On 2/9/23 16:11, Mikulas Patocka wrote: > Hi > > I discussed this uevent thing with Zdenek and we concluded that it is > sufficient to send just one uevent on resize, not two. > > So, I made a second patch that fixes the uevent-while-suspended bug. It > replaces set_capacity_and_notify with set_capacity (so that no uevent is > set at this point) and it detects capacity change in do_resume - and if > the capacity was changed, it adds "RESIZE=1" to an uevent that is being > sent. > > Zdenek and Peter, please test it. > OK, I've tested this and this works correctly as expected. If we could go with this patch, that would much better than having 2 independent uevents. With just one uevent we don't need any special treatment in udev rules for the extra "RESIZE=1" uevent and also we don't need to process all the udev rules twice uselessly when those 2 uevents (RESIZE="1" and "DM RESUME") always go together anyway. Thanks! -- Peter -- dm-devel mailing list dm-devel@xxxxxxxxxx https://listman.redhat.com/mailman/listinfo/dm-devel
