Below, please find the 3nd submitted version of this patch, now with some more
issues resolved.
Regards,
Abe
From 87af575347e216672e322bbc1b4ae0a5ab93507f Mon Sep 17 00:00:00 2001
From: Abe
Date: Mon, 18 May 2015 14:26:29 -0500
Subject: [PATCH] fix PR46029: reimplement if conversion of lo
On Thu, Jul 09, 2015 at 05:52:20PM +0200, Thomas Schwinge wrote:
> --- gcc/builtins.def
> +++ gcc/builtins.def
> @@ -182,7 +182,9 @@ along with GCC; see the file COPYING3. If not see
> #define DEF_GOMP_BUILTIN(ENUM, NAME, TYPE, ATTRS) \
>DEF_BUILTIN (ENUM, "__builtin_" NAME, BUILT_IN_NORMAL,
Hi,
the DR got resolved in time for C++11 and Jonathan noticed that we
should remove the pedwarn, not a big deal. Tested x86_64-linux.
Thanks,
Paolo.
//
/cp
2015-07-09 Paolo Carlini
PR c++/61491
* pt.c (maybe_process_partial_specialization): Allow enum temp
On 07/09/2015 07:56 AM, mliska wrote:
gcc/ChangeLog:
2015-07-09 Martin Liska
* dbgcnt.def: Add new debug counter.
* ipa-icf-gimple.c (func_checker::compare_ssa_name): Add flag
for strict mode.
(func_checker::compare_memory_operand): Likewise.
(func_che
On 07/08/2015 08:42 PM, Andrew MacLeod wrote:
blah, not so trivial. One of the primary things predict.h does is
create enum br_predictor by including predict,def.. so moving that enum
doesnt really make sense.
Fixing gimple,h isn't too bad, I could split the prediction stuff out
into gimple-pre
graphite-scop-detection.c contained a copy of graphite.c
print_graphite_statistics() and print_graphite_scop_statistics().
The latter gained a parameter to distinguish
\nBefore limit_scops SCoP statistics (
from
\nSCoP statistics (
Note that previously the version in gimple.c was never called (AF
On 07/09/2015 12:09 PM, Paolo Carlini wrote:
the DR got resolved in time for C++11 and Jonathan noticed that we
should remove the pedwarn, not a big deal. Tested x86_64-linux.
How about adding the testcase from the DR as well? OK with that change.
Jason
On 07/09/2015 09:41 AM, Jakub Jelinek wrote:
On Thu, Jul 09, 2015 at 09:34:25AM -0600, Jeff Law wrote:
On 07/09/2015 08:13 AM, Jakub Jelinek wrote:
On Thu, Jul 09, 2015 at 03:56:35PM +0200, mliska wrote:
---
gcc/testsuite/g++.dg/ubsan/vptr-1.C | 2 +-
gcc/testsuite/g++.dg/ubsan/vptr-2.C | 2
On 07/09/2015 03:19 AM, Tom de Vries wrote:
On 09/07/15 05:33, Jeff Law wrote:
On 07/07/2015 09:58 AM, Tom de Vries wrote:
[Big snip]
0001-Add-rewrite_virtuals_into_loop_closed_ssa.patch
Add rewrite_virtuals_into_loop_closed_ssa
2015-07-07 Tom de Vries
* tree-cfg.c (get_virtual_phi):
On 07/09/2015 10:30 AM, Bernhard Reutner-Fischer wrote:
Thanks,
gcc/ChangeLog
2015-07-09 Bernhard Reutner-Fischer
* graphite.h: New file.
(print_graphite_statistics): Extern declaration.
* graphite-scop-detection.c (print_graphite_scop_statistics,
print_grap
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* hash-set.h (remove): New function.
(iterator): New iteration class for hash_set.
