Messages in this thread | | | Date | Tue, 20 Feb 2007 12:03:35 -0500 | From | Chris Mason <> | Subject | Re: [PATCH 2/2] aio: propogate post-EIOCBQUEUED errors to completion event |
| |
On Tue, Feb 20, 2007 at 07:57:49PM +0300, Ananiev, Leonid I wrote: > Zach> This addresses an oops reported by Leonid Ananiev > <leonid.i.ananiev@intel.com> > Zach> as archived at http://lkml.org/lkml/2007/2/8/337. > .... > Zach> This was tested by running O_DIRECT aio-stress concurrently with > buffered reads > > The oops was with aio-stress only running in the loop > WITHOUT buffered or mmaped IO which are patched and discussed now. > Actually 47% aio is finished with EIO after patch.
I looked through the thread and couldn't find the aio-stress command line used for the whole test. Could you please post it here?
If aio+dio is being used to extend the file, you'll get pages in the page cache, which is hopefully why you're able to trigger the EIOs.
-chris - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |