On Fri, 24 Sep 2004, Oliver Neukum wrote: > > > Presumably the state would be the state at power up. > > > > What if, as you suggest, there are other hosts on the bus using the > > device? Aren't they liable to alter the device from its power-up state? > > The same problem is facing probe(). The driver must be able to deal with it.
True. In any case, deciding on an "idle" state is really an issue for each driver to solve for itself; the PM core shouldn't have to worry about it. > > No -- I'm not talking about the memory image; I'm talking about the bootup > > kernel. There may be modules stored in the memory image that aren't yet > > loaded into the bootup kernel when the resume-from-disk operation begins. > > Yes. They will get devices in their virgin state. Statically compiled drivers > will have initialised their devices. The problem is solveable with > #ifdef MODULE Drivers will need to deal with this possibility. It also would be good if they could place a suspended device into an "idle" state without having to resume it first. I still think it's worthwhile trying to avoid resuming the entire system just to write the memory image to a single swap device. > > > A problem for higher layers. Higher layers should checksum media. > > > > I agree, but they don't always do it and it's not always practical. > > If upper layers can't find out, neither can we. But the upper layers don't necessarily know whether the transport is hot-pluggable. If they knew that, and if they were unable to use a media checksum, then they could make some intelligent decision. I guess you're saying that they should be responsible for making the decision in any case. > > > That negates the purpose of suspension. If you unmount everything > > > you are doing a conventional power cycle. > > > > Not true. User programs would remain running across a suspend. > > Then the filesystems they are mmapped from must remain mounted. > A forced unmount would break the connection between memory mappings > and filesystems. Is there no way to unmount filesystems temporarily while preserving the memory mappings? Especially if no user processes are running at the time? > > It seems to me you've just indicated a case (i-scsi) where a tree is > > insufficient, because the network adapters that must be turned on before > > using an i-scsi host aren't among the ancestors of that host. The > > ordering constraints are _more_ strict than can be expressed using a tree. > > Please don't ask me to come up with a better structure. I wasn't going to! Alan Stern ------------------------------------------------------- This SF.Net email is sponsored by: YOU BE THE JUDGE. Be one of 170 Project Admins to receive an Apple iPod Mini FREE for your judgement on who ports your project to Linux PPC the best. Sponsored by IBM. Deadline: Sept. 24. Go here: http://sf.net/ppc_contest.php _______________________________________________ [EMAIL PROTECTED] To unsubscribe, use the last form field at: https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
