Re: [PATCH 3] sata: restore sg on setup failure

[Tejun Heo]

 I forgot to restore sg->length on setup failure.  This patch adds it.


Signed-off-by: Tejun Heo <htejun@xxxxxxxxx>

Index: work/drivers/scsi/libata-core.c
===================================================================
--- work.orig/drivers/scsi/libata-core.c	2005-08-07 15:13:20.000000000 +0900
+++ work/drivers/scsi/libata-core.c	2005-08-07 15:15:27.000000000 +0900
@@ -2401,8 +2401,11 @@ static int ata_sg_setup_one(struct ata_q
 
 	dma_address = dma_map_single(ap->host_set->dev, qc->buf_virt,
 				     sg->length, dir);
-	if (dma_mapping_error(dma_address))
+	if (dma_mapping_error(dma_address)) {
+		/* restore sg */
+		sg->length += qc->pad_len;
 		return -1;
+	}
 
 	sg_dma_address(sg) = dma_address;
 	sg_dma_len(sg) = sg->length;
@@ -2473,8 +2476,11 @@ static int ata_sg_setup(struct ata_queue
 
 	dir = qc->dma_dir;
 	n_elem = dma_map_sg(ap->host_set->dev, sg, qc->n_elem, dir);
-	if (n_elem < 1)
+	if (n_elem < 1) {
+		/* restore last sg */
+		lsg->length += qc->pad_len;
 		return -1;
+	}
 
 	DPRINTK("%d sg elements mapped\n", n_elem);
 
-
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/