Re: [PATCHv3 0/2] mm: map few pages around fault address if they are in page cache

[Kirill A. Shutemov]

On Thu, Feb 27, 2014 at 01:28:22PM -0800, Linus Torvalds wrote:
>  (a) could you test this on a couple of different architectures?


Here's data from 4-socket Westmere. The advantage of faultaround is not that
obvious here.

FAULT_AROUND_ORDER              baseline        2               4               6

Linux build (make -j40)
minor-faults                    297,626,681     245,934,795     227,800,211     221,052,532
time, seconds                   293.4322        291.4602        292.9198        295.2355

Linux rebuild (make -j40)
minor-faults                    5,903,936       3,802,148       3,018,728       2,735,602
time, seconds                   41.5657         41.0301         40.7621         41.1161

Git test suite
minor-faults                    171,314,056     109,187,718     81,955,503      70,172,157
time, seconds                   223.2327        220.2623        223.8355        231.7843

I don't have time to test on Haswell now. Probably later.

-- 
 Kirill A. Shutemov
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/