https://bugs.dpdk.org/show_bug.cgi?id=1337
Bug ID: 1337
Summary: [21.11.5] iavf driver stuck in rte_eal_alarm_cancel
Product: DPDK
Version: 21.11
Hardware: All
OS: All
Status: UNCONFIRMED
Severity: normal
Priority: Normal
Component: ethdev
Assignee: dev@dpdk.org
Reporter: ktray...@redhat.com
Target Milestone: ---
With iavf device in OVS and dpdk 21.11.5.
OVS main thread is calling rte_eth_dev_configure()...iavf_queues_req_reset()
and is getting stuck in rte_eal_alarm_cancel() waiting for alarms to be
cancelled.
eal-intr-thread is stuck in application callback as it is requesting
reconfiguration and waiting for current configure to be completed in the main
thread.
The issue was introduced with the patch below. There are some related patches
in this area that were not previously backported and newer ones to come.
commit b965aceffdb9dd8d2501314c174716617371d752
Author: Mingjin Ye <mingjinx...@intel.com>
Date: Wed Jun 14 09:53:03 2023 +0000
net/iavf: fix abnormal disable HW interrupt
[ upstream commit 675a104e2e940ec476e8b469725e8465d01c0098 ]
For command VIRTCHNL_OP_REQUEST_QUEUES, polling access to the admin
queue has the issue of access overruns after disabling interrupt. That
results in FW disabling HW interrupt for protection purposes.
The updates/changes in this patch:
1. Remove the polling admin queue processing and use the generic interrupt
processing instead.
2. Release redundant queue resource before stopping processing interrupt
events.
Fixes: 22b123a36d07 ("net/avf: initialize PMD")
Fixes: ef807926e148 ("net/iavf: support requesting additional queues from
PF")
Fixes: 84108425054a ("net/iavf: support asynchronous virtual channel
message")
Signed-off-by: Mingjin Ye <mingjinx...@intel.com>
Acked-by: Qi Zhang <qi.z.zh...@intel.com>
--
You are receiving this mail because:
You are the assignee for the bug.
