1 files changed, 239 insertions, 71 deletions

diff --git a/drivers/spi/spi-tegra114.c b/drivers/spi/spi-tegra114.c
index a76acedd7e2f..b1f31bb16659 100644
--- a/drivers/spi/spi-tegra114.c
+++ b/drivers/spi/spi-tegra114.c
@@ -149,6 +149,8 @@
 
 #define SPI_TX_FIFO				0x108
 #define SPI_RX_FIFO				0x188
+#define SPI_INTR_MASK				0x18c
+#define SPI_INTR_ALL_MASK			(0x1fUL << 25)
 #define MAX_CHIP_SELECT				4
 #define SPI_FIFO_DEPTH				64
 #define DATA_DIR_TX				(1 << 0)
@@ -161,6 +163,10 @@
 #define MAX_HOLD_CYCLES				16
 #define SPI_DEFAULT_SPEED			25000000
 
+struct tegra_spi_soc_data {
+	bool has_intr_mask_reg;
+};
+
 struct tegra_spi_data {
 	struct device				*dev;
 	struct spi_master			*master;
@@ -211,6 +217,7 @@ struct tegra_spi_data {
 	u32					*tx_dma_buf;
 	dma_addr_t				tx_dma_phys;
 	struct dma_async_tx_descriptor		*tx_dma_desc;
+	const struct tegra_spi_soc_data		*soc_data;
 };
 
 static int tegra_spi_runtime_suspend(struct device *dev);
@@ -259,7 +266,8 @@ static unsigned tegra_spi_calculate_curr_xfer_param(
 
 	tspi->bytes_per_word = DIV_ROUND_UP(bits_per_word, 8);
 
-	if (bits_per_word == 8 || bits_per_word == 16) {
+	if ((bits_per_word == 8 || bits_per_word == 16 ||
+	     bits_per_word == 32) && t->len > 3) {
 		tspi->is_packed = 1;
 		tspi->words_per_32bit = 32/bits_per_word;
 	} else {
@@ -307,10 +315,16 @@ static unsigned tegra_spi_fill_tx_fifo_from_client_txbuf(
 				x |= (u32)(*tx_buf++) << (i * 8);
 			tegra_spi_writel(tspi, x, SPI_TX_FIFO);
 		}
+
+		tspi->cur_tx_pos += written_words * tspi->bytes_per_word;
 	} else {
+		unsigned int write_bytes;
 		max_n_32bit = min(tspi->curr_dma_words,  tx_empty_count);
 		written_words = max_n_32bit;
 		nbytes = written_words * tspi->bytes_per_word;
+		if (nbytes > t->len - tspi->cur_pos)
+			nbytes = t->len - tspi->cur_pos;
+		write_bytes = nbytes;
 		for (count = 0; count < max_n_32bit; count++) {
 			u32 x = 0;
 
@@ -319,8 +333,10 @@ static unsigned tegra_spi_fill_tx_fifo_from_client_txbuf(
 				x |= (u32)(*tx_buf++) << (i * 8);
 			tegra_spi_writel(tspi, x, SPI_TX_FIFO);
 		}
+
+		tspi->cur_tx_pos += write_bytes;
 	}
-	tspi->cur_tx_pos += written_words * tspi->bytes_per_word;
+
 	return written_words;
 }
 
@@ -344,20 +360,27 @@ static unsigned int tegra_spi_read_rx_fifo_to_client_rxbuf(
 			for (i = 0; len && (i < 4); i++, len--)
 				*rx_buf++ = (x >> i*8) & 0xFF;
 		}
-		tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 		read_words += tspi->curr_dma_words;
+		tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 	} else {
 		u32 rx_mask = ((u32)1 << t->bits_per_word) - 1;
+		u8 bytes_per_word = tspi->bytes_per_word;
+		unsigned int read_bytes;
 
+		len = rx_full_count * bytes_per_word;
+		if (len > t->len - tspi->cur_pos)
+			len = t->len - tspi->cur_pos;
+		read_bytes = len;
 		for (count = 0; count < rx_full_count; count++) {
 			u32 x = tegra_spi_readl(tspi, SPI_RX_FIFO) & rx_mask;
 
