]> err.no Git - linux-2.6/commitdiff
[PATCH] cciss: fix XFER_READ/XFER_WRITE in do_cciss_request
authorMike Miller <mike.miller@hp.com>
Mon, 18 Dec 2006 10:00:14 +0000 (11:00 +0100)
committerJens Axboe <jens.axboe@oracle.com>
Mon, 18 Dec 2006 10:00:14 +0000 (11:00 +0100)
This patch fixes a stupid bug. Sometime during the 2tb enhancement I ended up
replacing the macros XFER_READ and XFER_WRITE with h->cciss_read and
h->cciss_write respectively. It seemed to work somehow at least on x86_64 and
ia64. I don't know how. But people started complaining about command timeouts
on older controllers like the 64xx series and only on ia32. This resolves the
issue reproduced in our lab. Please consider this for inclusion.

Thanks,
mikem

Signed-off-by: Mike Miller <mike.miller@hp.com>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
drivers/block/cciss.c

index 67a6d4a9343a660d038cfd4c0bf297a5cac82cf8..9d2ddb209343ac7cd8d91ca5aaf2742df14df17c 100644 (file)
@@ -2492,7 +2492,7 @@ static void do_cciss_request(request_queue_t *q)
        c->Request.Type.Type = TYPE_CMD;        // It is a command.
        c->Request.Type.Attribute = ATTR_SIMPLE;
        c->Request.Type.Direction =
-           (rq_data_dir(creq) == READ) ? h->cciss_read : h->cciss_write;
+           (rq_data_dir(creq) == READ) ? XFER_READ : XFER_WRITE;
        c->Request.Timeout = 0; // Don't time out
        c->Request.CDB[0] =
            (rq_data_dir(creq) == READ) ? h->cciss_read : h->cciss_write;