[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Thu, 2 Feb 2017 12:47:27 +0000
From: Will Deacon <will.deacon@....com>
To: Sodagudi Prasad <psodagud@...eaurora.org>
Cc: peterz@...radead.org, mingo@...hat.com, acme@...nel.org,
mark.rutland@....com, catalin.marinas@....com,
linux-kernel@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
tsoni@...cinc.com, pfay@....qualcomm.com
Subject: Re: irq_work_run called from armv8pmu_handle_irq
On Wed, Feb 01, 2017 at 12:15:01PM -0800, Sodagudi Prasad wrote:
> we would like to understand why irq_work_run() is getting called from
> armv8pmu_handle_irq() irq handler for arm SoC.
>
> Is this irq_work frame work designed to run works from ARMPMU irq handlers?
> If there are no of drivers using the IRQ WORK frame work, we may see that
> ARMPMU irq handler taking long time, right?
The perf_event_overflow callback can schedule IRQ work to e.g. disable the
PMU and want to make sure this runs before we EOI the PMU IRQ.
Will
Powered by blists - more mailing lists