Re: [PATCH][CFT] blk-fair-batches vs 2.4.20-rc6

Nick Piggin (piggin@cyberone.com.au)
Tue, 03 Jun 2003 10:10:30 +1000


Rik van Riel wrote:

> On Mon, 2 Jun 2003, Nick Piggin wrote:
>
>> Previously:
>> * request queue fills up
>> * process 1 calls get_request, sleeps
>> * a couple of requests are freed
>> * process 2 calls get_request, proceeds
>> * a couple of requests are freed
>> * process 2 calls get_request, proceeds
>> ...
>
>
> In an early 2.4 kernel I've caught a few processes sleeping
> in get_request_wait for 5 minutes or so, while other processes
> were allocating new requests at exactly the speed they were
> processed.
>
> Of course, a patch to fix the problem was shot down due to
> lower dbench performance ... good thing Andrew Morton has
> more sense than that.

Mmm... unfortunately nearly everywhere >1 threads compete for
a resource, it comes down to throughput vs latency.

I think this patch is valid. It does not seem to be the sole
cause of the hangs though... err, and its against 21-rc6,
sorry :P

-
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/