On 12/06/2017 02:04 PM, Jakub Jelinek wrote:
> Hi!
>
> Aggregate assignments and clears aren't in GIMPLE represented as calls,
> and while often they expand inline, sometimes we emit libcalls for them.
> This patch allows us to tail call those libcalls if there is nothing
> after them. The patch
On 06/22/2018 03:34 PM, Maya Rashish wrote:
> That isn't a default any more because the default is 80486 with FPU.
>
> ---
> gcc/doc/invoke.texi | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
Thanks. Installed on the trunk.
jeff
On 06/20/2018 04:32 AM, Richard Sandiford wrote:
> vect_recog_rotate_pattern had code to prevent operations
> on invariants being vectorised unnecessarily:
>
> if (dt == vect_external_def
> && TREE_CODE (oprnd1) == SSA_NAME
> && is_a (vinfo))
> {
> struct loop *loop =
On 06/18/2018 05:00 PM, Martin Sebor wrote:
> Attached is an updated patch with the additional text as
> discussed below.
OK with a ChangeLog entry.
jeff
On 06/18/2018 09:07 AM, Richard Sandiford wrote:
> This patch adds an overload of vect_reassociating_reduction_p
> that checks for a vectorizable associative reduction,
> since the check was duplicated in three functions.
>
> Tested on aarch64-linux-gnu and x86_64-linux-gnu. OK to install?
>
>
On 05/29/2018 06:08 PM, Qing Zhao wrote:
> Hi, Jeff,
>
> Thanks a lot for your review and comments.
>
> I have updated my patch based on your suggestion, and retested this whole
> patch on both X86 and aarch64.
>
> please take a look at the patch again.
>
> thanks.
>
> Qing
>
>> On May 25,
On 05/22/2018 07:40 PM, Martin Sebor wrote:
> Here's another small refinement to -Wstringop-truncation to
> avoid diagnosing more arguably "safe" cases of strncat() that
> match the expected pattern of
>
> strncat (d, s, sizeof d - strlen (d) - 1);
>
> such as
>
> extern char a[4];
>
This patch by Cherry Zhang fixes the libgo runtime code to
consistently use #ifdef instead of #if for USING_SPLIT_STACK.
Bootstrapped and ran Go testsuite on x86_64-pc-linux-gnu. Committed
to mainline.
Ian
Index: gcc/go/gofrontend/MERGE
Hi Carl,
On Fri, Jun 22, 2018 at 07:32:47AM -0700, Carl Love wrote:
> The following patch changes the word selected when extracting the word
> from the second vector to insert into the first vector by the
> vec_insert() builtin.
>
> Specifically, the test case
>
> vector float
> fn2 (float a,
On Thu, Jun 21, 2018 at 07:11:37PM -0400, Michael Meissner wrote:
> On Thu, Jun 21, 2018 at 06:07:36PM -0500, Segher Boessenkool wrote:
> > On Wed, Jun 20, 2018 at 10:54:18AM -0400, Michael Meissner wrote:
> > > This patch fixes the tests in the testsuite that implicitly were
> > > expecting long
On Thu, Jun 21, 2018 at 10:43:16PM +, Joseph Myers wrote:
> On Thu, 21 Jun 2018, Segher Boessenkool wrote:
>
> > The comment doesn't make much sense. If long double is IBM, the long
> > double complex mode is ICmode, not KCmode. "To get the IEEE128 complex
> > type", perhaps? And can't you
On Fri, Jun 22, 2018 at 04:46:19PM -0400, Fritz Reese wrote:
>
> With the patch, the trunk compiler does not exhibit the bugs in the
> mentioned PRs (82972, 83088, 85851). I did encounter some issues
> building and regression testing the trunk compiler which I am certain
> are unrelated. The
Hi Segher,
After waiting a few days for this newly committed patch to settle, is it ok to
backport to gcc 6, gcc 7, and gcc 8?
Thanks.