OK.
jeff
On 07/08/2015 10:42 PM, Andrew MacLeod wrote:
On 07/08/2015 06:43 PM, Jeff Law wrote:
predict.h is actually required by gimple.h for a few reasons, enum
be_predictor is used in parameter lists and a few inlines use the
TAKEN,
NOT_TAKEN macros
Its also needed by cfghooks.h, and betwen those 2 f
This patch enables builds with mips[32|64]r3 and mips[32|64]r5 in the
mips-mti-linux-gnu toolchain. t-mti-linux uses MULTILIB_MATCHES to
map these to r2 but SYSROOT_SUFFIX_SPEC was not being set properly
to find the sysroot (the r2 one) for these architectures. This patch
fixes that problem by up
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* cgraph.c (symbol_table::create_edge): Introduce summary_uid
for cgraph_edge.
* cgraph.h (struct GTY): Likewise.
* ipa-inline-analysis.c (estimate_function_body_sizes): Use
n
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-inline-analysis.c
(inline_edge_summaries): New data structure.
(redirect_to_unreachable): Use renamed
function get_inline_edge_summary.
(edge_set_predicate): Likewise.
gcc/ChangeLog
2015-07-09 Bernhard Reutner-Fischer
* ipa-icf.c (sem_item_optimizer::do_congruence_step): Fix typo
in dump message.
Ok for trunk if testing passes?
Hmz, that's obvious, will commit tomorrow after the regstrap during
night.
---
gcc/ipa-icf.c | 2 +-
1 file change
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-reference.c (ipa_ref_opt_summary_t): New class.
(get_reference_optimization_summary): Use it.
(set_reference_optimization_summary): Likewise.
(ipa_init): Remove hook holders usa
On 07/09/2015 11:32 AM, Bernhard Reutner-Fischer wrote:
gcc/ChangeLog
2015-07-09 Bernhard Reutner-Fischer
* ipa-icf.c (sem_item_optimizer::do_congruence_step): Fix typo
in dump message.
Ok for trunk if testing passes?
Hmz, that's obvious, will commit tomorrow after the regst
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-cp.c (struct edge_clone_summary): New structure.
(class edge_clone_summary_t): Likewise.
(edge_clone_summary_t::initialize): New method.
(edge_clone_summary_t::duplicate): Likew
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-pure-const.c (struct funct_state_d): New.
(funct_state_d::default_p): Likewise.
(has_function_state): Remove.
(get_function_state): Likewise.
(set_function_state): Likew
The first patch is a cleanup hoisted from the c++-concepts branch: on
the branch we want to be able to parse type/template in multiple places,
so Andrew factored that out, but it's a good code cleanup on the trunk
as well.
While looking at this and related code on the branch, I noticed that
t
On July 9, 2015 6:48:39 PM GMT+02:00, Jeff Law wrote:
>On 07/09/2015 10:30 AM, Bernhard Reutner-Fischer wrote:
>>
>> Thanks,
>>
>> gcc/ChangeLog
>>
>> 2015-07-09 Bernhard Reutner-Fischer
>>
>> * graphite.h: New file.
>> (print_graphite_statistics): Extern declaration.
>> * graphi
On 07/09/2015 11:06 AM, Andrew MacLeod wrote:
On 07/08/2015 10:42 PM, Andrew MacLeod wrote:
On 07/08/2015 06:43 PM, Jeff Law wrote:
predict.h is actually required by gimple.h for a few reasons, enum
be_predictor is used in parameter lists and a few inlines use the
TAKEN,
NOT_TAKEN macros
Its al
On 07/09/2015 11:48 AM, Bernhard Reutner-Fischer wrote:
On July 9, 2015 6:48:39 PM GMT+02:00, Jeff Law wrote:
On 07/09/2015 10:30 AM, Bernhard Reutner-Fischer wrote:
Thanks,
gcc/ChangeLog
2015-07-09 Bernhard Reutner-Fischer
* graphite.h: New file.
(print_graphite_statist
There's no reason we should have the logic for whether a TRAIT_EXPR is
value-dependent in two places.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit fcd3802b2b88251b6c0dea63b38e6f269be7713d
Author: Jason Merrill
Date: Tue Jul 7 23:49:21 2015 -0400
* pt.c (instantiation_dependent_
On Thu, Jul 09, 2015 at 12:25:18PM +0200, Andre Vehreschild wrote:
>
> I need your help on how to interpret the standard(s) or how to
> implement handling an allocatable function's result, when that
> result is not allocated by the function. Imagine the simple
> (albeit artificial) case:
>
> int
On 07/09/2015 01:49 PM, Jeff Law wrote:
On 07/09/2015 11:06 AM, Andrew MacLeod wrote:
The total result affect 227 files.
