I've committed this patch to rectify the misinterpretation of the specification.
The pragma comes in two forms:
#pragma acc routine clauses
fn-decl-or-defn
and
#pragma acc routine (name) clauses
The latter form names a function already in scope, which makes it fairly
straight
On Aug 15, 2015, at 9:19 AM, Richard Sandiford rdsandif...@googlemail.com
wrote:
Mike Stump mikest...@comcast.net writes:
On Aug 15, 2015, at 3:32 AM, Richard Sandiford
rdsandif...@googlemail.com wrote:
The port is only buggy if they have define_peephole2s with match_scratches
and those
When a vtable uses a deprecated virtual function, that's outside the
user's control and so shouldn't be warned about.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 3e78f42251b5a82bc3aa2c00b85e40abb54fc70f
Author: Jason Merrill ja...@redhat.com
Date: Fri Aug 14 17:41:24 2015 +0100
FAIL: gcc.target/aarch64/target_attr_crypto_ice_1.c (internal compiler error)
In file included from
/opt/gcc/gcc-20150815/gcc/testsuite/gcc.target/aarch64/target_attr_crypto_ice_1.c:4:0:
/opt/gcc/gcc-20150815/Build/gcc/include/arm_neon.h: In function
'test_vsha1cq_u32':
/opt/gcc/gcc-20150815
On Sat, Aug 15, 2015 at 2:47 AM, Kai Zhao loversz...@gmail.com wrote:
Hi,
There are some useless variables, class Gt, and struct CompLast in
testsuite/25_algorithms/*
This patch is to remove those useless variables, class Gt and struct
CompLast.
The patch is attached.
commit
Robert Suchanek robert.sucha...@imgtec.com writes:
You also need to do the same thing for TARGET_HARD_REGNO_SCRATCH_OK,
to stop peephole2 from using unsaved registers as scratch registers.
I should dig out my patches to clean up this interface. It's just
too brittle to have two macros
On August 14, 2015 8:20:18 PM GMT+02:00, Jeff Law l...@redhat.com wrote:
On 08/13/2015 06:55 PM, Mikhail Maltsev wrote:
Hi all.
These two patches are refactoring of dominator-related code.
The comment in dominance.c says: We work in a poor-mans object
oriented
fashion, and carry an instance
This patch resets the lambda scope based sequence numbering used to assign
numbers to lambdas during parsing when we pop a module scope. This resets
the numbering for subsequent modules imported during LIPO compilation.
It also renames cp_clear_deferred_fns to reset_parsing_state, as that
routine
On 08/14/2015 11:02 AM, Richard Biener wrote:
So the last time I did similar refactoring I wondered if we can somehow avoid
the noise in non-IPA passes. Ideas I came up with are
a) Inherit gimple/rtl pass classes from a class which is initialized with
the
function the pass
All:
Please find the updated patch with suggestion and feedback incorporated.
Thanks Jeff and Richard for the review comments.
Following changes were done based on the feedback on RFC comments. and the
review for the previous patch.
1. Both tracer and path splitting pass are separate passes
The symbol _ZNSt8ios_base7failureB5cxx11C1EPKcRKSt10error_code, which
appears in libstdc++, was being demangled as
std::ios_base::failure[abi:cxx11]::cxx11(char const*, std::error_code const)
That is clearly incorrect: std::ios_base::failure does not have a
method cxx11, and anyhow if you look
Hi Ian,
On 08/15/2015 03:23 PM, Ian Lance Taylor wrote:
The symbol _ZNSt8ios_base7failureB5cxx11C1EPKcRKSt10error_code, which
appears in libstdc++, was being demangled as
std::ios_base::failure[abi:cxx11]::cxx11(char const*, std::error_code const)
That is clearly incorrect:
On Aug 15, 2015, at 3:32 AM, Richard Sandiford rdsandif...@googlemail.com
wrote:
The port is only buggy if they have define_peephole2s with match_scratches
and those match_scratches require a register that would be saved by
an interrupt function. In other cases defining T_H_R_S_O would have
On Aug 14, 2015, at 3:24 PM, Paolo Bonzini bonz...@gnu.org wrote:
There are plenty of targets that do not require -fPIC because they always
generate position independent code, but none of them feels the need to
complain with the user about an unnecessary but perfectly valid option,
on each and
ok.
David
On Fri, Aug 14, 2015 at 11:13 PM, Teresa Johnson tejohn...@google.com wrote:
This patch resets the lambda scope based sequence numbering used to assign
numbers to lambdas during parsing when we pop a module scope. This resets
the numbering for subsequent modules imported during LIPO
I've committed this to the gomp4 branch. It extends the 'oacc function'
attribute's dimension handling to deal with routines. With the latter we now
use TREE_PURPOSE of the dimension list to indicate whether the routine may or
may not spawn a partitioned loop on a particular axis. I had to
... I think your patch fixes c++/63887. The libiberty testsuite passes
with the below applied.
Paolo.
//
Index: testsuite/demangle-expected
===
--- testsuite/demangle-expected (revision 226910)
+++
Mike Stump mikest...@comcast.net writes:
On Aug 15, 2015, at 3:32 AM, Richard Sandiford
rdsandif...@googlemail.com wrote:
The port is only buggy if they have define_peephole2s with match_scratches
and those match_scratches require a register that would be saved by
an interrupt function. In
18 matches
Mail list logo