https://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
Bug 51500 depends on bug 52344, which changed state.
Bug 52344 Summary: _Jv_platform_nanotime of win32.cc is wrong
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52344
What|Removed |Added
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
Kai Tietz ktietz at gcc dot gnu.org changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
gee jojelino at gmail dot com changed:
What|Removed |Added
Attachment #26574|0 |1
is obsolete|
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #20 from gee jojelino at gmail dot com 2012-02-04 11:32:07 UTC ---
Created attachment 26573
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=26573
proposed patch for workaround Throw_2 failure
it was call %ebx for sig_pat. and
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #21 from gee jojelino at gmail dot com 2012-02-05 01:22:36 UTC ---
Created attachment 26574
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=26574
proposed patch for workaround TestEarlyGC failure
it deals raw closure stub with
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #22 from gee jojelino at gmail dot com 2012-02-05 02:11:12 UTC ---
Breakpoint 9, _Unwind_RaiseException (exc=0xa92820)
at ../.././libgcc/unwind.inc:88
88uw_init_context (this_context);
(gdb) bt
#0 _Unwind_RaiseException
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #23 from gee jojelino at gmail dot com 2012-02-05 02:43:41 UTC ---
(In reply to comment #22)
Breakpoint 9, _Unwind_RaiseException (exc=0xa92820)
at ../.././libgcc/unwind.inc:88
88uw_init_context (this_context);
(gdb) bt
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #19 from Kai Tietz ktietz at gcc dot gnu.org 2012-02-01 10:46:40
UTC ---
Author: ktietz
Date: Wed Feb 1 10:46:36 2012
New Revision: 183794
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=183794
Log:
PR target/51500
*
Ok, here is a variant treating stack-alignment for closure.
ChangeLog
2012-01-31 Kai Tietz kti...@redhat.com
* src/prep_cif.c (ffi_prep_cif): Allow for X86_WIN32
also FFI_THISCALL.
* src/x86/ffi.c (ffi_closure_THISCALL): Add prototype.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #17 from ktietz70 at googlemail dot com 2012-01-31 09:47:08 UTC ---
Ok, here is a variant treating stack-alignment for closure.
ChangeLog
2012-01-31 Kai Tietz kti...@redhat.com
* src/prep_cif.c (ffi_prep_cif): Allow for
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
gee jojelino at gmail dot com changed:
What|Removed |Added
Depends on||51901
--- Comment #18
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #11 from gee jojelino at gmail dot com 2012-01-30 15:10:48 UTC ---
Comment on attachment 26513
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=26513
proposed patch for handling thiscall r3
@@ -505,6 +508,23 @@
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #12 from Uros Bizjak ubizjak at gmail dot com 2012-01-30 16:39:04
UTC ---
(In reply to comment #11)
+ *(unsigned int*) __tramp[24] = 0x8304244c; /* xchgl 4(%esp),%ecx*/ \
Don't use this insn, it implies lock prefix and costs
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #13 from Kai Tietz ktietz at gcc dot gnu.org 2012-01-30 19:20:16
UTC ---
Right the following sequence seems to be better IMHO:
pop %eax
push %ecx
push %eax
mov __ctx, %eax
call fun
pop %ecx
mov %ecx, %(esp)
ret
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #14 from Uros Bizjak ubizjak at gmail dot com 2012-01-30 19:58:43
UTC ---
(In reply to comment #13)
Right the following sequence seems to be better IMHO:
pop %eax
push %ecx
push %eax
mov __ctx, %eax
call fun
pop %ecx
mov
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #15 from Kai Tietz ktietz at gcc dot gnu.org 2012-01-30 22:21:16
UTC ---
So this should be better then:
#define FFI_INIT_TRAMPOLINE_THISCALL(TRAMP,FUN,CTX,SIZE) \
{ unsigned char *__tramp = (unsigned char*)(TRAMP); \
unsigned int
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #16 from Uros Bizjak ubizjak at gmail dot com 2012-01-30 23:04:54
UTC ---
(In reply to comment #15)
#define FFI_INIT_TRAMPOLINE_THISCALL(TRAMP,FUN,CTX,SIZE) \
{ unsigned char *__tramp = (unsigned char*)(TRAMP); \
unsigned int
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
gee jojelino at gmail dot com changed:
What|Removed |Added
Attachment #26338|0 |1
is obsolete|
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
gee jojelino at gmail dot com changed:
What|Removed |Added
Attachment #26499|0 |1
is obsolete|
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #10 from Kai Tietz ktietz at gcc dot gnu.org 2012-01-30 07:29:28
UTC ---
Hmm, not sure if you already noticed it, but ffi_call supports by recent patch
thiscall/fastcall convention.
So parts of you patch won't apply mainline.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #7 from gee jojelino at gmail dot com 2012-01-18 13:35:15 UTC ---
after applying pr50057 patch
http://gcc.gnu.org/bugzilla/attachment.cgi?id=26365
=== libjava Summary ===
# of expected passes2324
# of
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #6 from gee jojelino at gmail dot com 2012-01-16 13:17:09 UTC ---
Created attachment 26338
-- http://gcc.gnu.org/bugzilla/attachment.cgi?id=26338
proposed patch for handling thiscall
adopted from
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #5 from gee jojelino at gmail dot com 2011-12-18 22:02:07 UTC ---
(In reply to comment #4)
Hmm, this might be another instance for already solved PR/50053. Could you
recheck, if error is still present for current 4.7 trunk version.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
Kai Tietz ktietz at gcc dot gnu.org changed:
What|Removed |Added
CC||ktietz at gcc dot
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
Richard Guenther rguenth at gcc dot gnu.org changed:
What|Removed |Added
Priority|P3 |P4
Target
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #2 from gee jojelino at gmail dot com 2011-12-12 14:46:26 UTC ---
please fix java::lang::Class::initializeClass for mingw32 ia-32 by adding
__thiscall before being involved this bug. surely PR50053 is fixed but i leaved
a kludge toward
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #3 from gee jojelino at gmail dot com 2011-12-12 14:48:52 UTC ---
it seems that libffi handles calling convention of method which is cdecl or
stdcall.
but not for thiscall for mingw32 ia-32.
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51500
--- Comment #1 from gee jojelino at gmail dot com 2011-12-10 18:59:19 UTC ---
stacktrace
Starting program:
/tmp/gcc/i686-pc-mingw32/libjava/testsuite/.libs/TestClosureGC
[New Thread 3600.0xf5c]
[New Thread 3600.0x1338]
Program received signal
28 matches
Mail list logo
