Re: Block device request merging - repeated requests.

From: Stephen C. Tweedie (sct@redhat.com)
Date: Wed Feb 02 2000 - 09:56:43 EST


Hi,

On Tue, 01 Feb 2000 11:33:06 +0000, David Woodhouse
<dwmw2@infradead.org> said:

> axboe@suse.de said:
>> You don't need to increase CURRENT->sector, you just need to keep
>> track of it. I still think we should sector to refer to the current
>> cluster, but you can always look at CURRENT->b_rsector and b_blocknr
>> is you want.

> Which? What's the difference? They are documented as "Real buffer
> location on disk" and "block number" respectively, which doesn't
> really help me choose between them.

Ignore the block number: it is only used by the upper layers to find
blocks in the buffer cache. The rsector is what you want. If you have
a remapping layer like raid0 or LVM present, the blocknr will refer to
the block on the logical device, but the rsector will always give you an
index onto the real physical device you are going to do IO on (the LVM
remapping will be included in the calculation of rsector).

--Stephen

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Mon Feb 07 2000 - 21:00:08 EST