cxgb4/cxgb4vf: For T5 use Packing and Padding Boundaries for SGE DMA transfers

T5 introduces the ability to have separate Packing and Padding Boundaries
for SGE DMA transfers from the chip to Host Memory. This change set takes
advantage of that to set up a smaller Padding Boundary to conserve PCI Link
and Memory Bandwidth with T5.

Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

3 files changed, 53 insertions, 2 deletions

diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
index a18830d8aa6d..cd5b7896cb67 100644
--- a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
+++ b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c
@@ -2436,6 +2436,7 @@ int t4vf_sge_init(struct adapter *adapter)
 	u32 fl0 = sge_params->sge_fl_buffer_size[0];
 	u32 fl1 = sge_params->sge_fl_buffer_size[1];
 	struct sge *s = &adapter->sge;
+	unsigned int ingpadboundary, ingpackboundary;
 
 	/*
 	 * Start by vetting the basic SGE parameters which have been set up by
@@ -2460,8 +2461,34 @@ int t4vf_sge_init(struct adapter *adapter)
 	s->stat_len = ((sge_params->sge_control & EGRSTATUSPAGESIZE_MASK)
 			? 128 : 64);
 	s->pktshift = PKTSHIFT_GET(sge_params->sge_control);
-	s->fl_align = 1 << (INGPADBOUNDARY_GET(sge_params->sge_control) +
-			    SGE_INGPADBOUNDARY_SHIFT);
+
+	/* T4 uses a single control field to specify both the PCIe Padding and
+	 * Packing Boundary.  T5 introduced the ability to specify these
+	 * separately.  The actual Ingress Packet Data alignment boundary
+	 * within Packed Buffer Mode is the maximum of these two
+	 * specifications.  (Note that it makes no real practical sense to
+	 * have the Pading Boudary be larger than the Packing Boundary but you
+	 * could set the chip up that way and, in fact, legacy T4 code would
+	 * end doing this because it would initialize the Padding Boundary and
+	 * leave the Packing Boundary initialized to 0 (16 bytes).)
+	 */
+	ingpadboundary = 1 << (INGPADBOUNDARY_GET(sge_params->sge_control) +
+			       X_INGPADBOUNDARY_SHIFT);
+	if (is_t4(adapter->params.chip)) {
+		s->fl_align = ingpadboundary;
+	} else {
+		/* T5 has a different interpretation of one of the PCIe Packing
+		 * Boundary values.
+		 */
+		ingpackboundary = INGPACKBOUNDARY_G(sge_params->sge_control2);
+		if (ingpackboundary == INGPACKBOUNDARY_16B_X)
+			ingpackboundary = 16;
+		else
+			ingpackboundary = 1 << (ingpackboundary +
+						INGPACKBOUNDARY_SHIFT_X);
+
+		s->fl_align = max(ingpadboundary, ingpackboundary);
+	}
 
 	/*
 	 * Set up tasklet timers.
diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_common.h b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_common.h
index 95df61dcb4ce..b5c301d9565e 100644
--- a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_common.h
+++ b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_common.h
@@ -134,6 +134,7 @@ struct dev_params {
  */
 struct sge_params {
 	u32 sge_control;		/* padding, boundaries, lengths, etc. */
+	u32 sge_control2;		/* T5: more of the same */
 	u32 sge_host_page_size;		/* RDMA page sizes */
 	u32 sge_queues_per_page;	/* RDMA queues/page */
 	u32 sge_user_mode_limits;	/* limits for BAR2 user mode accesses */
diff --git a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c
index e984fdc48ba2..dc30d2852850 100644
--- a/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c
+++ b/drivers/net/ethernet/chelsio/cxgb4vf/t4vf_hw.c
@@ -468,6 +468,29 @@ int t4vf_get_sge_params(struct adapter *adapter)
 	sge_params->sge_timer_value_2_and_3 = vals[5];
 	sge_params->sge_timer_value_4_and_5 = vals[6];
 
+	/* T4 uses a single control field to specify both the PCIe Padding and
+	 * Packing Boundary.  T5 introduced the ability to specify these
+	 * separately with the Padding Boundary in SGE_CONTROL and and Packing
+	 * Boundary in SGE_CONTROL2.  So for T5 and later we need to grab
+	 * SGE_CONTROL in order to determine how ingress packet data will be
+	 * laid out in Packed Buffer Mode.  Unfortunately, older versions of
+	 * the firmware won't let us retrieve SGE_CONTROL2 so if we get a
+	 * failure grabbing it we throw an error since we can't figure out the
+	 * right value.
+	 */
+	if (!is_t4(adapter->params.chip)) {
+		params[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
+			     FW_PARAMS_PARAM_XYZ(SGE_CONTROL2_A));
+		v = t4vf_query_params(adapter, 1, params, vals);
+		if (v != FW_SUCCESS) {
+			dev_err(adapter->pdev_dev,
+				"Unable to get SGE Control2; "
+				"probably old firmware.\n");
+			return v;
+		}
+		sge_params->sge_control2 = vals[0];
+	}
+
 	params[0] = (FW_PARAMS_MNEM(FW_PARAMS_MNEM_REG) |
 		     FW_PARAMS_PARAM_XYZ(SGE_INGRESS_RX_THRESHOLD));
 	v = t4vf_query_params(adapter, 1, params, vals);