Re: hrtimer_interrupt time sync issues across cores

[valdis . kletnieks]

On Wed, 13 Dec 2017 23:01:57 -0800, Rajasekaran Chandrasekaran said:

> In our multi-core x86 based system that is running 3.4.19 version of

As Greg already pointed out, that's ancient history suitable only for kernel
archaeologists and masochists.

> Problem:
>
> Inside hrtimer_interrupt function, basenow.tv64 in CPU-3 is 1.6 seconds
> ahead of other CPU’s (we have 4 cores

I may be mis-remembering, but I think the timer code has undergone at least
2 major reworkings in the 6 years since 3.4.  So there's a good chance the bug
has already been fixed.  Of course, back-porting the fix may be close to 
impossible.

Good luck, you will need it.


pgpa5TCmmwqfk.pgp
Description: PGP signature
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: hrtimer_interrupt time sync issues across cores

[Greg KH]

On Wed, Dec 13, 2017 at 11:01:57PM -0800, Rajasekaran Chandrasekaran wrote:
> Hi,
> 
> 
> In our multi-core x86 based system that is running 3.4.19 version of
> kernel, hrtimer_interrupt (called from apic_timer_interrupt) keeps looping
> in hardirq for atleast 1.6 seconds.  We use tsc as our clock source. The
> issue happens very rarely in our system and hard to reproduce.

Please note, 3.4.19 is _very_ old, almost 300 thousand kernel changes
have been done since then.  Also, your kernel is missing an untold
number of security fixes, making your systems totally insecure.

If you are stuck at this kernel version, please work with the company
that you are paying to support it, as they are the only ones that can do
this, not the community, sorry.  Also, you are paying for that support,
please use it!

best of luck!

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies