Date: Mon, 7 May 2001 08:02:51 +1000 From: "Jan Mikkelsen" <janm@transactionware.com> To: "Jan Conrad" <conrad@th.physik.uni-bonn.de> Cc: "Doug Russell" <drussell@saturn-tech.com>, "Matt Dillon" <dillon@earth.backplane.com>, "freebsd-stable" <freebsd-stable@FreeBSD.ORG> Subject: Re: Tagged Queueing and ATA driver (was soft update should be default) Message-ID: <004f01c0d678$4c2c11b0$0901a8c0@haym.transactionsite.com>
next in thread | raw e-mail | index | archive | help
Jan Conrad <conrad@th.physik.uni-bonn.de> wrote: >Well, if I read the sources correctly, enabling tagged queueuing in the >ata driver *IMPLIES* write caching!!! > >So it's pretty clear why tagged queueing is as good as write cacheing (or >as bad :-) > [...] >What I really do not understand here - maybe it's trivial, as I am not an >expert here - is why does tags imply wc for ATA but not for SCSI? > >The SCSI driver says 'Tagged Queueing Enabled' even on my SCSI disk (IBM >DDYS-T18350N) where WCE is off! Interesting. The real issue is when the notification of a completed write arrives. For the drive to have multiple outstanding writes, it must have a buffer to store the data (the "write cache"). If it tells the host that the write is completed when it actually is completely written to disk, rather than when it hits the buffer in the drive, I don't mind using the write cache memory. The real question is this: Does TCQ notify completion once the write is on the disk, or once the data is in the write cache? The second one doesn't make much sense to me, as I don't see where the win would be over plain write caching. Jan Mikkelsen To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?004f01c0d678$4c2c11b0$0901a8c0>