how long wait_for_completion_interruptible_timeout() blocks
Hi All, I’ve two Linux systems, I tried to measure how long wait_for_completion_interruptible_timeout() blocked after completion is signaled by complete() call. I noticed two different results in both systems. In one system I noticed this duration in between 10 to 60 micro seconds, where as in other system it is in between 20 to 500 micro seconds. I’d like to know what factors can affect this latency? Also, is there any solution to reduce this latency? Thanks,
On Wed, Jun 19, 2019 at 10:47 PM Subhashini Rao Beerisetty < subhashbeerisetty@gmail.com> wrote:
Hi All,
I’ve two Linux systems, I tried to measure how long wait_for_completion_interruptible_timeout() blocked after completion is signaled by complete() call. I noticed two different results in both systems.
In one system I noticed this duration in between 10 to 60 micro seconds, where as in other system it is in between 20 to 500 micro seconds. I’d like to know what factors can affect this latency? Also, is there any solution to reduce this latency?
This is a problem which is fixed by RT Linux patches. It is taking time as Linux is not RTOS and scheduler took time to put your task in running state which is expected and depends on many factors: such as load, interrupts and so on.
Thanks,
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
On Thu, Jun 20, 2019 at 11:38 AM anish singh <anish198519851985@gmail.com> wrote:
On Wed, Jun 19, 2019 at 10:47 PM Subhashini Rao Beerisetty <subhashbeerisetty@gmail.com> wrote:
Hi All,
I’ve two Linux systems, I tried to measure how long wait_for_completion_interruptible_timeout() blocked after completion is signaled by complete() call. I noticed two different results in both systems.
In one system I noticed this duration in between 10 to 60 micro seconds, where as in other system it is in between 20 to 500 micro seconds. I’d like to know what factors can affect this latency? Also, is there any solution to reduce this latency?
This is a problem which is fixed by RT Linux patches.
Can these patches be applied to Linux kernel source? If so can you please give more information on this... where to find these patches?
It is taking time as Linux is not RTOS and scheduler took time to put your task in running state which is expected and depends on many factors: such as load, interrupts and so on.
Thanks,
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@kernelnewbies.org https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
participants (2)
-
anish singh -
Subhashini Rao Beerisetty