>> @@ -257,6 +257,8 @@ struct dentry *mount_pseudo_xattr(struct file_system_type *fs_type, char *name, >> + s->s_time_min = TIME64_MIN; >> + s->s_time_max = TIME64_MAX; >> @@ -518,6 +520,8 @@ int simple_fill_super(struct super_block *s, unsigned long magic, >> + s->s_time_min = TIME64_MIN; >> + s->s_time_max = TIME64_MAX; >> @@ -247,6 +247,8 @@ static struct super_block *alloc_super(struct file_system_type *type, int flags, >> + s->s_time_min = TIME64_MIN; >> + s->s_time_max = TIME64_MAX; > > If we always initialize them that way, why does libfs.c code bother > reassign the same values? Initially, I considered the idea of having invalid range initialization for all super blocks: https://patchwork.kernel.org/patch/8488641/ At that point, it made sense to override the initializations in libfs.c. The problem with this is we will have to change all filesystems at the same time. So I decided to change the strategy to include valid limits so that we don't break any filesystems and can first merge the basic infrastructure. But, I was hoping we could go back to invalid range eventually which is why I left the libfs.c initializations in. But, you are right. This is redundant for now and maybe makes sense to leave it out. I will post an update. Thanks, Deepa
