On (02/02/15 11:44), Minchan Kim wrote: > > sure, I did think about this. and I actually didn't find any reason not > > to use ->refcount there. if user wants to reset the device, he first > > should umount it to make bdev->bd_holders check happy. and that's where > > IOs will be failed. so it makes sense to switch to ->refcount there, IMHO. > > If we use zram as block device itself(not a fs or swap) and open the > block device as !FMODE_EXCL, bd_holders will be void. > hm. I don't mind to use ->disksize there, but personally I'd maybe prefer to use ->refcount, which just looks less hacky. zram's most common use cases are coming from ram swap device or ram device with fs. so it looks a bit like we care about some corner case here. just my opinion, no objections against ->disksize != 0. I need to check fs/block_dev. can we switch away from ->bd_holders? > Another topic: As I didn't see enough fs/block_dev.c bd_holders in zram > would be mess. I guess we need to study hotplug of device and implement > it for zram reset rather than strange own konb. It should go TODO. :( ok, need to investigate this later. let's land current activities first. -ss -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>