Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-16 Thread J. R. Okajima
Linus Torvalds: > Yes. That would be lovely. And trivial for most filesystems to support. > > Sure, you could have an inode if you need to (not all filesystems may > have a flag in the directory entry), so it would look like "mknod()" > for the filesystem. But the filesystem might decide to never

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-16 Thread J. R. Okajima
Al Viro: > Sure - btrfs happens to have an interesting limit on the number of > links to the same object located in one directory. It doesn't matter. On every filesystem, the link count has its upper limit eventually. When vfs_link() for whiteout returns EMLINK, aufs removes the whiteout-src file

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread Al Viro
On Sat, Mar 16, 2013 at 03:57:18AM +0900, J. R. Okajima wrote: > > Al Viro: > > The trouble with such mechanisms is that they tend to end up depending on > > fairly non-trivial properties of underlying fs. Try aufs one on btrfs, > > see how soon you spot the problem. It's nice when a method turn

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread Erez Zadok
I tend to agree with Al's and Linus's POV regarding whiteouts. There are three general techniques to implementing whiteouts: 1. namespace: special file names, hard/symlinks, or special "hidden" dot files. 2. extended attributes. 3. DT_WHT dirent flags. (there's actually a 4th method I've tried be

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread Linus Torvalds
On Thu, Mar 14, 2013 at 10:09 PM, J. R. Okajima wrote: > > "no inodes at all"? > Are you assuming the implementation in dcache only (with a new d_type > flag)? And it is up to the real fs (layer or branch) whether it consumes > inode or not? Yes. That would be lovely. And trivial for most filesys

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread J. R. Okajima
Al Viro: > The trouble with such mechanisms is that they tend to end up depending on > fairly non-trivial properties of underlying fs. Try aufs one on btrfs, > see how soon you spot the problem. It's nice when a method turns out > to be really redundant and implementable in uniform way via other

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread Al Viro
On Fri, Mar 15, 2013 at 08:15:18AM +, James Bottomley wrote: > It's the fact that a directory entry based whiteout limits the amount of > change to the VFS, but has to be supported by underlying filesystems. > The generic_dirent_fallthrough() mechanism is a nice way of hiding it, > but there a

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-15 Thread James Bottomley
On Fri, 2013-03-15 at 05:13 +, Al Viro wrote: > On Fri, Mar 15, 2013 at 02:09:14PM +0900, J. R. Okajima wrote: > > > If so, it has a big disadvantage for the layer-fs (or branch-fs) to have > > to implement a new method for whiteout. > > > > Overlayfs implements whiteout as symlink+xattr whic

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread Al Viro
On Fri, Mar 15, 2013 at 02:09:14PM +0900, J. R. Okajima wrote: > If so, it has a big disadvantage for the layer-fs (or branch-fs) to have > to implement a new method for whiteout. > > Overlayfs implements whiteout as symlink+xattr which consumes an > inode. And you don't like it, right? > What I

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread J. R. Okajima
Al Viro: > > +- whiteout is hardlinked in order to reduce the consumption of inodes > > + on branch > > *blink* Whiteouts have no inodes at all. Filesystem has an additional > kind of directory entries, recognizable as whiteouts. How they are > done is up to filesystem in question. "no inodes

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread Al Viro
On Fri, Mar 15, 2013 at 01:15:36PM +0900, J. R. Okajima wrote: > +- whiteout is hardlinked in order to reduce the consumption of inodes > + on branch *blink* Whiteouts have no inodes at all. Filesystem has an additional kind of directory entries, recognizable as whiteouts. How they are done is

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread J. R. Okajima
Al Viro: > Umm... I would prefer it to go through vfs.git, with serious modifications. > I really don't like the idea of xattr-based whiteouts, for example. I'd ask you how do you think another whiteout approach? Here is a part of a posts which describes about aufs2 in 2009

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread Al Viro
On Thu, Mar 14, 2013 at 02:43:32PM +0100, Miklos Szeredi wrote: > > Either way, I suggest the next step is to ask Stephen to line this up > > in linux-next. > > Al, are you okay with that? Or do you want it to go though -vfs? Umm... I would prefer it to go through vfs.git, with serious modific

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-14 Thread Miklos Szeredi
On Thu, Mar 14, 2013 at 12:08 AM, Andrew Morton wrote: > On Wed, 13 Mar 2013 15:16:24 +0100 Miklos Szeredi wrote: > >> Here's another version with the comments addressed plus a small bugfix and >> some >> checkpatch cleanups. > > It all looks nice to my rusty eye. But then, I like anything whic

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Andrew Morton
On Wed, 13 Mar 2013 15:16:24 +0100 Miklos Szeredi wrote: > Here's another version with the comments addressed plus a small bugfix and > some > checkpatch cleanups. It all looks nice to my rusty eye. But then, I like anything which has comments and passes checkpatch ;) I see that quite a few p

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Phillip Lougher
On Wed, Mar 13, 2013 at 4:10 PM, Sedat Dilek wrote: > On Wed, Mar 13, 2013 at 4:53 PM, Sedat Dilek wrote: >> On Wed, Mar 13, 2013 at 4:26 PM, Sedat Dilek wrote: >>> On Wed, Mar 13, 2013 at 4:18 PM, Miklos Szeredi wrote: > Looks like this is missing (or intended?): > > diff --git a/f

