This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
RE: Input stream crash with gcc 3.1/cygwin (stack trace incl.)
- From: "Billinghurst, David (CRTS)" <David dot Billinghurst at riotinto dot com>
- To: "Benjamin Kosnik" <bkoz at redhat dot com>,<gcc at gcc dot gnu dot org>,<dylan at q-games dot com>
- Date: Tue, 26 Mar 2002 10:28:48 +1100
- Subject: RE: Input stream crash with gcc 3.1/cygwin (stack trace incl.)
That was it. With it I get
=== libstdc++-v3 tests ===
Schedule of variations:
unix
Running target unix
Using /usr/share/dejagnu/baseboards/unix.exp as board description file for target.
Using /usr/share/dejagnu/config/unix.exp as generic interface file for target.
Using /usr/local/src/gcc-3.1/libstdc++-v3/testsuite/config/default.exp as tool-and-target-specific interface file.
Running /usr/local/src/gcc-3.1/libstdc++-v3/testsuite/libstdc++-v3.dg/dg.exp ...
XPASS: 22_locale/collate_members_wchar_t.cc execution test
XPASS: 22_locale/money_get_members_wchar_t.cc execution test
XPASS: 22_locale/money_put_members_wchar_t.cc execution test
XPASS: 22_locale/moneypunct_members_wchar_t.cc execution test
XPASS: 22_locale/num_get_members_wchar_t.cc execution test
XPASS: 22_locale/num_put_members_wchar_t.cc execution test
XPASS: 22_locale/numpunct_members_wchar_t.cc execution test
XPASS: 22_locale/time_get_members_wchar_t.cc execution test
XPASS: 22_locale/time_put_members_wchar_t.cc execution test
FAIL: 27_io/filebuf_members.cc execution test
=== libstdc++-v3 Summary ===
# of expected passes 358
# of unexpected failures 1
# of unexpected successes 9
# of expected failures 16
-----Original Message-----
From: Benjamin Kosnik [mailto:bkoz@redhat.com]
Sent: Tuesday, 26 March 2002 6:17
To: gcc@gcc.gnu.org; dylan@q-games.com; Billinghurst, David (CRTS)
Subject: Re: Input stream crash with gcc 3.1/cygwin (stack trace incl.)
Okay. This has been an ongoing issue for a bit. David and I tried to
tackle this on the libstdc++ list, on and off last month. I don't have
full-time access to a cygwin setup, and so it's kind of hard for me to
debug this.
However.
Your email had a pretty good backtrace, and I went through the CVS logs
for the cygwin files, and cross checked this with David's postings on
gcc-testresults. It looks like my 2002-01-20 patch killed ctype support
for this platform.
Can you please try this:
2002-03-25 Benjamin Kosnik <bkoz@redhat.com>
* config/os/newlib/bits/ctype_noninline.h (ctype::classic): Set to
_ctype_.
Index: ctype_noninline.h
===================================================================
RCS file: /cvs/gcc/gcc/libstdc++-v3/config/os/newlib/bits/ctype_noninline.h,v
retrieving revision 1.8
diff -c -p -r1.8 ctype_noninline.h
*** ctype_noninline.h 2002/01/21 04:08:51 1.8
--- ctype_noninline.h 2002/03/25 19:12:38
***************
*** 35,41 ****
const ctype_base::mask*
ctype<char>::classic_table() throw()
! { return 0; }
ctype<char>::ctype(__c_locale, const mask* __table, bool __del,
size_t __refs)
--- 35,41 ----
const ctype_base::mask*
ctype<char>::classic_table() throw()
! { return _ctype_; }
ctype<char>::ctype(__c_locale, const mask* __table, bool __del,
size_t __refs)