This is the mail archive of the
gcc-regression@gcc.gnu.org
mailing list for the GCC project.
Re: build failure between 108150 and 108164
- From: Geoff Keating <geoffk at geoffk dot org>
- To: Richard Sandiford <rsandifo at redhat dot com>
- Cc: gcc-regression at gcc dot gnu dot org
- Date: Fri, 9 Dec 2005 14:43:22 -0800
- Subject: Re: build failure between 108150 and 108164
- References: <570128AE-611E-473F-8012-09675322021A@geoffk.org>
Hi Richard,
The trunk now builds, but many PCH tests are failing. For instance,
gcc.dg/pch/decl-1.c crashes here:
#0 0x90003008 in strlen ()
#1 0x9000d830 in __vfprintf$LDBL128 ()
#2 0x900ff038 in vfprintf_l$LDBL128 ()
#3 0x90101768 in fprintf$LDBL128 ()
#4 0x001721f0 in switch_to_section (new_section=0x41e11df0) at /
Network/Servers/hills/Volumes/capanna/gkeating/co/gcc-trunk/gcc/
varasm.c:5616
#5 0x0008cecc in machopic_output_indirection (slot=0xa5a5a5a5,
data=0xa000e790) at /Network/Servers/hills/Volumes/capanna/gkeating/
co/gcc-trunk/gcc/config/darwin.c:982
#6 0x004d3874 in htab_traverse_noresize (htab=0xa5a5a5a5,
callback=0x8cc70 <machopic_output_indirection>, info=0xa000e790) at /
Network/Servers/hills/Volumes/capanna/gkeating/co/gcc-trunk/libiberty/
hashtab.c:729
#7 0x0008cbf4 in darwin_file_end () at /Network/Servers/hills/
Volumes/capanna/gkeating/co/gcc-trunk/gcc/config/darwin.c:1445
(Not all of the variable values in the backtrace are correct.)
It looks like section variables, like
"*machopic_nl_symbol_ptr_section", are being allocated with gcc_alloc
but their roots are not being marked.
Attachment:
smime.p7s
Description: S/MIME cryptographic signature