On Sat Jun 24, 2023 at 2:39 AM CEST, наб wrote: > Quoting myself from #musl: > 01:59:40 hm, I think this was just invented for symmetry with bfree/bavail > 02:00:46 FFS has minfree for space but nothing equivalent for inodes > 02:32:31 (this is mirrored in ext4; > a global grep over DragonFlyBSD and the illumos gate > showed just NFSv3 forwarding from the server; > OpenBSD always sets it to the same thing as f_ffree; > oddly, NetBSD /does/ calculate it differently > for LFS and FFS but due to queued writes or > w/e not because of root reservation; > and as expected a lot of "/* what to put in here? */" > and "// XXX same??") > > Link: https://lore.kernel.org/linux-man/f54kudgblgk643u32tb6at4cd3kkzha6hslahv24szs4raroaz@ogivjbfdaqtb/t/#u > Signed-off-by: Ahelenia Ziemiańska <nabijaczleweli@xxxxxxxxxxxxxxxxxx> > --- > man3/statvfs.3 | 8 ++++++++ > 1 file changed, 8 insertions(+) > > diff --git a/man3/statvfs.3 b/man3/statvfs.3 > index b1f0e7545..272ee5391 100644 > --- a/man3/statvfs.3 > +++ b/man3/statvfs.3 > @@ -227,6 +227,14 @@ .SH NOTES > .BR statvfs () > with the argument > .IR path . > +.PP > +Under Linux, > +.I f_favail > +is always the same as > +.IR f_ffree , > +and there's no way for a file-system to report otherwise. s/file-system/filesystem/ as that's what's used elsewhere. > +This is not an issue, since no filesystems with an i-node > +root reservation exist. s/i-node/inode/ for the same reason. > .SH STANDARDS > POSIX.1-2008. > .SH HISTORY -- Best Regards, Tom Schwindl
