
error messages - but I think it's a step in the right direction: it
moves the declaration of
On 6/12/19 3:33 PM, Rainer Orth wrote:
Richard Biener writes:
On Mon, Jun 10, 2019 at 10:37 PM Martin Sebor wrote:
On 6/10/19 1:29 PM, Jakub Jelinek wrote:
On Mon, Jun 10, 2019 at 01:23:28PM -0600, Martin Sebor wrote:
+ else if (integer_zerop (TREE_OPERAND (node, 1))
+ /* Dump the
Hi,
On 12/06/19 22:24, Jason Merrill wrote:
On 6/6/19 9:00 AM, Paolo Carlini wrote:
Hi,
only minor functional changes here - more precise locations for two
error messages - but I think it's a step in the right direction: it
moves the declaration of id_loc way up, near typespec_loc and as su
On 6/11/19 3:59 PM, Marek Polacek wrote:
On Tue, Jun 11, 2019 at 03:05:26PM -0400, Jason Merrill wrote:
On 6/11/19 2:28 PM, Marek Polacek wrote:
On Tue, Jun 11, 2019 at 08:37:27AM -0400, Jason Merrill wrote:
On 6/11/19 7:47 AM, Jakub Jelinek wrote:
On Mon, Jun 10, 2019 at 09:59:46PM -0400, Ma
Richard Biener writes:
> On Mon, Jun 10, 2019 at 10:37 PM Martin Sebor wrote:
>>
>> On 6/10/19 1:29 PM, Jakub Jelinek wrote:
>> > On Mon, Jun 10, 2019 at 01:23:28PM -0600, Martin Sebor wrote:
>> >> + else if (integer_zerop (TREE_OPERAND (node, 1))
>> >> + /* Dump the types of INTEGER_CSTs
On 6/12/19 7:07 AM, Richard Biener wrote:
On Tue, 11 Jun 2019, Jason Merrill wrote:
On Tue, Jun 11, 2019, 6:45 AM Richard Biener wrote:
The following fixes the documented(!) quadraticness in
split_nonconstant_init_1 by simply marking to be removed
constructor elements and doing that in a se
On 6/12/19 5:33 AM, Richard Biener wrote:
On Wed, Jun 12, 2019 at 1:57 AM Martin Sebor wrote:
On 6/11/19 3:02 PM, Aldy Hernandez wrote:
On 6/11/19 12:52 PM, Martin Sebor wrote:
On 6/11/19 10:26 AM, Aldy Hernandez wrote:
On 6/11/19 12:17 PM, Martin Sebor wrote:
On 6/11/19 9:05 AM, Aldy
On 6/5/19 10:17 AM, Richard Sandiford wrote:
Thanks for the patch and sorry that there was no response.
I've added the C++ maintainers to cc:
Thanks. In the future, mentioning C++ in the subject line will help me
see it.
On 2019-04-08 7:20 p.m., Nicholas Krause wrote:
This fixes the calle
On 6/6/19 9:00 AM, Paolo Carlini wrote:
Hi,
only minor functional changes here - more precise locations for two
error messages - but I think it's a step in the right direction: it
moves the declaration of id_loc way up, near typespec_loc and as such
id_loc is immediately used. Then its value
This is related to changes I made to readings.html one-and-a-half
days ago. Essentially fortran.org underwent quite some changes,
breaking links left and right, and frankly looking a little (too)
commercial.
In any case I could not find what looks like a proper replacement
for this link, so I had
This macro is defined in several palces, and we only fixed the
incorrect value in some of those places.
* include/std/algorithm (__cpp_lib_parallel_algorithm): Fix value.
* include/std/memory (__cpp_lib_parallel_algorithm): Likewise.
* include/std/numeric (__cpp_lib_parall
On 6/8/19 8:51 PM, Marek Polacek wrote:
This patch improves a diagnostic when parsing a lambda introducer; in
particular,
we should handle '&this' better than we currently do. clang/icc specifically
point out that 'this' cannot be captured by reference, let's do the same.
