On Mon, 10 Sep 2018 20:31:49 -0400
Julian Brown wrote:
> [...] I think the handling of references can and should match between
> the two APIs (though implementation details of the patch to make that
> so need a little work still).
Here's a new version of the patch, somewhat simplified and
On 9/11/18 4:09 AM, Aldy Hernandez wrote:
> We can calculate wide_int_range_shift_undefined_p() without having to
> pass down the sign bit of the operand. Also, vrp_shift_undefined_p is a
> brain dead wrapper so I'm removing it.
>
> OK for trunk?
>
> curr.patch
>
> commit
On 9/5/18 5:48 AM, a...@codesourcery.com wrote:
>
> The HSA GPU drivers can't cope with binaries that have the same symbol defined
> multiple times, even though the names are not exported. This happens whenever
> there are file-scope static variables with matching names. I believe it's
> also
On 9/7/18 3:12 AM, Aldy Hernandez wrote:
> Semi obvious patch here...
>
> As the comment for alloca_type_and_limit says:
>
> // For ALLOCA_BOUND_MAYBE_LARGE and ALLOCA_BOUND_DEFINITELY_LARGE
> // types, this field indicates the assumed limit if known or
> // integer_zero_node if unknown.
On 9/5/18 5:51 AM, a...@codesourcery.com wrote:
>
> This patch contains the GCN port of libgfortran. We use the minimal
> configuration created for NVPTX. That's all that's required, besides the
> target-independent bug fixes posted already.
>
> 2018-09-05 Andrew Stubbs
> Kwok
On 9/5/18 5:49 AM, a...@codesourcery.com wrote:
>
> GCN's 64-lane vectors tend to make RTL dumps very long. This patch makes them
> far more bearable by eliding long sequences of the same element into
> "repeated"
> messages.
>
> 2018-09-05 Andrew Stubbs
> Jan Hubicka
>
On 9/5/18 5:49 AM, a...@codesourcery.com wrote:
>
> The vec_select operator is documented to require a const_int for the lane
> selector operand, but GCN has an instruction that can select the lane at
> runtime, so it seems reasonable to remove this restriction.
>
> This patch simply replaces
On 9/5/18 5:48 AM, a...@codesourcery.com wrote:
>
> GCN has some registers which are special purpose, but not "fixed" because we
> want the register allocator to track their usage and select alternatives that
> use different special registers (e.g. scalar cc vs. vector cc).
>
> Sometimes this
On Tue, 2018-09-11 at 13:20 -0500, Segher Boessenkool wrote:
> Hi again,
>
> On Tue, Sep 11, 2018 at 10:04:45AM -0500, Will Schmidt wrote:
> > On Fri, 2018-09-07 at 16:54 -0500, Segher Boessenkool wrote:
> > > On Fri, Sep 07, 2018 at 04:27:34PM -0500, Will Schmidt wrote:
> > > > -/* { dg-final {
On 9/11/18 3:58 PM, Joseph Myers wrote:
I'd expect all the cases where you pass a string containing '%s' to the
generic diagnostic functions to use %qs instead of '%s' for quoting, to
get appropriate locale-specific quotes.
yes of course, slipped my mind.
Committed the attached after manual
(just some very quick comments, they may be off, and I haven't looked
closely)
On Tue, 11 Sep 2018, MCC CS wrote:
+/* (~x & y) | ~(x | y) -> ~x */
+(simplify
+ (bit_ior:c (bit_and:c (bit_not @0) @1) (bit_not (bit_ior:s @0 @1)))
+ (bit_not @0))
Did you mean :c instead of :s?
You should write
Hi, Alexander,
thanks for your comments.
> On Sep 11, 2018, at 12:48 PM, Alexander Monakov wrote:
>
> On Tue, 11 Sep 2018, Qing Zhao wrote:
>> '-finline-visibility=[all|extern|static]'
>> By default, GCC inlines functions without considering their
>> visibility. This flag allows finer
On 9/11/18 7:38 AM, Vlad Lazar wrote:
> Hi.
