Re: 2.6.19-rc4: known unfixed regressions (v3)

From: Eric W. Biederman
Date: Sun Nov 05 2006 - 10:23:39 EST

Adrian Bunk <bunk@xxxxxxxxx> writes:

> Subject : x86_64: NR_IRQ increase causes 11.5% slowdown
> in lmbench's fork benchmark
> References :
> Submitter : Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>
> Caused-By : Eric W. Biederman <ebiederm@xxxxxxxxxxxx>
> commit 550f2299ac8ffaba943cf211380d3a8d3fa75301
> Handled-By : Eric W. Biederman <ebiederm@xxxxxxxxxxxx>
> Andi Kleen <ak@xxxxxxx>
> Status : problem is being debugged

Currently I'm at a loss why the cross cpu fork lm_bench numbers should get
worse when you simply double NR_IRQS. As far as I can determine nothing
on that code path is directly affected by that change.

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at