-			for (i = 0; (i < tspi->bytes_per_word); i++)
+			for (i = 0; len && (i < bytes_per_word); i++, len--)
 				*rx_buf++ = (x >> (i*8)) & 0xFF;
 		}
-		tspi->cur_rx_pos += rx_full_count * tspi->bytes_per_word;
 		read_words += rx_full_count;
+		tspi->cur_rx_pos += read_bytes;
 	}
+
 	return read_words;
 }
 
@@ -372,12 +395,17 @@ static void tegra_spi_copy_client_txbuf_to_spi_txbuf(
 		unsigned len = tspi->curr_dma_words * tspi->bytes_per_word;
 
 		memcpy(tspi->tx_dma_buf, t->tx_buf + tspi->cur_pos, len);
+		tspi->cur_tx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 	} else {
 		unsigned int i;
 		unsigned int count;
 		u8 *tx_buf = (u8 *)t->tx_buf + tspi->cur_tx_pos;
 		unsigned consume = tspi->curr_dma_words * tspi->bytes_per_word;
+		unsigned int write_bytes;
 
+		if (consume > t->len - tspi->cur_pos)
+			consume = t->len - tspi->cur_pos;
+		write_bytes = consume;
 		for (count = 0; count < tspi->curr_dma_words; count++) {
 			u32 x = 0;
 
@@ -386,8 +414,9 @@ static void tegra_spi_copy_client_txbuf_to_spi_txbuf(
 				x |= (u32)(*tx_buf++) << (i * 8);
 			tspi->tx_dma_buf[count] = x;
 		}
+
+		tspi->cur_tx_pos += write_bytes;
 	}
-	tspi->cur_tx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 
 	/* Make the dma buffer to read by dma */
 	dma_sync_single_for_device(tspi->dev, tspi->tx_dma_phys,
@@ -405,20 +434,28 @@ static void tegra_spi_copy_spi_rxbuf_to_client_rxbuf(
 		unsigned len = tspi->curr_dma_words * tspi->bytes_per_word;
 
 		memcpy(t->rx_buf + tspi->cur_rx_pos, tspi->rx_dma_buf, len);
+		tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 	} else {
 		unsigned int i;
 		unsigned int count;
 		unsigned char *rx_buf = t->rx_buf + tspi->cur_rx_pos;
 		u32 rx_mask = ((u32)1 << t->bits_per_word) - 1;
+		unsigned consume = tspi->curr_dma_words * tspi->bytes_per_word;
+		unsigned int read_bytes;
 
+		if (consume > t->len - tspi->cur_pos)
+			consume = t->len - tspi->cur_pos;
+		read_bytes = consume;
 		for (count = 0; count < tspi->curr_dma_words; count++) {
 			u32 x = tspi->rx_dma_buf[count] & rx_mask;
 
-			for (i = 0; (i < tspi->bytes_per_word); i++)
+			for (i = 0; consume && (i < tspi->bytes_per_word);
+							i++, consume--)
 				*rx_buf++ = (x >> (i*8)) & 0xFF;
 		}
+
+		tspi->cur_rx_pos += read_bytes;
 	}
-	tspi->cur_rx_pos += tspi->curr_dma_words * tspi->bytes_per_word;
 
 	/* Make the dma buffer to read by dma */
 	dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
@@ -470,22 +507,39 @@ static int tegra_spi_start_rx_dma(struct tegra_spi_data *tspi, int len)
 	return 0;
 }
 
-static int tegra_spi_start_dma_based_transfer(
-		struct tegra_spi_data *tspi, struct spi_transfer *t)
+static int tegra_spi_flush_fifos(struct tegra_spi_data *tspi)
 {
-	u32 val;
-	unsigned int len;
-	int ret = 0;
+	unsigned long timeout = jiffies + HZ;
 	u32 status;
 