Now, predict.h is *still* more pervasive than it needs to be, but thats
a different patch :-). There are a set of routines in there like
optimize_{fucntion,loop,edge,bb}_f
We reduce code size for IA MCU by adjusting variable shift costs for IA
MCU. OK for trunk?
Thanks.
H.J.
---
PR target/66821
* config/i386/i386.c (iamcu_cost): Adjust variable shift costs.
---
gcc/config/i386/i386.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --g
Hi!
On Thu, Jul 09, 2015 at 11:24:44AM -0700, Aldy Hernandez wrote:
Thanks for working on it.
> + wide_int offset = wi::neg (addend, &overflow);
> + addend = wide_int_to_tree (TREE_TYPE (addend), offset);
> + if (overflow)
> + warning_at (c_parser_peek_t
Hi Steve,
Thanks for your knowledge. Can you support your statement that an allocatable
function has to return an allocated object by a part of the standard? I totally
agree with you that this code is ill-designed, but IMO is it not the task of
the compiler to address ill design. The compiler h
ix86_split_long_move can optimize floating point constant move, which
can be used to optimize SFmode move for IA MCU.
OK for trunk if there is no regression?
H.J.
---
gcc/
PR target/66824
* config/i386/i386.c (ix86_split_to_parts): Allow SFmode move
for IA MCU.
(
The following patch goes along with Jakub's parsing of ordered(n) loops.
With it, we can now parse his testcase, along with a variety of other
tests with appropriate diagnostics.
The lowering to gimple is still not done, as we should agree on what
needs to be emitted first.
I'll follow up w
Hi,
On 07/09/2015 06:33 PM, Jason Merrill wrote:
On 07/09/2015 12:09 PM, Paolo Carlini wrote:
the DR got resolved in time for C++11 and Jonathan noticed that we
should remove the pedwarn, not a big deal. Tested x86_64-linux.
How about adding the testcase from the DR as well? OK with that cha
> 2015-07-09 Steve Ellcey
>
> * config/mips/mti-linux.h (MIPS_SYSVERSION_SPEC): Update
> to handle mips[32|64]r3 and mips[32|64]r5.
OK, thanks.
Matthew
On Thu, Jul 09, 2015 at 05:22:33PM +0200, Uros Bizjak wrote:
> Hello!
>
> This ICE was caused by a peephole2 pattern that allowed non-general
> regs arguments.
>
> 2015-07-08 Uros Bizjak
>
> PR target/66814
> * config/i386/predicates.md (nonimmediate_gr_operand): New predicate.
>
On Thu, Jul 09, 2015 at 08:59:08PM +0200, Andre Vehreschild wrote:
> Hi Steve,
>
> Thanks for your knowledge. Can you support your statement that an allocatable
> function has to return an allocated object by a part of the standard? I
> totally agree with you that this code is ill-designed, but
On 06/01/2015 06:26 AM, Ilya Enkovich wrote:
2015-05-29 1:15 GMT+03:00 Jeff Law :
Right, but you're blindly propagating. The right thing to do is look at
some kind of metric to estimate when it's profitable to propagate the
constant back in vs leave it hoisted out.
No, the patch is not to bl
On Thu, Jul 9, 2015 at 9:38 PM, Jakub Jelinek wrote:
> On Thu, Jul 09, 2015 at 05:22:33PM +0200, Uros Bizjak wrote:
>> Hello!
>>
>> This ICE was caused by a peephole2 pattern that allowed non-general
>> regs arguments.
>>
>> 2015-07-08 Uros Bizjak
>>
>> PR target/66814
>> * config/i386/
On Thu, Jul 09, 2015 at 05:23:36PM +0300, Maxim Blumental wrote:
> Now the patch is corrected (according to the previous letter) and
> ready to be reviewed. I'm looking forward to your feedback.