>
> This patch makes the scheduler prefer instructions with higher cost if
> two given instructions are equally good.
> Issuing more restricted instructions first is particularly useful on
> in-order cores because it increases the
> number of dual issue
On 9/11/18 9:00 AM, Kyrill Tkachov wrote:
> Hi Ramana,
>
> On 11/09/18 15:55, Ramana Radhakrishnan wrote:
>> On Tue, 11 Sep 2018, 14:38 Vlad Lazar, wrote:
>>
>> > Hi.
>> >
>> > This patch makes the scheduler prefer instructions with higher cost
>> if two
>> > given instructions are equally good.
On Sun, 2 Sep 2018, Gerald Pfeifer wrote:
> Committed.
Actually there still was one warning, though technically not a
validation error.
Fixed thusly, using CSS and using instead of on the
way, so improving this even beyond that.
Gerald
Index: projects/tree-ssa/vectorization.html
Hello!
The mode of clobbers was changed from QImode to reg_raw_mode from a
long time ago...
2018-09-11 Uros Bizjak
* reg-stack.c (subst_asm_stack_regs): Call replace_reg also
for clobbers. Remove obsolete comment.
Bootstrapped and regression tested on x86_64-linux-gnu {,-m32}.
Applied.
Gerald
Index: gcc-3.3/gcj-status.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-3.3/gcj-status.html,v
retrieving revision 1.11
diff -u -r1.11 gcj-status.html
--- gcc-3.3/gcj-status.html 2 Sep 2018 19:10:52 -
Applied, after visual inspection before/after.
Gerald
Index: c99status.html
===
RCS file: /cvs/gcc/wwwdocs/htdocs/c99status.html,v
retrieving revision 1.66
diff -u -r1.66 c99status.html
--- c99status.html 1 Sep 2018 23:41:59
Hi again,
On Tue, Sep 11, 2018 at 10:04:45AM -0500, Will Schmidt wrote:
> On Fri, 2018-09-07 at 16:54 -0500, Segher Boessenkool wrote:
> > On Fri, Sep 07, 2018 at 04:27:34PM -0500, Will Schmidt wrote:
> > > -/* { dg-final { scan-assembler-times "lvx" 4 } } */
> > > -/* { dg-final {
On Tue, 11 Sep 2018, Qing Zhao wrote:
> '-finline-visibility=[all|extern|static]'
> By default, GCC inlines functions without considering their
> visibility. This flag allows finer control of inlining based on
> their visibility.
>
> The value 'extern' tells the compiler to
Fixed as 'obvious' on trunk and 8-branch in revisions 264210 and
264213 respectively.
Paul
2018-09-11 Paul Thomas
PR fortran/87277
* expr.c (is_subref_array): Add the check of dimensionality for
class, dummy, pointer arrays.
2018-09-11 Paul Thomas
PR fortran/87277
*
On Wed, Sep 5, 2018 at 4:38 AM, Kyrill Tkachov
wrote:
> On 04/09/18 17:52, Kyrill Tkachov wrote:
>>
>> On 04/09/18 15:31, Richard Biener wrote:
>>> On Tue, 4 Sep 2018, Kyrill Tkachov wrote:
Hi Richard,
On 31/08/18 12:07, Richard Biener wrote:
> On Thu, 30 Aug 2018, Kyrill
2018-09-11 Uros Bizjak
* config/i386/i386.md (define_attr "type"): Remove mpxmov, mpxmk,
mpxchk, mpxld and mpxst types.
(define_attr length_immediate): Remove all processing of mpx types.
(define_attr prefix_0f): Ditto.
(define_attr memory): Ditto.
Patch was bootstrapped
On Mon, Aug 06, 2018 at 11:14:17AM -0500, Vlad Lazar wrote:
> Hi,
>
> The patch adds support for the TARGET_COMPUTE_FRAME_LAYOUT hook on AArch64
> and removes unneeded frame layout recalculation.
>
> The removed aarch64_layout_frame calls are unnecessary because the functions
> in which
> they
Hello!
