I''m running OpenSolaris 2008.05 in VirtualBox on a Windows XP host. Playing around with various probes, I found that trying to load any probe associated with bdev_strategy dumps core. I can think of one or two likely and reasonable causes for this, but am assuming it''s undesirable behavior. Anyone know what''s happening here? -- This message posted from opensolaris.org
On Mon, Jun 09, 2008 at 01:27:11PM -0700, Michael Ernest wrote:> I''m running OpenSolaris 2008.05 in VirtualBox on a Windows XP host. Playing > around with various probes, I found that trying to load any probe associated > with bdev_strategy dumps core.Dumps a dtrace(1M) core or crashes the system? In either case, can you provide the core/crash dump? Adam -- Adam Leventhal, Fishworks http://blogs.sun.com/ahl
Oops, I spoke too soon. I thought it was dumping core, owing to the significant delay after locking. However I haven''t yet persuaded this setup to recover the dump files (dumpadm shows proper config). I''ll provide one as soon as I get that going. -- This message posted from opensolaris.org
Sorry for the triple-clutch on that last post. These crash files are monsters; the core image is 83M *after* gzipping, and I''m only dumping kernel pages. What''s the limit on file upload to this site? M -- This message posted from opensolaris.org
Don''t upload it to the list. Let''s talk offline about how to transfer the crash dump. Adam On Mon, Jun 09, 2008 at 03:31:17PM -0700, Michael Ernest wrote:> Sorry for the triple-clutch on that last post. > > These crash files are monsters; the core image is 83M *after* gzipping, and I''m only dumping kernel pages. > > What''s the limit on file upload to this site? > > M > > > -- > This message posted from opensolaris.org > _______________________________________________ > dtrace-discuss mailing list > dtrace-discuss at opensolaris.org-- Adam Leventhal, Fishworks http://blogs.sun.com/ahl
Anyone interested in replicating? Here''s the dtrace one-liner I used:
dtrace -n ''::bdev_strategy:entry {}''
And here''s the panic string from one attempt:
panic[cpu0]/thread=d3c2bde0:
BAD TRAP: type=e (#pf=Page Fault) rp=d3c2bc30 address=0 occurred in module
"<unknown>" due to a NULL pointer dereference.
--
This message posted from opensolaris.org
> Anyone interested in replicating? Here''s the dtrace one-liner I used: > > dtrace -n ''::bdev_strategy:entry {}'' > > And here''s the panic string from one attempt: > > panic[cpu0]/thread=d3c2bde0: > > BAD TRAP: type=e (#pf=Page Fault) rp=d3c2bc30 address=0 occurred in module "<unknown>" due to a NULL pointer dereference.Hmm... Using VirtualBox 1.6.2 on an Amd64 X2 machine running SXCE (b85, bfu''ed to snv_92), and an OpenSolaris 2008.05 guest there is no panic. My OS2008.05 guest has 384 MB of memory, ACPI and IO-APIC are enabled, PAE/NX is disabled. I tried both with "VT-x/AMD-V" enabled and disabled. -- This message posted from opensolaris.org