Ok, I guess while the filenames are still different, it is quite easily
possible to find corresponding
On Thu, Jul 09, 2015 at 10:13:49PM +0200, Uros Bizjak wrote:
> I was under impression that peephole2 pass doesn't see subregs of hard
> regs (all x86 predicates are written in this way). Even documentation
> somehow agrees with this:
>
> '(subreg:M1 REG:M2 BYTENUM)'
>
> 'subreg' expressions a
Hello!
> The patch was bootstrapped and tested on x86/x86-64.
>
> Committed as rev. 225618.
>
> 2015-07-09 Vladimir Makarov
>
> PR rtl-optimization/66782
> * lra-int.h (struct lra_insn_recog_data): Add comment about
> clobbered hard regs for arg_hard_regs.
> * lr
On Thu, Jul 9, 2015 at 10:17 PM, Jakub Jelinek wrote:
> On Thu, Jul 09, 2015 at 10:13:49PM +0200, Uros Bizjak wrote:
>> I was under impression that peephole2 pass doesn't see subregs of hard
>> regs (all x86 predicates are written in this way). Even documentation
>> somehow agrees with this:
>>
>>
On Thu, Jul 9, 2015 at 8:05 PM, H.J. Lu wrote:
> We reduce code size for IA MCU by adjusting variable shift costs for IA
> MCU. OK for trunk?
IMO, tuning patches should fall into "obvious" category. I don't have
any data to to do any meaningful review of a cost metric for a new
target.
So, inst
On Thu, Jul 9, 2015 at 1:30 PM, Uros Bizjak wrote:
> On Thu, Jul 9, 2015 at 8:05 PM, H.J. Lu wrote:
>> We reduce code size for IA MCU by adjusting variable shift costs for IA
>> MCU. OK for trunk?
>
> IMO, tuning patches should fall into "obvious" category. I don't have
> any data to to do any m
On 07/09/2015 07:15 PM, Jeff Law wrote:
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* cgraph.c (symbol_table::create_edge): Introduce summary_uid
for cgraph_edge.
* cgraph.h (struct GTY): Likewise.
* ipa-inline-analysis.c (estimate_function_b
On 07/09/2015 07:35 PM, Jeff Law wrote:
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-reference.c (ipa_ref_opt_summary_t): New class.
(get_reference_optimization_summary): Use it.
(set_reference_optimization_summary): Likewise.
(ipa_init)
On 07/09/2015 07:44 PM, Jeff Law wrote:
On 07/09/2015 03:13 AM, mliska wrote:
gcc/ChangeLog:
2015-07-03 Martin Liska
* ipa-pure-const.c (struct funct_state_d): New.
(funct_state_d::default_p): Likewise.
(has_function_state): Remove.
(get_function_state): Likewise.
(set_f
On 07/09/2015 03:36 PM, Paolo Carlini wrote:
Finally, we do *not* reject, as we should, the line:
template<> enum A::E : char { echar };
Then, overall, is it Ok to simply suppress the pedwarn in C++11, and
xfail for now the error? Should I open a new, separate bug report about
the latter? (note
OK, thanks.
Jason
On Thu, Jul 09, 2015 at 12:13:38PM -0700, H.J. Lu wrote:
> ix86_split_long_move can optimize floating point constant move, which
> can be used to optimize SFmode move for IA MCU.
>
> OK for trunk if there is no regression?
>
>
> H.J.
> ---
> gcc/
>
> PR target/66824
> * config/i386/
On 09.07.15 17:30, Jeff Law wrote:
On 07/09/2015 07:53 AM, John Marino wrote:
DragonFly supports PIE but the tests for it are disabled.
The attached patch for the trunk will enable these checks on DragonFly.
Thanks for considering this!
John
p.s. suggested gcc/testsuite/changelog entry:
2015-
On Thu, Jul 09, 2015 at 11:21:01AM +0200, Jakub Jelinek wrote:
> > Hmm, this looks like it could be split out to a function.
>
> Ok, will try that.
Here is what I've committed after another bootstrap/regtest on x86_64-linux
and i686-linux. You're right, the separate function cleaned stuff up.
2
Hi!
