Skip to contentSkip to conversation

spi: dw: DW SPI DMA Driver updates

PATCH Superseded Joy Chakraborty 2 participants 9 messages Apr 14, 2023 – Apr 18, 2023 SPI

v6 of 11 Superseded Latest is v11 →

lore
Share & tools
lore
#0 lore raw

v6 cover spi: dw: DW SPI DMA Driver updates

This Patch series adds support for 32 bits per word trasfers using DMA and some defensive checks around dma controller capabilities.
---
V1 Changes : Add support for AxSize=4 bytes to support 32bits/word.
---
V1->V2 Changes : Add dma capability check to make sure address widths are supported.
---
V2->V3 Changes : Split changes , add DMA direction check and other cosmetic chnages.
---
V3->V4 Changes : Fix Sparce Warning
| Reported-by: kernel test robot <[email protected]>
| Link: https://lore.kernel.org/oe-kbuild-all/[email protected]/
---
V4->V5 Changes : Preserve reverse xmas Tree order, move direction check before initalisation of further capabilities, remove zero initialisations, remove error OR'ing.
---
V5->V6 Changes :
-Remove case of n_bytes=3 using 4_bytes buswidth
-Avoid forward decaration
-Break capability check patch into 2
-round n_bytes to power of 2 ( Bug Fix)
-Add more explanation in commit text.
---

Joy Chakraborty (5):
spi: dw: Add 32 bpw support to SPI DW DMA driver
spi: dw: Move "dw_spi_can_dma" function
spi: dw: Add DMA directional capability check
spi: dw: Add DMA address widths capability check
spi: dw: Round of n_bytes to power of 2

 drivers/spi/spi-dw-core.c |  2 +-
 drivers/spi/spi-dw-dma.c  | 75 +++++++++++++++++++++++++++++----------
 drivers/spi/spi-dw.h      |  1 +
 3 files changed, 59 insertions(+), 19 deletions(-)
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog
#1 lore raw

PATCH v6 1/5 spi: dw: Add 32 bpw support to SPI DW DMA driver

Add Support for AxSize = 4 bytes configuration from dw dma driver if n_bytes i.e. number of bytes per write to fifo is 4.

Number of bytes written to fifo per write is depended on the bits/word configuration being used which the DW core driver translates to n_bytes. Hence, for bits per word values between 17 and 32 n_bytes should be equal to 4.

---
 drivers/spi/spi-dw-dma.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/spi/spi-dw-dma.c +8 -4
index ababb910b391..c1b42cb59965 100644
@@ -208,12 +208,16 @@ static bool dw_spi_can_dma(struct spi_controller *master,
 
 static enum dma_slave_buswidth dw_spi_dma_convert_width(u8 n_bytes)
 {
-	if (n_bytes == 1)
+	switch (n_bytes) {
+	case 1:
 		return DMA_SLAVE_BUSWIDTH_1_BYTE;
-	else if (n_bytes == 2)
+	case 2:
 		return DMA_SLAVE_BUSWIDTH_2_BYTES;
-
-	return DMA_SLAVE_BUSWIDTH_UNDEFINED;
+	case 4:
+		return DMA_SLAVE_BUSWIDTH_4_BYTES;
+	default:
+		return DMA_SLAVE_BUSWIDTH_UNDEFINED;
+	}
 }
 
 static int dw_spi_dma_wait(struct dw_spi *dws, unsigned int len, u32 speed)
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog
#2 lore raw

PATCH v6 2/5 spi: dw: Move "dw_spi_can_dma" function

In reply to Joy Chakraborty

Move "dw_spi_can_dma" function implementation below "dw_spi_dma_convert_width" function for handing compile dependency in future patches.

