Hello everyone, Btrfs v0.10 is now available for download from: http://oss.oracle.com/projects/btrfs/ Btrfs is still in an early alpha state, and the disk format is not finalized. v0.10 introduces a new disk format, and is not compatible with v0.9. The core of this release is explicit back references for all metadata blocks, data extents, and directory items. These are a crucial building block for future features such as online fsck and migration between devices. The back references are ...

Andrew, This patchset mainly polishes the writeback queuing policies. The main goals are: (1) small files should not be starved by big dirty files (2) sync as fast as possible for not-blocked inodes/pages - don't leave them out; no congestion_wait() in between them (3) avoid busy iowait for blocked inodes - retry them in the next go of s_io(maybe at the next wakeup of pdflush) The role of the queues: s_dirty: park for dirtied_when expiration s_io: park for io submission s_m...

Convert to requeue_io_wait() for case: pages skipped due to locked buffers. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn> --- fs/fs-writeback.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) --- linux-mm.orig/fs/fs-writeback.c +++ linux-mm/fs/fs-writeback.c @@ -456,7 +456,7 @@ int generic_sync_sb_inodes(struct super_ * writeback is not making progress due to locked ...

Make the if-else straight in __sync_single_inode(). No behavior change. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn> --- fs/fs-writeback.c | 15 +++++++-------- 1 files changed, 7 insertions(+), 8 deletions(-) --- linux-mm.orig/fs/fs-writeback.c +++ linux-mm/fs/fs-writeback.c @@ -254,8 +254,13 @@ __sync_single_inode(struct inode *inode, spin_lock(&inode_lock); inode->i_state ...

Introduce writeback_control.more_io to indicate that more I/O is scheduled for this wakeup of pdflush. Note that more_io is only updated on the _visited_ superblocks, which prevents pdflush deamons from interfering with one another. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: ...

Introduce super_block.s_more_io_wait to park inodes that for some reason cannot be synced immediately. They will be revisited in the next s_io enqueue time(<=5s). The new data flow after this patchset: s_dirty --> s_io --> s_more_io/s_more_io_wait --+ ^ | | | +----------------------------------+ - to fill s_io: s_more_io + s_dirty(expired) + s_more_io_wait ---> s_io - to drain s...

Replace redirty_tail() with queue_dirty() on memory backed bdi. It makes no difference - only simpler. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn> --- fs/fs-writeback.c | 2 +- 1 files changed, 1 insertion(+), 1 deletion(-) --- linux-mm.orig/fs/fs-writeback.c +++ linux-mm/fs/fs-writeback.c @@ -407,7 +407,7 @@ int generic_sync_sb_inodes(struct super_ int err; if (!bdi_cap_writ...

Merge duplicate code from background_writeout() and wb_kupdate() into writeback_some_pages(). The pages_skipped in background_writeout() is ignored. The inode cannot be written now will be retried in the next run of pdflush, typically in 5s. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Fengguang Wu <wfg@mail.ustc.edu.cn> --- mm/page-writeback.c | 43 +++++++++++++++++++++--------------------- 1 files changed, 22 insertions(...

Convert to requeue_io_wait() for case: - kupdate cannot write all pages due to some blocking condition; - during sync, a file is being written to too fast, starving other files. In the case of sync, requeue_io_wait() can break the starvation because the inode requeued into s_more_io_wait will be served _after_ normal inodes, hence won't stand in the way of other inodes in the next run. Cc: Michael Rubin <mrubin@google.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-...

Seems my ext3 version doesn't work, it craps out in ext3_get_blocks_handle() triggering this bug: J_ASSERT(handle != NULL || create == 0); I'll see if I can fix that, being fairly fs ignorant... -- Jens Axboe -

On Tue, 15 Jan 2008 11:07:40 +0100 You can use DIO_CREDITS instead of len >> 9, just like the ext3 O_DIRECT code does. Your current patch is fine, except it breaks data=ordered rules. My plan to work within data=ordered: 1) Inside ext3_map_extent (while the transaction was running), increment a counter in the ext3 journal for number of pending IOs. Then end the transaction handle. 2) Drop this counter inside the IO completion call 3) Change the ext3 commit code to wait for the IO ...

Ok, should something like this be added to the documentation? It would be cool to be able to include few examples (modern SATA disks support bariers so are safe, any IDE from 1989 is unsafe), but I do not know enough about hw... Signed-off-by: Pavel Machek <pavel@suse.cz> diff --git a/Documentation/filesystems/ext3.txt b/Documentation/filesystems/ext3.txt index b45f3c1..adfcc9d 100644 --- a/Documentation/filesystems/ext3.txt +++ b/Documentation/filesystems/ext3.txt @@ -183,6 +183,18 @...

Nice FAQ, yep. Perhaps you should move parts of it to Documentation/ , and I could then make ext3 FAQ point to it? I had write cache enabled on my main computer. Oops. I guess that means we do need better documentation. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html -

Ah yes, this is indeed confusing. Last time dup_mnt_ns() returned a namespace pointer or NULL. But now I see it returns an ERR_PTR(error) instead, which means it's cleaner to just propagate the error value. I'll fix this. Thanks, Miklos -

Hmm, Part of me wants to say the safest thing for now would be to refuse mounts propagation from non-user mounts to user mounts. I assume you're thinking about a fully user-mounted chroot, where the user woudl still want to be able to stick in a cdrom and have it automounted under /mnt/cdrom, propagated from the root mounts ns? But then are there no devices which the user could create on a floppy while inserted into his own laptop, owned by his own uid, then insert into this machine, and use...

Yeah, of course, what I'm saying is no different whether the upper mount And really this shouldn't be an issue at all - the usermount chroot would be set up under something like /share/hallyn/root, so the admin would have to purposely set up propagation into that tree, so this Thanks for humoring me, -serge -

That sounds good. -serge -

Why not "nosubmnt"? -- A. C. Censi accensi [em] gmail [ponto] com accensi [em] montreal [ponto] com [ponto] br -

Well it really should have 'user' or 'unpriv' in the name somewhere. 'nosubmnt' is more confusing than 'nomnt' because it no submounts really sounds like a reasonable thing in itself... But I never win naming arguments, so I accept that I have poor naming judgement :) -serge -

Something like the loop above is not going to go in for sure. Once we get rid of the sb->s_files we can put the list_head in struct file to new use eventually if we don't want to get rid of it. E.g. and per-inode list would be much better than the per-superblock one and would regularize what the tty driver is doing. But I'm not too interesting in hashing out these details currently, my primary concern is to get the per-mount r/o plus fallout like the correct remount r/o and file_list_lock re...

OK, thanks. Hopefully we will get back to this--it'll be nice to finally make progress on it. --b. -

If we did those two steps separately, would the result be two simpler The return value is the reverse of what I'd naively expect--I don't expect something named flock_find_conflict to return something exactly when conflict is *not* found, but I don't see a better way to do it. Perhaps there's a better name? -

Makes sense, thanks. So the assumption we were depending on the BKL for was that we could count on the wake-up not coming till after we block, so we could skip a check ->fl_next that's normally needed to resolve the usual sleeping-on-some-condition race? -

OK, thanks, applied just with the "few assumptions" replaced by a description of that particular problem: "interruptible_sleep_on_locked() is just an open-coded wait_event_interruptible_timeout(), with the one difference that interruptible_sleep_on_locked() doesn't bother to check the condition on which it waits, depending instead on the BKL to avoid the case where it blocks after the wakeup has already been called. locks_block_on_timeout() is only used in one place, so it's actually ...