Bootstrapped/regteste
On 6/9/19 7:37 PM, Marek Polacek wrote:
This testcase used to ICE but was fixed by r259067. We now issue
67533.C:5:26: error: conversion from ‘void’ to non-scalar type ‘Tls’ requested
5 | thread_local Tls mytls = mytls; // { dg-error "" }
| ^
whereas cl
On вторник, 11 юни 2019 г. 14:04:41 EEST Jeff Law wrote:
> On 6/9/19 2:01 PM, Dimitar Dimitrov wrote:
> > This is the latest patch set for adding TI PRU I/O processor backend to
> > GCC. Comments from all previous series have been addressed [1], [2], [3],
> > [4], [5].>
> > Test results can be dow
Committed as obvious.
ChangeLog:
2019-06-12 Dimitar Dimitrov
* MAINTAINERS (Write After Approval): Add myself.
Signed-off-by: Dimitar Dimitrov
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 14fa95ed38b..b8d703c535d 100644
--- a/M
On 6/9/19 8:45 PM, Marek Polacek wrote:
The problem here is that we're getting "requested alignment is not an
integer constant" since r261971, because of this part of the patch:
@ -4676,7 +4685,11 @@ cxx_eval_constant_expression (const constexpr_ctx *ctx,
tree t,
conversion. */
On 6/10/19 6:36 AM, Paolo Carlini wrote:
Hi,
this one requires my last patch, uses id_loc in a few additional places
in the last part of the function.
Most of the changes should be straightforward, I only want to mention
the "typedef name may not be a nested-name-specifier" change, where, no
On Wed, 12 Jun 2019 13:57:22 +0200
Thomas Schwinge wrote:
> Hi!
>
> First, thanks for picking this up, and improving the patch you
> inherited.
Thanks for review!
> I understand right that this will address some aspects of PR90115
> "OpenACC: predetermined private levels for variables declared
This PR flags a section of confusing but correct code.
The attach patch rewrites the code to hopefully provide
some clarity. It has lived my tree for around 6 months,
so has sufferred through numerous regression tests.
OK to commit?
2019-06-12 Steven G. Kargl
PR fortran/88810
The attach patch has lived in my tree for 4 months.
It's time to submit it. Passed regression testing
for a long time.
An INTENT(in) entity that has CLASS(*) dummy argument
should not use SELECT TYPE to then try to assign to the
entity. OK to commit?
2019-06-12 Steven G. Kargl
PR fo
Thanks.
Committed as revision 272201.
--
steve
On Wed, Jun 12, 2019 at 06:14:11AM +0100, Paul Richard Thomas wrote:
> OK, Steve
>
> Thanks
>
> Paul
>
> On Wed, 12 Jun 2019 at 02:04, Steve Kargl
> wrote:
> >
> > The attached patch fixes an ICE when freeing an array-spec
> > that involves a
Patch fixes following PRs:
c++/90816 - -finstrument-functions-exclude-function-list improperly
handles namespace/class definitions
c++/90809 - -finstrument-functions-exclude-function-list mishandles
comma escaping
Fixes as follows:
At flag_instrument_functions_exclude_p [gimplify.c]
Using lang_hoo
On Tue, Jun 11, 2019 at 11:50:40AM +0200, Jakub Jelinek wrote:
> On Tue, Jun 11, 2019 at 10:30:59AM +0100, Mark Eggleston wrote:
> > Jim MacArthur
> > Mark Eggleston
>
> Two spaces before < instead of one.
>
> This is not a patch review, just comments:
Mark, do you plan to address any
On Wed, Jun 12, 2019 at 09:24:39AM +0200, Jakub Jelinek wrote:
> On Tue, Jun 11, 2019 at 12:49:32PM +0200, Jakub Jelinek wrote:
> > Tested on x86_64-linux with check-gfortran, ok for trunk if full
> > bootstrap/regtest
> > passes?
