[lkml]   [2010]   [Oct]   [13]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: [PATCH 15/18] fs: icache remove inode_lock
> +[mandatory]
> + inode_lock is gone, replaced by fine grained locks. See fs/inode.c
> +for details of what locks to replace inode_lock with in order to protect
> +particular things. Most of the time, a filesystem only needs ->i_lock, which
> +protects *all* the inode state and its membership on lists that was
> +previously protected with inode_lock.

Actually in general filesystem don't need to know anything of the inode
locking, I suspect we could just drop this blurb. inode_lock wasn't
exported so the only thing that changed for filesystems is that the
atomic i_count counter was replaced by i_ref.

Maybe replace the above with:

The i_count field in the inode is replaced with i_ref, which is
a regular integer instead of an atomic_t. Filesystems should
not manipulate it directly but use helpers like iref, igrab
and iput.

And btw, Documentation/filesystems/vfs.txt and include/linux/fs.h
still mention i_count, and arch/powerpc/platforms/cell/spufs/file.c
still has a reference to it in code.

> @@ -1261,9 +1234,7 @@ int sync_inode(struct inode *inode, struct writeback_control *wbc)
> {
> int ret;
> - spin_lock(&inode_lock);
> ret = writeback_single_inode(inode, wbc);
> - spin_unlock(&inode_lock);
> return ret;
> }
> EXPORT_SYMBOL(sync_inode);

At this point writeback_single_inode and sync_inode are the same.
I'd just rename writeback_single_inode to sync_inode and kill the

> * Lock orders
> - * inode_lock
> * inode hash bucket lock
> * inode->i_lock
> *
> - * inode_lock
> * sb inode lock
> * inode_lru_lock
> * wb->b_lock


Otherwise looks good,

Reviewed-by: Christoph Hellwig <>

 \ /
  Last update: 2010-10-13 15:45    [W:0.293 / U:0.120 seconds]
©2003-2017 Jasper Spaans. hosted at Digital OceanAdvertise on this site