Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Steven Rostedt
On Thu, 25 Feb 2016 15:54:13 -0800 "Shi, Yang" wrote: > Can we disable irqs in tracepoints since spin_lock_irqsave is used by > kernfs_* functions. Disabling preemption or irqs is fine a tracepoint. You just can't sleep, which spin_lock_irqsave() would do on the -rt kernel. -- Steve

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Shi, Yang
On 2/25/2016 3:47 PM, Shi, Yang wrote: On 2/25/2016 3:31 PM, Steven Rostedt wrote: On Thu, 25 Feb 2016 15:16:54 -0800 "Shi, Yang" wrote: Actually, regardless whether this is the right fix for the splat, it makes me be wondering if the spin lock which protects the whole for loop is really nec

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Shi, Yang
On 2/25/2016 3:31 PM, Steven Rostedt wrote: On Thu, 25 Feb 2016 15:16:54 -0800 "Shi, Yang" wrote: Actually, regardless whether this is the right fix for the splat, it makes me be wondering if the spin lock which protects the whole for loop is really necessary. It sounds feasible to move it in

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Steven Rostedt
On Thu, 25 Feb 2016 15:16:54 -0800 "Shi, Yang" wrote: > Actually, regardless whether this is the right fix for the splat, it > makes me be wondering if the spin lock which protects the whole for loop > is really necessary. It sounds feasible to move it into the for loop and > just protect the

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Shi, Yang
On 2/25/2016 11:54 AM, Steven Rostedt wrote: On Thu, 25 Feb 2016 11:38:48 -0800 "Shi, Yang" wrote: On 2/24/2016 6:40 PM, Steven Rostedt wrote: On Wed, 24 Feb 2016 14:47:23 -0800 Yang Shi wrote: commit 5634cc2aa9aebc77bc862992e7805469dcf83dac ("writeback: update writeback tracepoints to rep

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Steven Rostedt
On Thu, 25 Feb 2016 11:38:48 -0800 "Shi, Yang" wrote: > On 2/24/2016 6:40 PM, Steven Rostedt wrote: > > On Wed, 24 Feb 2016 14:47:23 -0800 > > Yang Shi wrote: > > > >> commit 5634cc2aa9aebc77bc862992e7805469dcf83dac ("writeback: update > >> writeback > >> tracepoints to report cgroup") made w

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-25 Thread Shi, Yang
On 2/24/2016 6:40 PM, Steven Rostedt wrote: On Wed, 24 Feb 2016 14:47:23 -0800 Yang Shi wrote: commit 5634cc2aa9aebc77bc862992e7805469dcf83dac ("writeback: update writeback tracepoints to report cgroup") made writeback tracepoints report cgroup writeback, but it may trigger the below bug on -r

Re: [PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-24 Thread Steven Rostedt
On Wed, 24 Feb 2016 14:47:23 -0800 Yang Shi wrote: > commit 5634cc2aa9aebc77bc862992e7805469dcf83dac ("writeback: update writeback > tracepoints to report cgroup") made writeback tracepoints report cgroup > writeback, but it may trigger the below bug on -rt kernel due to the list_lock > held for

[PATCH] writeback: call writeback tracepoints withoud holding list_lock in wb_writeback()

2016-02-24 Thread Yang Shi
commit 5634cc2aa9aebc77bc862992e7805469dcf83dac ("writeback: update writeback tracepoints to report cgroup") made writeback tracepoints report cgroup writeback, but it may trigger the below bug on -rt kernel due to the list_lock held for the for loop in wb_writeback(). BUG: sleeping function calle