--
pcarlini at suse dot de changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |pcarlini at suse dot de
|dot org |
--- Comment #31 from bkoz at gcc dot gnu dot org 2007-04-06 16:38 ---
For generic targets, can't you just
-errno = 0;
+static __thread int errno = 0;
on releases of gcc (4.3+) that have the wherewithal to fudge the issue for all
systems, even the ones without builtin-TLS?
-benjamin
--- Comment #32 from pcarlini at suse dot de 2007-04-06 16:46 ---
(In reply to comment #31)
For generic targets, can't you just
-errno = 0;
+static __thread int errno = 0;
on releases of gcc (4.3+) that have the wherewithal to fudge the issue for all
systems, even the ones
--- Comment #33 from bkoz at gcc dot gnu dot org 2007-04-06 17:15 ---
Nice idea, but I'm not sure it would work (i'm not very familiar with __thread,
etc.): are we sure that with that kind of declaration the strtof co functions
themselves automatically use that new improved errno
--- Comment #34 from paolo at gcc dot gnu dot org 2007-04-07 00:37 ---
Subject: Bug 31117
Author: paolo
Date: Sat Apr 7 00:36:56 2007
New Revision: 123635
URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=123635
Log:
2007-04-06 Paolo Carlini [EMAIL PROTECTED]
PR
--- Comment #35 from pcarlini at suse dot de 2007-04-07 00:37 ---
Fixed for 4.3.0.
--
pcarlini at suse dot de changed:
What|Removed |Added
Status|ASSIGNED
--- Comment #28 from craig dot lawson at centrify dot com 2007-03-22 23:44
---
For HP-UX, SunOS, and Irix:
#if !defined(_POSIX_C_SOURCE)
# define _POSIX_C_SOURCE 199506L
#elif _POSIX_C_SOURCE199506L
# error Conflict
#endif
(why would anyone compile C++ with POSIX
--- Comment #29 from pcarlini at suse dot de 2007-03-23 00:29 ---
Thanks a lot Craig. Touching _POSIX_C_SOURCE makes me a little nervous, in
particular if that ends up changing _REENTRANT, as happens on HP-UX. I'd like
to ask Dave opinion about it, whether to his best knowledge there is
--- Comment #30 from dave at hiauly1 dot hia dot nrc dot ca 2007-03-23
00:56 ---
Subject: Re: c++locale.o thread-unsafe in libstdc++
--- Comment #29 from pcarlini at suse dot de 2007-03-23 00:29 ---
Thanks a lot Craig. Touching _POSIX_C_SOURCE makes me a little nervous, in
--- Comment #14 from pcarlini at suse dot de 2007-03-15 09:38 ---
Thanks Craig for all your additional info, I'll try to make further progress
myself over the next days...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #15 from pcarlini at suse dot de 2007-03-15 12:05 ---
(In reply to comment #13)
Exactly which CPP symbol: On this platform, to select the errno function,
/usr/include/errno.h requires
defined(_REENTRANT) || defined(_TS_ERRNO) || _POSIX_C_SOURCE - 0 =
199506L
--- Comment #16 from craig dot lawson at centrify dot com 2007-03-15 18:15
---
(In reply to comment #15)
Tried it. Did not work.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #17 from pcarlini at suse dot de 2007-03-15 18:27 ---
(In reply to comment #16)
Tried it. Did not work.
Why?
--
pcarlini at suse dot de changed:
What|Removed |Added
--- Comment #18 from craig dot lawson at centrify dot com 2007-03-15 18:39
---
(In reply to comment #17)
Why?
Don't know. The file and change were included as expected.
Without fully understanding how CPP is evaluating these files, we're shooting
in the dark. gcc's -E and -CC
--- Comment #19 from pinskia at gcc dot gnu dot org 2007-03-15 18:40
---
Try adding -dD which causes the preprocessor to also emit all #defines.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #20 from pcarlini at suse dot de 2007-03-15 18:45 ---
Thanks Andrew. I'd like to know what happens if you just compile a small
snippet like:
#define _TS_ERRNO
#include cerrno
int main()
{
return errno;
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #21 from craig dot lawson at centrify dot com 2007-03-15 19:07
---
(In reply to comment #17)
Why?
OK, now I know why. Reason is that os_defines.h is included after errno.h.
Here's the include sequence:
c++locale.cc
cerrno
/usr/include/errno.h-- extern
--- Comment #22 from pcarlini at suse dot de 2007-03-15 19:17 ---
(In reply to comment #21)
(In reply to comment #17)
Why?
OK, now I know why. Reason is that os_defines.h is included after errno.h.
