This libgo patch by Tony Reix passes a -X option to ar on AIX, so that
it does the right object file handling. Bootstrapped and ran Go
testsuite on x86_64-pc-linux-gnu. Committed to mainline.
Ian
Index: gcc/go/gofrontend/MERGE
===
-
Clients of the evrp range analysis may not have initialized the SCEV
infrastructure, and in fact my not want to (DOM for example).
Yet inside both vr-values.c and gimple-ssa-evrp-analyze.c we have calls
into SCEV (that will fault/abort if SCEV is not properly initialized).
This patch allows clie
On 11/21/2017 12:07 PM, Martin Sebor wrote:
> On 11/21/2017 09:55 AM, Jeff Law wrote:
>> On 11/19/2017 04:28 PM, Martin Sebor wrote:
>>> On 11/18/2017 12:53 AM, Jeff Law wrote:
On 11/17/2017 12:36 PM, Martin Sebor wrote:
> The attached patch enhances -Wstringop-overflow to detect more
On 11/16/2017 02:29 PM, Martin Sebor wrote:
> Ping.
>
> I've fixed the outstanding false positive exposed by the Linux
> kernel. The kernel builds with four instances of the warning,
> all of them valid (perfect overlap in memcpy).
>
> I also built Glibc. It shows one instance of the warning, a
On 11/22/2017 02:01 PM, Marc Glisse wrote:
> On Sat, 18 Nov 2017, Jeff Law wrote:
>
>> * gimple-ssa-evrp-analyze.c: New file pulled from gimple-ssa-evrp.c.
>
> With the move to C++, wasn't there a policy to name new files *.cc
> instead of *.c?
I'd wondered about that myself. The only three file
On 11/22/2017 03:57 PM, Jakub Jelinek wrote:
Hi!
My reading of insert_late_enum_def_bindings etc. is that it intentionally
appends stuff into the vector without checking if the trees are already
present there (on the testcase below in particular CONST_DECLs for the e1
and e2 enumerators which ar
This is a regression present on mainline for SPARC under the form of the
failure of g++.dg/tree-prof/partition1.C. The delayed-branch scheduling pass
happily deletes a CROSSING_JUMP_P jump insn, which was precisely added to
bridge the gap between the hot and cold section.
It turns out that CRO
On Wed, Nov 22, 2017 at 10:07:04PM +0100, Jakub Jelinek wrote:
> On Mon, Nov 13, 2017 at 11:32:14AM +0100, Marek Polacek wrote:
> > * g++.dg/abi/pr68355.C: New test.
>
> This testcase FAILs on i686-linux, my understanding is that the ABI
> changed only if TARGET_64BIT on x86_64/i686-linux.
C
(please try to convince your mailer to use something other than
Content-Type: application/octet-stream for a patch)
On Tue, 21 Nov 2017, Dominik Inführ wrote:
this patch tries to extend tree-ssa-dce.c to remove unnecessary
new/delete-pairs (it already does that for malloc/free). Clang does it
Jakub Jelinek writes:
> On Wed, Nov 22, 2017 at 05:46:26PM +, Richard Sandiford wrote:
>> 2017-11-22 Richard Sandiford
>>
>> gcc/
>> * cselib.c (expand_loc, cselib_expand_value_rtx_1): Change
>> justification for checking for the stack and hard frame pointers.
>> Check them
jit.dg/test-error-array-bounds.c is a test of how libgccjit
handles errors coming from the middle-end, using -Warray-bounds as
an example.
It started failing as of r254830 (aka 987b3316c4e7c908e3c7d2151c20971f9707ad66),
due to that commit changing the wording of the message.
This commit fixes the
On Mon, Nov 13, 2017 at 11:32:14AM +0100, Marek Polacek wrote:
> * g++.dg/abi/pr68355.C: New test.
This testcase FAILs on i686-linux, my understanding is that the ABI
changed only if TARGET_64BIT on x86_64/i686-linux.
The following patch fixes it for me, ok for trunk?
2017-11-22 Jakub Jel
Hi!
In the bootstrap today I've noticed another testcase that FAILs, dunno
how it happened I've missed it earlier.
