Review needed

[Warner Losh]

On Jun 3, 2014, at 12:22 PM, John-Mark Gurney <jmg at funkthat.com> wrote:

> Warner Losh wrote this message on Mon, Jun 02, 2014 at 10:46 -0600:
>> The code that combines adjacent ranges for BIO_DELETEs to optimize
>> trims to the device assumes the list is sorted. Don't apply the
>> optimization of not sorting the queue when we have SSDs to the
>> delete_queue, since it causes more discard traffic to the drive. While
> 
> I'm puzzled by this statement, if I remove the double negative, I get:
> Do apply the optimization of sorting the queue when we have SSDs to the
> delete_queue, since it (what the new optimizations?) causes more discard
> traffic to the drive.
> 
> Do you mean previously?

Yes, it should reduce traffic because sorting allows ranges to be collapsed. it == the lack of sorting

I’ll tweak the commit message to be less opaque.

>> one could argue that the higher levels should coalesce the trims,
>> that's not done today, so some optimization at this level is needed.
>> 
>> Diffs and a (hopefully) cool review tool: https://phabric.freebsd.org/D142
> 
> Please include a raw diff...  A Phabric is only for FreeBSD committers,
> it takes the rest of the community out of the ability to review.

I thought people w/o an account could at least see the diff. Looks like that’s not the case (and somewhat lame).

You can get the same thing from my hg patch queue:

http://people.freebsd.org/~imp/patch-queue/bio_delete

Warner


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 842 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.freebsd.org/pipermail/freebsd-scsi/attachments/20140603/3545c2f3/attachment.sig>