---
 drivers/spi/spi-dw-dma.c | 17 +++++++++--------
 1 file changed, 9 insertions(+), 8 deletions(-)
drivers/spi/spi-dw-dma.c +9 -8
index c1b42cb59965..75e1b11af7c9 100644
@@ -198,14 +198,6 @@ static irqreturn_t dw_spi_dma_transfer_handler(struct dw_spi *dws)
 	return IRQ_HANDLED;
 }
 
-static bool dw_spi_can_dma(struct spi_controller *master,
-			   struct spi_device *spi, struct spi_transfer *xfer)
-{
-	struct dw_spi *dws = spi_controller_get_devdata(master);
-
-	return xfer->len > dws->fifo_len;
-}
-
 static enum dma_slave_buswidth dw_spi_dma_convert_width(u8 n_bytes)
 {
 	switch (n_bytes) {
@@ -220,6 +212,15 @@ static enum dma_slave_buswidth dw_spi_dma_convert_width(u8 n_bytes)
 	}
 }
 
+static bool dw_spi_can_dma(struct spi_controller *master,
+			   struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+
+	return xfer->len > dws->fifo_len;
+}
+
+
 static int dw_spi_dma_wait(struct dw_spi *dws, unsigned int len, u32 speed)
 {
 	unsigned long long ms;
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog
#3 lore raw

Re: [PATCH v6 2/5] spi: dw: Move "dw_spi_can_dma" function

Quoted 3 lines · Joy Chakraborty

Move "dw_spi_can_dma" function implementation below "dw_spi_dma_convert_width" function for handing compile dependency in future patches.

We refer to the functions like func().

...

drivers/spi/spi-dw-dma.c static enum dma_slave_buswidth dw_spi_dma_convert_width(u8 n_bytes) · line 223

Quoted 10 lines
drivers/spi/spi-dw-dma.c +9 -0
+static bool dw_spi_can_dma(struct spi_controller *master,
+			   struct spi_device *spi, struct spi_transfer *xfer)
+{
+	struct dw_spi *dws = spi_controller_get_devdata(master);
+
+	return xfer->len > dws->fifo_len;
+}
 
+
+

Single blank line is enough.

Signature · 2 lines
With Best Regards,
Andy Shevchenko
#4 lore raw

Re: [PATCH v6 2/5] spi: dw: Move "dw_spi_can_dma" function

Quoted 18 lines · Andy Shevchenko
Quoted 3 lines · Joy Chakraborty

Move "dw_spi_can_dma" function implementation below "dw_spi_dma_convert_width" function for handing compile dependency in future patches.

We refer to the functions like func().

...

Quoted 7 lines
drivers/spi/spi-dw-dma.c +7 -0
+static bool dw_spi_can_dma(struct spi_controller *master,
+                        struct spi_device *spi, struct spi_transfer *xfer)
+{
+     struct dw_spi *dws = spi_controller_get_devdata(master);
+
+     return xfer->len > dws->fifo_len;
+}
Quoted 4 lines

Review

Quoted 2 lines
(quoted diff) +2 -0
+
+

Single blank line is enough.

Sure,I will make these changes and send V7.
Will wait for some time to see if Serge(y) has any other comments on the patch series.

Quoted 6 lines

--
With Best Regards,
Andy Shevchenko

#5 lore raw

PATCH v6 3/5 spi: dw: Add DMA directional capability check

In reply to Joy Chakraborty

Check capabilities of DMA controller during init to make sure it is capable of handling MEM2DEV for tx channel, DEV2MEM for rx channel.

Current DW DMA driver requires both tx and rx channel to be configured and functional for any kind of transfers to take effect including half duplex. Hence, check for both tx and rx direction and fail on unavailbility of either.

---
 drivers/spi/spi-dw-dma.c | 39 ++++++++++++++++++++++++++++++---------
 1 file changed, 30 insertions(+), 9 deletions(-)
drivers/spi/spi-dw-dma.c +30 -9
index 75e1b11af7c9..e1dd13fe4fd0 100644
@@ -72,12 +72,22 @@ static void dw_spi_dma_maxburst_init(struct dw_spi *dws)
 	dw_writel(dws, DW_SPI_DMATDLR, dws->txburst);
 }
 
-static void dw_spi_dma_sg_burst_init(struct dw_spi *dws)
+static int dw_spi_dma_caps_init(struct dw_spi *dws)
 {
-	struct dma_slave_caps tx = {0}, rx = {0};
+	struct dma_slave_caps tx, rx;
+	int ret;
+
+	ret = dma_get_slave_caps(dws->txchan, &tx);
+	if (ret)
+		return ret;
+
+	ret = dma_get_slave_caps(dws->rxchan, &rx);
+	if (ret)
+		return ret;
 