-	/* Make sure that Rx and Tx fifo are empty */
 	status = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
 	if ((status & SPI_FIFO_EMPTY) != SPI_FIFO_EMPTY) {
-		dev_err(tspi->dev, "Rx/Tx fifo are not empty status 0x%08x\n",
-			(unsigned)status);
-		return -EIO;
+		status |= SPI_RX_FIFO_FLUSH | SPI_TX_FIFO_FLUSH;
+		tegra_spi_writel(tspi, status, SPI_FIFO_STATUS);
+		while ((status & SPI_FIFO_EMPTY) != SPI_FIFO_EMPTY) {
+			status = tegra_spi_readl(tspi, SPI_FIFO_STATUS);
+			if (time_after(jiffies, timeout)) {
+				dev_err(tspi->dev,
+					"timeout waiting for fifo flush\n");
+				return -EIO;
+			}
+
+			udelay(1);
+		}
 	}
 
+	return 0;
+}
+
+static int tegra_spi_start_dma_based_transfer(
+		struct tegra_spi_data *tspi, struct spi_transfer *t)
+{
+	u32 val;
+	unsigned int len;
+	int ret = 0;
+	u8 dma_burst;
+	struct dma_slave_config dma_sconfig = {0};
+
 	val = SPI_DMA_BLK_SET(tspi->curr_dma_words - 1);
 	tegra_spi_writel(tspi, val, SPI_DMA_BLK);
 
@@ -496,23 +550,40 @@ static int tegra_spi_start_dma_based_transfer(
 		len = tspi->curr_dma_words * 4;
 
 	/* Set attention level based on length of transfer */
-	if (len & 0xF)
+	if (len & 0xF) {
 		val |= SPI_TX_TRIG_1 | SPI_RX_TRIG_1;
-	else if (((len) >> 4) & 0x1)
+		dma_burst = 1;
+	} else if (((len) >> 4) & 0x1) {
 		val |= SPI_TX_TRIG_4 | SPI_RX_TRIG_4;
-	else
+		dma_burst = 4;
+	} else {
 		val |= SPI_TX_TRIG_8 | SPI_RX_TRIG_8;
+		dma_burst = 8;
+	}
 
-	if (tspi->cur_direction & DATA_DIR_TX)
-		val |= SPI_IE_TX;
+	if (!tspi->soc_data->has_intr_mask_reg) {
+		if (tspi->cur_direction & DATA_DIR_TX)
+			val |= SPI_IE_TX;
 
-	if (tspi->cur_direction & DATA_DIR_RX)
-		val |= SPI_IE_RX;
+		if (tspi->cur_direction & DATA_DIR_RX)
+			val |= SPI_IE_RX;
+	}
 
 	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
 	tspi->dma_control_reg = val;
 
+	dma_sconfig.device_fc = true;
 	if (tspi->cur_direction & DATA_DIR_TX) {
+		dma_sconfig.dst_addr = tspi->phys + SPI_TX_FIFO;
+		dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.dst_maxburst = dma_burst;
+		ret = dmaengine_slave_config(tspi->tx_dma_chan, &dma_sconfig);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"DMA slave config failed: %d\n", ret);
+			return ret;
+		}
+
 		tegra_spi_copy_client_txbuf_to_spi_txbuf(tspi, t);
 		ret = tegra_spi_start_tx_dma(tspi, len);
 		if (ret < 0) {
@@ -523,6 +594,16 @@ static int tegra_spi_start_dma_based_transfer(
 	}
 
 	if (tspi->cur_direction & DATA_DIR_RX) {
+		dma_sconfig.src_addr = tspi->phys + SPI_RX_FIFO;
+		dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
+		dma_sconfig.src_maxburst = dma_burst;
+		ret = dmaengine_slave_config(tspi->rx_dma_chan, &dma_sconfig);
+		if (ret < 0) {
+			dev_err(tspi->dev,
+				"DMA slave config failed: %d\n", ret);
+			return ret;
+		}
+
 		/* Make the dma buffer to read by dma */
 		dma_sync_single_for_device(tspi->dev, tspi->rx_dma_phys,
 				tspi->dma_buf_size, DMA_FROM_DEVICE);
@@ -570,8 +651,9 @@ static int tegra_spi_start_cpu_based_transfer(
 
 	tspi->is_curr_dma_xfer = false;
 
-	val |= SPI_DMA_EN;
-	tegra_spi_writel(tspi, val, SPI_DMA_CTL);
+	val = tspi->command1_reg;
+	val |= SPI_PIO;
+	tegra_spi_writel(tspi, val, SPI_COMMAND1);
 	return 0;
 }
 