On 6/22/18 5:34 PM, Kelvin Nilsen wrote:
> Thanks for feedback. It turns out that the vmrgew and vmrgow instructions
> require power 8.
>
> After
Thanks for feedback. It turns out that the vmrgew and vmrgow instructions
require power 8.
After coordinating with Segher on minor refinements to the test cases, I have
committed the patch as quoted below to the trunk.
On 6/19/18 5:37 PM, Segher Boessenkool wrote:
> Hi!
>
> On Tue, Jun 19,
On 06/18/2018 01:15 PM, Martin Sebor wrote:
> While looking into opportunities to detect strnlen/strlen coding
> mistakes (pr86199) I noticed a bug in the strnlen implementation
> I committed earlier today that lets a strnlen() result be saved
> and used in subsequent calls to strlen() with the
The problem in this testcase was that the initializer looked constant
to is_nondependent_constant_expression, but instantiating it found the
conversion error, but because fold_non_dependent_expr passed tf_none
to the instantiation, we didn't give an error. store_init_value then
happily stuck the
GCC Maintainers:
The following patch addresses test failures on AIX.
The patch has been tested on
powerpc64le-unknown-linux-gnu (Power 8 LE)
powerpc64-unknown-linux-gnu (Power 8 BE)
AIX 7200-00-01-1543 (Power 8 BE)
With no regressions.
Please let me know if the patch looks OK
Hi!
The following testcase got fixed in 8/trunk with r253924 part of
PR82370 enhancements, but that is not IMHO something we should backport.
So instead the following patch adds something simpler, use Yv constraint
for the DImode shift count in instructions that need AVX512VL when EVEX
encoded if
On 03/03/2018 05:48 AM, David Malcolm wrote:
> PR demangler/84668 reports this failure of c++filt (found by fuzzing):
>
> $ c++filt '__H5z55_H5z555'
> c++filt: out of memory allocating 18446744071696285694 bytes after a total of
> 135168 bytes
>
>
Hi!
I've backported following 61 trunk commits to 7.x branch (2 patches contain
more than one commit), bootstrapped/regtested them on x86_64-linux and
i686-linux and committed to 7.x branch.
Jakub
2018-06-22 Jakub Jelinek
Backported from mainline
2018-03-02 Jakub
That isn't a default any more because the default is 80486 with FPU.
---
gcc/doc/invoke.texi | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/gcc/doc/invoke.texi b/gcc/doc/invoke.texi
index 9bb1a4910b2..f231da3cde2 100644
--- a/gcc/doc/invoke.texi
+++ b/gcc/doc/invoke.texi
On Thu, Jun 21, 2018 at 02:22:50PM -0500, Segher Boessenkool wrote:
> On Thu, Jun 21, 2018 at 01:26:19PM -0400, Michael Meissner wrote:
> > On Thu, Jun 21, 2018 at 12:09:12PM -0500, Segher Boessenkool wrote:
> > > > * config.gcc (powerpc64le*): Remove multilib support for IEEE
> > > > and
All,
The attached patch generates special initializers for derived-type
c_ptr and c_funptr symbols so they are handled correctly by the
translation phase. Previously, an EXPR_STRUCTURE expression was
generated for both types, as c_ptr and c_funptr are structure types
with a c_address field. The
On 06/21/2018 11:06 AM, Olivier Hainque wrote:
> Hello Joseph,
>
> Thanks for getting back to me on this!
