On Sun, 29 Apr 2018, Richard Biener wrote:
> On April 29, 2018 1:06:47 AM GMT+02:00, David Edelsohn
> wrote:
> >Hi, Richi
> >
> >I had been using two source trees to speed the bisection and didn't
> >realize
> >that one defaulted to DWARF debugging and the other defaulted to XCOFF
> >debugging,
Hi!
I'd like to ping following patches for stage1:
- PR78420 __builtin_early_constant_p
http://gcc.gnu.org/ml/gcc-patches/2018-03/msg00355.html
- use --push-state --as-needed and --pop-state around -lgcc_s
http://gcc.gnu.org/ml/gcc-patches/2018-04/msg00567.html
- PR85466 next{aft
> It seems like we're likely to need the same thing when we get to
> make_decl_one_only by other paths; perhaps we should put this
> recalculation in a cxx_make_decl_one_only.
There are 4 calls to make_decl_one_only in the cp/ directory: the one at stake
here in comdat_linkage, 1 in maybe_make_on
Hi all,
attached patch partially reverts r259385 preventing an ICE. At the same time
the check_dependency routine is parameterized correctly so that variable
dependencies in coarray expressions are determined more precisely.
Bootstrapped and regtested ok on x86_64-linux-gnu/f27. Ok for trunk,
gcc
Bootstrapped on x86_64-unknown-linux-gnu, applied.
Richard.
2018-04-30 Richard Biener
* tree-chrec.h (evolution_function_is_constant_p): Remove
redundant check.
* tree-cfg.c (tree_node_can_be_shared): Re-order checks.
Index: gcc/tree-chrec.h
=
>
> FAIL: gcc.dg/lto/20081204-2 c_lto_20081204-2_0.o-c_lto_20081204-2_0.o
> link, -w -flto -fPIC -r -nostdlib (internal compiler error)
> FAIL: gcc.dg/lto/20090914-2 c_lto_20090914-2_0.o-c_lto_20090914-2_0.o
> link, -O2 -flto -fuse-linker-plugin (internal compiler error)
> FAIL: gcc.dg/lto/2009101
On Sat, Apr 28, 2018 at 11:20 AM, Eric Botcazou wrote:
>> This looks not a generic enough fix to me - consider
>>
>> void foo(void) { int a[10]; a[0] = 1; a[9] = 1; }
>> int main() { try { foo (); } catch (...) {} }
>>
>> with -fnon-call-exceptions. If we'd like to catch the SEGV from sta
>
> Honza, do you have any suggestions here for the issue of detecting possible
> recursion as opposted to just direct recursion?
Well, you can look for strongly connected components but you must assume that
every call out of the current unit that is not leaf may call any function that
is externa
On 04/28/2018 11:33 AM, Sandra Loosemore wrote:
On 04/27/2018 07:12 AM, Nathan Sidwell wrote:
Hmmm, I'd like to wordsmith this a bit and clean up the markup, etc. How
about this?
looks good, this is what I'm committing, thanks!
nathan
--
Nathan Sidwell
2018-04-30 Nathan Sidwell
San
On Sat, 28 Apr 2018, Will Hawkins wrote:
> +{
> + warning (OPT_Wignored_qualifiers, "asm-specifier is ignored in "
> + "typedef declaration");
This does not match the documented semantics of -Wignored-qualifiers. I
don't think it's appropriate to expand those semantics to inc
Fixes a singular vs plural mixup.
Committed to trunk.
commit e72e2d5538d11734cd4c9ac2b4edcf179062cf6e
Author: Jonathan Wakely
Date: Mon Apr 30 13:10:35 2018 +0100
* doc/invoke.texi (-fdebug-types-section): Fix grammar.
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
index
DWARF5 defines a small header for .debug_str_offsets. Since we only use
it for split dwarf .dwo files we don't need to keep track of the actual
index offset in an attribute.
gcc/ChangeLog:
* dwarf2out.c (count_index_strings): New function.