Fixed thusly, committed as obvious to trunk.
2017-11-22 Jakub Jelinek
* g++.dg/graphite/pr41562.C: Add -Wno-return-type.
--- gcc/testsuite/g++.dg/graphite/pr41562.C.jj
On Sat, 18 Nov 2017, Jeff Law wrote:
* gimple-ssa-evrp-analyze.c: New file pulled from gimple-ssa-evrp.c.
With the move to C++, wasn't there a policy to name new files *.cc
instead of *.c?
--
Marc Glisse
Hi!
store_expr which is called if to_rtx is a CONCAT and from has complex mode
covering the whole to_rtx can handle the case when from expands to a CONCAT
or if it has the same complex mode, but if e.g. one mode is CTImode and
the other mode is TCmode and from expands to something other than a CON
Hi!
My reading of insert_late_enum_def_bindings etc. is that it intentionally
appends stuff into the vector without checking if the trees are already
present there (on the testcase below in particular CONST_DECLs for the e1
and e2 enumerators which are already there before), then qsorts the vector
On Mon, Nov 20, 2017 at 02:58:49PM +0100, Martin Liška wrote:
> Hello.
>
> This is fix of compilation error I see with
> --enable-offload-targets=nvptx-none=. It's explained
> in very detail way here:
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83007#c1
>
> Ready for trunk?
> Martin
>
> libg
Looks like this got dropped on the floor back in June.
Jason approved an earlier version of this, so I've gone ahead and
updated it, and committed it to trunk.
Here's the version I've committed to trunk (as r255076), having verified
bootstrap®rtest (on x86_64-pc-linux-gnu).
Changed in v5:
* te
On Wed, Nov 22, 2017 at 10:25 PM, Thomas Koenig wrote:
> Am 22.11.2017 um 21:06 schrieb Janne Blomqvist:
>>
>> Is that better? Ok for trunk?
>
>
> Yes, OK.
>
> Thanks a lot for the patch!
Thanks, committed as r255077!
--
Janne Blomqvist
These are the same as bswap effective target. Also, all listed targets
are capable of int32plus.
2017-11-22 Uros Bizjak
* lib/target-supports.exp (check_effective_target_bswap16): Remove
(check_effective_target_bswap32): Ditto.
(check_effective_target_bswap64): Ditto.
* gcc.dg/
On 11/20/2017 03:06 AM, Richard Biener wrote:
> On Sat, Nov 18, 2017 at 9:39 AM, Jeff Law wrote:
>>
>> A batch of random cleanups in the various vrp related bits.
>>
>> The goal here is to start banging some of the classes into shape.
>> Included in this patch:
>>
>>
>> For evrp_folder:
>> Accep
On 22 November 2017 at 09:44, Richard Biener wrote:
>
> The following fixes if-conversion to free SCEV/niter estimates because
> it DCEs stmts.
>
> Bootstrapped and tested on x86_64-unknown-linux-gnu, applied.
>
> Richard.
>
> 2017-11-22 Richard Biener
>
> PR tree-optimization/83089
>
Am 22.11.2017 um 21:06 schrieb Janne Blomqvist:
Is that better? Ok for trunk?
Yes, OK.
Thanks a lot for the patch!
Regards
Thomas
On Wed, Nov 22, 2017 at 3:58 PM, Shalnov, Sergey
wrote:
> Hi,
> This patch making –mprefer-vector-width= option inclusive. This means that
> if we use –mprefer-vector-width=128 it should switch TARGET_PREFER_AVX128=ON
> and TARGET_PREFER_AVX256=ON also.
> It is minor change to generate “xmm” with
On Wed, Nov 22, 2017 at 8:03 PM, Thomas Koenig wrote:
> Hi Janne,
>
>> Regtested on x86_64-pc-linux-gnu, Ok for trunk?
>>
>> libgfortran/ChangeLog:
>>
>> 2017-11-22 Janne Blomqvist
>>
>> * intrinsics/eoshift0.c (eoshift0): Make shift an index_type.