As discussed on IRC, we have a problem because fold_stmt can call
force_gimple_operand_1, which in turn does
push_gimplify_context/pop_gimplify_context (NULL), even when inside
of some gimplify_ctxp and thus pushes the decls into function context
rather than the current gimplify_ctxp.
So far
On Mon, Jun 29, 2015 at 11:23:24AM +0200, Eric Botcazou wrote:
> > Don't you need to handle convert_nonlocal_omp_clauses similarly (need_chain
> > in that case)?
> > At least looking at your r211308 commit, for !optimize you force not just
> > the frame, but also chain.
>
> You're very likely righ
Hi,
On 07/09/2015 10:53 PM, Jason Merrill wrote:
On 07/09/2015 03:36 PM, Paolo Carlini wrote:
Finally, we do *not* reject, as we should, the line:
template<> enum A::E : char { echar };
Then, overall, is it Ok to simply suppress the pedwarn in C++11, and
xfail for now the error? Should I open
On 07/03/2015 06:18 PM, Richard Sandiford wrote:
Hi Martin,
Martin Liška writes:
On 07/03/2015 03:07 PM, Richard Sandiford wrote:
Martin Jambor writes:
On Fri, Jul 03, 2015 at 09:55:58AM +0100, Richard Sandiford wrote:
Trevor Saunders writes:
On Thu, Jul 02, 2015 at 09:09:31PM +0100, Ric
On 08/07/15 00:41, Jeff Law wrote:
> On 07/07/2015 06:50 AM, Kugan wrote:
>>
>> Thanks for the review. I have addressed your comments above in the
>> attached patch.
>>
>> I have one question with respect to unary operation. For generic unary
>> operation with INTEGER_CST, do we skip this or do w
This is the patch I committed. Bernd pointed out that I didn't need to be so
coy about the branches in the middle of blocks at that point of the compilation
anyway. So we remove a couple of unneeded insn patterns.
nathan
2015-07-09 Nathan Sidwell
Infrastructure:
* gimple.
On 08.07.2015 13:55, Ian Lance Taylor wrote:
> I don't know of anybody who actually uses the DMGL_TYPES support. I
> don't know why anybody would.
>
> Ian
Thanks for pointing that out. I updated the testcases, so that now they
don't depend on DMGL_TYPES being used.
> But better still is to consi
On 06/18/2015 08:32 AM, Yuri Rumyantsev wrote:
Richard,
Here is updated patch which does not include your proposal related to
the target hook deletion.
You wrote:
I still don't understand why you need the new target hook. If we have a masked
load/store then the mask is computed by an assignmen
On Thu, Jul 09, 2015 at 03:46:08PM +0200, Richard Biener wrote:
> On Thu, 9 Jul 2015, Bernhard Reutner-Fischer wrote:
>
> > gcc/ChangeLog
> >
> > 2015-07-09 Bernhard Reutner-Fischer
> >
> > * builtins.c (fold_builtin_tolower, fold_builtin_toupper): New
> > static functions.
> > (f
> Sorry about that. Does the attached Patch fix the problem?
Yup. Thanks!
On Wed, 8 Jul 2015, H.J. Lu wrote:
> On Wed, Jul 8, 2015 at 4:46 AM, Richard Biener wrote:
> >
> > Passes do not expect post-dominators being around and thus forget
> > to invalidate them properly. Thus passes computing them have to
> > free them. The patch fixes path-isolation and adds an asse
On Wed, Jul 8, 2015 at 5:32 PM, H.J. Lu wrote:
> There is no need to try different alignment on variable of
> error_mark_node.
>
> OK for trunk if there is no regression?
Can't we avoid calling align_variable on error_mark_node type decls
completely? That is, punt earlier when we try to emit it.
On Wed, Jul 8, 2015 at 6:37 PM, Mike Stump wrote:
> I’d like to merge in the fix from https://gcc.gnu.org/PR66523 into the
> gcc-5-branch.
>
> RM Ok?
Ok.
Richard.
> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35773:
> diff --git a/gcc/config/darwin.c b/gcc/config/darwin.c
> index 40804b8..0
Hi Jim!
