[PATCH 4.9 065/130] ibmvscsis: Fix max transfer length

[Greg Kroah-Hartman]

4.9-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Bryant G. Ly <bryantly@xxxxxxxxxxxxxxxxxx>

commit 387b978cb0d12cf3720ecb17e652e0a9991a08e2 upstream.

Current code incorrectly calculates the max transfer length, since
it is assuming a 4k page table, but ppc64 all run on 64k page tables.

Reported-by: Steven Royer <seroyer@xxxxxxxxxxxxxxxxxx>
Tested-by: Steven Royer <seroyer@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Bryant G. Ly <bryantly@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Bart Van Assche <bart.vanassche@xxxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

---
 drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

--- a/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
+++ b/drivers/scsi/ibmvscsi_tgt/ibmvscsi_tgt.c
@@ -45,6 +45,7 @@
 
 #define	INITIAL_SRP_LIMIT	800
 #define	DEFAULT_MAX_SECTORS	256
+#define MAX_TXU			1024 * 1024
 
 static uint max_vdma_size = MAX_H_COPY_RDMA;
 
@@ -1291,7 +1292,7 @@ static long ibmvscsis_adapter_info(struc
 	info->mad_version = cpu_to_be32(MAD_VERSION_1);
 	info->os_type = cpu_to_be32(LINUX);
 	memset(&info->port_max_txu[0], 0, sizeof(info->port_max_txu));
-	info->port_max_txu[0] = cpu_to_be32(128 * PAGE_SIZE);
+	info->port_max_txu[0] = cpu_to_be32(MAX_TXU);
 
 	dma_wmb();
 	rc = h_copy_rdma(sizeof(*info), vscsi->dds.window[LOCAL].liobn,