--- Comment #24 from howarth at nitro dot med dot uc dot edu 2006-11-02
03:19 ---
The previous comment should indicate the crash in nan_inf_fmt disappears when n
is a double.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29302
--- Comment #23 from howarth at nitro dot med dot uc dot edu 2006-11-02
03:16 ---
One more observation. If I change...
Index: write.c
===
--- write.c (revision 118343)
+++ write.c (working copy)
@@ -893,7 +893,7 @@
--- Comment #22 from howarth at nitro dot med dot uc dot edu 2006-11-02
02:51 ---
A couple other observations. The latest Xcode 2.4.1 release doesn't fix this
problem. Also, shorter version of the testcase...
implicit none
character*40 l
character*12 fmt
rea
--- Comment #21 from howarth at nitro dot med dot uc dot edu 2006-11-01
14:00 ---
This bug is rather confusing. It appears that the actual macros for isfinite on
Darwin PPC look like...
static __inline__ int __inline_isinff( float __x ) { return
__builtin_fabsf(__x) == __builtin_inff
--- Comment #20 from howarth at nitro dot med dot uc dot edu 2006-11-01
06:06 ---
Created an attachment (id=12525)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12525&action=view)
assembly file from libgfortran/io/write.c on Darwin PPC
--
http://gcc.gnu.org/bugzilla/show_bug.
--- Comment #19 from howarth at nitro dot med dot uc dot edu 2006-11-01
05:10 ---
Created an attachment (id=12524)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12524&action=view)
preprocessed file for libgfortran/io/write.c on Darwin PPC
--
http://gcc.gnu.org/bugzilla/show_b
--- Comment #18 from howarth at nitro dot med dot uc dot edu 2006-11-01
05:05 ---
Andrew,
I'll post the preprocessed source for write.c shortly. I did a quick test
though moving...
#define isfinite(x) (fpclassify(x) != FP_NAN && fpclassify(x) != FP_INFINITE)
outside of the prepro
--- Comment #17 from jvdelisle at gcc dot gnu dot org 2006-11-01 04:56
---
Jack is right about one thing and I should have noticed this. For that test
case we should never get into output_float where we are segfaulting. The
handling of nan and inf are handled in the calling function w
--- Comment #16 from pinskia at gcc dot gnu dot org 2006-11-01 04:48
---
(In reply to comment #15)
> I think it we need to cast both x variables in the macro to unsigned long.
No, that would be incorrect.
Jack, can you attach the preprocessed source for write.c?
--
pinskia at gcc d
--- Comment #15 from howarth at nitro dot med dot uc dot edu 2006-11-01
04:33 ---
I have pinned down the problem. On Darwin PPC with Xcode 2.4, at -O1 or
higher we have problems with the isfinite() macro from libgfortran.h. For the
nan_inf_fmt testcase it should always return 0, but
--- Comment #14 from jvdelisle at gcc dot gnu dot org 2006-10-31 23:24
---
Jack,
I don't think this is the final fix. However, after you test that other
possibility I sent you I think I can finalize a patch. If zero terminating
buffer does not solve the problem, then I think it confi
--- Comment #13 from howarth at nitro dot med dot uc dot edu 2006-10-31
14:14 ---
As I mentioned to Jerry in a private email, his suggestion of...
Index: io/write.c
===
--- io/write.c (revision 118229)
+++ io/write.c (wo
--- Comment #12 from jvdelisle at gcc dot gnu dot org 2006-10-31 03:18
---
Geoff,
I am trying to understand what you are saying.
I could extract a segment of the C code that is being used implement here so
that you can see why it is failing.
I think it is because Apple code is return
--- Comment #11 from geoffk at gcc dot gnu dot org 2006-10-31 01:03 ---
Please do not add '[EMAIL PROTECTED]' to the CC list of any bugzilla bug.
--
geoffk at gcc dot gnu dot org changed:
What|Removed |Added
--- Comment #10 from geoffk at gcc dot gnu dot org 2006-10-31 01:02 ---
Hi Jack,
I have no idea what's causing this, and I don't know why you think I do. I am
the one asking you to provide a testcase in the Radar. You're claiming that
this is a bug in Xcode 2.4 (which is supported by
--- Comment #9 from howarth at nitro dot med dot uc dot edu 2006-10-06
01:32 ---
Geoff,
This bug is also radar://4759173. I've been asked to...
Please debug and produce a small self-contained (that is, not including
gfortran) testcase showing what goes wrong.
Could you expand in
--- Comment #8 from howarth at nitro dot med dot uc dot edu 2006-10-05
02:03 ---
Geoff,
Can you expand on this? Assuming it is a atoi() tail-calls strtol_l
why is Darwin the only arch effected and what recourse do we have
in fixing this? I tried a brute force replacement of the ato
--- Comment #7 from geoffk at gcc dot gnu dot org 2006-10-04 23:57 ---
Probably atoi() tail-calls strtol_l.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29302
--- Comment #6 from kargl at gcc dot gnu dot org 2006-10-01 00:40 ---
(In reply to comment #5)
> Jerry,
> I've run this with guardmalloc...
>
>
Jack, This is an Apple library problem. Please report the
inappropriate use of strtol_l to Apple.
--
http://gcc.gnu.org/bugzilla/sh
--- Comment #5 from howarth at nitro dot med dot uc dot edu 2006-09-30
22:50 ---
Jerry,
I've run this with guardmalloc...
http://developer.apple.com/documentation/Darwin/Reference/ManPages/man3/libgmalloc.3.html
using all the permutations and nothing gets triggered.
--- Comment #4 from jvdelisle at gcc dot gnu dot org 2006-09-30 21:53
---
Jack, I would forward this one to Apple.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29302
--- Comment #3 from jvdelisle at gcc dot gnu dot org 2006-09-30 21:36
---
Valgrind shows tight as a drum on i686-linux. Do you have something equivalent
to valgrind for Darwin PPC?
I am not seeing anything obvious in the code. On i686 the call at line 496 is
to atoi(). Your system is
--- Comment #2 from jvdelisle at gcc dot gnu dot org 2006-09-30 21:09
---
I will see if I can spot something here. Jack, I will need your help testing.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=29302
--- Comment #1 from howarth at nitro dot med dot uc dot edu 2006-09-30
20:46 ---
Created an attachment (id=12362)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=12362&action=view)
assembly file for nan_inf_fmt that seqfaults on Darwin PPC with Xcode 2.4
--
http://gcc.gnu.org/b
24 matches
Mail list logo
