| To: | Dave Chinner <david@xxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: [PATCH 12/16] xfs: implement batched inode lookups for AG walking |
| From: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
| Date: | Wed, 22 Sep 2010 13:33:06 -0400 |
| Cc: | xfs@xxxxxxxxxxx |
| In-reply-to: | <1285137869-10310-13-git-send-email-david@xxxxxxxxxxxxx> |
| References: | <1285137869-10310-1-git-send-email-david@xxxxxxxxxxxxx> <1285137869-10310-13-git-send-email-david@xxxxxxxxxxxxx> |
| User-agent: | Mutt/1.5.20 (2009-08-17) |
> + int (*grab)(struct xfs_inode *ip), I think we can do without this callback. The inode verification that needs to be done is always the same. The only difference is that the dqrele code skips the quota inodes - but this can easily be done in the execute callback, and given that it's a slow path the additional two igrab calls won't hurt either. To be symmetic with that the IRELE call should also be moved from the execute callbacks into the caller. All this is stuff enabled by splitting out the reclaim code earlier which was pretty different in this area. In fact just moving the validation + igrab and IRELE into common code might just be done in a separate patch between the last one and this. |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH 11/16] xfs: split inode AG walking into separate code for reclaim, Christoph Hellwig |
|---|---|
| Next by Date: | Re: [PATCH 13/16] xfs: batch inode reclaim lookup, Christoph Hellwig |
| Previous by Thread: | [PATCH 12/16] xfs: implement batched inode lookups for AG walking, Dave Chinner |
| Next by Thread: | Re: [PATCH 12/16] xfs: implement batched inode lookups for AG walking, Dave Chinner |
| Indexes: | [Date] [Thread] [Top] [All Lists] |