Re: [RFC] split kiobuf/buffer-head IO code Re: xfs file system

[Marcelo Tosatti <marcelo@xxxxxxxxxxxxxxxx>]

On Mon, 20 Nov 2000, Chaitanya Tumuluri wrote:

<snip>

> The other reason (they seem to multiply with every email!) I don't want to 
> make too
> many changes currently, is that Eric Youngdale is thinking in terms of a 
> major rewrite
> to the block-queueing layer (including getting rid of the old scsi 
> error-handling code
> and the drivers that break with it!). 

I see. 

> So, perhaps, we should revisit this discussion when 2.5 does open up
> and then rework the queueing layer as you suggest along with everyone
> else that wants to improve it! :^)

Agreed. 

For now, I'll finish the RAID1/5 kiobuf IO support and look a bit more at
the difference between bh based XFS and kiobuf based XFS.

Do you have any results with different types of sequential/random IO on
both IO methods?