>
>> On 19 Jun 2018, at 17:50, Joseph Myers wrote:
>>
>> On Thu, 7 Jun 2018, Olivier Hainque wrote:
>>
>>> An updated version of the patch is attached, accounting for
>>> your two comments
On 06/20/2018 07:56 AM, Eric Botcazou wrote:
> Hi,
>
> the fix for PR target/81736 unconditionally overrules
> -fno-omit-frame-pointer,
> including at -O0 where it is implicit. That seems brutal and unnecessary so
> the attached patch reverts it at -O0 and thus fixes 135 failures in the GDB
On 05/16/2018 05:53 AM, Martin Liška wrote:
> On 12/21/2017 10:13 AM, Martin Liška wrote:
>> On 12/20/2017 06:45 PM, Jakub Jelinek wrote:
>>> Another thing is that the "/" in there is wrong, so
>>> const char dir_separator_str[] = { DIR_SEPARATOR, '\0' };
>>> char *b = concat
On 06/12/2018 08:02 PM, Alexandre Oliva wrote:
>
> We didn't split cross-partition ranges in loclists to output a
> whole-function location expression, but with nonzero locviews, we
> force loclists, and then we have to split to avoid cross-partition
> list entries.
>
> From: Alexandre Oliva
>
On 06/19/2018 09:13 AM, dave.pa...@oracle.com wrote:
> This patch fixes spurious failure for C test added with original patch
> for bug 55976.
>
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=55976
>
> Added dg-prune-output for extraneous message causing unexpected test
> failure for test
NightStrike and I were chatting on IRC last week about
issues with trying to vectorize the following code:
#include
std::size_t f(std::vector> const & v) {
std::size_t ret = 0;
for (auto const & w: v)
ret += w.size();
return ret;
}
icc could vectorize it,
On Fri, Jun 22, 2018 at 02:08:27PM -0600, Jeff Law wrote:
> On 06/12/2018 04:46 PM, Adam Borowski wrote:
> > Here's a patch for Bugzilla 80271, to add support for env CLICOLOR_FORCE.
> >
> > Automated build environments -- and even some which you run by hand --
> > redirect the output somewhere
On 06/12/2018 04:46 PM, Adam Borowski wrote:
> Hi!
> Here's a patch for Bugzilla 80271, to add support for env CLICOLOR_FORCE.
>
> Automated build environments -- and even some which you run by hand --
> redirect the output somewhere to save a log, even if they tee it to the
> terminal anyway.
On петък, 22 юни 2018 г. 10:55:29 EEST Jeff Law wrote:
> On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> > The floating point support has been borrowed from C6X libgcc port
> > to help with TI PRU toolchain ABI compatibility.
> >
> > libgcc/ChangeLog:
> >
> > 2018-06-13 Dimitar Dimitrov
> >
On Thu, Jun 21, 2018 at 06:12:52PM -0500, Segher Boessenkool wrote:
> On Wed, Jun 20, 2018 at 10:49:41AM -0400, Michael Meissner wrote:
> > These patches fix the tests in the testsuite that check whether
> > -mno-float128
> > works properly. In these cases, I explicitly run them with long double
OK.
On Fri, Jun 22, 2018 at 1:18 PM, Paolo Carlini wrote:
> Hi,
>
> this rather straightforward patch simply uses available location information
> - in the form of declspecs->locations - in diagnostic messages. In order to
> do that, the information is also propagated to bad_specifiers and
>
On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * configure: Regenerate.
> * configure.ac: Add PRU target.
>
> gcc/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * config/pru/pru-ldst-multiple.md: Generate using
On Fri, Jun 22, 2018 at 11:33:06AM -0600, Jeff Law wrote:
> On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> > The PRU load/store instructions can access memory with byte
> > granularity for all 30 of its 32-bit GP registers. Examples:
> >
> ># Load 17 bytes from address r0[0] into registers
On Fri, Jun 22, 2018 at 08:30:12AM -0500, Kelvin Nilsen wrote:
> This has been committed to trunk.
>
> Is this ok to backport to gcc6, gcc7, and gcc8?
Yes, but wait a few days please. Thanks!
Segher
On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> The PRU load/store instructions can access memory with byte
> granularity for all 30 of its 32-bit GP registers. Examples:
>
># Load 17 bytes from address r0[0] into registers r10.b1-r14.b2
>lbbo r10.b1, r0, 0, 17
>
># Load 100 bytes
Hi!
