https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Francois-Xavier Coudert fxcoudert at gcc dot gnu.org changed:
What|Removed |Added
Status|NEW
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #36 from Dominique d'Humieres dominiq at lps dot ens.fr ---
Please file test suite failures as new PR (one new PR per test failure
with a different backtrace, ideally). It's very hard to track as such
(the subject is not accurate
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #27 from Martin Liška marxin at gcc dot gnu.org ---
Author: marxin
Date: Tue Nov 11 14:58:21 2014
New Revision: 217355
URL: https://gcc.gnu.org/viewcvs?rev=217355root=gccview=rev
Log:
PR ipa/63622
PR ipa/63795
*
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #28 from howarth at bromo dot med.uc.edu ---
At r217355, the failures for the gfortran.dg/assumed_rank_[8/9/10].f90
execution tests , first noted in Comment 7, still remain on x86_64 darwin.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #29 from howarth at bromo dot med.uc.edu ---
(In reply to howarth from comment #28)
At r217355, the failures for the gfortran.dg/assumed_rank_[8/9/10].f90
execution tests , first noted in Comment 7, still remain on x86_64 darwin.
A
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #30 from howarth at bromo dot med.uc.edu ---
Created attachment 33936
-- https://gcc.gnu.org/bugzilla/attachment.cgi?id=33936action=edit
assembly for gfortran.dg/assumed_rank_10.f90 -O3 -fomit-frame-pointer
-funroll-loops execution
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #31 from howarth at bromo dot med.uc.edu ---
(In reply to howarth from comment #28)
As noted before, passing -fno-ipa-icf to these failing fortran test cases
suppresses the execution failure.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #32 from howarth at bromo dot med.uc.edu ---
Created attachment 33939
-- https://gcc.gnu.org/bugzilla/attachment.cgi?id=33939action=edit
assumed_rank_10.f90.003t.original output from -fdump-ipa-icf-details
Generated using
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #33 from howarth at bromo dot med.uc.edu ---
Created attachment 33940
-- https://gcc.gnu.org/bugzilla/attachment.cgi?id=33940action=edit
assumed_rank_10.f90.051i.icf output from -fdump-ipa-icf-details
Generated using
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #34 from Dominique d'Humieres dominiq at lps dot ens.fr ---
Reduced test case for assumed_rank_10.f90
program test
implicit none
type t
integer :: aa
end type t
integer, pointer :: iip
type(t), pointer :: jjp
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Francois-Xavier Coudert fxcoudert at gcc dot gnu.org changed:
What|Removed |Added
CC|
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #26 from Iain Sandoe iains at gcc dot gnu.org ---
OK - so two more data points.
( i)
With the alias patch, I can bootstrap x86_64-darwin12
However, there are lots of new fails in Ada (for example) that go away when I
force ipa-icf
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Igor Zamyatin izamyatin at gmail dot com changed:
What|Removed |Added
CC||izamyatin at
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #21 from Martin Liška marxin at gcc dot gnu.org ---
(In reply to Igor Zamyatin from comment #20)
This is mentioned here -
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63534#c9
Fix for this is under review, start of the discussion
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #22 from howarth at bromo dot med.uc.edu ---
(In reply to Martin Liška from comment #21)
(In reply to Igor Zamyatin from comment #20)
This is mentioned here -
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63534#c9
Fix for
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #23 from howarth at bromo dot med.uc.edu ---
(In reply to Martin Liška from comment #21)
(In reply to Igor Zamyatin from comment #20)
This is mentioned here -
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63534#c9
Fix for
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #24 from howarth at bromo dot med.uc.edu ---
The error seen in Comment 19 appears to be identical to that reported in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63534#c9.
If I also add in the proposed patch
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #25 from Dominique d'Humieres dominiq at lps dot ens.fr ---
The error seen in Comment 19 appears to be identical to that reported in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63534#c9.
This is fixed by the patch at
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #8 from howarth at bromo dot med.uc.edu ---
At r217144 with https://gcc.gnu.org/bugzilla/attachment.cgi?id=33897 to fix
PR63750, gcc trunk now bootstraps up to the illegal text reloc error. Is this
problem understood well enough to
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #9 from howarth at bromo dot med.uc.edu ---
The darwin linker developer's analysis of this failing linkage is as follows...
