On 02/15/17 15:31, Andrey Drobyshev wrote:
> From: Nikita Edward Baruzdin <[email protected]>
>
> This commit is supposed to resolve the issue with hard lockups on systems using
> jiffies as their clock source. Namely, it sets limits on number of iterations
> clock source may remain unchanged (i. e. not being updated for one reason or
> another, as it is with jiffies clock source), and on unsuccessful
> getnstimeofday() polls as well. In case limit is reached, we consider clock
> source incompatible with this driver or unstable.
>
> Considering this issue to be fixed, un-BROKEN pps_gen_parport.
>
> For explanation of the problem see this thread on lkml:
> https://lkml.org/lkml/2011/2/18/310
>
> Signed-off-by: Nikita Edward Baruzdin <[email protected]>
> Signed-off-by: Alexander GQ Gerasiov <[email protected]>
Acked-by: Rodolfo Giometti <[email protected]>
Hello Rodolfo,
I should say, that I found this solution is not the best. For example on
high speed PCs (I think over 3GHz) 5 iteration is not enough and module
stops working after several hours. That makes it not user-friendly,
cause user had to calibrate and set failure_iterations manually.
It would be better to use pre-measured loops_per_jiffy (from delay.h)
some way.
On Sat, 4 Mar 2017 08:44:03 +0100
Rodolfo Giometti <[email protected]> wrote:
> On 02/15/17 15:31, Andrey Drobyshev wrote:
> > From: Nikita Edward Baruzdin <[email protected]>
> >
> > This commit is supposed to resolve the issue with hard lockups on
> > systems using jiffies as their clock source. Namely, it sets limits
> > on number of iterations clock source may remain unchanged (i. e.
> > not being updated for one reason or another, as it is with jiffies
> > clock source), and on unsuccessful getnstimeofday() polls as well.
> > In case limit is reached, we consider clock source incompatible
> > with this driver or unstable.
> >
> > Considering this issue to be fixed, un-BROKEN pps_gen_parport.
> >
> > For explanation of the problem see this thread on lkml:
> > https://lkml.org/lkml/2011/2/18/310
> >
> > Signed-off-by: Nikita Edward Baruzdin <[email protected]>
> > Signed-off-by: Alexander GQ Gerasiov <[email protected]>
>
> Acked-by: Rodolfo Giometti <[email protected]>
--
Best regards,
Alexander Gerasiov
Contacts:
e-mail: [email protected] Homepage: http://gerasiov.net Skype: gerasiov
PGP fingerprint: 04B5 9D90 DF7C C2AB CD49 BAEA CA87 E9E8 2AAC 33F1
On 03/04/17 10:19, Alexander Gerasiov wrote:
> Hello Rodolfo,
>
> I should say, that I found this solution is not the best. For example on
> high speed PCs (I think over 3GHz) 5 iteration is not enough and module
> stops working after several hours. That makes it not user-friendly,
> cause user had to calibrate and set failure_iterations manually.
>
> It would be better to use pre-measured loops_per_jiffy (from delay.h)
> some way.
I see... however I cannot test the patch so if you tried it and you noticed
these issues we should consider to add it but with the experimental warning?
Ciao,
Rodolfo
--
HCE Engineering e-mail: [email protected]
GNU/Linux Solutions [email protected]
Linux Device Driver [email protected]
Embedded Systems phone: +39 349 2432127
UNIX programming skype: rodolfo.giometti
Cosino Project - the quick prototyping embedded system - http://www.cosino.io
Freelance ICT Italia - Consulente ICT Italia - http://www.consulenti-ict.it