Re: [PATCH 3/3] readahead: scale max readahead size depending on memory size

[Fengguang Wu]

On Sun, Jul 22, 2007 at 10:59:11AM +0200, Peter Zijlstra wrote:
> On Sun, 2007-07-22 at 16:45 +0800, Fengguang Wu wrote:
> > How about the following rules?
> > - limit it under 1MB: we have to consider latencies
> 
> readahead is done async and we have these cond_resched() things
> sprinkled all over, no?

Yeah, it should not be a big problem.

> > - make them alignment-friendly, i.e. 128K, 256K, 512K, 1M.
> 
> Would that actually matter? but yeah, that seems like a sane suggestion.
> roundup_pow_of_two() comes to mind.

E.g. RAID stride size, and the max_sectors_kb.
Typically they are power-of-two.

> > My original plan is to simply do the following:
> > 
> > - #define VM_MAX_READAHEAD        128     /* kbytes */
> > + #define VM_MAX_READAHEAD        512     /* kbytes */
> 
> Yeah, the trouble I have with that is that it might adversely affect
> tiny systems (although the trash detection might mitigate that impact)

I'm also OK with the scaling up scheme. It's reasonable.

> > I'd like to post some numbers to back-up the discussion:
> > 
> >   readahead   readahead
> >        size        miss
> >        128K         38%
> >        512K         45%
> >       1024K         49%
> > 
> > The numbers are measured on a fresh booted KDE desktop.
> > 
> > The majority misses come from the larger mmap read-arounds.
> 
> the mmap code never gets into readahead unless madvise(MADV_SEQUENTIAL)
> is used afaik.

Sadly mmap read-around reuses the same readahead size.
- for read-around, VM_MAX_READAHEAD is the _real_ readahead size
- for readahead, VM_MAX_READAHEAD is the _max_ readahead size
If we simply increasing VM_MAX_READAHEAD, tiny systems can be
immediately hurt by large read-arounds. That's the problem.

-
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/