On Aug 9, 2013, Ed Cashin wrote:
> On Aug 8, 2013, at 9:05 PM, Kent Overstreet wrote:
> ...
> > It's in the for-jens branch now.
>
>
> Just examining the patches, I like the way it cleans up the aoe code. I
> had a question about a new BUG added by the for-jens branch the
> read-response handling path of the aoe driver.
The aoe driver in linux-bcache/for-jens commit 4c36c973a8f45 is
passing my tests.
Here is a patch against that branch illustrating my suggestion for
handling bad target responses gracefully.
commit 2c39f50b1ee02e2ac07fd072a883a91713da53cc
Author: Ed Cashin <ecas...@coraid.com>
Date: Tue Aug 13 10:50:28 2013 -0400
aoe: bad AoE responses fail I/O without BUG
Instead of having a BUG when the AoE target does something wrong,
just fail the I/O and log the problem with rate limiting.
diff --git a/drivers/block/aoe/aoecmd.c b/drivers/block/aoe/aoecmd.c
index cacd48e..b9916a6 100644
--- a/drivers/block/aoe/aoecmd.c
+++ b/drivers/block/aoe/aoecmd.c
@@ -1096,7 +1096,6 @@ bvcpy(struct sk_buff *skb, struct bio *bio, struct
bvec_iter iter, long cnt)
int soff = 0;
struct bio_vec bv;
- BUG_ON(cnt > iter.bi_size);
iter.bi_size = cnt;
__bio_for_each_segment(bv, bio, iter, iter) {
@@ -1196,6 +1195,14 @@ noskb: if (buf)
clear_bit(BIO_UPTODATE, &buf->bio->bi_flags);
break;
}
+ if (n > f->iter.bi_size) {
+ pr_err_ratelimited("%s e%ld.%d. bytes=%ld need=%u\n",
+ "aoe: too-large data size in read from",
+ (long) d->aoemajor, d->aoeminor,
+ n, f->iter.bi_size);
+ clear_bit(BIO_UPTODATE, &buf->bio->bi_flags);
+ break;
+ }
bvcpy(skb, f->buf->bio, f->iter, n);
case ATA_CMD_PIO_WRITE:
case ATA_CMD_PIO_WRITE_EXT:
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