>
>
> Looks good in principle. Is it pos
Test case that requires 4 GB RAM that exercises eoshift0:
program main
implicit none
integer(1) :: line(2_8**31+10)
line = 42
line = eoshift(line, 2_8**31+7)
print *, line(1:4)
print *, line((ubound(line, 1, kind=8)-4):ubound(line, 1, kind=8))
end program main
Before patch:
❯ ./a.out
On Wed, 22 Nov 2017, Jakub Jelinek wrote:
On Wed, Nov 22, 2017 at 06:34:08PM +0100, Marc Glisse wrote:
Hello,
I hadn't implemented this simplification because I think it is invalid code,
but apparently people do write it, so we might as well handle it sensibly.
This also happens to work around
On Wed, Nov 22, 2017 at 8:16 PM, Thomas Koenig wrote:
> Hi janne,
>
>> Regtested on x86_64-pc-linux-gnu, Ok for trunk?
>
>
> Jerry already OK'd this, so you can commit if you want.
> What you could do is to hide the macro invocation behind
> a macro in libgfortran.h, something like
>
> #define BIG
Added myself as GCC maintainer with r255071
*** ChangeLog ***
2017-11-22 Sudakshina Das
* MAINTAINERS (Write After Approval): Add myself.
Thanks
Sudi
diff --git a/ChangeLog b/ChangeLog
index 13b0321..adaec62 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2017-11-22 Sudaks
On Nov 22 2017, Thomas Koenig wrote:
> Jerry already OK'd this, so you can commit if you want.
> What you could do is to hide the macro invocation behind
> a macro in libgfortran.h, something like
>
> #define BIG_ENDIAN (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
That can easily be confused with th
On 11/21/2017 04:57 AM, James Greenhalgh wrote:
> I've finally built up enough courage to start getting my head around this...
Can't blame you for avoiding :-) This stuff isn't my idea of fun either.
>
> I see one outstanding issue sitting on this patch version:
>
> On Sat, Oct 28, 2017 at 05:08
Hi janne,
Regtested on x86_64-pc-linux-gnu, Ok for trunk?
Jerry already OK'd this, so you can commit if you want.
What you could do is to hide the macro invocation behind
a macro in libgfortran.h, something like
#define BIG_ENDIAN (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)
Also, I have opened
Richard Sandiford writes:
> This patch adds support for the SVE bitwise reduction instructions
> (ANDV, ORV and EORV). It's a fairly mechanical extension of existing
> REDUC_* operators.
>
> Tested on aarch64-linux-gnu (with and without SVE), x86_64-linux-gnu
> and powerpc64le-linux-gnu.
Here's
Hi Janne,
So, attached is a new version of the patch. No update
on the ChangeLog. OK for trunk?
Yup, just really fix the copyright and string length stuff first. Thanks!
Committed as rev 255070 with the fixes.
There are still some files which mention Fortran 95, that can be fixed
later.
Tha
Richard Sandiford writes:
> Two things stopped us using SLP reductions with variable-length vectors:
>
> (1) We didn't have a way of constructing the initial vector.
> This patch does it by creating a vector full of the neutral
> identity value and then using a shift-and-insert function
>
Richard Sandiford writes:
> This patch makes vectorizable_reduction cope with variable-length vectors.
> We can handle the simple case of an inner loop reduction for which
> the target has native support for the epilogue operation. For now we
> punt on other cases, but patches after the main SVE
Hi Janne,
Regtested on x86_64-pc-linux-gnu, Ok for trunk?
libgfortran/ChangeLog:
2017-11-22 Janne Blomqvist
* intrinsics/eoshift0.c (eoshift0): Make shift an index_type.
Looks good in principle. Is it possible to come up with a test case?
This might not be executable on all systems
On Wed, Nov 22, 2017 at 05:46:26PM +, Richard Sandiford wrote:
> 2017-11-22 Richard Sandiford
>
> gcc/
> * cselib.c (expand_loc, cselib_expand_value_rtx_1): Change
> justification for checking for the stack and hard frame pointers.
