Re: [Patch] invalidate range of pages after direct IO write

[Andrew Morton]

Zach Brown <zach.brown@xxxxxxxxxx> wrote:
>
> > I'll make that change and plop the patch into -mm, but we need to think
>  > about the infinite-loop problem..
> 
>  I can try hacking together that macro and auditing pagevec_lookup()
>  callers..

I'd be inclined to hold off on the macro until we actually get the
open-coded stuff right..  Sometimes the page lookup loops take an end+1
argument and sometimes they take an inclusive `end'.  I think.  That might
make it a bit messy.

How's this look?



- Don't look up more pages than we're going to use

- Don't test page->index until we've locked the page

- Check for the cursor wrapping at the end of the mapping.

--- 25/mm/truncate.c~invalidate-range-of-pages-after-direct-io-write-fix	2005-02-03 18:20:22.000000000 -0800
+++ 25-akpm/mm/truncate.c	2005-02-03 18:28:24.627796400 -0800
@@ -264,18 +264,14 @@ int invalidate_inode_pages2_range(struct
 	pagevec_init(&pvec, 0);
 	next = start;
 	while (next <= end &&
-	       !ret && pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE)) {
+	       !ret && pagevec_lookup(&pvec, mapping, next,
+			min(end - next, (pgoff_t)PAGEVEC_SIZE - 1) + 1)) {
 		for (i = 0; !ret && i < pagevec_count(&pvec); i++) {
 			struct page *page = pvec.pages[i];
 			int was_dirty;
 
-			if (page->index > end) {
-				next = page->index;
-				break;
-			}
-
 			lock_page(page);
-			if (page->mapping != mapping) {	/* truncate race? */
+			if (page->mapping != mapping || page->index > end) {
 				unlock_page(page);
 				continue;
 			}
@@ -311,6 +307,8 @@ int invalidate_inode_pages2_range(struct
 		}
 		pagevec_release(&pvec);
 		cond_resched();
+		if (next == 0)
+			break;		/* The pgoff_t wrapped */
 	}
 	return ret;
 }
_

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