Messages in this thread | | | From | "Dilger, Andreas" <> | Subject | Re: [PATCH v2] staging/lustre: lloop depends on BLOCK | Date | Thu, 1 Aug 2013 19:57:22 +0000 |
| |
On 2013/08/01 2:45 AM, "Christoph Hellwig" <hch@infradead.org> wrote:
>On Tue, Jul 30, 2013 at 08:29:51AM +0800, Xiong Zhou wrote: >> From: Xiong Zhou <jencce.kernel@gmail.com> >> >> In the lustre client driver, lloop depends on BLOCK. Add an >> option for this dependence. Last version of this patch makes >> LUSTRE_FS depends on BLOCK. >> Remove unnecessary jdb head files which depends on BLOCK. > >The driver should be removed, a filesystem has no business bringing >its own loop driver.
It provides significant performance improvement for network IO on Lustre. It bypasses DLM locking in Lustre and the VFS layer on the client, copying in the loop driver, and page-by-page IO submission in the normal IO path.
Cheers, Andreas -- Andreas Dilger
Lustre Software Architect Intel High Performance Data Division
| |