> >
> > 2019-06-11 Jakub Jelinek
> >
> > * io.c (check_for
Hi,
If we compile the openacc testcase with -fopenacc -O2, we run into a SIGSEGV
or assert. The root cause for this is that pass_thread_jumps breaks the
invariant that OACC_FORK and OACC_JOIN mark the start and end of a
single-entry-single-exit region.
Fix this by bailing out when encountering a
Initialize CTF container when -gtLEVEL is specified. Generate CTF debug info
for global decls. Import the CTF header from binutils.
[Changes from V1]
- Instead of using the debug hooks infrastructure, the generation and
emission of CTF is done by exposing the CTF APIs to the rest of the
c
-gtLEVEL is used to request CTF debug information and also to specify how much
CTF debug information.
[Changes from V1]
None
gcc/ChangeLog:
* common.opt: Add CTF debug info options.
* doc/invoke.texi: Document the CTF debug info options.
* flag-types.h (enum ctf_debug_
Hello,
Thanks for the feedback on the previous patch set.
This is the second posting of the RFC patch for CTF support in GCC. This patch
set does not rely on debug hooks, but it keeps CTF and DWARF debug info
generation separated in the compiler.
For CTF generation, callsites in symbol_table::fi
[Changes from V1]
None
gcc/ChangeLog:
* langhooks.c (lang_GNU_GIMPLE): New Function.
* langhooks.h: New Prototype.
---
gcc/ChangeLog | 5 +
gcc/langhooks.c | 9 +
gcc/langhooks.h | 1 +
3 files changed, 15 insertions(+)
diff --git a/gcc/langhooks.c b/gcc/langhoo
On 6/12/19 10:40 AM, Jakub Jelinek wrote:
On Wed, Jun 12, 2019 at 10:13:57AM -0600, Martin Sebor wrote:
But GCC doesn't support such an implementation, does it?
Why would that be relevant?
Obviously because it makes no sense to cater to all conceivable
extensions provided by all sorts of imp
On Wed, Jun 12, 2019 at 10:13:57AM -0600, Martin Sebor wrote:
> But GCC doesn't support such an implementation, does it?
Why would that be relevant? The warning would cause people to make portable
code less portable (by removing the alloca (0) calls that were added there
for portability reasons),
I think copyright assignment is done. Thanks for bearing with me.
I noticed the version I submitted in April is missing some changes we
discussed on October 2018.
I took the patch from then and removed -matpcs too, the unnecessary
change to libgcc t-netbsd (which is the OABI configuration anyway)
On 6/12/19 9:25 AM, Michael Matz wrote:
Hi,
On Wed, 12 Jun 2019, Martin Sebor wrote:
Otherwise LGTM as the patch, but I'd like to hear from others whether
it is kosher to add such a special case to the warn_unused_result
attribute warning. And if the agreement is yes, I think it should be
doc
Hi,
On Wed, 12 Jun 2019, Martin Sebor wrote:
> > Otherwise LGTM as the patch, but I'd like to hear from others whether
> > it is kosher to add such a special case to the warn_unused_result
> > attribute warning. And if the agreement is yes, I think it should be
> > documented somewhere that a
Hi,
This patch adds support in the vectorizer for masking fold left reductions.
This avoids the need to insert a conditional assignment with some identity
value.
For example, this C code:
double
f (double *restrict x, int n)
{
double res = 0.0;
for (int i = 0; i < n; i++)
{
res +=
Hi,
With the test-case of PR90009 we run into a SIGSEGV in nvptx_neuter_pars,
because par->join_insn and par->join_block are NULL.
Detect this problem earlier, by adding an assert in nvptx_find_par that
asserts that a fork cannot be paired with more than one join.
Build and tested on x86_64 with
On 6/12/19 5:37 AM, Jakub Jelinek wrote:
On Wed, Jun 12, 2019 at 01:30:14PM +0200, Martin Liška wrote:
@@ -9447,10 +9448,19 @@ do_warn_unused_result (gimple_seq seq)
location_t loc = gimple_location (g);
if (fdecl)
- warning_at (loc, OPT_Wunused_result,
-
Also fix a warning with -Wunused-parameter -Wsystem-headers.