-	dma_get_slave_caps(dws->txchan, &tx);
-	dma_get_slave_caps(dws->rxchan, &rx);
+	if (!(tx.directions & BIT(DMA_MEM_TO_DEV) &&
+	      rx.directions & BIT(DMA_DEV_TO_MEM)))
+		return -ENXIO;
 
 	if (tx.max_sg_burst > 0 && rx.max_sg_burst > 0)
 		dws->dma_sg_burst = min(tx.max_sg_burst, rx.max_sg_burst);
@@ -95,6 +105,7 @@ static int dw_spi_dma_init_mfld(struct device *dev, struct dw_spi *dws)
 	struct dw_dma_slave dma_rx = { .src_id = 0 }, *rx = &dma_rx;
 	struct pci_dev *dma_dev;
 	dma_cap_mask_t mask;
+	int ret = -EBUSY;
 
 	/*
 	 * Get pci device for DMA controller, currently it could only
@@ -124,20 +135,25 @@ static int dw_spi_dma_init_mfld(struct device *dev, struct dw_spi *dws)
 
 	init_completion(&dws->dma_completion);
 
-	dw_spi_dma_maxburst_init(dws);
+	ret = dw_spi_dma_caps_init(dws);
+	if (ret)
+		goto free_txchan;
 
-	dw_spi_dma_sg_burst_init(dws);
+	dw_spi_dma_maxburst_init(dws);
 
 	pci_dev_put(dma_dev);
 
 	return 0;
 
+free_txchan:
+	dma_release_channel(dws->txchan);
+	dws->txchan = NULL;
 free_rxchan:
 	dma_release_channel(dws->rxchan);
 	dws->rxchan = NULL;
 err_exit:
 	pci_dev_put(dma_dev);
-	return -EBUSY;
+	return ret;
 }
 
 static int dw_spi_dma_init_generic(struct device *dev, struct dw_spi *dws)
@@ -163,12 +179,17 @@ static int dw_spi_dma_init_generic(struct device *dev, struct dw_spi *dws)
 
 	init_completion(&dws->dma_completion);
 
-	dw_spi_dma_maxburst_init(dws);
+	ret = dw_spi_dma_caps_init(dws);
+	if (ret)
+		goto free_txchan;
 
-	dw_spi_dma_sg_burst_init(dws);
+	dw_spi_dma_maxburst_init(dws);
 
 	return 0;
 
+free_txchan:
+	dma_release_channel(dws->txchan);
+	dws->txchan = NULL;
 free_rxchan:
 	dma_release_channel(dws->rxchan);
 	dws->rxchan = NULL;
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog
#6 lore raw

PATCH v6 4/5 spi: dw: Add DMA address widths capability check

In reply to Joy Chakraborty

Store address width capabilities of DMA controller during init and check the same per transfer to make sure the bits/word requirement can be met.

Current DW DMA driver requires both tx and rx channel to be configured and functional hence a subset of both tx and rx channel address width capability is checked with the width requirement(n_bytes) for a transfer.

---
 drivers/spi/spi-dw-dma.c | 17 +++++++++++++++--
 drivers/spi/spi-dw.h     |  1 +
 2 files changed, 16 insertions(+), 2 deletions(-)
drivers/spi/spi-dw-dma.c +15 -2
index e1dd13fe4fd0..45980c46946d 100644
@@ -97,6 +97,14 @@ static int dw_spi_dma_caps_init(struct dw_spi *dws)
 		dws->dma_sg_burst = rx.max_sg_burst;
 	else
 		dws->dma_sg_burst = 0;
+
+	/*
+	 * Assuming both channels belong to the same DMA controller hence the
+	 * address width capabilities most likely would be the same.
+	 */
+	dws->dma_addr_widths = tx.dst_addr_widths & rx.src_addr_widths;
+
+	return 0;
 }
 
 static int dw_spi_dma_init_mfld(struct device *dev, struct dw_spi *dws)
@@ -237,10 +245,15 @@ static bool dw_spi_can_dma(struct spi_controller *master,
 			   struct spi_device *spi, struct spi_transfer *xfer)
 {
 	struct dw_spi *dws = spi_controller_get_devdata(master);
+	enum dma_slave_buswidth dma_bus_width;
 
