From: Jonathan L. <la...@us...> - 2001-09-10 23:13:43
|
Christoph Hellwig [hc...@ns...] wrote: > On Fri, Sep 07, 2001 at 02:36:06PM -0700, Hanna Linder wrote: > > II. Jonathan Lahr talked about his io_request_lock patch and > > some performance results should come out soon. > > Any news on what this patch is actually supposed to do? It takes an > q->queze_lock in some places where io_request_lock is already held > and sometimes kiils io_request_lock without replacement. > > Doesn't make much sense to me.. The patch changes the i/o subsystem to use q->queue_lock instead of io_request_lock to protect the request queue. The queue_lock is taken in addition to the io_request_lock to preserve existing io_request_lock serialization while also protecting queue access by __make_request. Jonathan -- Jonathan Lahr IBM Linux Technology Center Beaverton, Oregon la...@us... 503-578-3385 |