spi: Update speed/mode on change
Subsystems (3)
PATCH v2 1/1 spi: Update speed/mode on change
The spi_get_bus_and_cs() may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Call spi_set_speed_mode() in case bus frequency or bus mode changed to let the controller update the configuration.
The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With this patch, the later transfer happens correctly at 2000 Hz.
---
V2: Set plat in case it is NULL
--- drivers/spi/spi-uclass.c | 12 ++++++++++++ 1 file changed, 12 insertions(+)
drivers/spi/spi-uclass.c +12 -0
@@ -391,6 +391,8 @@ int spi_get_bus_and_cs(int busnum, int cs, int speed, int mode, }else if (ret) {dev_err (bus,"Invalid chip select %d:%d (err=%d)\n" , busnum, cs, ret);return ret; + }else if (dev) { + plat =dev_get_parent_plat (dev); }if (!device_active (dev)) {
@@ -416,12 +418,22 @@ int spi_get_bus_and_cs(int busnum, int cs, int speed, int mode, goto err; } +/* In case bus frequency or mode changed, update it. */ +if ((speed && bus_data->speed && bus_data->speed != speed) || + (plat && plat->mode != mode)) { + ret =spi_set_speed_mode (bus, speed, mode); +if (ret) + goto err_speed_mode; + } + *busp = bus; *devp = slave;log_debug ("%s: bus=%p, slave=%p\n" , __func__, bus, *devp);return 0 ; +err_speed_mode: +spi_release_bus (slave); err:log_debug ("%s: Error path, created=%d, device '%s'\n" , __func__, created, dev->name);
Signature · 1 line
2.30.2
Quoted 12 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With this patch, the later transfer happens correctly at 2000 Hz.
Quoted 3 lines
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>
Applied to u-boot/master, thanks!
Signature · 1 line
Tom
Hi Marek,
This patch breaks meson_spifc:
SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiBmeson_spifc spi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)
Best,
Da
Quoted 23 lines · Tom Rini
Quoted 12 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With this patch, the later transfer happens correctly at 2000 Hz.Quoted 3 lines
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
Tom
Quoted 9 lines · Da Xue
Hi Marek,
This patch breaks
meson_spifc:SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)
Quoted 28 lines
Best,
Da
Quoted 22 lines · Tom Rini
Quoted 12 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With this patch, the later transfer happens correctly at 2000 Hz.Quoted 3 lines
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
Tom
Seems like you're hitting this code in drivers/spi/meson_spifc.c
250 static int meson_spifc_set_mode(struct udevice *dev, uint mode) 251 {
252 struct meson_spifc_priv *spifc = dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA | SPI_RX_QUAD | SPI_RX_DUAL |
255 SPI_TX_QUAD | SPI_TX_DUAL))
256 return -ENODEV;
(the -ENODEV code doesn't look right, it should be some -EOPNOTSUP or so)
Can you check which of the mode bits is set and triggers the condition ?
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.
A message that is not in the archive (sent off-list or not yet received): CACdvmAh4MZ-rACRAY1nv1d0myCMyNnXjR5htkHTpe3A-wm=B1g@mail.gmail.com
Quoted 6 lines · Da Xue
SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
SPI mode: 3meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)
Quoted 61 lines
I added spi-rx-bus-width = <1> and spi-tx-bus-width = <1>
Quoted 57 lines · Marek Vasut
Quoted 9 lines · Da Xue
Hi Marek,
This patch breaks
meson_spifc:SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Quoted 28 lines
Best,
Da
Quoted 22 lines · Tom Rini
Quoted 12 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With this patch, the later transfer happens correctly at 2000 Hz.Quoted 3 lines
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
TomSeems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPor so)Can you check which of the mode bits is set and triggers the condition ?
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.
Can you check which of the mode bits is set and triggers the condition ?
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
Quoted 72 lines · Marek Vasut
Quoted 6 lines · Da Xue
SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
SPI mode: 3meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Quoted 10 lines
I added spi-rx-bus-width = <1> and spi-tx-bus-width = <1>
Quoted 6 lines · Marek Vasut
Quoted 5 lines · Da Xue
Hi Marek,
This patch breaks
meson_spifc:SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB,
total
Quoted 18 lines
Quoted 18 lines
Quoted 18 lines
16 MiB
meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Best,
Da
Quoted 8 lines · Tom Rini
Quoted 6 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With thispatch,
Quoted 21 lines
Quoted 21 lines
Quoted 10 lines
Quoted 10 lines
Quoted 5 lines
the later transfer happens correctly at 2000 Hz.
