On Tuesday 24 July 2001 22:24, Patrick Dreker wrote:
> I just decided to give this patch a try, as I have written a little
> application which does some statistics over traces dumped by another
> program by mmap()ing a large file and reading it sequentially. The
> trace file to be analyzed is about 240 megs in size and consists of
> records each 249 bytes long. The analysis program opens and the
> mmap()s the trace file doing some small calculations on the data
> (basically it adds up fields from the records to get overall values).
>
> I have tested this on my Athlon 600 with 128 Megs of RAM, and it does
> not make any difference whether I use plain 2.4.7 or 2.4.5-use-once.
> The program always takes about 1 minute 6 seconds (+- 2 seconds) to
> complete, and the machine starts swapping out stuff
In this case that's an excellent result:
- The optimization doesn't include mmap's (yet)
- It doesn't break swap. (Good, I didn't check that myself)
This is a case of "no news is good news".
> (thus I have
> omitted further stats like vmstat output). I have just taken another
> look into my program to verify it does not do something silly, like
> keeping old data around, but the program cycle is always the same:
> copy a record from the mmap into a struct, perform analysis, and copy
> next record. The struct is always reused for the next struct (so
> there is only one struct at any time).
>
> I can do further tests, if someone asks me to. I could even modify
> the analysis program to check changes in behaviour...
(Already read your mail where you picked up the 20% improvement,
thanks, it warms my heart:-)
-- Daniel- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Tue Jul 31 2001 - 21:00:20 EST