Re: [PATCH GCC]Improve how we handle overflow for type conversion in scev/ivopts, part I

2015-05-23 Thread Bin.Cheng
On Fri, May 22, 2015 at 7:45 PM, Richard Biener wrote: > On Wed, May 20, 2015 at 11:41 AM, Bin Cheng wrote: >> Hi, >> As we know, GCC is too conservative when checking overflow behavior in SCEV >> and loop related optimizers. Result is some variable can't be recognized as >> scalar evolution and

Remove method_class_type

2015-05-23 Thread Jan Hubicka
Hi, when I introduced method_class_type I was not aware that TYPE_METHOD_BASETYPE gives previsely the same information. Bootstrapped/regtested x86_64-linux, will commit it shortly. * ipa-utils.h (method_class_type): Remove. * cgraphunit.c (walk_polymorphic_call_targets): Use TYP

Constify few functions in tree.c

2015-05-23 Thread Jan Hubicka
Hi, this patch turns some paramters to const_tree. Bootstrapped/regtested ppc64-linux, comitted as obvious. Honza * tree.c (prototype_p, virtual_method_call_p, obj_type_ref_class, is_typedef_decl, typedef_variant_p): Constify. * tree.h (prototype_p, virtual_method_call_p

[PATCH 6/7] add default for HAVE_store_multiple

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_store_multiple): New function. (HAVE_store_multiple): Add default value. * expr.c (move_block_from_reg): Adjust. --- gcc/ChangeLog | 6 ++ gcc/defaults.h | 10 ++ gcc/expr

[PATCH 4/7] provide default for HAVE_mem_signal_fence

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_mem_signal_fence): New function. (HAVE_mem_signal_fence): Add default value. * optabs.c: Adjust. --- gcc/ChangeLog | 6 ++ gcc/defaults.h | 10 ++ gcc/optabs.c | 5

[PATCH 5/7] add default for HAVE_load_multiple

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_load_multiple): New function. (HAVE_load_multiple): Add default value. * expr.c (move_block_to_reg): Adjust. --- gcc/ChangeLog | 6 ++ gcc/defaults.h | 10 ++ gcc/expr.c

[PATCH 7/7] add default for HAVE_tablejump

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_tablejump): New function. (HAVE_tablejump): Add default value. * expr.c: Adjust. * stmt.c: Likewise. --- gcc/ChangeLog | 7 +++ gcc/defaults.h | 10 ++ gcc/expr.c

[PATCH 2/7] provide default for HAVE_mem_thread_fence

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_mem_thread_fence): New function. (HAVE_mem_thread_fence): Add default definition. * optabs.c: Adjust. --- gcc/ChangeLog | 6 ++ gcc/defaults.h | 10 ++ gcc/optabs.c | 4

[PATCH 0/7] More ifdef reduction

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders Hi, yet more of the same. each individually bootstrapped + regtested on x86_64-linux-gnu, and made sure config-list.mk was fine at the end. I expect this stuff is all still preapproved so committing to trunk. Trev Trevor Saunders (7): always define HAVE_lo_sum provi

[PATCH 3/7] always define HAVE_memory_barrier

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * defaults.h (gen_memory_barrier): New function. (HAVE_memory_barrier): Add default value. * optabs.c: Adjust. --- gcc/ChangeLog | 6 ++ gcc/defaults.h | 10 ++ gcc/optabs.c | 5 - 3

[PATCH 1/7] always define HAVE_lo_sum

2015-05-23 Thread tbsaunde+gcc
From: Trevor Saunders gcc/ChangeLog: 2015-05-23 Trevor Saunders * combine.c (find_split_point): Check the value of HAVE_lo_sum instead of if it is defined. (combine_simplify_rtx): Likewise. * lra-constraints.c (process_address_1): Likewise. * config/da

[Patch, fortran] PR66257 [5/6 regression] elemental typebound calls rejected as actual argument

2015-05-23 Thread Mikael Morin
Hello, For PR 63727, a check was introduced, rejecting procedure pointer components used as actual arguments: foo(obj%proc_comp) but it had the side effect of also rejecting foo(obj%proc_comp(arg)) Fixed by the attached patch. Tested on x86_64-linux. OK for 6/5 ? Mikael 2015-0

Re: [match-and-simplify] reject expanding operator-list to implicit 'for'

2015-05-23 Thread Prathamesh Kulkarni
On 20 May 2015 at 17:39, Prathamesh Kulkarni wrote: > On 20 May 2015 at 17:01, Richard Biener wrote: >> On Wed, 20 May 2015, Prathamesh Kulkarni wrote: >> >>> On 20 May 2015 at 16:17, Prathamesh Kulkarni >>> wrote: >>> > Hi, >>> > This patch rejects expanding operator-list to implicit 'for'. >>>

Re: [C++ PATCH] fix canonical type ICE

2015-05-23 Thread Jason Merrill
OK, thanks. Jason

Re: [Patch, fortran] PR66079 - [6 Regression] memory leak with source allocation in internal subprogram

