* How is jiffies upation while holding spinlock in Uniprocessor or SMP environment?
@ 2015-10-26 6:49 raghu MG
2015-10-26 10:48 ` Russell King - ARM Linux
0 siblings, 1 reply; 3+ messages in thread
From: raghu MG @ 2015-10-26 6:49 UTC (permalink / raw)
To: linux-arm-kernel, linux-kernel, Andrew Lunn
Hello,
I have couple of questions.
1) In case of UP how is jiffies(timer) get updated while holding spin
lock using spinlock_irq_save?
2)My understanding is in smp environment jiffies updation cannot be
done on processor holding spinlock irrespective of spinlock API.So
timer interrupt should be enabled across all cores in SoC. Imagine a
scenario in quad core processor where 4 cores are holding 4 different
spinlocks & how is the timer interrupt for jiffies handled?
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: How is jiffies upation while holding spinlock in Uniprocessor or SMP environment?
2015-10-26 6:49 How is jiffies upation while holding spinlock in Uniprocessor or SMP environment? raghu MG
@ 2015-10-26 10:48 ` Russell King - ARM Linux
2015-10-30 10:59 ` raghu MG
0 siblings, 1 reply; 3+ messages in thread
From: Russell King - ARM Linux @ 2015-10-26 10:48 UTC (permalink / raw)
To: raghu MG; +Cc: linux-arm-kernel, linux-kernel, Andrew Lunn
On Mon, Oct 26, 2015 at 12:19:31PM +0530, raghu MG wrote:
> Hello,
> I have couple of questions.
> 1) In case of UP how is jiffies(timer) get updated while holding spin
> lock using spinlock_irq_save?
It isn't, and...
>
> 2)My understanding is in smp environment jiffies updation cannot be
> done on processor holding spinlock irrespective of spinlock API.So
> timer interrupt should be enabled across all cores in SoC. Imagine a
> scenario in quad core processor where 4 cores are holding 4 different
> spinlocks & how is the timer interrupt for jiffies handled?
You can't rely on jiffies being updated while holding a spinlock.
Expecting jiffies to update while holding a spinlock is a bug.
--
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
according to speedtest.net.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: How is jiffies upation while holding spinlock in Uniprocessor or SMP environment?
2015-10-26 10:48 ` Russell King - ARM Linux
@ 2015-10-30 10:59 ` raghu MG
0 siblings, 0 replies; 3+ messages in thread
From: raghu MG @ 2015-10-30 10:59 UTC (permalink / raw)
To: Russell King - ARM Linux; +Cc: linux-arm-kernel, linux-kernel, Andrew Lunn
Thank you for clarifying.
On Mon, Oct 26, 2015 at 4:18 PM, Russell King - ARM Linux
<linux@arm.linux.org.uk> wrote:
> On Mon, Oct 26, 2015 at 12:19:31PM +0530, raghu MG wrote:
>> Hello,
>> I have couple of questions.
>> 1) In case of UP how is jiffies(timer) get updated while holding spin
>> lock using spinlock_irq_save?
>
> It isn't, and...
>>
>> 2)My understanding is in smp environment jiffies updation cannot be
>> done on processor holding spinlock irrespective of spinlock API.So
>> timer interrupt should be enabled across all cores in SoC. Imagine a
>> scenario in quad core processor where 4 cores are holding 4 different
>> spinlocks & how is the timer interrupt for jiffies handled?
>
> You can't rely on jiffies being updated while holding a spinlock.
> Expecting jiffies to update while holding a spinlock is a bug.
>
> --
> FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up
> according to speedtest.net.
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2015-10-30 10:59 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2015-10-26 6:49 How is jiffies upation while holding spinlock in Uniprocessor or SMP environment? raghu MG
2015-10-26 10:48 ` Russell King - ARM Linux
2015-10-30 10:59 ` raghu MG
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®