mtd/fs/jffs2 nodemgmt.c,1.91,1.92

David Woodhouse dwmw2 at infradead.org
Wed Jan 22 09:40:29 EST 2003


Update of /home/cvs/mtd/fs/jffs2
In directory phoenix.infradead.org:/tmp/cvs-serv30191

Modified Files:
	nodemgmt.c 
Log Message:
Remove BUG() check on alloc_sem in jffs2_mark_node_obsolete()

Index: nodemgmt.c
===================================================================
RCS file: /home/cvs/mtd/fs/jffs2/nodemgmt.c,v
retrieving revision 1.91
retrieving revision 1.92
diff -u -r1.91 -r1.92
--- nodemgmt.c	21 Jan 2003 18:11:28 -0000	1.91
+++ nodemgmt.c	22 Jan 2003 14:40:26 -0000	1.92
@@ -424,6 +424,10 @@
 		if (c->wbuf_len) {
 			D1(printk(KERN_DEBUG "...and adding to erasable_pending_wbuf_list\n"));
 			list_add_tail(&jeb->list, &c->erasable_pending_wbuf_list);
+#if 0 /* This check was added to allow us to find places where we added nodes to the lists
+	 after dropping the alloc_sem, and it did that just fine. But it also caused us to
+	 lock the alloc_sem in other places, like clear_inode(), when we wouldn't otherwise
+	 have needed to. So I suspect it's outlived its usefulness. Thomas? */
 
 			/* We've changed the rules slightly. After
 			   writing a node you now mustn't drop the
@@ -443,6 +447,7 @@
 				printk(KERN_CRIT "jffs2_mark_node_obsolete() called with wbuf active but alloc_sem not locked!\n");
 				BUG();
 			}
+#endif
 		} else {
 			if (jiffies & 127) {
 				/* Most of the time, we just erase it immediately. Otherwise we





More information about the linux-mtd-cvs mailing list