2015-05-23 Thread Paul Richard Thomas
Dear Andre, To answer your fist question - no, it doesn't. I was working with my laptop, which is over slow when it comes to updating. I should have realised that since you are working in this area that there might be a problem. I discovered it when I did an update on my workstation this afternoon

Re: [Patch, fortran] PR66079 - [6 Regression] memory leak with source allocation in internal subprogram

2015-05-23 Thread Andre Vehreschild
Hi Paul, does this patch apply to current trunk cleanly? I get an issue with the last hunk, because all of the prerequisites are gone since r223445. The string copy is completely handled by the trans_assignment at the bottom of the if (code->expr3) block. Therefore I doubt the patches last hunk is

Re: C/C++ PATCH to allow deprecating enum values (PR c/47043)

2015-05-23 Thread Ed Smith-Rowland
On 05/22/2015 06:19 PM, Mikhail Maltsev wrote: On 22.05.2015 12:10, Marek Polacek wrote: Thanks, applied. Here's the final version. By the way, we have a feature test macro, __cpp_attributes=200809 which can be used to determine, whether C++11 attribute syntax is supported by the compiler. I

Re: [PATCH] Don't combine param and return value copies

2015-05-23 Thread Segher Boessenkool
Hi Alan, On Sat, May 23, 2015 at 06:03:05PM +0930, Alan Modra wrote: > This stops combine messing with parameter and return value copies > from/to hard registers. Bootstrapped and regression tested > powerpc64le-linux, powerpc64-linux and x86_64-linux. In looking at a > number of different power

[Patch, fortran] PR66079 - [6 Regression] memory leak with source allocation in internal subprogram

2015-05-23 Thread Paul Richard Thomas
Dear All, This patch started out fixing a single source of memory leak and then went on to fix various other issues that I found upon investigation. The fortran ChangeLog entry is sufficiently descripive that I do not think that there is a need to say more. Bootstrapped and regtested on x86_64/F

Re: PR fortran/44054 Convert all gfc_error_1 calls to gfc_error

2015-05-23 Thread Mikael Morin
Le 23/05/2015 01:04, Manuel López-Ibáñez a écrit : > PING: https://gcc.gnu.org/ml/gcc-patches/2015-05/msg01511.html > > This only needs approval from Fortran maintainers. > I approved: https://gcc.gnu.org/ml/fortran/2015-05/msg00135.html But it seems I replied to the fortran list only. Mikael

Re: [patch] libstdc++/64657 support iterators with overloaded comma operator

2015-05-23 Thread François Dumont
On 23/05/2015 13:19, François Dumont wrote: On 29/04/2015 17:22, Jonathan Wakely wrote: I think this covers all the places in the library where we do: ++i, ++j Tested powerpc64le-linux, committed to trunk. I just committed 2 missing places in debug mode. 2015-05-23 François Dumont fdum...

Re: [patch] libstdc++/64657 support iterators with overloaded comma operator

2015-05-23 Thread François Dumont
On 29/04/2015 17:22, Jonathan Wakely wrote: I think this covers all the places in the library where we do: ++i, ++j Tested powerpc64le-linux, committed to trunk. I just committed 2 missing places in debug mode. 2015-05-23 François Dumont fdum...@gcc.gnu.org> PR libstdc++/64657 * i

add names for constants in line-map.c

2015-05-23 Thread Manuel López-Ibáñez
This patch adds descriptive names to various constants in line-map.c. There were some differences between the constants used a various places, but my understanding is that these differences are arbitrary and it is easier to understand the code if they are harmonized. Bootstrapped and regression t

Re: [PATCH] Don't combine param and return value copies

2015-05-23 Thread Andrew Pinski
On Sat, May 23, 2015 at 1:33 AM, Alan Modra wrote: > This stops combine messing with parameter and return value copies > from/to hard registers. Bootstrapped and regression tested > powerpc64le-linux, powerpc64-linux and x86_64-linux. In looking at a > number of different powerpc64le gcc/*.o fil

[PATCH] Don't combine param and return value copies

2015-05-23 Thread Alan Modra
This stops combine messing with parameter and return value copies from/to hard registers. Bootstrapped and regression tested powerpc64le-linux, powerpc64-linux and x86_64-linux. In looking at a number of different powerpc64le gcc/*.o files, I noticed a few code generation improvements. There wer

Re: match.pd: (x | y) & ~x -> y & ~x

2015-05-23 Thread Richard Biener
On May 22, 2015 10:54:02 PM GMT+02:00, Marc Glisse wrote: >On Mon, 18 May 2015, Richard Biener wrote: > >> On Fri, May 15, 2015 at 7:22 PM, Marc Glisse >wrote: >> >>> we already have the more complicated: x & ~(x & y) -> x & ~y (which >I am >>> reindenting by the way) and the simpler: (~x | y) &