On Wed, 8 Jul 2015 13:00:16 -0500, James Norris
wrote:
> This patch adds handling of the deviceptr clause when
> used within a Fortran program.
Please motivate such non-obvious code changes by a test case. At least
to me, it's not at all obvious what's going on here...
Is that the occ
On Wed, Jun 10, 2015 at 11:38:25AM +, Joseph Myers wrote:
> This patch, for gomp-4_0-branch, adds support for C++ "this" in
> OpenACC directives. (This patch does not do anything to handle OpenMP
> differently from OpenACC; that - bug 66053 - will need to be resolved
> for mainline, either dec
Hi Jim!
On Tue, 7 Jul 2015 10:19:39 -0500, James Norris
wrote:
> This patch fixes an issue where the deviceptr clause in an outer
> directive was being ignored during implicit variable definition
> on a nested directive.
> Committed to gomp-4_0-branch.
> --- a/gcc/gimplify.c
> +++ b/gcc/gimpli
On Wed, 8 Jul 2015, Marek Polacek wrote:
> On Fri, Jul 03, 2015 at 03:41:29PM +0200, Richard Biener wrote:
> > On Fri, 3 Jul 2015, Marek Polacek wrote:
> >
> > > This patch implements a new pass, called laddress, which deals with
> > > lowering ADDR_EXPR assignments. Such lowering ought to help
On Thu, Jul 9, 2015 at 8:35 AM, Paolo Bonzini wrote:
>
>
> On 07/07/2015 11:08, Richard Biener wrote:
>> Also I am not sure ceil_div and floor_div can be handled this way.
>> (5 /[ceil] 2) * 2 == 6 but you compute it as 4. So I am only convinced
>> trunc_div works this way.
>
> Of course also flo
On Thu, Jul 09, 2015 at 10:53:30AM +0200, Richard Biener wrote:
> > I renamed the file to gimple-laddress.c then and adjusted the timevar.
> > Another change is that for x86_64 we don't need -mavx at all, so I dropped
> > that. The test is now restricted to x86_64/i?86; on aarch64/ppc64 we aren't
On Thu, Jul 09, 2015 at 11:04:38AM +0200, Marek Polacek wrote:
> On Thu, Jul 09, 2015 at 10:53:30AM +0200, Richard Biener wrote:
> > > I renamed the file to gimple-laddress.c then and adjusted the timevar.
> > > Another change is that for x86_64 we don't need -mavx at all, so I dropped
> > > that.
On Sat, 4 Jul 2015, Jakub Jelinek wrote:
> On Fri, Jul 03, 2015 at 04:06:26PM +0200, Jakub Jelinek wrote:
> > In the pr59984.c testcase, with Marek's patch and this patch, one loop in
> > test is already vectorized (the ICE was on the other one), I'll work on
> > recognizing multiples of GOMP_SIMD
On 09/07/15 05:33, Jeff Law wrote:
On 07/07/2015 09:58 AM, Tom de Vries wrote:
[Big snip]
0001-Add-rewrite_virtuals_into_loop_closed_ssa.patch
Add rewrite_virtuals_into_loop_closed_ssa
2015-07-07 Tom de Vries
* tree-cfg.c (get_virtual_phi): New function.
* tree-cfg.h (get_virtual_
On Thu, Jul 09, 2015 at 11:14:01AM +0200, Richard Biener wrote:
> > @@ -2764,6 +2769,93 @@ vectorizable_simd_clone_call (gimple stm
> > || thisarginfo.dt == vect_external_def)
> >&& POINTER_TYPE_P (TREE_TYPE (op)))
> > thisarginfo.align = get_pointer_alignment (op) / BIT
Hi,
given that the current warn_args_num only outputs errors, never
warnings, I think we really want to rename it. Unless it's the wrong
time for such changes...?!?
Thanks,
Paolo.
