From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Zeuthen Date: Sat, 12 Jun 2004 12:21:27 +0000 Subject: Re: How to detect whether media is present in a reader? Message-Id: <1087042887.6345.12.camel@ixus.fubar.dk> List-Id: References: <200406101053.33097.michael@gentoo.co.nz> In-Reply-To: <200406101053.33097.michael@gentoo.co.nz> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: linux-hotplug@vger.kernel.org On Sat, 2004-06-12 at 23:39 +1200, Michael Hamilton wrote: > Thanks for the pointers. However it has just occurred to me - > and please correct me if I'm wrong - that frequently polling flash > media might prematurely wear it out. It is my understanding that flash media (NOR and NAND flash) only wears out when doing erase/write operations, think of a lot of embedded devices using flash. But I'm not sure. Second, all that the polling for media entails is a readonly nonblocking open(2) on the top level block device, e.g. /dev/sda. Whether this means that the device actually accesses the flash memory or it's handled in some logic another place, I'm not sure either. I think Greg's point was that the cheap devices doesn't give the kernel a signal when media is inserted/removed, not that it was impossible to poll for it - and I guess polling is frowned upon in the kernel. But now I'm just speculating; I don't know really. > I guess the media should > only be polled when the user tries to access it - something like > the approach taken by submount - so I'm probably going to look > at virtual file systems (maybe using the fuse library). > There's something called volumagic that goes a step further and mounts/ unmounts devices based on whether the device is accessed - btw, it is all implemented in user space through a NFS server. This solves the problems of user yanking out a USB device and locking of optical drives. IMHO, It would be good, for desktop Linux at least, if functionality like this was implemented in the kernel without changing the interface that userspace has today, but it doesn't seem like it's going to happy anytime soon. Oh well. Cheers, David ------------------------------------------------------- This SF.Net email is sponsored by the new InstallShield X. >From Windows to Linux, servers to mobile, InstallShield X is the one installation-authoring solution that does it all. Learn more and evaluate today! http://www.installshield.com/Dev2Dev/0504 _______________________________________________ Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net Linux-hotplug-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel