[PATCH V2 RESEND] fsl-sata: I/O load balancing
Liu Qiang-B32616
B32616 at freescale.com
Fri Feb 17 14:30:36 EST 2012
> -----Original Message-----
> From: Benjamin Herrenschmidt [mailto:benh at kernel.crashing.org]
> Sent: Friday, February 17, 2012 11:18 AM
> To: Liu Qiang-B32616
> Cc: jgarzik at pobox.com; linux-ide at vger.kernel.org; linuxppc-
> dev at lists.ozlabs.org; linux-kernel at vger.kernel.org
> Subject: RE: [PATCH V2 RESEND] fsl-sata: I/O load balancing
>
> On Fri, 2012-02-17 at 01:54 +0000, Liu Qiang-B32616 wrote:
> > The default will be set in a common interface
> > fsl_sata_set_irq_coalescing when initialize the controller. This
> > interface will check the range of intr count and ticks and make sure
> the values are reasonably.
>
> Allright, but the current defaults are basically no coalescing right ?
>
> > It's hard to find a aggressive value to adapt all scenarios, so I use
> > echo to adjust the value. I remember P5020 have some performance issue,
> I will check it.
> > BTW, which filesystem do you use? Ext2 is lower than ext4 because
> > metadata is continuously wrote to disk. You can try ext4 or xfs.
>
> ext3 at the moment, I plan to switch to ext4 when I finish that fsck pass
> which is taking hours... I am not aware of the 5020 performance issues,
> is this something documented and/or fixable ?
>
My patch may not meet your requirement:(
For your performance requirement, I suggest use ext4 filesystem and SSD.
Thanks.
> Cheers,
> Ben.
>
>
More information about the Linuxppc-dev
mailing list