On Fri, Jul 13, 2012 at 12:05 AM, Pekka Enberg <penberg@xxxxxxxxxx> wrote: > On Thu, Jul 12, 2012 at 6:18 PM, Asias He <asias.hejun@xxxxxxxxx> wrote: >> With Direct I/O, file reads and writes go directly from the applications >> to the storage device, bypassing the operating system read and write >> caches. This is useful for applications that manage their own caches. >> >> Open a disk image with O_DIRECT: >> $ lkvm run -d ~/img/test.img,direct >> >> The original readonly flag is still supported. >> Open a disk image with O_DIRECT and readonly: >> $ lkvm run -d ~/img/test.img,direct,ro >> >> Signed-off-by: Asias He <asias.hejun@xxxxxxxxx> > > Well, what's the use case for this? With O_DIRECT, host page cache will be skipped. It saves the copying of data between page cache and user space, thus reduces cpu utilization and memory usage. This might be useful for users with high-end storage. It's also useful when we benchmark the performance of the storage stack in kvm tool side because it excludes one factor we are not interested in. -- Asias He -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