@@ -582,7 +664,6 @@ static int tegra_spi_init_dma_param(struct tegra_spi_data *tspi,
 	u32 *dma_buf;
 	dma_addr_t dma_phys;
 	int ret;
-	struct dma_slave_config dma_sconfig;
 
 	dma_chan = dma_request_slave_channel_reason(tspi->dev,
 					dma_to_memory ? "rx" : "tx");
@@ -603,19 +684,6 @@ static int tegra_spi_init_dma_param(struct tegra_spi_data *tspi,
 	}
 
 	if (dma_to_memory) {
-		dma_sconfig.src_addr = tspi->phys + SPI_RX_FIFO;
-		dma_sconfig.src_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
-		dma_sconfig.src_maxburst = 0;
-	} else {
-		dma_sconfig.dst_addr = tspi->phys + SPI_TX_FIFO;
-		dma_sconfig.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES;
-		dma_sconfig.dst_maxburst = 0;
-	}
-
-	ret = dmaengine_slave_config(dma_chan, &dma_sconfig);
-	if (ret)
-		goto scrub;
-	if (dma_to_memory) {
 		tspi->rx_dma_chan = dma_chan;
 		tspi->rx_dma_buf = dma_buf;
 		tspi->rx_dma_phys = dma_phys;
@@ -625,11 +693,6 @@ static int tegra_spi_init_dma_param(struct tegra_spi_data *tspi,
 		tspi->tx_dma_phys = dma_phys;
 	}
 	return 0;
-
-scrub:
-	dma_free_coherent(tspi->dev, tspi->dma_buf_size, dma_buf, dma_phys);
-	dma_release_channel(dma_chan);
-	return ret;
 }
 
 static void tegra_spi_deinit_dma_param(struct tegra_spi_data *tspi,
@@ -696,6 +759,16 @@ static u32 tegra_spi_setup_transfer_one(struct spi_device *spi,
 		else if (req_mode == SPI_MODE_3)
 			command1 |= SPI_CONTROL_MODE_3;
 
+		if (spi->mode & SPI_LSB_FIRST)
+			command1 |= SPI_LSBIT_FE;
+		else
+			command1 &= ~SPI_LSBIT_FE;
+
+		if (spi->mode & SPI_3WIRE)
+			command1 |= SPI_BIDIROE;
+		else
+			command1 &= ~SPI_BIDIROE;
+
 		if (tspi->cs_control) {
 			if (tspi->cs_control != spi)
 				tegra_spi_writel(tspi, command1, SPI_COMMAND1);
@@ -728,8 +801,15 @@ static int tegra_spi_start_transfer_one(struct spi_device *spi,
 
 	total_fifo_words = tegra_spi_calculate_curr_xfer_param(spi, tspi, t);
 
+	if (t->rx_nbits == SPI_NBITS_DUAL || t->tx_nbits == SPI_NBITS_DUAL)
+		command1 |= SPI_BOTH_EN_BIT;
+	else
+		command1 &= ~SPI_BOTH_EN_BIT;
+
 	if (tspi->is_packed)
 		command1 |= SPI_PACKED;
+	else
+		command1 &= ~SPI_PACKED;
 
 	command1 &= ~(SPI_CS_SEL_MASK | SPI_TX_EN | SPI_RX_EN);
 	tspi->cur_direction = 0;
@@ -748,6 +828,9 @@ static int tegra_spi_start_transfer_one(struct spi_device *spi,
 	dev_dbg(tspi->dev, "The def 0x%x and written 0x%x\n",
 		tspi->def_command1_reg, (unsigned)command1);
 
+	ret = tegra_spi_flush_fifos(tspi);
+	if (ret < 0)
+		return ret;
 	if (total_fifo_words > SPI_FIFO_DEPTH)
 		ret = tegra_spi_start_dma_based_transfer(tspi, t);
 	else
@@ -774,6 +857,12 @@ static int tegra_spi_setup(struct spi_device *spi)
 		return ret;
 	}
 