///
2015-07-09 Paolo Carlini
* typeck.c (warn_args_num): Rename to error_args_nu
On Tue, Jul 07, 2015 at 07:47:50AM +0200, Marc Glisse wrote:
> On Tue, 7 Jul 2015, Prathamesh Kulkarni wrote:
>
> >+/* a * (1 << b) -> a << b */
> >+(simplify
> >+ (mult:c @a (lshift integer_onep @b))
> >+ (if (!FLOAT_TYPE_P (type))
> >+(lshift @a @b)))
Just a nit: the last line is wrongly
Jeff Law wrote:
On 07/08/2015 03:43 AM, Richard Biener wrote:
On Wed, Jul 8, 2015 at 12:07 AM, Jeff Law wrote:
On 07/07/2015 06:37 AM, Alan Lawrence wrote:
As per https://gcc.gnu.org/ml/gcc-patches/2015-04/msg01346.html. Fixes
FAIL of advsimd-intrinsics vcreate.c on aarch64_be-none-elf from
p
Hi,
When I going through the code, I spot this minor issue. When
start_cand/orig_cand/third_cand have overall cost in order like "start_cand
< third_cand < orig_cand", GCC chooses the third_cand instead of start_cand
because we haven't set best_cost for start_cand. This is an obvious fix to
it.
Hi,
This patch refactors codes adding iv candidates in IVO. It renames
functions using straightforward names, it also factors function call to
add_autoinc_candidates from add_candidate to add_iv_candidate_for_use.
Before this patch, we tried to add autoinc candidates for every call to
add_candidat
Ping.
On Fri, Jun 26, 2015 at 4:47 PM, Bin Cheng wrote:
> Hi,
> The canonical form of subtract of immediate is (add op0 minus_imm), which is
> supported with addsi3_aarch64 pattern on aarch64. Unfortunately wrong cost
> (8 rather than 4) is computed by aarch64_rtx_cost because it doesn't honor
>
On Thu, Jul 9, 2015 at 11:34 AM, Alan Lawrence wrote:
> Jeff Law wrote:
>>
>> On 07/08/2015 03:43 AM, Richard Biener wrote:
>>>
>>> On Wed, Jul 8, 2015 at 12:07 AM, Jeff Law wrote:
On 07/07/2015 06:37 AM, Alan Lawrence wrote:
>
> As per https://gcc.gnu.org/ml/gcc-patches/2015-04
On Thu, Jul 9, 2015 at 11:37 AM, Bin Cheng wrote:
> Hi,
> When I going through the code, I spot this minor issue. When
> start_cand/orig_cand/third_cand have overall cost in order like "start_cand
> < third_cand < orig_cand", GCC chooses the third_cand instead of start_cand
> because we haven't s
On Thu, Jul 09, 2015 at 10:16:38AM +0200, Richard Biener wrote:
> On Wed, Jul 8, 2015 at 5:32 PM, H.J. Lu wrote:
> > There is no need to try different alignment on variable of
> > error_mark_node.
> >
> > OK for trunk if there is no regression?
>
> Can't we avoid calling align_variable on error_m
On Thu, Jul 9, 2015 at 11:40 AM, Bin Cheng wrote:
> Hi,
> This patch refactors codes adding iv candidates in IVO. It renames
> functions using straightforward names, it also factors function call to
> add_autoinc_candidates from add_candidate to add_iv_candidate_for_use.
> Before this patch, we t
On Thu, Jul 9, 2015 at 11:52 AM, H.J. Lu wrote:
> On Thu, Jul 09, 2015 at 10:16:38AM +0200, Richard Biener wrote:
>> On Wed, Jul 8, 2015 at 5:32 PM, H.J. Lu wrote:
>> > There is no need to try different alignment on variable of
>> > error_mark_node.
>> >
>> > OK for trunk if there is no regressio
Richard Biener wrote:
I wonder why wi::from_buffer doesn't have the same issue though
for HImode ints. It's structured differently, without magic '4's as well.
I don't claim to understand the rest of wi::from_buffer and why it is different.
However, wrt. HImode, I think the key line is:
o
Hi Ilya!
On Tue, 21 Oct 2014 21:20:34 +0400, Ilya Verbin wrote:
> This patch contains liboffloadmic library.
>
> It is used by ICC for offloading. The sources are imported from upstream
> ( https://www.openmprtl.org/sites/default/files/liboffload_oss.tgz )
> Configure and makefiles are new.
