[OpenAFS-devel] fileserver profiling

[Derrick J Brashear]

On Mon, 7 Mar 2005, Troy Benjegerdes wrote:

> How many CPU's did this system have, and what version of Linux? If this 
> is still reproduceable on a 2.4 or 2.6 kernel, it should be fixed. 
> (Also, was this x86, or some other architecture)

I don't remember, but conveniently the guy who might is also at IETF this 
week so when I see him later I will ask him. IIRC 2 cpus, x86, 2.4.x

> If we have some process updating the time in a mmap()'ed file, wouldn't that
> have about the same cost as getting a cached time? If we are paranoid
> about time going backwards, the process updating the time could check
> for that as well, and spit out some verbose warning to syslog.

The mmap() code presumably needs another thread, and would need to be 
different for lwp. Why have 2 sets of code, and use up another thread?