+	if (tspi->soc_data->has_intr_mask_reg) {
+		val = tegra_spi_readl(tspi, SPI_INTR_MASK);
+		val &= ~SPI_INTR_ALL_MASK;
+		tegra_spi_writel(tspi, val, SPI_INTR_MASK);
+	}
+
 	spin_lock_irqsave(&tspi->lock, flags);
 	val = tspi->def_command1_reg;
 	if (spi->mode & SPI_CS_HIGH)
@@ -799,6 +888,33 @@ static void tegra_spi_transfer_delay(int delay)
 	udelay(delay % 1000);
 }
 
+static void tegra_spi_transfer_end(struct spi_device *spi)
+{
+	struct tegra_spi_data *tspi = spi_master_get_devdata(spi->master);
+	int cs_val = (spi->mode & SPI_CS_HIGH) ? 0 : 1;
+
+	if (cs_val)
+		tspi->command1_reg |= SPI_CS_SW_VAL;
+	else
+		tspi->command1_reg &= ~SPI_CS_SW_VAL;
+	tegra_spi_writel(tspi, tspi->command1_reg, SPI_COMMAND1);
+	tegra_spi_writel(tspi, tspi->def_command1_reg, SPI_COMMAND1);
+}
+
+static void tegra_spi_dump_regs(struct tegra_spi_data *tspi)
+{
+	dev_dbg(tspi->dev, "============ SPI REGISTER DUMP ============\n");
+	dev_dbg(tspi->dev, "Command1:    0x%08x | Command2:    0x%08x\n",
+		tegra_spi_readl(tspi, SPI_COMMAND1),
+		tegra_spi_readl(tspi, SPI_COMMAND2));
+	dev_dbg(tspi->dev, "DMA_CTL:     0x%08x | DMA_BLK:     0x%08x\n",
+		tegra_spi_readl(tspi, SPI_DMA_CTL),
+		tegra_spi_readl(tspi, SPI_DMA_BLK));
+	dev_dbg(tspi->dev, "TRANS_STAT:  0x%08x | FIFO_STATUS: 0x%08x\n",
+		tegra_spi_readl(tspi, SPI_TRANS_STATUS),
+		tegra_spi_readl(tspi, SPI_FIFO_STATUS));
+}
+
 static int tegra_spi_transfer_one_message(struct spi_master *master,
 			struct spi_message *msg)
 {
@@ -838,21 +954,32 @@ static int tegra_spi_transfer_one_message(struct spi_master *master,
 		if (WARN_ON(ret == 0)) {
 			dev_err(tspi->dev,
 				"spi transfer timeout, err %d\n", ret);
+			if (tspi->is_curr_dma_xfer &&
+			    (tspi->cur_direction & DATA_DIR_TX))
+				dmaengine_terminate_all(tspi->tx_dma_chan);
+			if (tspi->is_curr_dma_xfer &&
+			    (tspi->cur_direction & DATA_DIR_RX))
+				dmaengine_terminate_all(tspi->rx_dma_chan);
 			ret = -EIO;
+			tegra_spi_dump_regs(tspi);
+			tegra_spi_flush_fifos(tspi);
+			reset_control_assert(tspi->rst);
+			udelay(2);
+			reset_control_deassert(tspi->rst);
 			goto complete_xfer;
 		}
 
 		if (tspi->tx_status ||  tspi->rx_status) {
 			dev_err(tspi->dev, "Error in Transfer\n");
 			ret = -EIO;
+			tegra_spi_dump_regs(tspi);
 			goto complete_xfer;
 		}
 		msg->actual_length += xfer->len;
 
 complete_xfer:
 		if (ret < 0 || skip) {
-			tegra_spi_writel(tspi, tspi->def_command1_reg,
-					SPI_COMMAND1);
+			tegra_spi_transfer_end(spi);
 			tegra_spi_transfer_delay(xfer->delay_usecs);
 			goto exit;
 		} else if (list_is_last(&xfer->transfer_list,
@@ -860,13 +987,11 @@ complete_xfer:
 			if (xfer->cs_change)
 				tspi->cs_control = spi;
 			else {
-				tegra_spi_writel(tspi, tspi->def_command1_reg,
-						SPI_COMMAND1);
+				tegra_spi_transfer_end(spi);
 				tegra_spi_transfer_delay(xfer->delay_usecs);
 			}
 		} else if (xfer->cs_change) {
-			tegra_spi_writel(tspi, tspi->def_command1_reg,
-					SPI_COMMAND1);
+			tegra_spi_transfer_end(spi);
 			tegra_spi_transfer_delay(xfer->delay_usecs);
 		}
 