> Check them by pointer rather than regis
On 11/22/2017 03:31 AM, Martin Jambor wrote:
> On Tue, Nov 21 2017, Jeff Law wrote:
>> On 11/21/2017 04:14 AM, Dominik Inführ wrote:
>>> Hi,
>>>
>>> this patch tries to extend tree-ssa-dce.c to remove unnecessary
>>> new/delete-pairs (it already does that for malloc/free). Clang does it too
>>> a
One of the effects of:
https://gcc.gnu.org/ml/gcc-patches/2017-10/msg02066.html
is that we now emit:
(set (hard-frame-pointer) (stack-pointer))
instead of the previous non-canonical:
(set (hard-frame-pointer) (plus (stack-pointer) (const_int 0)))
However, recent changes to aarch64_ex
On Wed, Nov 22, 2017 at 06:34:08PM +0100, Marc Glisse wrote:
> Hello,
>
> I hadn't implemented this simplification because I think it is invalid code,
> but apparently people do write it, so we might as well handle it sensibly.
> This also happens to work around PR 83104 (already fixed).
>
> boot
On 11/22/2017 04:31 AM, Alan Hayward wrote:
>
>> On 21 Nov 2017, at 03:13, Jeff Law wrote:
>>>
You might also look at TARGET_HARD_REGNO_CALL_PART_CLOBBERED. I'd
totally forgotten about it. And in fact it seems to come pretty close
to what you need…
>>>
>>> Yes, some of the c
Hello,
I hadn't implemented this simplification because I think it is invalid
code, but apparently people do write it, so we might as well handle it
sensibly. This also happens to work around PR 83104 (already fixed).
bootstrap+regtest on powerpc64le-unknown-linux-gnu.
2017-11-22 Marc Gliss
Hello,
The Vxworks ports have specific default settings controlling the
dwarf output to accomodate the possible use of old non-gdb debuggers
provided with the VxWorks environment.
Those settings are hardcoded today, even though different versions
of VxWorks come with different constraints.
This
This patch version does not appear to address my comment that you're
leaving behind comments in c-parser.c relating to Cilk array notations
while removing the subsequent code. (Or in one case actually reindenting
the comment that is no longer relevant, rather than removing it.)
--
Joseph S. M
On 05/03/2016 08:50 AM, Oleg Endo wrote:
Hi,
The attached patch cleans up the SH div and mul patterns a bit. Almost
no changes in CSiBE, except for a few minor improvements.
Tested on sh-elf with
make -k check RUNTESTFLAGS="--target_board=sh-sim\{-m2/-ml,-m2/-mb,
-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/
Thank you both for your comments.
I have added the check to ensure the index vector won't cause an
overflow. I also added tests to the testsuite in order to check that the
loop is vectorized for UINT_MAX - 1 iterations but not UINT_MAX
iterations. I was not able to write code that triggers
INTEGE
Hello,
with POINTER_DIFF_EXPR, checking if the result is integral does not imply
that the arguments are as well, so this check needed a tweak.
Bootstrap+regtest on gcc112, pre-approved by Richard.
2017-11-22 Marc Glisse
PR tree-optimization/83104
gcc/
* vr-values.c (simpli
On 11/22/2017 04:34 AM, Janne Blomqvist wrote:
> By using the __BYTE_ORDER__ predefined macro we don't need the
> determine_endianness function anymore.
>
> Regtested on x86_64-pc-linux-gnu, Ok for trunk?
>
Looks good, thanks for patch!
Jerry
On 11/21/2017 03:18 AM, Eric Botcazou wrote:
First of all, the structuring in this section is screwed up. The
discussion and examples for the previous item (#pragma ivdep) should be
moved inside the @table so that you don't have to introduce another
@table here, just insert another entry into th
Hi,
On 21 November 2017 at 18:23, Jeff Law wrote:
> On 11/21/2017 06:52 AM, Jakub Jelinek wrote:
>> Hi!
>>
>> The C++ FE now emits __builtin_unreachable () with BUILTINS_LOCATION
>> on spots that return from functions/methods returning non-void without
>> proper return. This breaks the -Wreturn
On 22/11/17 15:21, Kyrill Tkachov wrote:
On 22/11/17 11:25, Sudi Das wrote:
Hi Kyrill and Christophe
In case of soft fp testing, there are other assembly directives apart
from the vmov one which are also failing. The directives probably
make more sense in the hard fp context so instead of
lookup_name_fuzzy can offer some reserved words as suggestions for
misspelled words, helping with "singed"/"signed" typos.
