On 2008-10-16 19:57, Hans Morten Kind wrote: > When vba-detection in libclamav/ole2_extract.c was implemented in 0.93.1, > our solaris servers started dumping cores. We are experiencing the same > behavour with simple .doc-files also on 0.94 and the latest devel-release. > > hdr.m_area is memcpy'd to &hdr with hdr_size in the entry of > cli_ole2_extract(), but when hdr->m_area is tested in ole2_read_block() or > even in ole2_walk_property_tree() we get a signal 10, Bus error > We have tried to find the problem but have not succeeded. > > The fault can be forced by clamscan on the unix-prompt as by submitting > the ole2-file to clamd. Is anybody using clamd on solaris-10/sparc64? > > I can make a gdb backtrace for your bugzilla if this is not a known > problem occuring only on our servers. >
Hi, Please open a bug on bugs.clamav.net, and attach your gdb backtrace. Also please attach one sample that reproduce the bug to the bugzilla. Is this reproducible on sparc32 [1], or x86-32 / solaris-10? [1] Try building with CFLAGS="-m32" and see if the bug is still there Best regards, --Edwin _______________________________________________ Help us build a comprehensive ClamAV guide: visit http://wiki.clamav.net http://www.clamav.net/support/ml