That link line contains “-undefined dynamic_lookup” which is almost always a
bad idea. Removing it shows the
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #10 from howarth at bromo dot med.uc.edu ---
Note that removal of the -undefined dynamic_lookup produces...
Undefined symbols for architecture x86_64:
__ZNKSt11logic_error4whatEv, referenced from:
__ZTVSt11logic_error in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #11 from howarth at bromo dot med.uc.edu ---
It looks like we aren't compiling some object files in libstdc++-v3/src/c++98
for gcc trunk which are are on gcc 4.9.2. In particular, we are missing...
ctype.o
ctype_configure_char.o
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #12 from howarth at bromo dot med.uc.edu ---
This looks like fallout from...
https://gcc.gnu.org/ml/libstdc++-cvs/2014-q4/msg00080.html
https://gcc.gnu.org/ml/libstdc++-cvs/2014-q3/msg00181.html
where the files moved to c++11 in gcc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #13 from howarth at bromo dot med.uc.edu ---
Created attachment 33904
-- https://gcc.gnu.org/bugzilla/attachment.cgi?id=33904action=edit
preprocessed source for libstdc++-v3/src/c++11/streambuf-inst.cc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #14 from howarth at bromo dot med.uc.edu ---
The attached preprocessed source was produced from gcc trunk with...
# /sw/src/fink.build/gcc50-5.0.0-1000/darwin_objdir/./gcc/xgcc -shared-libgcc
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #15 from Jonathan Wakely redi at gcc dot gnu.org ---
(In reply to howarth from comment #14)
which produces as set of warnings that don't appear in the build of
libstdc++-v3/src/c++98/streambuf-inst.cc. These are...
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #16 from howarth at bromo dot med.uc.edu ---
(In reply to Jonathan Wakely from comment #15)
(In reply to howarth from comment #14)
which produces as set of warnings that don't appear in the build of
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #17 from howarth at bromo dot med.uc.edu ---
These warnings are emitted from this code in gcc/config/darwin.h
/* Mach-O supports 'weak imports', and 'weak definitions' in coalesced
sections. machopic_select_section ensures that
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #18 from howarth at bromo dot med.uc.edu ---
Interestingly, these errors don't occur if the compilation of
libstdc++-v3/src/c++11/streambuf-inst.cc is done at -O0, -O1 or if -fno-ipa-icf
is appended at -O2 or higher.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #19 from howarth at bromo dot med.uc.edu ---
Apparently we need something along the lines of the previously proposed IPA
ICF aliasing restriction patch, however when applied to current gcc trunk the
bootstrap fails at...
libtool:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #6 from Martin Liška marxin at gcc dot gnu.org ---
(In reply to Dominique d'Humieres from comment #4)
Created attachment 33843 [details]
IPA ICF aliasing restriction
Can you please try to apply following patch?
It should
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #7 from Dominique d'Humieres dominiq at lps dot ens.fr ---
That looks like duplicate of PR63580. I hope the PR will be fixed today.
It looks like: I have bootstrapped r216914 with the patches in comment 3 and in
PR3580 comment 4.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #3 from Martin Liška marxin at gcc dot gnu.org ---
Created attachment 33843
-- https://gcc.gnu.org/bugzilla/attachment.cgi?id=33843action=edit
IPA ICF aliasing restriction
Can you please try to apply following patch?
It should
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
--- Comment #4 from Dominique d'Humieres dominiq at lps dot ens.fr ---
Created attachment 33843 [details]
IPA ICF aliasing restriction
Can you please try to apply following patch?
It should disable alias creation for targets that do not
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
howarth at bromo dot med.uc.edu changed:
What|Removed |Added
CC||howarth at bromo dot
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Richard Biener rguenth at gcc dot gnu.org changed:
What|Removed |Added
Target Milestone|--- |5.0
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Dominique d'Humieres dominiq at lps dot ens.fr changed:
What|Removed |Added
Status|UNCONFIRMED |NEW
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63622
Dominique d'Humieres dominiq at lps dot ens.fr changed:
What|Removed |Added
CC|
38 matches
Mail list logo