Attached patch removes unnecessary mixed-mode log patterns. These are
not needed, because all x87 float_extend RTXes degenerate to a plain
move (or a no-op move). The patch also includes a couple of cleanups
with no functional changes.
2018-09-11 Uros Bizjak
* config/i386/i386.md
On Tue, 11 Sep 2018, Andrew Stubbs wrote:
> On 05/09/18 15:44, Joseph Myers wrote:
> > On Wed, 5 Sep 2018, Andrew Stubbs wrote:
> >
> > > I suppose someone might choose to compile things on an alternative system
> > > for
> > > running on a compatible system, in which case we'd want to simply
On 05/09/18 15:44, Joseph Myers wrote:
On Wed, 5 Sep 2018, Andrew Stubbs wrote:
I suppose someone might choose to compile things on an alternative system for
running on a compatible system, in which case we'd want to simply skip this
binary.
How does one normally do this?
I'd expect a
On Sat, Sep 1, 2018 at 7:10 PM, Tom de Vries wrote:
> On 08/31/2018 11:19 PM, Jason Merrill wrote:
>> On Fri, Aug 24, 2018 at 11:38 AM, Tom de Vries wrote:
>>> [ was: Re: [PATCH][debug] Add -gdescriptive-dies ]
>>> On Fri, Aug 24, 2018 at 12:44:38PM +0200, Richard Biener wrote:
On Wed, 22
On Mon, Sep 10, 2018 at 08:31:49PM -0400, Julian Brown wrote:
> #pragma omp target enter data map(to:arr[0:32])
>
> //#pragma omp target // works, mapped as zero-length array section (*)
> //#pragma omp target map(tofrom:myptr) // crashes (**)
In this case OpenMP doesn't allow the implementation
On Tue, Sep 4, 2018 at 9:01 AM, H.J. Lu wrote:
> On Tue, Aug 28, 2018 at 11:04 AM, H.J. Lu wrote:
>> With -mavx, for
>>
>> [hjl@gnu-cfl-1 skx-2]$ cat foo.i
>> extern float f;
>> extern double d;
>> extern int i;
>>
>> void
>> foo (void)
>> {
>> d = f;
>> f = i;
>> }
>>
>> we need to generate
I'd expect all the cases where you pass a string containing '%s' to the
generic diagnostic functions to use %qs instead of '%s' for quoting, to
get appropriate locale-specific quotes.
--
Joseph S. Myers
jos...@codesourcery.com
On Fri, Sep 07, 2018 at 11:03:28AM -0500, Tamar Christina wrote:
> Hi Richard,
>
> The 08/28/2018 21:58, Richard Sandiford wrote:
> > Tamar Christina writes:
> > > + HOST_WIDE_INT guard_used_by_caller = STACK_CLASH_CALLER_GUARD;
> > > + /* When doing the final adjustment for the outgoing
Hi Richard,
> -Original Message-
> From: Richard Sandiford
> Sent: Tuesday, September 11, 2018 15:49
> To: Tamar Christina
> Cc: Jeff Law ; gcc-patches@gcc.gnu.org; nd
> ; James Greenhalgh ;
> Richard Earnshaw ; Marcus Shawcroft
>
> Subject: Re: [PATCH][GCC][AArch64] Updated
On 09/01/2018 08:10 PM, Tom de Vries wrote:
>> Please add more of this description to the one-line documentation
>> patch you have now;
> Done.
>
>> there are many DIEs that have no name because they
>> don't need one, and this patch doesn't add names to all of them.
>>
> I've added the
* Vineet Gupta [2018-09-10 11:26:35 -0700]:
>* config/arc/arc.c: object attributes for core4 not reflected correctly
>* config/arc/arc.h: Don't restrict DBNZ to core3 (core4 includes core3)
>
> Signed-off-by: Vineet Gupta
> ---
> gcc/ChangeLog| 7 +++
>
Tamar Christina writes:
>> > +
>> > + /* No probe leave. */
>> > + ASM_OUTPUT_INTERNAL_LABEL (asm_out_file, loop_end_lab);
>> > + return "";
>>
>> With the CFA stuff and constant load, I think this works out as:
>>
>> -
>> # 12 insns
>> mov
On Tue, Sep 04, 2018 at 10:13:43AM -0500, Sam Tebbs wrote:
> Hi James,
>
> Thanks for the feedback. Here is an update with the changes you proposed
> and an updated changelog.