On the 7.x branch, without the fix offsetof2.C testcase passes at -O0,
which is how it is compiled in the testsuite, and only fails at -O2.
Thus, this patch adds a copy of that testcase to be compiled for -O2, so
that it is tested too, and I'll backport then both the original fix and this
On Thu, 21 Jun 2018, Eric Gallager wrote:
> On 6/21/18, Jeff Law wrote:
> > On 06/12/2018 11:21 AM, Joseph Myers wrote:
> >> On Tue, 12 Jun 2018, Martin Sebor wrote:
> >>
> >>> The proposal to enable -Wstrict-prototypes discussed below
> >>> was considered too late for GCC 8. I'd like to revive
Hi,
this rather straightforward patch simply uses available location
information - in the form of declspecs->locations - in diagnostic
messages. In order to do that, the information is also propagated to
bad_specifiers and mark_inline_variable. Tested x86_64-linux.
Thanks, Paolo.
On Thu, 21 Jun 2018, Jeff Law wrote:
> I think all this implies that the setting of -fno-math-errno by default
> really depends on the math library in use since it's the library that
> has to arrange for either errno to get set or for an exception to be raised.
If the library does not set errno,
On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> The floating point support has been borrowed from C6X libgcc port
> to help with TI PRU toolchain ABI compatibility.
>
> libgcc/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * config.host: Add PRU target.
> * config/pru/asri.c:
On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * g++.old-deja/g++.abi/ptrmem.C: Add PRU to list.
OK once the rest of the kit is approved.
jeff
On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * gcc.dg/tree-ssa/20040204-1.c: XFAIL on pru.
> * gcc.dg/tree-ssa/reassoc-33.c: Ditto.
> * gcc.dg/tree-ssa/reassoc-34.c: Ditto.
> *
On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * gcc.c-torture/execute/20101011-1.c: Define DO_TEST to 0 for PRU.
> * gcc.dg/20020312-2.c: No PIC register for PRU.
This is OK once the rest of the kit is approved.
On 06/13/2018 12:58 PM, Dimitar Dimitrov wrote:
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * gcc.dg/stack-usage-1.c: Define PRU stack usage.
This is OK once the rest of the kit is approved.
Jeff
On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> Not all C language features are supported when -mabi=ti option is
> used for PRU target.
>
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * lib/gcc-dg.exp: Filter unsupported features in PRU's TI ABI mode.
> *
On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> PRU architecture supports maximum 256k program memory (IMEM). Some GCC
> test cases manage to produce executables bigger than that.
>
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * lib/gcc-dg.exp: Bail on region
On 06/13/2018 12:57 PM, Dimitar Dimitrov wrote:
> gcc/testsuite/ChangeLog:
>
> 2018-06-13 Dimitar Dimitrov
>
> * gcc.target/pru/abi-arg-struct.c: New test.
> * gcc.target/pru/ashiftrt.c: New test.
> * gcc.target/pru/builtins-1.c: New test.
> *
On 06/21/2018 10:01 PM, Dimitar Dimitrov wrote:
> On четвъртък, 21 юни 2018 г. 17:03:55 EEST Jeff Law wrote:
>> On 06/21/2018 11:44 AM, Vladimir Makarov wrote:
>>> On 06/13/2018 02:58 PM, Dimitar Dimitrov wrote:
From: Dimitar Dimitrov
For some targets, Pmode != UNITS_PER_WORD. Take
Hi Jeff,
> On 22 Jun 2018, at 18:13, Jeff Law wrote:
>
>>> * gimplify.c (gimplify_function_tree): Prevent creation
>>> of a trampoline for the address of the current function
>>> passed to entry/exit instrumentation hooks.
> I was a little concerned about the TREE_NO_TRAMPOLINE
Hello,
getenv_spec_function features code to return a fake
value for an undefined variable when we know that we
shouldn't complain about the variable being undefined
and that the variable value doesn't matter.
This code is there to prevent errors about undefined
variables referenced in specs
On 06/22/2018 03:23 AM, Martin Liška wrote:
> Hi.
