Messages in this thread | | | Date | Wed, 19 Mar 2014 20:34:07 +0100 | From | Peter Zijlstra <> | Subject | Re: [RFC PATCH] cifs: Fix possible deadlock with cifs and work queues |
| |
On Wed, Mar 19, 2014 at 03:12:52PM -0400, Steven Rostedt wrote: > My question to Tejun is, if we create another workqueue, to add the > rdata->work to, would that prevent the above problem? Or what other > fixes can we do?
The way I understand workqueues is that we cannot guarantee concurrency like this. It tries, but there's no guarantee.
WQ_MAX_ACTIVE seems to be a hard upper limit of concurrent workers. So given 511 other blocked works, the described problem will always happen.
Creating another workqueue doesn't actually create more threads.
There is the kthread_work stuff for if you want a guaranteed worker thread.
| |