spi: imx: mx51-ecspi: Fix low-speed CONFIGREG delay calculation
Subsystems (2)
PATCH 1/1 spi: imx: mx51-ecspi: Fix low-speed CONFIGREG delay calculation
The spi_imx->spi_bus_clk may be uninitialized and thus also zero in mx51_ecspi_prepare_message(), which would lead to division by zero in kernel. Since bitbang .setup_transfer callback which initializes the spi_imx->spi_bus_clk is called after bitbang prepare_message callback, iterate over all the transfers in spi_message, find the one with lowest bus frequency, and use that bus frequency for the delay calculation.
Note that it is not possible to move this CONFIGREG delay back into the .setup_transfer callback, because that is invoked too late, after the GPIO chipselects were already configured.
---
Sigh ...
--- drivers/spi/spi-imx.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
drivers/spi/spi-imx.c +9 -1
@@ -505,7 +505,9 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, struct spi_message *msg) {struct spi_device *spi = msg->spi; +struct spi_transfer *xfer;u32 ctrl =MX51_ECSPI_CTRL_ENABLE ; +u32 min_speed_hz = ~0U ;u32 testreg, delay;u32 cfg =readl (spi_imx->base +MX51_ECSPI_CONFIG );
@@ -578,7 +580,13 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx,* the SPI communication as the device on the other end would consider * the change of SCLK polarity as a clock tick already. */ - delay = (2 *1000000 ) / spi_imx->spi_bus_clk; +list_for_each_entry (xfer, &msg->transfers, transfer_list) { +if (!xfer->speed_hz) +continue ; + min_speed_hz =min (xfer->speed_hz, min_speed_hz); + } + + delay = (2 *1000000 ) / min_speed_hz;if (likely (delay <10 ))/* SCLK is faster than 100 kHz */ udelay (delay);else /* SCLK is _very_ slow */
Signature · 1 line
2.30.2
Quoted 37 lines · Marek Vasut
The
spi_imx->spi_bus_clkmay be uninitialized and thus also zero inmx51_ecspi_prepare_message(), which would lead to division by zero in kernel. Since bitbang .setup_transfer callback which initializes thespi_imx->spi_bus_clkis called after bitbangprepare_messagecallback, iterate over all the transfers inspi_message, find the one with lowest bus frequency, and use that bus frequency for the delay calculation.Note that it is not possible to move this CONFIGREG delay back into the .setup_transfer callback, because that is invoked too late, after the GPIO chipselects were already configured.
Fixes 135cbd378eab ("spi: imx: mx51-ecspi: Reinstate low-speed CONFIGREG delay")Signed-off-by Marek Vasut <[email protected]>Cc Uwe Kleine-König <[email protected]>Cc Mark Brown <[email protected]>---
Sigh ...--- drivers/spi/spi-imx.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)drivers/spi/spi-imx.c +2
-0@@ -505,7 +505,9 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, struct spi_message *msg) {struct spi_device *spi = msg->spi; +struct spi_transfer *xfer;u32 ctrl =MX51_ECSPI_CTRL_ENABLE ; +u32 min_speed_hz = ~0U ;u32 testreg, delay;u32 cfg =readl (spi_imx->base +MX51_ECSPI_CONFIG );
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 587
Quoted 9 lines
drivers/spi/spi-imx.c +5
-1@@ -578,7 +580,13 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx,* the SPI communication as the device on the other end would consider * the change of SCLK polarity as a clock tick already. */ - delay = (2 *1000000 ) / spi_imx->spi_bus_clk; +list_for_each_entry (xfer, &msg->transfers, transfer_list) { +if (!xfer->speed_hz) +continue ; + min_speed_hz =min (xfer->speed_hz, min_speed_hz); + }
Can it happen that all transfer's spped_hz are zero?
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 589
Quoted 2 lines
drivers/spi/spi-imx.c +2
-0+ + delay = (2 *1000000 ) / min_speed_hz;
Orthogonal to your change: I wonder if we need to round up the division here.
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 592
Quoted 3 lines
drivers/spi/spi-imx.c +0
-0if (likely (delay <10 ))/* SCLK is faster than 100 kHz */ udelay (delay);else /* SCLK is _very_ slow */
Also the comments are wrong here. Is SCLK is 150 kHz we have min_speed_hz = 150000, right? Then delay becomes 13 and the slow freq path is entered. The right comment (when keeping delay = (2 * 1000000) / min_speed_hz) would be
if (likely(delay < 10)) /* SCLK is faster than 181.818 kHz */
Best regards
Uwe
Signature · 2 lines
Pengutronix e.K. | Uwe Kleine-König | Industrial Linux Solutions | https://www.pengutronix.de/ |
On 7/19/21 10:20 AM, Uwe Kleine-König wrote:
[...]