PR c++/81610 and PR c++/80567 report problems where the C++ frontend
suggested "if", "for" and "else" as corrections for misspelled variable
names.
The root cause is that in
On Sat, 4 Nov 2017 21:39:14 -0700
Kevin Buettner wrote:
> On Tue, 31 Oct 2017 08:03:22 +0100
> Jakub Jelinek wrote:
>
> > On Mon, Oct 30, 2017 at 04:06:15PM -0700, Kevin Buettner wrote:
> > > This patch adds a new member named "pthread_id" to the gomp_thread
> > > struct. It is initialized i
On 22/11/17 11:25, Sudi Das wrote:
Hi Kyrill and Christophe
In case of soft fp testing, there are other assembly directives apart from the
vmov one which are also failing. The directives probably make more sense in the
hard fp context so instead of removing the vmov, I have added the
-mfloa
On Wed, Nov 22, 2017 at 3:06 AM, Jonathan Wakely wrote:
> On 22/11/17 10:56 +, Jonathan Wakely wrote:
>>
>> On 22/11/17 11:23 +0100, Rainer Orth wrote:
>>>
>>> Hi Jonathan,
>>>
This uses static_assert to improve the errors when attempting to
instantiate invalid specializations of con
On 22/11/17 15:07, Thomas Preudhomme wrote:
On 22/11/17 14:45, Kyrill Tkachov wrote:
Hi Thomas,
On 15/11/17 17:12, Thomas Preudhomme wrote:
Hi,
Functions cmse_nonsecure_call_clear_caller_saved and
cmse_nonsecure_entry_clear_before_return both contain very similar code
to clear registers. W
On 22/11/17 14:45, Kyrill Tkachov wrote:
Hi Thomas,
On 15/11/17 17:12, Thomas Preudhomme wrote:
Hi,
Functions cmse_nonsecure_call_clear_caller_saved and
cmse_nonsecure_entry_clear_before_return both contain very similar code
to clear registers. What's worse, they differ slightly at times so
Hi,
This patch making –mprefer-vector-width= option inclusive. This means that
if we use –mprefer-vector-width=128 it should switch TARGET_PREFER_AVX128=ON
and TARGET_PREFER_AVX256=ON also.
It is minor change to generate “xmm” with –mprefer-vector-width=128
on the platform with “zmm”.
Sergey
2
On Wed, Nov 22, 2017 at 12:01 PM, Alan Hayward wrote:
>
>> On 22 Nov 2017, at 09:14, Richard Biener wrote:
>>
>> On Tue, Nov 21, 2017 at 5:43 PM, Kilian Verhetsel
>> wrote:
>>>
This is PR81179 I think, please mention that in the changelog.
>>>
>>> Correct, my bad for missing that.
>>>
Hi,
I received a private report from a customer that gcc was putting out
calls to __divdf3 when compiling with +nosimd. When the reciprocal math
support was added this was probably an oversight or a typo.
The canonical examples is :
double
foo (double x, double y)
{
ret
Hi,
this ICE on valid is triggered by the existing
cpp0x/lambda/lambda-ice20.C when -Wshadow is requested. Today I
confirmed that it can be reproduced only after r251433, the "Reimplement
handling of lambdas in templates." patch: then, as part of
tsubst_lambda_expr, do_pushdecl calls check_lo
On Wed, Nov 22, 2017 at 3:17 PM, Richard Sandiford
wrote:
> Jakub Jelinek writes:
>> On Wed, Nov 22, 2017 at 02:51:09PM +0100, Richard Biener wrote:
>>> On Wed, Nov 22, 2017 at 10:30 AM, Richard Sandiford
>>> wrote:
>>> > r254589 was supposed to leave tests unchanged for the default setting
>>>
Hi Thomas,
On 15/11/17 17:12, Thomas Preudhomme wrote:
Hi,
Functions cmse_nonsecure_call_clear_caller_saved and
cmse_nonsecure_entry_clear_before_return both contain very similar code
to clear registers. What's worse, they differ slightly at times so if a
bug is found in one careful thoughts is
On Wed, Nov 22, 2017 at 02:17:59PM +, Richard Sandiford wrote:
> Jakub Jelinek writes:
> > On Wed, Nov 22, 2017 at 02:51:09PM +0100, Richard Biener wrote:
> >> On Wed, Nov 22, 2017 at 10:30 AM, Richard Sandiford
> >> wrote:
> >> > r254589 was supposed to leave tests unchanged for the default
On 11/22/2017 08:59 AM, Richard Biener wrote:
Anything else can be done as followup and need not be done as part of
this patch. An enum for
this would work I guess.
