re-committing on b_devel:

b=1997
r=alex
My previous fix for truncate/write lock inversion was close, but badly
flawed. I failed to remember that truncate will internally restart
the transaction -- start a new one, for all intents and purposes -- so
the ordering was backwards. "i_sem before transaction" is the
cardinal rule.

I tried to avoid that because I didn't want to hold the i_sem across
the entire disk I/O in filter_commitrw_write. After some discussion
with Alex, however, we decided that the i_sem need only be held for
the block allocation, and that the BKL suffices for updating i_size.
So the ordering in the write path is now: take i_sem, start transaction,
call filter_direct_io, f_d_io does block alloc, drops i_sem, does I/O.

diff --git a/lustre/obdfilter/filter_io.c b/lustre/obdfilter/filter_io.c
index a8d77c5..43f16f4 100644 (file)
--- a/lustre/obdfilter/filter_io.c
+++ b/lustre/obdfilter/filter_io.c
@@ -374,7 +374,6 @@ void flip_into_page_cache(struct inode *inode, struct page *new_page)
 {
         struct page *old_page;
         int rc;
-        ENTRY;
 
         do {
                 /* the dlm is protecting us from read/write concurrency, so we
@@ -412,8 +411,6 @@ void flip_into_page_cache(struct inode *inode, struct page *new_page)
                 rc = 0;
 #endif
         } while (rc != 0);
-
-        EXIT;
 }
 
 /* XXX needs to trickle its oa down */