Quoted 10 lines
Quoted 10 lines
drivers/spi/spi-imx.c +2
-0@@ -505,7 +505,9 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, struct spi_message *msg) {struct spi_device *spi = msg->spi; +struct spi_transfer *xfer;u32 ctrl =MX51_ECSPI_CTRL_ENABLE ; +u32 min_speed_hz = ~0U ;u32 testreg, delay;u32 cfg =readl (spi_imx->base +MX51_ECSPI_CONFIG );
Quoted 12 lines
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 587
Quoted 10 lines
drivers/spi/spi-imx.c +5
-1@@ -578,7 +580,13 @@ static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx,* the SPI communication as the device on the other end would consider * the change of SCLK polarity as a clock tick already. */ - delay = (2 *1000000 ) / spi_imx->spi_bus_clk; +list_for_each_entry (xfer, &msg->transfers, transfer_list) { +if (!xfer->speed_hz) +continue ; + min_speed_hz =min (xfer->speed_hz, min_speed_hz); + }Can it happen that all transfer's
spped_hzare zero?
I don't think so, what kind of spi_message would that be ?
And even if it was zero, the delay would be 2000000/~0U , so zero as well, which I suppose is the best we can do in such a case.
Quoted 5 lines
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 589
Quoted 2 lines
drivers/spi/spi-imx.c +2
-0+ + delay = (2 *1000000 ) / min_speed_hz;Orthogonal to your change: I wonder if we need to round up the division here.
That is not necessary, since the delay here is twice what it needs to be (because we really do not know what happens in the hardware internally).
Quoted 10 lines
drivers/spi/spi-imx.c static int mx51_ecspi_prepare_message(struct spi_imx_data *spi_imx, · line 592
Quoted 3 lines
drivers/spi/spi-imx.c +0
-0if (likely (delay <10 ))/* SCLK is faster than 100 kHz */ udelay (delay);else /* SCLK is _very_ slow */ Also the comments are wrong here. Is SCLK is 150 kHz we have
min_speed_hz= 150000, right? Then delay becomes 13 and the slow freq path is entered. The right comment (when keeping delay = (2 * 1000000) /min_speed_hz) would beif (likely(delay < 10)) /* SCLK is faster than 181.818 kHz */
This whole if/else is in fact based on Documentation/timers/timers-howto.rst , which says use usleep_range() for delays above 10uS or so.
The comment should be updated, but that's a separate patch.
Signature · 6 lines
Best regards, Marek Vasut DENX Software Engineering GmbH, Managing Director: Wolfgang Denk HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany Phone: (+49)-8142-66989-56 Email: [email protected]
Mark Brown
maintainerQuoted 8 lines · Marek Vasut
Quoted 1 line · Uwe Kleine-König
Can it happen that all transfer's
spped_hzare zero?I don't think so, what kind of
spi_messagewould that be ?And even if it was zero, the delay would be 2000000/~0U , so zero as well, which I suppose is the best we can do in such a case.
I can see that happening for a controller that didn't set any speed information with a driver that also didn't set any speed information, everything is just hoping that the default is fine but only the hardware is actually setting something. I've not gone and checked if anything ever insists there absolutely must be something specified in software.
Quoted 7 lines · Mark Brown
Quoted 5 lines · Marek Vasut
Quoted 1 line · Uwe Kleine-König
Can it happen that all transfer's
spped_hzare zero?I don't think so, what kind of
spi_messagewould that be ?
Quoted 8 lines
Quoted 2 lines
And even if it was zero, the delay would be 2000000/~0U , so zero as well, which I suppose is the best we can do in such a case.
I can see that happening for a controller that didn't set any speed information with a driver that also didn't set any speed information, everything is just hoping that the default is fine but only the hardware is actually setting something. I've not gone and checked if anything ever insists there absolutely must be something specified in software.
So, should I send a V2 here with any changes or is this one OK as-is ?
Mark Brown
maintainerQuoted 1 line · Marek Vasut
So, should I send a V2 here with any changes or is this one OK as-is ?
It wasn't clear that Uwe was OK with the current version, I don't mind.
Quoted 5 lines · Mark Brown
Quoted 1 line · Marek Vasut
So, should I send a V2 here with any changes or is this one OK as-is ?
It wasn't clear that Uwe was OK with the current version, I don't mind.
My concerns were mostly orthogonal to Marek's patch. The only (little critical) thing is: Can it happen that all transfer's speed_hz are zero? What happens with the code change is ok, when reading the code this looks this is by chance howver. So I would have added a comment describing that.
But all in all I'm convinced that the change is an improvement, so no further concerns from my side.
Best regards
Uwe
Signature · 2 lines
Pengutronix e.K. | Uwe Kleine-König | Industrial Linux Solutions | https://www.pengutronix.de/ |