On Wed, 22 Apr 2009 09:20:20 -0700 (PDT) Steve Gonczi <gonczi at comcast.net> wrote:
> Greetings, > > I have been using mdb for years but have not seen this problem before. > When I attempt to debug a few rather large-ish (2.5G) kernel crash dumps, > loading dmods fails thus: > > mdb -k 2 > mdb: failed to read .symtab header for 'unix', id=0: no mapping for address > mdb: failed to read .symtab header for 'genunix', id=1: no mapping for address > mdb: failed to read modctl at ffffff113d7aff08: no mapping for address > > ::dmods > krtld > mdb > mdb_kb > mdb_kproc > mdb_ks > mdb_kvm > mdb_proc > mdb_raw > > Subsequently, it is impossible to debug the crash because the most useful > commands > are missing. > What could be the reason? Perhaps corrupted unix.x and vmcore.x files? > mdb -k on the currently running kernel loads all dmods as expected, and > works fine. I've seen that most often when the dump has been foreshortened. Most recently it occurred to me when my dump device was too small. James C. McPherson -- Senior Kernel Software Engineer, Solaris Sun Microsystems http://blogs.sun.com/jmcp http://www.jmcp.homeunix.com/blog Kernel Conference Australia - http://au.sun.com/sunnews/events/2009/kernel
