[patch] use pdflush for unused inode writeback

Andrew Morton (akpm@zip.com.au)
Tue, 09 Apr 2002 23:52:53 -0700


This is pdflush's first application! The writeback of
the unused inodes list by keventd is removed, and a
pdflush thread is dispatched instead.

There is a need for exclusion - to prevent all the
pdflush threads from working against the same request
queue. This is implemented locally. And this is a
problem, because other pdflush threads can be dispatched
to writeback other filesystem objects, and they don't
know that there's already a pdflush thread working that
request queue.

So moving the exclusion into the request queue itself
is on my things-to-do-list. But the code as-is works
OK - under a `dbench 100' load the number of pdflush
instances can grow as high as four or five. Some fine
tuning is needed...

Patch is against 2.5.8-pre3+ratcache+readahead+pageprivate+pdflush

--- 2.5.8-pre3/fs/inode.c~dallocbase-45-pdflush_inodes Tue Apr 9 22:53:11 2002
+++ 2.5.8-pre3-akpm/fs/inode.c Tue Apr 9 22:53:11 2002
@@ -433,7 +433,7 @@ void sync_inodes(void)
}
}

-static void try_to_sync_unused_inodes(void * arg)
+static void try_to_sync_unused_inodes(unsigned long pexclusive)
{
struct super_block * sb;
int nr_inodes = inodes_stat.nr_unused;
@@ -450,10 +450,9 @@ static void try_to_sync_unused_inodes(vo
}
spin_unlock(&sb_lock);
spin_unlock(&inode_lock);
+ clear_bit(0, (unsigned long *)pexclusive);
}

-static struct tq_struct unused_inodes_flush_task;
-
/**
* write_inode_now - write an inode to disk
* @inode: inode to write to disk
@@ -746,8 +745,15 @@ void prune_icache(int goal)
* from here or we're either synchronously dogslow
* or we deadlock with oom.
*/
- if (goal)
- schedule_task(&unused_inodes_flush_task);
+ if (goal) {
+ static unsigned long exclusive;
+
+ if (!test_and_set_bit(0, &exclusive)) {
+ if (pdflush_operation(try_to_sync_unused_inodes,
+ (unsigned long)&exclusive))
+ clear_bit(0, &exclusive);
+ }
+ }
}
/*
* This is called from kswapd when we think we need some
@@ -1173,8 +1179,6 @@ void __init inode_init(unsigned long mem
NULL);
if (!inode_cachep)
panic("cannot create inode slab cache");
-
- unused_inodes_flush_task.routine = try_to_sync_unused_inodes;
}

static inline void do_atime_update(struct inode *inode)

-
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/