>
> For correctness of clustering algorithm:
> https://www.semanticscholar.org/paper/Short-Communication%3A-Correction-to-'Producing-Good-Kannan-Proebsting/311091fb9fb9d38f7b76e768a603c02acc799fe0
>
> one needs to allow single case clusters as
On 06/22/2018 05:06 AM, Martin Liška wrote:
> On 06/20/2018 04:41 PM, David Malcolm wrote:
>> On Mon, 2018-05-14 at 14:50 +0200, Martin Liška wrote:
>>> First part with introduction of auto_string_vec class.
>>>
>> FWIW, I'm fine with the changes to the jit subdir, but I don't think I
>> have
On 06/18/2018 02:34 PM, Marek Polacek wrote:
> This patch improves the checks in size_must_be_zero_p so that we don't
> call get_range_info with SIZE of a pointer type.
>
> Bootstrapped/regtested on x86_64-linux, ok for trunk and 8?
>
> 2018-06-18 Marek Polacek
>
> PR middle-end/86202
On 06/22/2018 09:47 AM, Olivier Hainque wrote:
>
>
>> On 22 Jun 2018, at 17:44, Olivier Hainque wrote:
>
>> Receiving the actual function address would seem more
>> useful for function-instrumentation purposes, and this
>> patch adjusts the compiler to do that instead.
>>
>> 2018-06-22
On 06/22/2018 05:25 AM, Martin Liška wrote:
> On 06/20/2018 05:27 PM, David Malcolm wrote:
>> On Mon, 2018-05-14 at 14:51 +0200, Martin Liška wrote:
>>> Main part where I still need to write ChangeLog file and
>>> gcc.sh needs to be moved to bash-completions project.
>>>
>>> Martin
>>
>> As
On 06/22/2018 05:07 AM, Martin Liška wrote:
> On 06/20/2018 04:57 PM, David Malcolm wrote:
>> On Mon, 2018-05-14 at 14:51 +0200, Martin Liška wrote:
>>> Second part refactors function from gcc.c into a new class
>>> option_proposer.
>>>
>>> Martin
>>
>> [...snip...]
>>
>> diff --git
On 06/22/2018 04:32 AM, Martin Liška wrote:
> On 06/20/2018 05:16 PM, Jeff Law wrote:
>> On 06/20/2018 05:25 AM, Steven Bosscher wrote:
>>> On Tue, Jun 12, 2018 at 10:44 PM, Jeff Law wrote:
On 06/05/2018 01:15 AM, marxin wrote:
> + The definition of "much bigger" depends on whether we
On 06/22/2018 03:11 AM, Kugan Vivekanandarajah wrote:
> When we set niter with maybe_zero, currently final_value_relacement
> will not happen due to expression_expensive_p not handling. Patch 1
> adds this.
>
> With that we have the following optimized gimple.
>
>[local count: 118111601]:
>
On 06/22/2018 05:38 AM, Tom de Vries wrote:
> Hi,
>
> when doing a combined build of gcc and binutils-gdb repos, guality still
> uses the gdb in PATH by default.
>
> This patch makes sure that quality uses the gdb from the combined build
> instead.
>
> Tested on x86_64.
>
> OK for trunk?
>
>
On 06/22/2018 05:35 AM, Tom de Vries wrote:
> Hi,
>
> by default, guality invokes gdb with --quiet to avoid being unnecessarily
> verbose for each test. But as a consequence, we don't see which version of
> gdb is used.