>
> gcc/
> 2018-09-04 Sam Tebbs
>
> PR target/85628
> * config/aarch64/aarch64.md
Hi,
This is a simple patch to add a new first-class option
-finline-visibility={all|extern|static}
to finer control inlining based on function’s visibility.
'-finline-visibility=[all|extern|static]'
By default, GCC inlines functions without considering their
visibility. This flag
On Tue, 11 Sep 2018, Martin Liška wrote:
> I've discussed the topic with Alexander on the Cauldron and we hoped
> that the issue with unique __gcov_root can be handled with DECL_COMMON in
> each DSO.
> Apparently this approach does not work as all DSOs in an executable have
> eventually
> just a
On Fri, 2018-09-07 at 16:54 -0500, Segher Boessenkool wrote:
> Hi!
>
> On Fri, Sep 07, 2018 at 04:27:34PM -0500, Will Schmidt wrote:
> > The expected instructions for this test (p8-vec-xl-*) were incorrect for
> > some
> > of the power targets. Add codegen variations as appropriate for
> > the
>
> > This to me feels like the wrong approach as it feels like you are assuming
> > INSN_COST is latency in some way ? Surely, we shouldn't be introducing
> > INSN_COST based stuff into the scheduler.
> >
> > Have you investigated using TARGET_SCHED_ADJUST_COST (IIRC, look for the
> > right
The driver gcc.c has 2 internal error reporting routines
perror_with_name & pfatal_with_name, which take a single string argument.
Firstly the latter forwards to the former, then calls exit after doing
some cleanup. However, that cleanup is already registered via atexit,
so it'll get done
Hi Ramana,
On 11/09/18 15:55, Ramana Radhakrishnan wrote:
On Tue, 11 Sep 2018, 14:38 Vlad Lazar, wrote:
> Hi.
>
> This patch makes the scheduler prefer instructions with higher cost if two
> given instructions are equally good.
> Issuing more restricted instructions first is particularly
On Tue, 11 Sep 2018, 14:38 Vlad Lazar, wrote:
> Hi.
>
> This patch makes the scheduler prefer instructions with higher cost if two
> given instructions are equally good.
> Issuing more restricted instructions first is particularly useful on
> in-order cores because it increases the
> number of
Tamar Christina writes:
> Hi Richard,
>
> The 08/28/2018 21:58, Richard Sandiford wrote:
>> Tamar Christina writes:
>> > + HOST_WIDE_INT guard_used_by_caller = STACK_CLASH_CALLER_GUARD;
>> > + /* When doing the final adjustment for the outgoing argument size
>> > we can't
>> > + assume that LR
On 05/09/18 13:43, Richard Biener wrote:
No. You might want to look into the x86 backend if there's maybe more tweaks
needed when using non-vector mask modes.
I tracked it down to the vector alignment configuration.
Apparently the vectorizer likes to build a "truth" vector, but is
perfectly
On 10.09.2018 19:53, Jakub Jelinek wrote:
> On Mon, Sep 10, 2018 at 04:05:26PM +0200, Andreas Krebbel wrote:
>> On 20.08.2018 16:30, Jeff Law wrote:
>>> On 08/18/2018 03:20 AM, Eric Botcazou wrote:
> Eric, didn't your patches explicitely handle this case of a non-constant
> inbetween?
On 08/29/2018 06:08 PM, Alexander Monakov wrote:
> On Wed, 29 Aug 2018, Martin Liška wrote:
>>> Can you shortly tell why the testcase in the PR segfaults? Does the issue
>>> only affect indirect call profiling?
