Skip to contentSkip to conversation

BUG at net/sunrpc/svc_xprt.c:921

DISCUSSION Tom Horsley 2 participants 2 messages Feb 7, 2013 – Feb 8, 2013

lore
Share & tools
lore
#0 lore raw

I noticed some previous messages with this subject, but the walkback I'm getting doesn't match exactly the ones shown in the threads I saw, so I figured I'd send this in.

This happens on both my Fedora 18 and Fedora 17 partitions when mounting filesystems from very old servers that need the proto=udp option to talk.

The redhat bugzilla is here:

https://bugzilla.redhat.com/show_bug.cgi?id=908451

There is a photo of the walkback in that bugzilla:

svc_delete_xprt+0x12a
svc_recv+0x101
? nfs_callback_authenticate+0x50 nfs4_callback_svc+0x3b
kthread+0xc0
? ftrace_raw_event_xen_mmu_flush_tlb_others+0x50
? kthread_create_on_node+0x120
ret_from_fork+0x7c
? kthread_create_on_node+0x120

#1 lore raw
Quoted 14 lines · Tom Horsley

I noticed some previous messages with this subject, but the walkback I'm getting doesn't match exactly the ones shown in the threads I saw, so I figured I'd send this in.

This happens on both my Fedora 18 and Fedora 17 partitions when mounting filesystems from very old servers that need the proto=udp option to talk.

The redhat bugzilla is here:

https://bugzilla.redhat.com/show_bug.cgi?id=908451

There is a photo of the walkback in that bugzilla:

Quoted 9 lines

svc_delete_xprt+0x12a
svc_recv+0x101
? nfs_callback_authenticate+0x50 nfs4_callback_svc+0x3b
kthread+0xc0
? ftrace_raw_event_xen_mmu_flush_tlb_others+0x50
? kthread_create_on_node+0x120
ret_from_fork+0x7c
? kthread_create_on_node+0x120

OK, yes this is the

	if (test_and_set_bit(XPT_DEAD, &xprt->xpt_flags))
	      BUG();

in svc_delete_xprt() and is a known (but unfixed) problem. Stanislav has some patches, that should be fixed soon....

--b.