Martin may be overplaying the performance angle. A previous patch took the adapter from 64K to 4MB transaction sizes across the board. This caused Martin's adapter and drive combination to tip-over. We had to scale back to 128KB sized transactions to get stability on his system. All systems handled the 4MB I/O size in our tests, but the tests that were done some time ago were not performed with the latest kernel, which contributed to a change in testing corners.
Future patches associated with the 'new comm' interface will be able to get finer grained performance tuning based on the adapter model rather than the coarse method that currently resides in the more stable kernels. Sincerely -- Mark Salyzyn -----Original Message----- From: Andrew Morton [mailto:[EMAIL PROTECTED] Sent: Friday, July 29, 2005 2:07 PM To: Martin Drab Cc: linux-kernel@vger.kernel.org; Salyzyn, Mark; [EMAIL PROTECTED] Subject: Re: AACRAID failure with 2.6.13-rc1 . . . ah, thanks. A temporary workaround which might affct performance sounds better than a dead box though. Mark, do you think that many systems are likely to be affected this way? Do you think we should do something temporary for 2.6.13? - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