(output_indirect_strings): Call count_in
We already emit DWARF5 attributes and tables for indirect addresses
and string offsets, but still use GNU forms. Add a new helper function
dwarf_FORM () for emitting the right form.
Currently we only use the uleb128 forms. But DWARF5 also allows
1, 2, 3 and 4 byte forms (DW_FORM_strx[1234] and DW_
On Fri, Apr 27, 2018 at 9:21 AM, graham stott via gcc-patches
wrote:
> All
> Just a heads the trunk has been broken since about Weds most files fail
> compare during bootstrap at O3 but pass at O2
> My last succesful boostrap at O3 was Tuesday
> I no idea which commit caused it
Should be fixed n
On Fri, Apr 27, 2018 at 11:43 PM, Jason Merrill wrote:
> 61982 notes that an explicit destructor call or delete expression ends
> the lifetime of an object, but we weren't clobbering affected objects
> if their destructors are trivial. This patch fixes that.
>
> The first commit just adds a helpe
This finally removes verify_expr and moves all its bells and whistles
(where they exist) elsewhere. There's one pice that didn't fit somewhere
and I put it where it fit most closely (but it still doesn't work because
of weird debug stmts so it's #if 0ed out).
This ends the series of verifier TLC
I noticed a couple of missing definite articles in the description of
-fpie and -fPIE, and found the last sentence unclear. This clarifies
that it means you should use these options during compilation if you
plan to use -pie during linking (rather than implying you use these
during linking as well
committed. Thank you for your review,
Claudiu
From: Andrew Burgess [andrew.burg...@embecosm.com]
Sent: Friday, April 27, 2018 11:27 PM
To: Claudiu Zissulescu
Cc: gcc-patches@gcc.gnu.org; francois.bed...@synopsys.com; Claudiu Zissulescu
Subject: Re: [PATCH 3/
committed. Thank you for your review,
Claudiu
From: Andrew Burgess [andrew.burg...@embecosm.com]
Sent: Friday, April 27, 2018 11:39 PM
To: Claudiu Zissulescu
Cc: gcc-patches@gcc.gnu.org; francois.bed...@synopsys.com; Claudiu Zissulescu
Subject: Re: [PATCH 4/
committed. Thank you for your review,
Claudiu
From: Andrew Burgess [andrew.burg...@embecosm.com]
Sent: Friday, April 27, 2018 11:40 PM
To: Claudiu Zissulescu
Cc: gcc-patches@gcc.gnu.org; francois.bed...@synopsys.com
Subject: Re: [PATCH 5/5] [ARC] Clear the i
Hi,
atm parallel-dims.c fails on Titan-V, with a cuda launch failure:
...
libgomp: cuLaunchKernel error: too many resources requested for launch
...
We've got a check in the libgomp nvptx plugin to prevent the cuda launch
failure and give a more informative error message:
...
/* Check if the
Hi,
atm parallel-dims.c fails on og7 with Titan V due to too few resources.
This patch reduces the amount of resources used for one offloading
region, and moves another to a link-only test-case.
This allows the test-case to pass.
Committed to og7.
Thanks,
- Tom
[openacc, testsuite] Reduce r
Hi All,
This patch adds the missing neon intrinsics for all 128 bit vector Integer
modes for the
three-way XOR and negate and xor instructions for Arm8.2-a to Armv8.4-a.
Bootstrapped and regtested on aarch64-none-linux-gnue and no issues.
Ok for master? And for backport to the GCC-8 branch?
gc
Hi Richard,
On 27/04/18 09:24, Richard Biener wrote:
On Thu, 26 Apr 2018, Richard Biener wrote:
>
> The following makes loop-header copying stop after the first exit test
> it copied. The reports rightfully complain about too much peeling.
> If some cases pop up which show we should peel up t
Successfully bootstrapped & regrtested on x86_64-pc-linux-gnu.
Committed to trunk as r259766.
gcc/ChangeLog:
* input.h (builtins_location_check): Convert to a STATIC_ASSERT.
