Oliver Neukum wrote: > > Hi, > > I think there's a small window during which a low level driver can be > unloaded while sg_open() is executing. > The attached patch should fix it by incrementing the usage counter before > sg_open() might sleep. Oliver, Yes that should be done. I intend to make some patches to sg to strengthen it when premature close()s and device detaches occur. On premature close()s with commands still "inflight", I intend to bump the module count until the response arrives. This will inhibit 'rmmod sg' causing havoc. Hence the module count thread was timely. As for device detaches, they should be permitted at any time, even with fds open and commands in flight. Doug Gilbert - To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to [EMAIL PROTECTED]
