Date: Fri, 02 Feb 2007 11:57:42 -0700 From: Scott Long <scottl@samsco.org> To: mjacob@freebsd.org Cc: scsi@freebsd.org, Nate Lawson <nate@root.org> Subject: Re: cvs commit: src/sys/cam/scsi scsi_da.c Message-ID: <45C389A6.1080606@samsco.org> In-Reply-To: <20070202080329.L17850@ns1.feral.com> References: <20070123173026.E692416A4CD@hub.freebsd.org> <45B65710.4060607@root.org> <20070123105009.G41619@ns1.feral.com> <45B67401.9070102@samsco.org> <20070201150111.B77236@ns1.feral.com> <45C27965.1010803@samsco.org> <45C2E7DB.30204@root.org> <20070202080329.L17850@ns1.feral.com>
next in thread | previous in thread | raw e-mail | index | archive | help
mjacob@freebsd.org wrote: > >> I think Windows actually never runs SYNC_CACHE unless you select >> "detach device". > > Maybe for pluggable devices, but otherwise Windows uses SYNC_CACHE and > FUA quite freely (and correctly). > > I'm uncomfortable with the notion that there is uncommitted data present > in a device after a close that can be lost due to power lossage (or > unpluggage). From a user application or filesystem point of view, this > is an axiom violation that no OS should ever allow. > > From a silly semantic point of view to get around this, we should still > support and require SYNC_CACHE on close except where devices don't > support it (and any device that hangs on a SYNC_CACHE doesn't support > it- period). The problem is that we don't know if the device will misbehave until it does, and then we don't know if we can reliably recover it. > On detach, devices that still need to have data commited > via an opcode that looks remarkably like SYNC_CACHE can and should have > that happen- with all the infrastructure changes that go along with > allowing devices to be detached (w/o complaint) with a live command. What instigates this problem is that the GEOM layer will open the device, read a few sectors, close it, then do that again a few more times, long before the user tries to mount/unmount it. It's the whole GEOM-taste thing where it tries to essentially auto-probe the storage. When we unconditionally send a SYNC_CACHE in daclose(), the misbehaving device is dead long before the user has a chance to do anything. One hack might be to track if any write command were done while the device was open, and only issue the SYNC_CACHE if so. Since the GEOM tasting will only read, it'll pass this test and avoid the problem. Scott
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?45C389A6.1080606>