---
gcc/input.h | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/gcc/input.h b/gcc/input.h
in
location_get_source_line returns a const char * that isn't 0-terminated,
writing back a length through an int * param.
This is error-prone, as all call-sites have to take into account the
lack of 0-termination, and respect the length of the buffer.
It's cleaner to bundle together this pointer+len
Successfully bootstrapped & regrtested on x86_64-pc-linux-gnu.
Committed to trunk as r259773.
gcc/ChangeLog:
* selftest.h: Fix alphabetization of per-source-file selftest
declarations.
---
gcc/selftest.h | 14 +++---
1 file changed, 7 insertions(+), 7 deletions(-)
diff -
GCC Maintainers:
The following patch adds tests for the vec_max builtin.
The patch for the test case was tested on
powerpc64le-unknown-linux-gnu (Power 8 LE)
powerpc64-unknown-linux-gnu (Power 8 BE)
powerpc64le-unknown-linux-gnu (Power 9 LE).
Please let me know if the patch looks
Our selftest assertions were of the form:
ASSERT_EQ (expected, actual)
and both Richard Sandiford and I find this "Yoda ordering" confusing.
Our existing tests aren't entirely consistent about this, and it doesn't make
sense for ASSERT_NE and its variants.
The ordering comes from googletest's
On Mon, Apr 30, 2018 at 3:50 AM Richard Biener wrote
> On Sun, 29 Apr 2018, Richard Biener wrote:
> > On April 29, 2018 1:06:47 AM GMT+02:00, David Edelsohn <
dje@gmail.com> wrote:
> > >Hi, Richi
> > >
> > >I had been using two source trees to speed the bisection and didn't
> > >realize
> >
On 03/29/2018 04:23 AM, Tom de Vries wrote:
> [ Fix ENOPATCH ]
>
> On 03/29/2018 12:17 PM, Tom de Vries wrote:
>> Hi,
>>
>> when we compile a function with attributes:
>> ...
>> int __attribute__((noinline, noclone))
>> foo (void)
>> {
>> return 2;
>> }
>> ...
>>
>> like this:
>> ...
>> gcc mai
On 04/30/2018 06:51 AM, Jonathan Wakely wrote:
> I noticed a couple of missing definite articles in the description of
> -fpie and -fPIE, and found the last sentence unclear. This clarifies
> that it means you should use these options during compilation if you
> plan to use -pie during linking (rat
GCC Maintainers:
The following patch adds a new test file for the vec_first_match_index,
vec_first_mismatch_index, vec_first_match_or_eos_index,
vec_first_mismatch_or_eos_index builtins.
The patch for the test case was tested on
powerpc64le-unknown-linux-gnu (Power 8 LE)
powerpc64-unkno
On 4/30/18 10:36 AM, Carl Love wrote:
> -/* We generate xxlor instructions for many reasons other than or'ing vector
> - operands or calling __builtin_vec_or(), which means we cannot rely on
> - their usage counts being stable. Therefore, we just ensure at least one
> - xxlor instruction wa
On 01/09/2018 11:39 AM, Richard Sandiford wrote:
> This patch generalises various places that used hwi tree accessors
> so that they can handle poly_ints instead. Earlier patches did
> this while updating interfaces; this patch just mops up some
> left-over pieces that weren't necessary to make th
GCC Maintainers:
I have re-worked the patch per Peter's comments. I retested the patch
on P8 BE, P8 LE and P9 LE.
Please let me know if the patch looks OK for GCC mainline.
Carl Love
gcc/testsuite/ChangeLog:
2018-
This removes paired single (used on the 750CL and friends). It was
deprecated in GCC 8. Removing it means we only have one vector model
to deal with (VMX+VSX, 16-byte vectors).
Tested on powerpc64-linux {-m32,-m64} and on powerpc64le-linux.
I plan to commit this later this week.
Segher
2018-
On 01/09/2018 02:41 PM, Martin Sebor wrote:
> I found a few problems in the previous revision:
>
> 1) It didn't handle the simple case of member arrays of const
> struct objects (only member arrays of const arrays of structs
> were handled).
