sysctl_ctx gets created only in non-recovery mode and is filled when device attach is completed. However, when device_attach fails, IFLIB internally calls into IFDI_QUEUES_FREE and attempts to parse sysctl_ctx list with no entries in it. As a result, it panics the node. So, check for non-empty list before attempting the free. Also, sysctl_ctx is a VSI resource rather a queue resource. So, move the sysctl_ctx free to DETACH() instead.
Details
- Reviewers
markj krzysztof.galazka_intel.com - Group Reviewers
NetApp Intel Networking
Added a dummy failpoint to simulate iflib_device_register() failure before IFDI_ATTACH_POST
Diff Detail
- Repository
- rS FreeBSD src repository - subversion
- Lint
Lint Passed - Unit
No Test Coverage - Build Status
Buildable 39878 Build 36767: arc lint + arc unit
Event Timeline
I think checking for TAILQ_EMPTY(&vsi->sysctl_ctx) is probably a hack. sysctl_ctx_free() handles an empty queue perfectly well. The problem seems to be that the list is not initialized (with sysctl_ctx_init()) in some paths where sysctl_ctx_free() may be called.
I would suggest instead calling sysctl_ctx_init(&vsi->sysctl_ctx) much earlier during ixl attach, in ixl_if_attach_pre().
sys/dev/ixl/if_ixl.c | ||
---|---|---|
867 | The formatting is wrong: continuing lines should be indendent by four spaces. |
Actually using separate sysctl_ctx was necessary in non-iflib version of the driver to allow releasing and re-allocating queues while handling EMP reset. In this version Rx queues memory is managed by the iflib with IFDI interface so using separate syssctl context may be not required. Let me double check that.