I've added this to https://gcc.gnu.org/wiki/ImprovementProjects:
Compress DECL flags
tree-core defines a number of bit fla
Jakub Jelinek writes:
> On Wed, Nov 22, 2017 at 02:51:09PM +0100, Richard Biener wrote:
>> On Wed, Nov 22, 2017 at 10:30 AM, Richard Sandiford
>> wrote:
>> > r254589 was supposed to leave tests unchanged for the default setting
>> > of VECTOR_BITS, but I must have got my sums wrong on pr81136.c.
On Wed, Nov 22, 2017 at 02:51:09PM +0100, Richard Biener wrote:
> On Wed, Nov 22, 2017 at 10:30 AM, Richard Sandiford
> wrote:
> > r254589 was supposed to leave tests unchanged for the default setting
> > of VECTOR_BITS, but I must have got my sums wrong on pr81136.c.
> > Sorry for the breakage.
>
On Wed, Nov 22, 2017 at 1:47 PM, Nathan Sidwell wrote:
> On 11/21/2017 06:14 AM, Dominik Inführ wrote:
>>
>> Hi,
>>
>> this patch tries to extend tree-ssa-dce.c to remove unnecessary
>> new/delete-pairs (it already does that for malloc/free). Clang does it too
>> and it seems to be allowed by
>> h
Hi,
Functions cmse_nonsecure_call_clear_caller_saved () and
cmse_nonsecure_entry_clear_before_return () use a separate variable
holding a pointer to padding_bits_to_clear array's first entry which is
used when calling function compute_not_to_clear_mask (). This does not
save space over using &pa
On Wed, Nov 22, 2017 at 12:15 PM, Richard Sandiford
wrote:
> Jakub Jelinek writes:
>> On Wed, Nov 22, 2017 at 10:09:08AM +, Richard Sandiford wrote:
>>> This patch replaces the REDUC_*_EXPR tree codes with internal functions.
>>> This is needed so that the support for in-order reductions can
On Wed, Nov 22, 2017 at 10:33 AM, Richard Sandiford
wrote:
> wi::add_large and wi::sub_large weren't setting the overflow bit
> correctly for unsigned operations if the result needed fewer HWIs
> than the precision.
>
> Tested on aarch64-linux-gnu, x86_64-linux-gnu and powerpc64le-linux-gnu.
> OK
On Wed, Nov 22, 2017 at 10:30 AM, Richard Sandiford
wrote:
> r254589 was supposed to leave tests unchanged for the default setting
> of VECTOR_BITS, but I must have got my sums wrong on pr81136.c.
> Sorry for the breakage.
>
> Tested on aarch64-linux-gnu, x86_64-linux-gnu and powerpc64le-linux-gnu
On Wed, Nov 22, 2017 at 10:26 AM, Richard Sandiford
wrote:
> This patch fixes a regression caused by r251469, where I'd incorrectly
> converted a call to mode_for_size that sometimes needs MODE_PARTIAL_INTs.
>
> Tested on aarch64-linux-gnu, x86_64-linux-gnu and powerpc64le-linux-gnu.
> Also spot-c
Thanks Kyrill.
Committed the attached rebased patch (same patch but without the last hunk
because a better fix was done in an earlier commit).
