[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: mmap002 execution time doubled... good or bad sign?
On Sat, 10 Feb 2001, Roger Larsson wrote:
> Hi,
>
> I have been running various stress tests on disk for some time.
> streaming write, copy, read, diff, dbench and mmap002
>
> This is what I have seen:
>
> >From 2.4.0 to 2.4.1 with Marcelos patch write were above 10 MB/s
> and read >13 MB/s, dbench > 10 MB/s, mmap took around 2m30.
>
> After 2.4.1-pre8 (did not test anything in between)
> Write is at 9-10 [lost 1 MB/s] read is down to 11-12 MB/s [lost 2 MB/s]
> dbench > 9 MB/s [one MB/s there too]
>
> But the really strange one - mmap002 now takes > 4m30
> Is this expected / good behaviour? mmap002 abuses mmaps...
These are probably Jens modifications to the block queuing mechanisms
since there were no VM changes from 2.4.1-pre8 to now in Linus tree.
Its not necessarily a bad thing, since Jens modifications make processes
throttle on IO sooner.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux.eu.org/Linux-MM/