@@ -889,11 +1014,14 @@ static irqreturn_t handle_cpu_based_xfer(struct tegra_spi_data *tspi)
 			tspi->status_reg);
 		dev_err(tspi->dev, "CpuXfer 0x%08x:0x%08x\n",
 			tspi->command1_reg, tspi->dma_control_reg);
+		tegra_spi_dump_regs(tspi);
+		tegra_spi_flush_fifos(tspi);
+		complete(&tspi->xfer_completion);
+		spin_unlock_irqrestore(&tspi->lock, flags);
 		reset_control_assert(tspi->rst);
 		udelay(2);
 		reset_control_deassert(tspi->rst);
-		complete(&tspi->xfer_completion);
-		goto exit;
+		return IRQ_HANDLED;
 	}
 
 	if (tspi->cur_direction & DATA_DIR_RX)
@@ -961,11 +1089,13 @@ static irqreturn_t handle_dma_based_xfer(struct tegra_spi_data *tspi)
 			tspi->status_reg);
 		dev_err(tspi->dev, "DmaXfer 0x%08x:0x%08x\n",
 			tspi->command1_reg, tspi->dma_control_reg);
+		tegra_spi_dump_regs(tspi);
+		tegra_spi_flush_fifos(tspi);
+		complete(&tspi->xfer_completion);
+		spin_unlock_irqrestore(&tspi->lock, flags);
 		reset_control_assert(tspi->rst);
 		udelay(2);
 		reset_control_deassert(tspi->rst);
-		complete(&tspi->xfer_completion);
-		spin_unlock_irqrestore(&tspi->lock, flags);
 		return IRQ_HANDLED;
 	}
 
@@ -1021,8 +1151,29 @@ static irqreturn_t tegra_spi_isr(int irq, void *context_data)
 	return IRQ_WAKE_THREAD;
 }
 
+static struct tegra_spi_soc_data tegra114_spi_soc_data = {
+	.has_intr_mask_reg = false,
+};
+
+static struct tegra_spi_soc_data tegra124_spi_soc_data = {
+	.has_intr_mask_reg = false,
+};
+
+static struct tegra_spi_soc_data tegra210_spi_soc_data = {
+	.has_intr_mask_reg = true,
+};
+
 static const struct of_device_id tegra_spi_of_match[] = {
-	{ .compatible = "nvidia,tegra114-spi", },
+	{
+		.compatible = "nvidia,tegra114-spi",
+		.data	    = &tegra114_spi_soc_data,
+	}, {
+		.compatible = "nvidia,tegra124-spi",
+		.data	    = &tegra124_spi_soc_data,
+	}, {
+		.compatible = "nvidia,tegra210-spi",
+		.data	    = &tegra210_spi_soc_data,
+	},
 	{}
 };
 MODULE_DEVICE_TABLE(of, tegra_spi_of_match);
@@ -1033,6 +1184,7 @@ static int tegra_spi_probe(struct platform_device *pdev)
 	struct tegra_spi_data	*tspi;
 	struct resource		*r;
 	int ret, spi_irq;
+	int bus_num;
 
 	master = spi_alloc_master(&pdev->dev, sizeof(*tspi));
 	if (!master) {
@@ -1047,16 +1199,28 @@ static int tegra_spi_probe(struct platform_device *pdev)
 		master->max_speed_hz = 25000000; /* 25MHz */
 
 	/* the spi->mode bits understood by this driver: */
-	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH;
+	master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LSB_FIRST |
+			    SPI_TX_DUAL | SPI_RX_DUAL | SPI_3WIRE;
+	master->bits_per_word_mask = SPI_BPW_RANGE_MASK(4, 32);
 	master->setup = tegra_spi_setup;
 	master->transfer_one_message = tegra_spi_transfer_one_message;
 	master->num_chipselect = MAX_CHIP_SELECT;
 	master->auto_runtime_pm = true;
+	bus_num = of_alias_get_id(pdev->dev.of_node, "spi");
+	if (bus_num >= 0)
+		master->bus_num = bus_num;
 
 	tspi->master = master;
 	tspi->dev = &pdev->dev;
 	spin_lock_init(&tspi->lock);
 