>
> This patch adds printing of the full path to the gdb used, as well as
On 06/22/2018 02:39 AM, Richard Biener wrote:
> On Thu, Jun 21, 2018 at 9:37 PM Jeff Law wrote:
>>
>> On 06/21/2018 11:04 AM, Eric Botcazou wrote:
>>> When code is compiled at -O0, the RTL middle-end makes sure that location
>>> information is preserved as much as possible by generating NOPs with
On Tue, 2018-06-19 at 15:22 +, Joseph Myers wrote:
>
>
> What should plain
>
> gcc -Wl,-rpath=/foo
>
> do? Whatever it should do regarding trying to link or not trying to link,
> it should do the same if you also name a header on the command line to be
> compiled to a precompiled
> On 22 Jun 2018, at 17:44, Olivier Hainque wrote:
> Receiving the actual function address would seem more
> useful for function-instrumentation purposes, and this
> patch adjusts the compiler to do that instead.
>
> 2018-06-22 Olivier Hainque
>
> * gimplify.c
Hello,
Execution of the example program below shows that when
a nested function is called by way of a stack trampoline,
the -finstrument-functions hooks receive the address of
the trampoline as the "this function" parameter.
Receiving the actual function address would seem more
useful for
GCC Maintainers:
The following patch adds tests for the vec_extract builtin. I also
adjusts the second argument on the existing tests so they match the
ABI, specifically an integer not a const integer.
The patch has been tested on
powerpc64le-unknown-linux-gnu (Power 9 LE)
With no
This patch fixes a number of test suite failures caused by data too large for
the address space or alignment larger than supported by this target.
paul
testsuite/ChangeLog:
2018-06-22 Paul Koning
* gcc.c-torture/execute/builtins/lib/chk.c: Use smaller alignment
if
GCC Maintainers:
The following patch changes the word selected when extracting the word
from the second vector to insert into the first vector by the
vec_insert() builtin.
Specifically, the test case
vector float
fn2 (float
Reportedly glibc 2.28 removes the ustat.h header file and the ustat
function. This breaks libgo, which provides syscall.Ustat that calls
the C ustat function. Fix this by removing syscall.Ustat from libgo.
Bootstrapped and ran Go testsuite on x86_64-pc-linux-gnu. Committed
to mainline, GCC 8
On 22/06/18 15:14 +0200, Christophe Lyon wrote:
Hi,
On Fri, 22 Jun 2018 at 14:54, Jonathan Wakely wrote:
On 22/06/18 14:51 +0200, Rainer Orth wrote:
>Hi Jonathan,
>
>> The SSO basic_string has a non-standard insert(iterator, initializer_list)
>> overload, from a C++0x draft. This adds the
On 22/06/18 14:13 +0100, Jonathan Wakely wrote:
On 22/06/18 13:54 +0100, Jonathan Wakely wrote:
On 22/06/18 14:51 +0200, Rainer Orth wrote:
Hi Jonathan,
The SSO basic_string has a non-standard insert(iterator, initializer_list)
overload, from a C++0x draft. This adds the correct overload,
On 22/06/18 14:56, Andre Vieira (lists) wrote:
On 17/05/18 09:29, Kyrill Tkachov wrote:
>
> On 16/05/18 10:22, Jérôme Lambourg wrote:
>> Hello Kyrill,
>>
>>> Thanks for the patch! To validate it your changes you can also look
>>> at the disassembly
>>> of the cmse.c binary in the build tree.
On 22/06/18 14:56, Andre Vieira (lists) wrote:
On 05/06/18 11:18, Kyrill Tkachov wrote:
>
> On 05/06/18 11:03, Andre Vieira (lists) wrote:
>> On 05/06/18 10:42, Kyrill Tkachov wrote:
>>> Hi Andre,
>>>
>>> On 05/06/18 10:40, Andre Vieira (lists) wrote:
Hi all,
This patch makes
On 17/05/18 09:29, Kyrill Tkachov wrote:
>
> On 16/05/18 10:22, Jérôme Lambourg wrote:
>> Hello Kyrill,
>>
>>> Thanks for the patch! To validate it your changes you can also look
>>> at the disassembly
>>> of the cmse.c binary in the build tree. If the binary changes with
>>> your patch then that
On 05/06/18 11:18, Kyrill Tkachov wrote:
>
> On 05/06/18 11:03, Andre Vieira (lists) wrote:
>> On 05/06/18 10:42, Kyrill Tkachov wrote:
>>> Hi Andre,
>>>
>>> On 05/06/18 10:40, Andre Vieira (lists) wrote:
Hi all,
This patch makes the arm_cmse.h header file ISO C compliant. We were
This has been committed to trunk.