Best regards,
Thomas
On 22/11/17 11:57, Kyrill Tkachov wrote:
Hi Thomas,
On 15/11/17 17:08, Thomas Preudhomme wrote:
Hi,
As part of r253256, cmse
On Wed, Nov 22, 2017 at 01:03:06PM +, Richard Sandiford wrote:
> Richard Biener writes:
> > On Wed, 22 Nov 2017, Jakub Jelinek wrote:
> >
> >> On Wed, Nov 22, 2017 at 10:41:19AM +0100, Richard Biener wrote:
> >> > How much churn would it be to pass down a mode alongside the operands
> >> > in
Richard Biener writes:
> On Wed, 22 Nov 2017, Jakub Jelinek wrote:
>
>> On Wed, Nov 22, 2017 at 10:41:19AM +0100, Richard Biener wrote:
>> > How much churn would it be to pass down a mode alongside the operands
>> > in expand_binop? Can't find it right now but didn't we introduce
>> > some rtx_wi
Jakub Jelinek writes:
> Hi!
>
> I've noticed that if we construct a VEC_SERIES and later propagate
> constants into both arguments (could happen e.g. in DEBUG_INSNs, or
> during combine etc.) we don't simplify that.
>
> The following patch adds that, ok for trunk if it passes bootstrap/regtest?
>
On 11/21/2017 06:14 AM, Dominik Inführ wrote:
Hi,
this patch tries to extend tree-ssa-dce.c to remove unnecessary
new/delete-pairs (it already does that for malloc/free). Clang does it too and
it seems to be allowed by
http://www.open-std.org/jtc1/sc22/wg21/docs/papers/2013/n3664.html. I’ve
By using the __BYTE_ORDER__ predefined macro we don't need the
determine_endianness function anymore.
Regtested on x86_64-pc-linux-gnu, Ok for trunk?
libgfortran/ChangeLog:
2017-11-22 Janne Blomqvist
* io/inquire.c (inquire_via_unit): Use __BYTE_ORDER__ predefined
macro.
On Wed, 22 Nov 2017, Jakub Jelinek wrote:
> Hi!
>
> On the following testcase we ICE, because eltsize is 0 and when we
> int_const_binop (TRUNC_DIV_EXPR, maxbound, size_int (0)), that of course
> returns NULL.
> The upper bound for zero sized elements is just not meaningful, if we use
> any index
Hi Thomas,
On 15/11/17 17:08, Thomas Preudhomme wrote:
Hi,
As part of r253256, cmse_nonsecure_entry_clear_before_return has been
rewritten to use auto_sbitmap instead of an integer bitfield to control
which register needs to be cleared. This commit continue this work in
cmse_nonsecure_call_clea
On Wed, 22 Nov 2017, Jakub Jelinek wrote:
> On Wed, Nov 22, 2017 at 11:41:24AM +0100, Richard Biener wrote:
> >
> > I am testing the following (old) patch to value number call lhs
> > according to the ERF_RETURNS_ARG annotation. This allows for
> > more expression simplification and also changes
Hi!
I've noticed that if we construct a VEC_SERIES and later propagate
constants into both arguments (could happen e.g. in DEBUG_INSNs, or
during combine etc.) we don't simplify that.
The following patch adds that, ok for trunk if it passes bootstrap/regtest?
Or should it go into simplify_consta
Regtested on x86_64-pc-linux-gnu, Ok for trunk?
libgfortran/ChangeLog:
2017-11-22 Janne Blomqvist
* intrinsics/eoshift0.c (eoshift0): Make shift an index_type.
---
libgfortran/intrinsics/eoshift0.c | 4 +---
1 file changed, 1 insertion(+), 3 deletions(-)
diff --git a/libgfortran/int
> On 21 Nov 2017, at 03:13, Jeff Law wrote:
>>
>>>
>>> You might also look at TARGET_HARD_REGNO_CALL_PART_CLOBBERED. I'd
>>> totally forgotten about it. And in fact it seems to come pretty close
>>> to what you need…
>>
>> Yes, some of the code is similar to the way
>> TARGET_HARD_REGNO_CALL
Hi Tamar,
On 06/11/17 16:53, Tamar Christina wrote:
Hi All,
This patch adds the NEON intrinsics for Dot product.
Dot product is available from ARMv8.2-a and onwards.
Regtested on arm-none-eabi, armeb-none-eabi,
aarch64-none-elf and aarch64_be-none-elf with no issues found.