Oh yes, you are totally right. The problem is that cerrno is one of the very
few
--- Comment #23 from craig dot lawson at centrify dot com 2007-03-15 19:26
---
(In reply to comment #22)
Oh yes, you are totally right. The problem is that cerrno is one of the very
few c headers not including bits/c++config.h first. We can easily fix that.
Why did cerrno not
--- Comment #24 from pcarlini at suse dot de 2007-03-15 19:37 ---
(In reply to comment #23)
(In reply to comment #22)
Oh yes, you are totally right. The problem is that cerrno is one of the
very
few c headers not including bits/c++config.h first. We can easily fix
that.
--- Comment #25 from craig dot lawson at centrify dot com 2007-03-15 20:28
---
(In reply to comment #24)
Oh, that's excellent news, because we have to include bits/c++config.h first
in cerrno anyway... But now you make me curious!!! If you look at
c++config.h
it doesn't appear do
--- Comment #26 from pcarlini at suse dot de 2007-03-15 20:44 ---
(In reply to comment #25)
Sorry, that news was too excellent. It's not true.
Ok.
The reason I suggested also including clocale in comment #10 was because I
found bits/c++config.h was not enough. I agree that clocale
--- Comment #27 from craig dot lawson at centrify dot com 2007-03-15 21:17
---
(In reply to comment #26)
I understand. I'll collect this information from the systems I have, and then
post it here ...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #10 from craig dot lawson at centrify dot com 2007-03-14 18:13
---
Found it.
The problem is that the library is not be ing properly compiled for
multi-threaded use, but that by some luck basic_file.cc happened to be compiled
that way anyway. c++locale.cc was not so lucky.
--- Comment #11 from pcarlini at suse dot de 2007-03-15 01:11 ---
(In reply to comment #10)
The problem is that the library is not be ing properly compiled for
multi-threaded use, but that by some luck basic_file.cc happened to be
compiled
that way anyway. c++locale.cc was not so
--- Comment #12 from pcarlini at suse dot de 2007-03-15 01:24 ---
(In reply to comment #11)
Honestly, I don't think that makes sense. Did you *really* test it? Because,
as
far as I can see, the cc file at issue includes the whole locale, which, in
turn, includes bits/c++locale.h,
--- Comment #13 from craig dot lawson at centrify dot com 2007-03-15 02:02
---
(reply to comment #11)
Shear luck: Fair enough. I don't claim to understand it all, either, especially
how it came to be this way.
What appears to be happening on Solaris 10 is that bits/c++config.h
--- Comment #4 from craig dot lawson at centrify dot com 2007-03-12 18:26
---
Perhaps there are different degrees of thread safety here.
If I write my multi-threaded program such that only one thread does stream I/O
with libstdc++, I would expect it to operate safely with respect to
--- Comment #5 from pcarlini at suse dot de 2007-03-12 18:51 ---
(In reply to comment #4)
Perhaps there are different degrees of thread safety here.
If I write my multi-threaded program such that only one thread does stream I/O
with libstdc++, I would expect it to operate safely
--- Comment #6 from craig dot lawson at centrify dot com 2007-03-12 19:23
---
Anyway, what happens in the GNU systems is that errno is a per-thread variable
and the __convert_to_v code has to do nothing special, just safely use it.
I agree. The problem is that this is not happening.
--- Comment #7 from pcarlini at suse dot de 2007-03-12 19:51 ---
(In reply to comment #6)
Anyway, what happens in the GNU systems is that errno is a per-thread
variable
and the __convert_to_v code has to do nothing special, just safely use it.
I agree. The problem is that this
--- Comment #8 from craig dot lawson at centrify dot com 2007-03-12 20:07
---
Not on Linux: correct.
I could give it a try, but I haven't built this library before. If you could
point me to a brief how-to, I could give it a try on a Solaris 10 SPARC.
Rather than code around it, I
--- Comment #9 from pcarlini at suse dot de 2007-03-12 20:14 ---
(In reply to comment #8)
Not on Linux: correct.
Ok.
I could give it a try, but I haven't built this library before. If you could
point me to a brief how-to, I could give it a try on a Solaris 10 SPARC.
Rather than
--- Comment #3 from pcarlini at suse dot de 2007-03-10 20:46 ---
Humm, I'm afraid this is basically a WONTFIX or at least WONTFIX-ANY-TIME-SOON,
because we talking about targets using the GENERIC locale model, where, at
variance with the GNU locale model, on which we are focused of
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-03-10 00:10 ---
Sounds like a define is not being added so errno turns into a thread safe
version ...
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
--- Comment #2 from craig dot lawson at centrify dot com 2007-03-10 00:15
---
Seems like that define should come from gcc's -pthread option so it applies
to all source files.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31117
36 matches
Mail list logo