> 2) The array_ctor_elt() function returned a nar
Hello,
this transformation was lacking symmetry, only handling & and not |.
It probably still fails to handle a < b & a <= 123, while it would handle
< 124, but that's for another day.
Bootstrap+testsuite on powerpc64le-unknown-linux-gnu.
2018-05-01 Marc Glisse
PR tree-optimizati
On 01/12/2018 06:21 AM, Wilco Dijkstra wrote:
> Hi,
>
> Here is the updated version:
>
> This patch implements some of the optimizations discussed in
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71026.
>
> Simplify (C / x >= 0.0) into x >= 0.0 with -fno-signed-zeros
> and -ffinite-math-only.
Hi all,
We can improve the performance of complex floating-point multiplications by
inlining the expansion a bit more aggressively.
We can inline complex x = a * b as:
x = (ar*br - ai*bi) + i(ar*bi + br*ai);
if (isunordered (__real__ x, __imag__ x))
x = __muldc3 (a, b); //Or __mulsc3 for singl
On 01/12/2018 02:30 PM, Martin Sebor wrote:
> A failure in a test for the recently enhanced -Warray-bounds
> warning exposed an unnecessarily broad restriction in the strlen
> pass that prevents it from tracking the length of a member string
> of locally defined and initialized struct:
>
> void
On 04/27/2018 02:41 AM, Alan Modra wrote:
> This patch adds ATTRIBUTE_NONSTRING, which will be used to curb
> -Wstringop-truncation warnings in binutils. OK to apply?
>
> * ansidecl.h (ATTRIBUTE_NONSTRING): Define.
+1, FWIW.
Thanks,
Pedro Alves
* Excludes: ld gas gdb gprof sim
* For
https://sourceware.org/git/gitweb.cgi?p=binutils-gdb.git;a=commit;h=fe944acf8f858cfe6bcfd00670a88847a464717c
Signed-off-by: Francois H. Theron
---
ChangeLog| 6 ++
configure| 4
configure.ac | 4
3 files changed, 14 insertions(+)
di
On Fri, 12 Jan 2018, Wilco Dijkstra wrote:
Hi,
Here is the updated version:
This patch implements some of the optimizations discussed in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71026.
Simplify (C / x >= 0.0) into x >= 0.0 with -fno-signed-zeros
and -ffinite-math-only. If C is negative t
On Mon, Apr 30, 2018 at 7:51 AM, Joseph Myers wrote:
> On Sat, 28 Apr 2018, Will Hawkins wrote:
>
>> +{
>> + warning (OPT_Wignored_qualifiers, "asm-specifier is ignored in "
>> + "typedef declaration");
>
> This does not match the documented semantics of -Wignored-qualifiers. I
Hi!
Before avx512vl we don't have a single instruction to do V2DImode and
V4DImode abs, but that isn't much different from say V4SImode before SSE3
where we also just emit a short sequence that is better than elementwise
expansion. Bootstrapped/regtested on x86_64-linux and i686-linux, ok for
tru
> +2018-04-30 Francois H. Theron
> +
> + * config.sub: Added "nfp" to basic_machine list.
> + * configure.ac: Added "nfp" target.
> + * configure: Regenerate.
I am not a maintainer, but I noticed that config.sub is not being
modified by this commit -- a small discrepancy between the
On Mon, 30 Apr 2018, Will Hawkins wrote:
> I agree! It was, however, the closest of all the categories that I
> could find that seemed to match the warning that I am trying to emit.
> I will go back and review the categories and see if there is something
> that I missed.
If there isn't a suitable
We parse the initializer for an init-capture as an initializer, which
checks for any unexpanded parameter packs. Normally we avoid that
error in a lambda by checking whether we're in a lambda, but at this
point we haven't entered the lambda scope yet.
Jakub, if we end up needing an RC3 it would b
Ping https://gcc.gnu.org/ml/gcc-patches/2017-11/msg02233.html
On Sat, 25 Nov 2017, Marc Glisse wrote:
Hello,
the way full masks are generated currently in avx2intrin.h is questionable:
opaque for the inline functions, weird/wrong for the macros.