-	return xfer->len > dws->fifo_len;
-}
+	if (xfer->len <= dws->fifo_len)
+		return false;
 
+	dma_bus_width = dw_spi_dma_convert_width(dws->n_bytes);
+
+	return dws->dma_addr_widths & BIT(dma_bus_width);
+}
 
 static int dw_spi_dma_wait(struct dw_spi *dws, unsigned int len, u32 speed)
 {
drivers/spi/spi-dw.h +1 -0
index 9e8eb2b52d5c..3962e6dcf880 100644
@@ -190,6 +190,7 @@ struct dw_spi {
 	struct dma_chan		*rxchan;
 	u32			rxburst;
 	u32			dma_sg_burst;
+	u32			dma_addr_widths;
 	unsigned long		dma_chan_busy;
 	dma_addr_t		dma_addr; /* phy address of the Data register */
 	const struct dw_spi_dma_ops *dma_ops;
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog
#7 lore raw

Re: [PATCH v6 4/5] spi: dw: Add DMA address widths capability check

Quoted 7 lines · Joy Chakraborty

Store address width capabilities of DMA controller during init and check the same per transfer to make sure the bits/word requirement can be met.

Current DW DMA driver requires both tx and rx channel to be configured and functional hence a subset of both tx and rx channel address width capability is checked with the width requirement(n_bytes) for a transfer.

...

drivers/spi/spi-dw-dma.c static int dw_spi_dma_caps_init(struct dw_spi *dws) · line 103

Quoted 3 lines
drivers/spi/spi-dw-dma.c +3 -0
+	/*
+	 * Assuming both channels belong to the same DMA controller hence the
+	 * address width capabilities most likely would be the same.

I would add something to explain the side of these address width, like

* Assuming both channels belong to the same DMA controller hence
* the peripheral side address width capabilities most likely would
* be the same.

Quoted 1 line
drivers/spi/spi-dw-dma.c +1 -0
+	 */
Signature · 2 lines
With Best Regards,
Andy Shevchenko
#8 lore raw

PATCH v6 5/5 spi: dw: Round of n_bytes to power of 2

In reply to Joy Chakraborty

n_bytes variable in the driver represents the number of bytes per word that needs to be sent/copied to fifo. Bits/word can be between 8 and 32 bits from the client but in memory they are a power of 2, same is mentioned in spi.h header:
"
* @bits_per_word: Data transfers involve one or more words; word sizes
* like eight or 12 bits are common. In-memory wordsizes are
* powers of two bytes (e.g. 20 bit samples use 32 bits).
* This may be changed by the device's driver, or left at the
* default (0) indicating protocol words are eight bit bytes.
* The spi_transfer.bits_per_word can override this for each transfer.
"

Hence, round of n_bytes to a power of 2 to avoid values like 3 which would generate unalligned/odd accesses to memory/fifo.

Fixes a51acc2400d4 ("spi: dw: Add support for 32-bits max xfer size")
---
 drivers/spi/spi-dw-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/spi/spi-dw-core.c +1 -1
index c3bfb6c84cab..a6486db46c61 100644
@@ -426,7 +426,7 @@ static int dw_spi_transfer_one(struct spi_controller *master,
 	int ret;
 
 	dws->dma_mapped = 0;
-	dws->n_bytes = DIV_ROUND_UP(transfer->bits_per_word, BITS_PER_BYTE);
+	dws->n_bytes = roundup_pow_of_two(DIV_ROUND_UP(transfer->bits_per_word, BITS_PER_BYTE));
 	dws->tx = (void *)transfer->tx_buf;
 	dws->tx_len = transfer->len / dws->n_bytes;
 	dws->rx = transfer->rx_buf;
Signature · 1 line
2.40.0.634.g4ca3ef3211-goog