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
TomSeems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;Quoted 1 line
Quoted 1 line
(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPorso)
Quoted 7 lines
Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.
Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
My apologies about the top-posting.
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts +2 -0
@@ -304,6 +304,8 @@ compatible = "jedec,spi-nor"; reg = <0>; spi-max-frequency = <80000000>; + spi-rx-bus-width = <1>; + spi-tx-bus-width = <1>; }; };
Quoted 87 lines · Da Xue
Quoted 72 lines · Marek Vasut
Quoted 6 lines · Da Xue
SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB, total 16 MiB
SPI mode: 3meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Quoted 10 lines
I added spi-rx-bus-width = <1> and spi-tx-bus-width = <1>
Quoted 6 lines · Marek Vasut
Quoted 5 lines · Da Xue
Hi Marek,
This patch breaks
meson_spifc:SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB,
total
Quoted 18 lines
Quoted 18 lines
Quoted 18 lines
16 MiB
meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Best,
Da
Quoted 8 lines · Tom Rini
Quoted 6 lines · Marek Vasut
The
spi_get_bus_and_cs()may be called on the same bus and chipselect with different frequency or mode. This is valid usecase, but the code fails to notify the controller of such a configuration change. Callspi_set_speed_mode()in case bus frequency or bus mode changed to let the controller update the configuration.Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With thispatch,
Quoted 21 lines
Quoted 21 lines
Quoted 10 lines
Quoted 10 lines
Quoted 5 lines
the later transfer happens correctly at 2000 Hz.
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
TomSeems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;Quoted 1 line
Quoted 1 line
(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPorso)
Quoted 7 lines
Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
My apologies about the top-posting.
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts line 310
Quoted 7 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts +2
-0@@ -304,6 +304,8 @@ compatible = "jedec,spi-nor"; reg = <0>; spi-max-frequency = <80000000>; + spi-rx-bus-width = <1>; + spi-tx-bus-width = <1>; }; };
That should do the trick. Can you check which of the mode bits is set in meson_spifc_set_mode() and triggers the ENODEV condition ?
Quoted 91 lines · Marek Vasut
Quoted 4 lines · Da Xue
Quoted 2 lines · Marek Vasut
Quoted 1 line · Da Xue
SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB,
total
Quoted 25 lines
Quoted 25 lines
Quoted 5 lines
16 MiB
SPI mode: 3meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Quoted 10 lines
I added spi-rx-bus-width = <1> and spi-tx-bus-width = <1>
Quoted 6 lines · Marek Vasut
Quoted 5 lines · Da Xue
Hi Marek,
This patch breaks
meson_spifc:SF: Detected gd25lq128 with page size 256 Bytes, erase size 64 KiB,
total
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
16 MiB
meson_spifcspi@8c80: Cannot set mode (err=-19)
Failed to initialize SPI flash at 0:0 (error -19)Best,
Da
On Wed, Jun 30, 2021 at 12:49 PM Tom Rini <[email protected]>
Quoted 4 lines · wrote:
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines · Marek Vasut
Quoted 2 lines
The
spi_get_bus_and_cs()may be called on the same bus andchipselect
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 1 line
with different frequency or mode. This is valid usecase, but the
code
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
fails to notify the controller of such a configuration change. Call
spi_set_speed_mode()in case bus frequency or bus mode changed tolet
Quoted 28 lines
Quoted 28 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
the controller update the configuration.
The problem can easily be triggered using the sspi command:
=> sspi 0:0@1000
=> sspi 0:0@2000
Without this patch, both transfers happen at 1000 Hz. With thispatch,
Quoted 21 lines
Quoted 21 lines
Quoted 10 lines
Quoted 10 lines
Quoted 5 lines
the later transfer happens correctly at 2000 Hz.
Signed-off-by Marek Vasut <[email protected]>Cc Jagan Teki <[email protected]>Cc Patrick Delaunay <[email protected]>Applied to u-boot/master, thanks!
--
TomSeems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;Quoted 4 lines
Quoted 4 lines
Quoted 1 line
Quoted 1 line
(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPorso)
Quoted 2 lines
Quoted 2 lines
Can you check which of the mode bits is set and triggers the
condition ?
Quoted 16 lines
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
My apologies about the top-posting.