* include/std/variant (get, get, get_if, get_if)
(variant::emplace): Change static_assert messages from "should be"
to "must be".
(hash::operator()): Remove name of unused parameter.
Tested x86_64-linux,
* include/std/mutex (scoped_lock::~scoped_lock()): Use fold
expression.
Thanks to Lars for pointing out this could be simplified.
Tested x86_64-linux, committed to trunk.
commit 7445abf1ee5e14e644efd0533fff9e0c125b1d2c
Author: redi
Date: Wed Jun 12 14:52:06 2019 +
Si
The std::to_chars functions from C++17 can be used to implement
std::to_string with much better performance than calling snprintf. Only
the __detail::__to_chars_len and __detail::__to_chars_10 functions are
needed for to_string, because it always outputs base 10 representations.
The return type o
This avoids loop_version () calls when if-conversion already versioned
the loops and simplifies vect_loop_versioning because we need not
do as much fixup. There's followup work to do for the profile scaling.
Honza - any suggestion on how to apply a different true/false
profile to an existing co
On 6/12/19 2:50 PM, Richard Biener wrote:
> New params should always go to the end
Ah, sorry, I'll take of that new time.
I've just changed the function signature and I can verify there are no
other calls (tested for --enable-languages=all).
Thanks,
Martin
On Wed, Jun 12, 2019 at 1:45 PM Martin Liška wrote:
>
> On 6/12/19 11:41 AM, Richard Biener wrote:
> > On Wed, Jun 12, 2019 at 11:15 AM Martin Liška wrote:
> >>
> >> On 6/12/19 10:02 AM, Martin Liška wrote:
> >>> On 6/12/19 9:59 AM, Richard Biener wrote:
> On Tue, Jun 11, 2019 at 9:02 PM Jas
Hi!
First, thanks for picking this up, and improving the patch you inherited.
Then, just a few individual comments, not a complete review.
(As far as I concerned, and as far as relevant, these can be addressed
later, incrementally, of course.)
I understand right that this will address some as
On 6/12/19 11:41 AM, Richard Biener wrote:
> On Wed, Jun 12, 2019 at 11:15 AM Martin Liška wrote:
>>
>> On 6/12/19 10:02 AM, Martin Liška wrote:
>>> On 6/12/19 9:59 AM, Richard Biener wrote:
On Tue, Jun 11, 2019 at 9:02 PM Jason Merrill wrote:
>
> On 6/11/19 9:16 AM, Martin Liška wro
On Wed, Jun 12, 2019 at 01:30:14PM +0200, Martin Liška wrote:
> @@ -9447,10 +9448,19 @@ do_warn_unused_result (gimple_seq seq)
> location_t loc = gimple_location (g);
>
> if (fdecl)
> - warning_at (loc, OPT_Wunused_result,
> - "ignoring
On 6/12/19 1:22 PM, Jakub Jelinek wrote:
> On Wed, Jun 12, 2019 at 01:11:09PM +0200, Martin Liška wrote:
>> 2019-06-12 Martin Liska
>>
>> * calls.c (special_function_p): Make it global.
>> * calls.h (special_function_p): Declare.
>
> Why?
Not needed any longer.
>
>> * tree-cfg
On Wed, Jun 12, 2019 at 01:11:09PM +0200, Martin Liška wrote:
> 2019-06-12 Martin Liska
>
> * calls.c (special_function_p): Make it global.
> * calls.h (special_function_p): Declare.
Why?
> * tree-cfg.c (do_warn_unused_result): Do not
> warn for alloca(0).
