On Tue, Sep 26, 2017 at 05:25:00PM +0200, Stanislav Brabec wrote: > Karel Zak wrote: > > On Tue, Sep 26, 2017 at 04:14:51PM +0200, Stanislav Brabec wrote: > > >> References: > >> [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/block/loop.c?id=89e4fdecb51cf5535867026274bc97de9480ade5 > > > > This kernel change has been reverted (1e6ec9ea89d30739b9447c1860fcb07fc29f3aef) after > > discussion on LKML. > > No, it was not. The reverted one is the previous implementation from > Hannes Reinecke. It introduced new flag LO_FLAGS_BLOCKSIZE. Ah, I've thought Omar postponed all the stuff for later. I see it in Linus' tree now. Sorry. > This losetup patch already uses the new implementation with the new > ioctl(,LOOP_SET_BLOCK_SIZE,), including syntax changes caused by the new > implementation. (The old implementation did not make possible "losetup > --logical-blocksize=4096 /dev/loop0", the new one does.) Yes, I've seen the patches in LKML, it's definitely better that the original attempt. I guess it's important enough and non-invasive that we can add the patch to -rc2. > Just an additional off topic note: > I reserved 4 characters for the blocksize. I am not sure, whether it is > sufficient, or whether kernel allows much larger logical blocks. (But > actually, the real column width is determined by width of string > "BLOCKSIZE"). The column width is hint only, libsmartcols should be smart enough to accept longer strings there. Thanks (and again sorry)! Karel -- Karel Zak <kzak@xxxxxxxxxx> http://karelzak.blogspot.com -- To unsubscribe from this list: send the line "unsubscribe util-linux" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