Is this ok to backport to gcc6, gcc7, and gcc8?
Thanks.
On 6/19/18 2:30 PM, Segher Boessenkool wrote:
> Hi!
>
> On Mon, Jun 18, 2018 at 11:29:55AM -0500, Kelvin Nilsen wrote:
>> +/* A single vpkudus matches twice because this is compiled with -dp,
>> +
On Fri, 2018-06-22 at 13:25 +0200, Martin Liška wrote:
> On 06/20/2018 05:27 PM, David Malcolm wrote:
> > On Mon, 2018-05-14 at 14:51 +0200, Martin Liška wrote:
[...snip...]
> > Thanks for working on this; the rest looks good to me (though as I
> > said, I'm not officially a reviewer for this).
Hi,
On Fri, 22 Jun 2018 at 14:54, Jonathan Wakely wrote:
>
> On 22/06/18 14:51 +0200, Rainer Orth wrote:
> >Hi Jonathan,
> >
> >> The SSO basic_string has a non-standard insert(iterator, initializer_list)
> >> overload, from a C++0x draft. This adds the correct overload, while also
> >>
On 22/06/18 13:54 +0100, Jonathan Wakely wrote:
On 22/06/18 14:51 +0200, Rainer Orth wrote:
Hi Jonathan,
The SSO basic_string has a non-standard insert(iterator, initializer_list)
overload, from a C++0x draft. This adds the correct overload, while also
preserving the old one so that the old
On 22/06/18 13:42 +0100, Jonathan Wakely wrote:
On 22/06/18 14:34 +0200, Rainer Orth wrote:
Hi Jonathan,
PR libstdc++/70940
* include/experimental/memory_resource (__resource_adaptor_common):
New base class.
(__resource_adaptor_common::_AlignMgr): Helper for
On 22/06/18 14:51 +0200, Rainer Orth wrote:
Hi Jonathan,
The SSO basic_string has a non-standard insert(iterator, initializer_list)
overload, from a C++0x draft. This adds the correct overload, while also
preserving the old one so that the old symbol is still exported from the
library.
The
Hi Jonathan,
> The SSO basic_string has a non-standard insert(iterator, initializer_list)
> overload, from a C++0x draft. This adds the correct overload, while also
> preserving the old one so that the old symbol is still exported from the
> library.
>
> The COW basic_string doesn't have any of
On 22/06/18 14:34 +0200, Rainer Orth wrote:
Hi Jonathan,
PR libstdc++/70940
* include/experimental/memory_resource (__resource_adaptor_common):
New base class.
(__resource_adaptor_common::_AlignMgr): Helper for obtaining aligned
pointer from unaligned,
Hi Jonathan,
> PR libstdc++/70940
> * include/experimental/memory_resource (__resource_adaptor_common):
> New base class.
> (__resource_adaptor_common::_AlignMgr): Helper for obtaining aligned
> pointer from unaligned, and vice versa.
>
Hi,
when doing a combined build of gcc and binutils-gdb repos, guality still
uses the gdb in PATH by default.
This patch makes sure that quality uses the gdb from the combined build
instead.
Tested on x86_64.
OK for trunk?
Thanks,
- Tom
[testsuite/guality] Use buildarea gdb by default in
On 06/21/2018 05:25 PM, Steve Kargl wrote:
> On Thu, Jun 21, 2018 at 09:03:47AM +0100, Paul Richard Thomas wrote:
>> Ping!
>>
>>> 2018-06-19 Paul Thomas
>>>
>>> PR fortran/49630
>>> * resolve.c (resolve_contained_fntype): Change standard ref.