>
On Thu, Jul 9, 2015 at 5:49 PM, Richard Biener
wrote:
> On Thu, Jul 9, 2015 at 11:37 AM, Bin Cheng wrote:
>> Hi,
>> When I going through the code, I spot this minor issue. When
>> start_cand/orig_cand/third_cand have overall cost in order like "start_cand
>> < third_cand < orig_cand", GCC choose
On Thu, 9 Jul 2015, Marek Polacek wrote:
> On Tue, Jul 07, 2015 at 07:47:50AM +0200, Marc Glisse wrote:
> > On Tue, 7 Jul 2015, Prathamesh Kulkarni wrote:
> >
> > >+/* a * (1 << b) -> a << b */
> > >+(simplify
> > >+ (mult:c @a (lshift integer_onep @b))
> > >+ (if (!FLOAT_TYPE_P (type))
> > >+
Hi all,
I need your help on how to interpret the standard(s) or how to implement
handling an allocatable function's result, when that result is not allocated by
the function. Imagine the simple (albeit artificial) case:
integer function read_input()
! Do whatever is needed to read an int.
rea
On Thu, Jul 09, 2015 at 12:02:19PM +0200, Richard Biener wrote:
> I've added support for non-digit names to allow more descriptive
> patterns. Like when we have
>
> /* Fold (X << C1) & C2 into (X << C1) & (C2 | ((1 << C1) - 1))
> (X >> C1) & C2 into (X >> C1) & (C2 | ~((type) -1 >> C1))
>
On 07/07/15 16:00, Michael Matz wrote:
Hi,
On Mon, 6 Jul 2015, Richard Biener wrote:
By doing so, we make the behaviour of gt_cleare_cache independent of the
order in which the entries are visited, turning:
- hard-to-trigger bugs which trigger for one visiting order but not for
another, int
Indirect sibcall with register arguments is OK when there is register
available for argument passing.
OK for trunk if there is no regression?
H.J.
---
gcc/
PR target/66819
* config/i386/i386.c (ix86_function_ok_for_sibcall): Allow
indirect sibcall with register arguments
On Wed, 8 Jul 2015, Richard Biener wrote:
>
> This introduces a :s flag to match expressions which enforces
> the expression to have a single-use if(!) the simplified
> expression is larger than one statement.
>
> Thus with that we for example allow
>
> tem = a + 1;
> x = tem - 3;
> foo (
On 07/07/15 17:58, Tom de Vries wrote:
If you can
handle one exit edge I also can't see the difficulty in handling
all exit edges.
Agreed, that doesn't look to complicated. I could call
rewrite_virtuals_into_loop_closed_ssa for all loops in
rewrite_virtuals_into_loop_closed_ssa, to get non-sin
On Thu, Jul 9, 2015 at 12:54 PM, H.J. Lu wrote:
> Indirect sibcall with register arguments is OK when there is register
> available for argument passing.
>
> OK for trunk if there is no regression?
>
>
> H.J.
> ---
> gcc/
>
> PR target/66819
> * config/i386/i386.c (ix86_function_ok
On Thu, 9 Jul 2015, Tom de Vries wrote:
> On 07/07/15 17:58, Tom de Vries wrote:
> > > If you can
> > > handle one exit edge I also can't see the difficulty in handling
> > > all exit edges.
> > >
> >
> > Agreed, that doesn't look to complicated. I could call
> > rewrite_virtuals_into_loop_close
Following series of patches continues where I stopped at the end of
previous stage 1. It ports places in IPA passes to function_summary container
and I introduce a new edge_summary, which is essentially very similar to
the aforementioned container.
Patches were tested together on x86_64-linux-gnu
gcc/ChangeLog:
2015-07-03 Martin Liska
* hash-set.h (remove): New function.
(iterator): New iteration class for hash_set.
---
gcc/hash-set.h | 39 +++
1 file changed, 39 insertions(+)
diff --git a/gcc/hash-set.h b/gcc/hash-set.h
index 2fb6c
1 - 100 of 160 matches
Mail list logo
