On Sep 21, 2015, at 1:24 PM, Brian King [off-list ref] wrote:
On 09/16/2015 04:30 PM, Matthew R. Ochs wrote:
quoted
Borrowing the TMF waitq's spinlock causes a stall condition when
waiting for the TMF to complete. To remedy, introduce our own spin
lock to serialize TMF and use the appropriate wait services.
Can you clarify what stall condition you were seeing. Its not obvious
to me what this fixes. Do you have soft lockup logs from the failure?
I believe we saw cascading RCU stalls.
I couldn't find any more details in my notes or development commits.
Unfortunately the logs are long gone as this was fixed in June.