A correction to the build description in my post: I left out 18 commits from 
master that we cherry-picked onto v26.06 earlier. Every build we've run since 
then has them, including the one both cores came from:

a2747c78 gre: fix off-by-one error in TEIB sw_if_index checks
5398814c interface: no-mfib table bind and create fixes
3f544b87 lldp: fix TLV validation
2e217922 vnet: fix reuse of deleted interfaces
6c15c5b7 ip6: fix potential small OOB read in extension header walk
3218185d vxlan: reset next_dpo on delete
193df5bb vlib: buffer: fix null ptr dereference
1576b357 dev: hold worker thread barrier during port stop
39b2a4ec vlib: fix timed event data parameter
d0ee8e57 vppinfra: make sparse vector lookup zero-safe
2158b113 buffers: return values; improve debug
1bba2c01 linux-cp: fix cleanup of special routes
35768573 vlib: punt: fix buffer reference after clone
ca4ef2e8 vppinfra: fifo: do not resize vector down to 0
b815e99f vlib: fix buffer ref_count accounting in partial clone
3294213d l2: fix buffer overflow while finding next node
2543d3fe l2: fix race condition and null ptr dereference in l2fib_scan
f9020c10 buffers: fix pool allocation on small pages

So two of the L2 fixes I said we would test (3294213d, 2543d3fe) were already 
running when both cores were taken, as were 3218185d and b815e99f. Of the three 
I listed, only 6ad56defd is missing from our build, and we'll add it.

No further crashes since Oct 5. Both cores are still available if anyone wants 
specific gdb output.

Regards,
Pieter
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#27240): https://lists.fd.io/g/vpp-dev/message/27240
Mute This Topic: https://lists.fd.io/mt/121592811/21656
Group Owner: [email protected]
Unsubscribe: https://lists.fd.io/g/vpp-dev/leave/14379924/21656/631435203/xyzzy 
[[email protected]]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to