+	tspi->soc_data = of_device_get_match_data(&pdev->dev);
+	if (!tspi->soc_data) {
+		dev_err(&pdev->dev, "unsupported tegra\n");
+		ret = -ENODEV;
+		goto exit_free_master;
+	}
+
 	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	tspi->base = devm_ioremap_resource(&pdev->dev, r);
 	if (IS_ERR(tspi->base)) {
@@ -1067,27 +1231,19 @@ static int tegra_spi_probe(struct platform_device *pdev)
 
 	spi_irq = platform_get_irq(pdev, 0);
 	tspi->irq = spi_irq;
-	ret = request_threaded_irq(tspi->irq, tegra_spi_isr,
-			tegra_spi_isr_thread, IRQF_ONESHOT,
-			dev_name(&pdev->dev), tspi);
-	if (ret < 0) {
-		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
-					tspi->irq);
-		goto exit_free_master;
-	}
 
 	tspi->clk = devm_clk_get(&pdev->dev, "spi");
 	if (IS_ERR(tspi->clk)) {
 		dev_err(&pdev->dev, "can not get clock\n");
 		ret = PTR_ERR(tspi->clk);
-		goto exit_free_irq;
+		goto exit_free_master;
 	}
 
 	tspi->rst = devm_reset_control_get_exclusive(&pdev->dev, "spi");
 	if (IS_ERR(tspi->rst)) {
 		dev_err(&pdev->dev, "can not get reset\n");
 		ret = PTR_ERR(tspi->rst);
-		goto exit_free_irq;
+		goto exit_free_master;
 	}
 
 	tspi->max_buf_size = SPI_FIFO_DEPTH << 2;
@@ -1095,7 +1251,7 @@ static int tegra_spi_probe(struct platform_device *pdev)
 
 	ret = tegra_spi_init_dma_param(tspi, true);
 	if (ret < 0)
-		goto exit_free_irq;
+		goto exit_free_master;
 	ret = tegra_spi_init_dma_param(tspi, false);
 	if (ret < 0)
 		goto exit_rx_dma_free;
@@ -1117,18 +1273,32 @@ static int tegra_spi_probe(struct platform_device *pdev)
 		dev_err(&pdev->dev, "pm runtime get failed, e = %d\n", ret);
 		goto exit_pm_disable;
 	}
+
+	reset_control_assert(tspi->rst);
+	udelay(2);
+	reset_control_deassert(tspi->rst);
 	tspi->def_command1_reg  = SPI_M_S;
 	tegra_spi_writel(tspi, tspi->def_command1_reg, SPI_COMMAND1);
 	pm_runtime_put(&pdev->dev);
+	ret = request_threaded_irq(tspi->irq, tegra_spi_isr,
+				   tegra_spi_isr_thread, IRQF_ONESHOT,
+				   dev_name(&pdev->dev), tspi);
+	if (ret < 0) {
+		dev_err(&pdev->dev, "Failed to register ISR for IRQ %d\n",
+			tspi->irq);
+		goto exit_pm_disable;
+	}
 
 	master->dev.of_node = pdev->dev.of_node;
 	ret = devm_spi_register_master(&pdev->dev, master);
 	if (ret < 0) {
 		dev_err(&pdev->dev, "can not register to master err %d\n", ret);
-		goto exit_pm_disable;
+		goto exit_free_irq;
 	}
 	return ret;
 
+exit_free_irq:
+	free_irq(spi_irq, tspi);
 exit_pm_disable:
 	pm_runtime_disable(&pdev->dev);
 	if (!pm_runtime_status_suspended(&pdev->dev))
@@ -1136,8 +1306,6 @@ exit_pm_disable:
 	tegra_spi_deinit_dma_param(tspi, false);
 exit_rx_dma_free:
 	tegra_spi_deinit_dma_param(tspi, true);
-exit_free_irq:
-	free_irq(spi_irq, tspi);
 exit_free_master:
 	spi_master_put(master);
 	return ret;