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 9:13 PM, Phillip Lougher wrote: > > > On Wed, Mar 13, 2013 at 4:10 PM, Sedat Dilek wrote: >> >> On Wed, Mar 13, 2013 at 4:53 PM, Sedat Dilek >> wrote: >> > On Wed, Mar 13, 2013 at 4:26 PM, Sedat Dilek >> > wrote: >> >> On Wed, Mar 13, 2013 at 4:18 PM, Miklos Szeredi >>

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 8:58 PM, Linus Torvalds wrote: > On Wed, Mar 13, 2013 at 12:54 PM, Eric W. Biederman > wrote: >>> >>> Hehe, I just checked my new kernel... that does not work (nothing in the >>> logs). >>> But I think it's good to see if the filesystem is registered/loaded. >> >> lsmod |

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Linus Torvalds
On Wed, Mar 13, 2013 at 12:54 PM, Eric W. Biederman wrote: >> >> Hehe, I just checked my new kernel... that does not work (nothing in the >> logs). >> But I think it's good to see if the filesystem is registered/loaded. > > lsmod | grep overlayfs How about the compiled-in case? What's wrong with

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Eric W. Biederman
Sedat Dilek writes: > On Wed, Mar 13, 2013 at 7:37 PM, Felix Fietkau wrote: >> On 2013-03-13 7:12 PM, Robin Holt wrote: >>> On Wed, Mar 13, 2013 at 05:51:33PM +0100, Sedat Dilek wrote: On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dile

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 7:37 PM, Felix Fietkau wrote: > On 2013-03-13 7:12 PM, Robin Holt wrote: >> On Wed, Mar 13, 2013 at 05:51:33PM +0100, Sedat Dilek wrote: >>> On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: >>> > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek >>> > wrote: >>> > >>> >

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Felix Fietkau
On 2013-03-13 7:12 PM, Robin Holt wrote: > On Wed, Mar 13, 2013 at 05:51:33PM +0100, Sedat Dilek wrote: >> On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: >> > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek wrote: >> > >> >> Anyway, with CONFIG_OVERLAYFS_FS=m I do not see any related message

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Robin Holt
On Wed, Mar 13, 2013 at 05:51:33PM +0100, Sedat Dilek wrote: > On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: > > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek wrote: > > > >> Anyway, with CONFIG_OVERLAYFS_FS=m I do not see any related messages > >> when the kernel-module is loaded. > >> S

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek wrote: > >> Anyway, with CONFIG_OVERLAYFS_FS=m I do not see any related messages >> when the kernel-module is loaded. >> So, is this intended? >> SquashFS prints into the logs, so what is it doin

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 5:21 PM, Miklos Szeredi wrote: > On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek wrote: > >> Anyway, with CONFIG_OVERLAYFS_FS=m I do not see any related messages >> when the kernel-module is loaded. >> So, is this intended? >> SquashFS prints into the logs, so what is it doin

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Miklos Szeredi
On Wed, Mar 13, 2013 at 5:10 PM, Sedat Dilek wrote: > Anyway, with CONFIG_OVERLAYFS_FS=m I do not see any related messages > when the kernel-module is loaded. > So, is this intended? > SquashFS prints into the logs, so what is it doing differently? Some modules do, some don't. It's not compulso

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 4:53 PM, Sedat Dilek wrote: > On Wed, Mar 13, 2013 at 4:26 PM, Sedat Dilek wrote: >> On Wed, Mar 13, 2013 at 4:18 PM, Miklos Szeredi wrote: Looks like this is missing (or intended?): diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index 482c26

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 4:26 PM, Sedat Dilek wrote: > On Wed, Mar 13, 2013 at 4:18 PM, Miklos Szeredi wrote: >>> Looks like this is missing (or intended?): >>> >>> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c >>> index 482c26f..f23ebfc 100644 >>> --- a/fs/overlayfs/super.c >>> +++ b/f

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 4:18 PM, Miklos Szeredi wrote: >> Looks like this is missing (or intended?): >> >> diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c >> index 482c26f..f23ebfc 100644 >> --- a/fs/overlayfs/super.c >> +++ b/fs/overlayfs/super.c >> @@ -684,3 +684,6 @@ static void __exit

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Miklos Szeredi
> Looks like this is missing (or intended?): > > diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c > index 482c26f..f23ebfc 100644 > --- a/fs/overlayfs/super.c > +++ b/fs/overlayfs/super.c > @@ -684,3 +684,6 @@ static void __exit ovl_exit(void) > > module_init(ovl_init); > module_exit(ovl_

Re: [PATCH 0/9] overlay filesystem: request for inclusion (v17)

2013-03-13 Thread Sedat Dilek
On Wed, Mar 13, 2013 at 3:31 PM, Sedat Dilek wrote: > On Wed, Mar 13, 2013 at 3:16 PM, Miklos Szeredi wrote: >> Here's another version with the comments addressed plus a small bugfix and >> some >> checkpatch cleanups. >> >> Changes in v17: >> >> - fix wrong return value in a failure path in ov