>>
>> What happens is that there will exist 2 instances of:
>> void *
Hi everyone,
this patch optimizes a few boolean expressions,
and fixes some unneeded whitespace.
Will do a full make check soon.
Thanks for your time.
2018-09-11 MCC CS
gcc/
PR tree-optimization/87261
* match.pd: Add boolean optimizations,
fix whitespace.
Hi.
This patch makes the scheduler prefer instructions with higher cost if two
given instructions are equally good.
Issuing more restricted instructions first is particularly useful on in-order
cores because it increases the
number of dual issue opportunities.
For example, on AArch64, instead
PR libstdc++/87278
* include/bits/shared_ptr.h (make_shared): Use remove_cv instead of
remove_const.
* testsuite/20_util/shared_ptr/creation/87278.cc: New test.
Tested x86_64-linux, committed to trunk and gcc-8-branch.
commit
ping!
Am Mo., 3. Sep. 2018 um 21:19 Uhr schrieb Janus Weil :
>
> Hi all,
>
> attached is a simple patch that fixes a regression which was
> introduced by one of my recent commits (spotted by Dominique).
>
> My first impulse to fix the spurious error was to check for the
> 'intrinsic' attribute,
LWG DR 2905 says that is_constructible_v, P, D const &>
should be false when D is not copy constructible. This commit implements
the changes from the DR and simplifies the signatures as per
https://github.com/cplusplus/draft/issues/1530
* include/bits/unique_ptr.h (__uniq_ptr_impl): Add
On 2018/9/11 1:22 AM, Cesar Philippidis wrote:
On 09/10/2018 08:04 AM, Chung-Lin Tang wrote:
GOACC_2.0 {
Index: libgomp/oacc-mem.c
===
--- libgomp/oacc-mem.c (revision 264192)
+++ libgomp/oacc-mem.c (working copy)
@@ -153,8
For addition and subtraction, we're currently punting on things like
[0,0] - VR_VARYING. However, if we normalize VR_VARYING to the entire
domain, we can see that the above is actually [-MIN+1, +MAX].
We would've normally caught this sort of things with my rewrite of the
binary operators,
On 05/09/18 14:41, David Malcolm wrote:
Please add the:
auto_diagnostic_group d;
line to the top of the function.
OK with that change.
Here's what I committed.
Andrew
Add sorry_at diagnostic function.
The plain "sorry" diagnostic only gives the "current" location, which is
typically
Doing this was approved by Richi at some point when I rewrote the
division code. I finally got around to testing it.
Committed.
gcc/
* tree-vrp (extract_range_from_binary_expr_1): Treat all divisions
by zero as VR_UNDEFINED.
diff --git a/gcc/tree-vrp.c b/gcc/tree-vrp.c
index
We can calculate wide_int_range_shift_undefined_p() without having to
pass down the sign bit of the operand. Also, vrp_shift_undefined_p is a
brain dead wrapper so I'm removing it.
OK for trunk?
commit 9aeb62d4c33b50bc007b07ec5097e8f3edd4b31b
Author: Aldy Hernandez
Date: Mon Sep 10
On 09/05/2018 09:28 PM, Indu Bhagat wrote:
> Patch for PR 86957 " gcc should warn about missing profiles for a compilation
> unit or a new function with -fprofile-use".
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86957
>
> The patch adds -Wmissing-profile warning flag to alert user about
On 09/05/2018 09:28 PM, Indu Bhagat wrote:
Hi.
Thanks for working on that. I believe it's useful enhancement. Note that I'm
not profile feedback maintainer,
but I'll provide some feedback:
> Patch for PR 86957 " gcc should warn about missing profiles for a compilation
> unit or a new function
Am Mo., 10. Sep. 2018 um 08:54 Uhr schrieb Paul Richard Thomas
:
>
> Hi Janus,
>
> That's fine - OK for both branches.
Committed to trunk as r264201. Will do 8-branch soon.
Thanks for the review,
Janus
> On 9 September 2018 at 21:34, Janus Weil wrote:
> > Hi all,
> >
> > the attached patch
61 matches
Mail list logo
