| To: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: [PATCH 2/5] xfs: use a cursor for bulk AIL insertion |
| From: | Alex Elder <aelder@xxxxxxx> |
| Date: | Thu, 7 Jul 2011 16:26:16 -0500 |
| Cc: | Dave Chinner <david@xxxxxxxxxxxxx>, <xfs@xxxxxxxxxxx> |
| In-reply-to: | <20110704212055.GA24500@xxxxxxxxxxxxx> |
| References: | <1309757260-5484-1-git-send-email-david@xxxxxxxxxxxxx> <1309757260-5484-3-git-send-email-david@xxxxxxxxxxxxx> <20110704212055.GA24500@xxxxxxxxxxxxx> |
| Reply-to: | <aelder@xxxxxxx> |
On Mon, 2011-07-04 at 17:20 -0400, Christoph Hellwig wrote:
> FYI: the following patch implementing my suggested cleanups survived a
> few rounds of xfsqa:
It also eliminates the concern I had about using the value
of "lip" after falling off the front of the list with a
list_for_each_entry_reverse(). A null pointer signifies
no entry has a lower lsn than the one provided, and
that covers both the "empty list" and the "lower than
all other lsn's in the list" cases.
I think you should base your solution on Christoph's
approach, Dave.
-Alex
|
| Previous by Date: | Re: [PATCH 5/5] xfs: add size update tracepoint to IO completion, Alex Elder |
|---|---|
| Next by Date: | Re: xfs_growfs doesn't resize, Keith Keller |
| Previous by Thread: | Re: [PATCH 2/5] xfs: use a cursor for bulk AIL insertion, Christoph Hellwig |
| Next by Thread: | Re: [PATCH 2/5] xfs: use a cursor for bulk AIL insertion, Dave Chinner |
| Indexes: | [Date] [Thread] [Top] [All Lists] |