Ok for trunk?
gcc
Hi Kyrill and Christophe
In case of soft fp testing, there are other assembly directives apart from the
vmov one which are also failing. The directives probably make more sense in the
hard fp context so instead of removing the vmov, I have added the
-mfloat-abi=hard option.
Is this ok for tr
Jakub Jelinek writes:
> On Wed, Nov 22, 2017 at 10:09:08AM +, Richard Sandiford wrote:
>> This patch replaces the REDUC_*_EXPR tree codes with internal functions.
>> This is needed so that the support for in-order reductions can also use
>> internal functions without too much complication.
>>
JonY <10wa...@gmail.com> writes:
> Is there a problem with using .so for internal libraries instead of
> "dll"...
I think not but I haven't tested it. The problem with using .so instead
of .dll is that producing this non-standard extension may not be easy
or possible depending on the build system
On 22/11/17 10:56 +, Jonathan Wakely wrote:
On 22/11/17 11:23 +0100, Rainer Orth wrote:
Hi Jonathan,
This uses static_assert to improve the errors when attempting to
instantiate invalid specializations of containers, e.g. set,
or unordered_set, hash> (which mixes up the
order of the hasher
> On 22 Nov 2017, at 09:14, Richard Biener wrote:
>
> On Tue, Nov 21, 2017 at 5:43 PM, Kilian Verhetsel
> wrote:
>>
>>> This is PR81179 I think, please mention that in the changelog.
>>
>> Correct, my bad for missing that.
>>
>>> This unconditionally pessimizes code even if there is no valid
On Wed, Nov 22, 2017 at 11:41:24AM +0100, Richard Biener wrote:
>
> I am testing the following (old) patch to value number call lhs
> according to the ERF_RETURNS_ARG annotation. This allows for
> more expression simplification and also changes downstream
> users of the argument to use the return
On 22/11/17 11:23 +0100, Rainer Orth wrote:
Hi Jonathan,
This uses static_assert to improve the errors when attempting to
instantiate invalid specializations of containers, e.g. set,
or unordered_set, hash> (which mixes up the
order of the hasher and equality predicate arguments).
This means i
Hi!
On the following testcase we ICE, because eltsize is 0 and when we
int_const_binop (TRUNC_DIV_EXPR, maxbound, size_int (0)), that of course
returns NULL.
The upper bound for zero sized elements is just not meaningful, if we use
any index we still won't reach maximum object size that way.
Whil
Hi,
this is a revised version of:
https://gcc.gnu.org/ml/gcc-patches/2017-11/msg01452.html
with the following changes:
1. integration of Bernhard's patch for the Fortran front-end,
2. Sandra's fix for the documentation,
3. minor tweaks to the C and C++ front-end,
4. change at the GIMPLE lev
On Wed, Nov 22, 2017 at 10:09:08AM +, Richard Sandiford wrote:
> This patch replaces the REDUC_*_EXPR tree codes with internal functions.
> This is needed so that the support for in-order reductions can also use
> internal functions without too much complication.
>
> This came out of the revie
I am testing the following (old) patch to value number call lhs
according to the ERF_RETURNS_ARG annotation. This allows for
more expression simplification and also changes downstream
users of the argument to use the return value which should help
register allocation and register lifetime in gene
On Wed, Nov 22, 2017 at 10:30:29AM +0100, Richard Biener wrote:
> --- a/gcc/tree-core.h
> +++ b/gcc/tree-core.h
> @@ -1787,7 +1787,9 @@ struct GTY(()) tree_function_decl {
>unsigned has_debug_args_flag : 1;
>unsigned tm_clone_flag : 1;
>unsigned versioned_function : 1;
> - /* No bits l
On Tue, Nov 21 2017, Jeff Law wrote:
> On 11/21/2017 04:14 AM, Dominik Inführ wrote:
>> Hi,
>>
>> this patch tries to extend tree-ssa-dce.c to remove unnecessary
>> new/delete-pairs (it already does that for malloc/free). Clang does it too
>> and it seems to be allowed by
>> http://www.open-std
1 - 100 of 139 matches
Mail list logo