>>> from F95 to F2003: C418. Correct a
Hi,
by default, guality invokes gdb with --quiet to avoid being unnecessarily
verbose for each test. But as a consequence, we don't see which version of
gdb is used.
This patch adds printing of the full path to the gdb used, as well as the
gdb version as printed by gdb.
Tested on x86_64.
OK
On 06/20/2018 05:27 PM, David Malcolm wrote:
> On Mon, 2018-05-14 at 14:51 +0200, Martin Liška wrote:
>> Main part where I still need to write ChangeLog file and
>> gcc.sh needs to be moved to bash-completions project.
>>
>> Martin
>
> As before, I'm not an official reviewer for it, but it
Hi,
this patch adds dumpfile support for dumps that come in multiple parts. This
is needed for WPA stream-out dump since we stream partitions in parallel and
the dumps would come up in random order. Parts are added by new parameter that
is initialzed by default to -1 (no parts).
One thing I
On 06/20/2018 04:57 PM, David Malcolm wrote:
> On Mon, 2018-05-14 at 14:51 +0200, Martin Liška wrote:
>> Second part refactors function from gcc.c into a new class
>> option_proposer.
>>
>> Martin
>
> [...snip...]
>
> diff --git a/gcc/c-family/cppspec.c b/gcc/c-family/cppspec.c
> index
On 06/20/2018 04:41 PM, David Malcolm wrote:
> On Mon, 2018-05-14 at 14:50 +0200, Martin Liška wrote:
>> First part with introduction of auto_string_vec class.
>>
>
> FWIW, I'm fine with the changes to the jit subdir, but I don't think I
> have approval rights on the vec.h changes.
Hi.
Good,
The following makes the vectorizer consider all vector sizes as advertised
by targetm.vectorize.autovectorize_vector_sizes and decide on which
vector size to use based on costs.
Given comparing costs is difficult if you do not know the number of
scalar iterations the patch simply uses the cost
This is the main part to make considering multiple vector sizes based on
costs less compile-time costly. It shares dataref analysis and
dependence analysis for loop vectorization (BB vectorization is only
adjusted to comply with the new APIs sofar).
Sharing means that DRs (and of course DDRs)
This is a prerequesite for sharing dataref and dependence analysis
results between different vector size vectorization attempts.
Bootstrapped on x86_64-unknown-linux-gnu (with 2/3 included),
testing in progress.
Richard.
>From 9b406b12269d2a0fe86591df862c2ffd0197df37 Mon Sep 17 00:00:00 2001
On Fri, Jun 22, 2018 at 11:21 AM Martin Liška wrote:
>
> Hello.
>
> Quite obvious patch where I accidentally removed check for:
> if (!targetm.have_casesi () && !targetm.have_tablejump ())
>
> Patch can bootstrap on ppc64le-redhat-linux and survives regression tests.
>
> Ready to be installed?
On 06/20/2018 05:16 PM, Jeff Law wrote:
> On 06/20/2018 05:25 AM, Steven Bosscher wrote:
>> On Tue, Jun 12, 2018 at 10:44 PM, Jeff Law wrote:
>>> On 06/05/2018 01:15 AM, marxin wrote:
+ The definition of "much bigger" depends on whether we are
+ optimizing for size or for
Am 2018-06-22 um 09:51 schrieb Rainer Orth:
Hi Franz,
No idea about possible problems, but isn't it usually recommended to use
either _FILE_OFFSET_BITS=64 or _LARGEFILE{64}_SOURCE=1, not both at the
same time?
quite the contrary: for regular largefile support, you're supposed to
use `getconf
Hello,
ping for https://gcc.gnu.org/ml/gcc-patches/2018-06/msg00228.html
> On 5 Jun 2018, at 16:13, Olivier Hainque wrote:
> The attached patch is a proposal to improve specs processing
> so %* works in spec function arguments (it doesn't as of today).
Thanks in advance!
With Kind Regards,
1 - 100 of 113 matches
Mail list logo