> --- a/gcc/
On 6/11/19 6:03 PM, Jakub Jelinek wrote:
> On Tue, Jun 11, 2019 at 03:58:27PM +, Michael Matz wrote:
>> On Tue, 11 Jun 2019, Martin Liška wrote:
>>
>>> I see 3 occurrences of the alloca (0) in libiberty/regex.c, but there are
>>> properly
>>> guarded within:
>>>
>>> # ifdef C_ALLOCA
>>>
On Tue, 11 Jun 2019, Jason Merrill wrote:
> On Tue, Jun 11, 2019, 6:45 AM Richard Biener wrote:
>
> >
> > The following fixes the documented(!) quadraticness in
> > split_nonconstant_init_1 by simply marking to be removed
> > constructor elements and doing that in a second run over
> > the const
On 12-06-19 12:22, Jakub Jelinek wrote:
> On Fri, Jun 07, 2019 at 03:08:37PM +0100, Julian Brown wrote:
>> diff --git a/gcc/omp-low.c b/gcc/omp-low.c
>> index a7f35ffe416..67e1e82ec00 100644
>> --- a/gcc/omp-low.c
>> +++ b/gcc/omp-low.c
>> @@ -9794,6 +9882,8 @@ lower_omp_target (gimple_stmt_iterato
On Fri, Jun 07, 2019 at 03:08:37PM +0100, Julian Brown wrote:
> diff --git a/gcc/omp-low.c b/gcc/omp-low.c
> index a7f35ffe416..67e1e82ec00 100644
> --- a/gcc/omp-low.c
> +++ b/gcc/omp-low.c
> @@ -9794,6 +9882,8 @@ lower_omp_target (gimple_stmt_iterator *gsi_p,
> omp_context *ctx)
>
>if (off
On Tue, Jun 11, 2019 at 4:40 AM Feng Xue OS wrote:
>
> Reformat to comply with gcc coding style.
OK for trunk.
Thanks,
Richard.
> Feng
>
> ---
> diff --git a/gcc/ChangeLog b/gcc/ChangeLog
> index 37aab79..87cc125 100644
> --- a/gcc/ChangeLog
> +++ b/gcc/ChangeLog
> @@ -1,3 +1,16 @@
> +2019-06-0
On Wed, Jun 12, 2019 at 11:15 AM Martin Liška wrote:
>
> On 6/12/19 10:02 AM, Martin Liška wrote:
> > On 6/12/19 9:59 AM, Richard Biener wrote:
> >> On Tue, Jun 11, 2019 at 9:02 PM Jason Merrill wrote:
> >>>
> >>> On 6/11/19 9:16 AM, Martin Liška wrote:
> On 6/11/19 2:27 PM, Jason Merrill wr
On Wed, Jun 12, 2019 at 1:57 AM Martin Sebor wrote:
>
> On 6/11/19 3:02 PM, Aldy Hernandez wrote:
> >
> >
> > On 6/11/19 12:52 PM, Martin Sebor wrote:
> >> On 6/11/19 10:26 AM, Aldy Hernandez wrote:
> >>>
> >>>
> >>> On 6/11/19 12:17 PM, Martin Sebor wrote:
> On 6/11/19 9:05 AM, Aldy Hernande
I'm looking into a regression between gcc-7 and gcc-8 that causes
compilation with -mfloat-abi=hard to fail on arm-eabi with:
$ arm-eabi-gcc -c -mfloat-abi=hard t.c
cc1: error: -mfloat-abi=hard: selected processor lacks an FPU
Per the documentation, -mfpu=auto was supposed to be the default, but
On Wed, 12 Jun 2019, Jakub Jelinek wrote:
> On Wed, Jun 12, 2019 at 11:00:43AM +0200, Richard Biener wrote:
> > > Bootstrapped/regtested on x86_64-linux and i686-linux and furthermore
> > > tested
> > > with x86_64-linux to nvptx-none offloading with the nvptx.c PR90811 fix
> > > reverted to veri
On 6/12/19 10:02 AM, Martin Liška wrote:
> On 6/12/19 9:59 AM, Richard Biener wrote:
>> On Tue, Jun 11, 2019 at 9:02 PM Jason Merrill wrote:
>>>
>>> On 6/11/19 9:16 AM, Martin Liška wrote:
On 6/11/19 2:27 PM, Jason Merrill wrote:
> On 6/11/19 3:41 AM, Martin Liška wrote:
>> On 6/10/19
On Wed, Jun 12, 2019 at 11:00:43AM +0200, Richard Biener wrote:
> > Bootstrapped/regtested on x86_64-linux and i686-linux and furthermore tested
> > with x86_64-linux to nvptx-none offloading with the nvptx.c PR90811 fix
> > reverted to verify the overaligned variables are gone. Ok for trunk?