It is possible we may want to add code so the
We were neither rejecting nor properly handling "unsigned typeof".
Since nobody is likely to be relying on a broken extension, this patch
makes it an error rather than the pedwarn it was previously.
While there I overhauled that section of diagnostics to be more
specific and give locations for the
On 04/30/2018 11:24 AM, Marc Glisse wrote:
> Hello,
>
> this transformation was lacking symmetry, only handling & and not |.
>
> It probably still fails to handle a < b & a <= 123, while it would
> handle < 124, but that's for another day.
>
> Bootstrap+testsuite on powerpc64le-unknown-linux-gnu
On 04/30/2018 09:48 AM, David Malcolm wrote:
> Our selftest assertions were of the form:
>
> ASSERT_EQ (expected, actual)
>
> and both Richard Sandiford and I find this "Yoda ordering" confusing.
>
> Our existing tests aren't entirely consistent about this, and it doesn't make
> sense for ASSE
On 01/25/2018 07:16 PM, Martin Sebor wrote:
> PR tree-optimization/83776 - [6/7/8 Regression] missing
> -Warray-bounds indexing past the end of a string literal,
> identified a not-so-recent improvement to constant propagation
> as the reason for GCC no longer being able to detect out-of-
> bounds
On 02/09/2018 05:42 AM, Roger Sayle wrote:
> The following patch implements a number of __builtin_popcount related
> optimizations.
> (i) popcount(x) == 0 can be simplified to x==0, and popcount(x) != 0 to
> x!=0.
> (ii) popcount(x&1) can be simplified to x&1, and for unsigned x,
> popcount(x>>31)
On 04/03/2018 06:29 AM, Martin Liška wrote:
> Hi.
>
> This helps the warning with --save-temps. Doing that one needs to preserve
> comments
> in preprocessed source file.
>
> Ready for trunk?
> Martin
>
> gcc/ChangeLog:
>
> 2018-04-03 Martin Liska
>
> PR preprocessor/78497
> *
We have some github repos that use git submodule to include other repos,
including gcc. When git submodule is used, .git is actually a file not a dir,
and contains the path to the modules file in the parent repo's .git dir. This
patch allows contrib/gcc_update to work in this situation.
OK?
Jim
On 13 April 2018 at 20:21, James Greenhalgh wrote:
> On Fri, Apr 13, 2018 at 03:39:32PM +0100, Sameera Deshpande wrote:
>> On Fri 13 Apr, 2018, 8:04 PM James Greenhalgh,
>> mailto:james.greenha...@arm.com>> wrote:
>> On Fri, Apr 06, 2018 at 08:55:47PM +0100, Christophe Lyon wrote:
>> > Hi,
>> >
>
This patch adds a support function to class gcc_rich_location
to make it easier for fix-it hints to use idiomatic C/C++
indentation, for use by the patch for PR c++/85523.
Successfully bootstrapped & regrtested on x86_64-pc-linux-gnu.
Committed to trunk as r259783.
gcc/ChangeLog:
PR c++/
Following on from the thread on the "gcc" list here:
https://gcc.gnu.org/ml/gcc/2018-04/msg00172.html
here's an updated version of Jonathan's patch, which:
* eliminates the separate "note" in favor of simplying having
the warning itself emit the "return *this;" fix-it hint
pr85523.C: In
On Mon, 2018-04-30 at 20:29 -0400, David Malcolm wrote:
[...]
> In doing so I guarded it with a call to:
> global_dc->option_enabled (OPT_Wreturn_type
Gah; typo, sorry:
> as per the insides of diagnostic.s, since adding a fix-it hint to
~~~^
This patch updates the edit-distance algorithm in spellcheck.c to
support transpositions as well as additions/deletions/substitutions,
so that a transposition error counts as a distance of 1 rather than 2.
This leads to saner suggestions for such cases.
Successfully bootstrapped & regrtested on x
63 matches
Mail list logo
