On 4/27/16 10:09 AM, Frederic Weisbecker wrote:
I first thought that this should be a tunable per event instead of a global sysctlYeah, I'll work on that too.There is no rush though. The sysfs limit will probably be enough for most users. Unless someone requested it?
I have. I spent time last winter (2015) looking into how to do it. The userspace syntax was more of a pain than passing the parameters to the kernel side as the intent is to specify N kernel frames and M user frames.