>
>
On Wed, 12 Jun 2019, Jakub Jelinek wrote:
> Hi!
>
> Apparently estimated_stack_frame_size -> expand_one_var -> add_stack_var
> -> align_local_variable changes DECL_ALIGN of stack variables from
> LOCAL_DECL_ALIGNMENT.
>
> That is unfortunately very undesirable for offloading, because this happen
The following patch aims to add support to android for aarch64 architecture.
diff --git a/gcc/config.gcc b/gcc/config.gcc
index 09fb9ecd2cd..a8b32e34b47 100644
--- a/gcc/config.gcc
+++ b/gcc/config.gcc
@@ -1022,8 +1022,9 @@ aarch64*-*-freebsd*)
tm_defines="${tm_defines} TARGET_DEFAULT_ASYNC_UNW
Hi Przemyslaw
On 6/7/19 10:09 AM, Przemyslaw Wirkus wrote:
Hi all,
This patch implements the usadv16qi and ssadv16qi standard names for arm.
The V16QImode variant is important as it is the most commonly used
pattern:
reducing vectors of bytes into an int.
The midend expects the optab to comp
Hi Kugan,
On 6/12/19 4:59 AM, Kugan Vivekanandarajah wrote:
AArch64 comment for ADDSUB iterator is a typo or copy-and-paste error.
Attached patch fixes this. I believe this falls under obvious
category. I will commit it after 48hrs unless comments should be
better worded.
Thanks,
Kugan
gcc/Ch
On 6/12/19 9:59 AM, Richard Biener wrote:
> On Tue, Jun 11, 2019 at 9:02 PM Jason Merrill wrote:
>>
>> On 6/11/19 9:16 AM, Martin Liška wrote:
>>> On 6/11/19 2:27 PM, Jason Merrill wrote:
On 6/11/19 3:41 AM, Martin Liška wrote:
> On 6/10/19 8:21 PM, Jason Merrill wrote:
>> On Mon, Jun
On Tue, Jun 11, 2019 at 9:02 PM Jason Merrill wrote:
>
> On 6/11/19 9:16 AM, Martin Liška wrote:
> > On 6/11/19 2:27 PM, Jason Merrill wrote:
> >> On 6/11/19 3:41 AM, Martin Liška wrote:
> >>> On 6/10/19 8:21 PM, Jason Merrill wrote:
> On Mon, Jun 10, 2019 at 3:08 AM Martin Liška wrote:
> >>
Hi!
While testing the second PR90811 patch with nvptx offloading, I've noticed 3
warnings in nvptx.c.
The first two hunks are obvious, the last warning is about unused mode_label
variable.
Either we can do what the patch does, or another option is throw away
both the mode_jump and mode_label tempo
On Tue, Jun 11, 2019 at 12:49:32PM +0200, Jakub Jelinek wrote:
> Tested on x86_64-linux with check-gfortran, ok for trunk if full
> bootstrap/regtest
> passes?
>
> 2019-06-11 Jakub Jelinek
>
> * io.c (check_format): Use G_(...) instead of _(...) for error values,
> append " in for
Hi!
Apparently estimated_stack_frame_size -> expand_one_var -> add_stack_var
-> align_local_variable changes DECL_ALIGN of stack variables from
LOCAL_DECL_ALIGNMENT.
That is unfortunately very undesirable for offloading, because this happens
early during IPA where we stream out LTO bytecode for t
77 matches
Mail list logo
