| To: | Barry Naujok <bnaujok@xxxxxxx> |
|---|---|
| Subject: | Re: REVIEW: fix up xfs_repair I/O size for filesystems with >= 16KB filesystem blocksize |
| From: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
| Date: | Sat, 31 May 2008 04:04:56 -0400 |
| Cc: | "xfs@xxxxxxxxxxx" <xfs@xxxxxxxxxxx> |
| In-reply-to: | <op.ubwjeftu3jf8g2@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> |
| References: | <op.ubwjeftu3jf8g2@xxxxxxxxxxxxxxxxxxxxxxxxxxxx> |
| Sender: | xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.5.17 (2007-11-01) |
On Thu, May 29, 2008 at 04:09:41PM +1000, Barry Naujok wrote: > For 4KB block size filesystems (the default), inode chunks are read and > written > in 8KB "clusters". With a 16KB block size (possible on ia64 boxes), the > inode > chunks are read/written in 16KB I/Os (matches the filesystem block size). > > This patch fixes up the I/Os for filesystems with a blocksize greater than > 8KB. Looks good. |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | [PATCH] attrmulti cleanup, Christoph Hellwig |
|---|---|
| Next by Date: | Re: [PATCH] always set a/c/mtime through ->setattr, Al Viro |
| Previous by Thread: | REVIEW: fix up xfs_repair I/O size for filesystems with >= 16KB filesystem blocksize, Barry Naujok |
| Next by Thread: | TAKE 907752 - XFSQA: Fix reisefs case in 032, Barry Naujok |
| Indexes: | [Date] [Thread] [Top] [All Lists] |