Quoted 14 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts line 310
Quoted 10 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts +2
-0@@ -304,6 +304,8 @@ compatible = "jedec,spi-nor"; reg = <0>; spi-max-frequency = <80000000>; + spi-rx-bus-width = <1>; + spi-tx-bus-width = <1>; }; };That should do the trick. Can you check which of the mode bits is set in
meson_spifc_set_mode()and triggers the ENODEV condition ?
SPI_CPHA seems to be the culprit. I tried adding spi-cpha = <0> to no avail.
On 7/2/21 9:35 PM, Da Xue wrote:
[...]
Quoted 46 lines
Quoted 31 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Seems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;Quoted 4 lines
Quoted 4 lines
Quoted 1 line
Quoted 1 line
(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPorso)
Quoted 2 lines
Quoted 2 lines
Can you check which of the mode bits is set and triggers the
condition ?
Quoted 16 lines
Quoted 6 lines
Quoted 5 lines
Quoted 5 lines
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.Quoted 7 lines
Can you check which of the mode bits is set and triggers the condition ?
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
My apologies about the top-posting.
Quoted 14 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts line 310
Quoted 10 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts +2
-0@@ -304,6 +304,8 @@ compatible = "jedec,spi-nor"; reg = <0>; spi-max-frequency = <80000000>; + spi-rx-bus-width = <1>; + spi-tx-bus-width = <1>; }; };That should do the trick. Can you check which of the mode bits is set in
meson_spifc_set_mode()and triggers the ENODEV condition ?
Quoted 2 lines
SPI_CPHAseems to be the culprit. I tried adding spi-cpha = <0> to no avail.
Can you find out what is setting the SPI_CPHA in the first place on your machine ?
Quoted 52 lines · Marek Vasut
On 7/2/21 9:35 PM, Da Xue wrote:
[...]
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Seems like you're hitting this code in
drivers/spi/meson_spifc.c250 static int
meson_spifc_set_mode(struct udevice *dev, uint mode)251 {
252 structmeson_spifc_priv*spifc =dev_get_priv(dev);
253
254 if (mode & (SPI_CPHA|SPI_RX_QUAD|SPI_RX_DUAL|
255SPI_TX_QUAD|SPI_TX_DUAL))
256 return-ENODEV;(the
-ENODEVcode doesn't look right, it should be some-EOPNOTSUPor
Quoted 11 lines
Quoted 11 lines
Quoted 3 lines
Quoted 3 lines
so)
Quoted 2 lines
Quoted 2 lines
Can you check which of the mode bits is set and triggers the
condition ?
Quoted 7 lines
Quoted 7 lines
Quoted 5 lines
Quoted 5 lines
I think you might be missing something like
spi-rx-bus-width = <1>;
spi-tx-bus-width = <1>;
in your DT, but that's a guess.Can you check which of the mode bits is set and triggers the
condition ?
Quoted 24 lines
Quoted 23 lines
Quoted 12 lines
Quoted 6 lines
Also, where in the DT did you add spi-rx-bus-width = <1> and spi-tx-bus-width = <1> ?
Finally, please do not top-post and keep the list on CC.
My apologies about the top-posting.
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts line 310
Quoted 7 lines
arch/arm/dts/meson-gxl-s805x-libretech-ac.dts +2
-0@@ -304,6 +304,8 @@ compatible = "jedec,spi-nor"; reg = <0>; spi-max-frequency = <80000000>; + spi-rx-bus-width = <1>; + spi-tx-bus-width = <1>; }; };That should do the trick. Can you check which of the mode bits is set in
meson_spifc_set_mode()and triggers the ENODEV condition ?
Quoted 6 lines
Quoted 2 lines
SPI_CPHAseems to be the culprit. I tried adding spi-cpha = <0> to no avail.Can you find out what is setting the
SPI_CPHAin the first place on your machine ?
The Kconfig was setting it to 3. I manually added the default mode (0) to my board's config and it worked. Thanks Marek.
On 7/2/21 10:10 PM, Da Xue wrote:
[...]
Quoted 7 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines
That should do the trick. Can you check which of the mode bits is set in
meson_spifc_set_mode()and triggers the ENODEV condition ?
SPI_CPHAseems to be the culprit. I tried adding spi-cpha = <0> to no avail.
Quoted 6 lines
Quoted 3 lines
Can you find out what is setting the
SPI_CPHAin the first place on your machine ?The
Kconfigwas setting it to 3. I manually added the default mode (0) to my board's config and it worked. Thanks Marek.
Sure, good thing this was fixed. Please send a patch for the board, so it can be added to this release.