dm: ns16550: Add support for reg-offset property

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

Quoted 14 lines · Simon Glass

Hi Michal,

Quoted 10 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Quoted 3 lines

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 10 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

It looks right to me.

Regards,
Simon
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 22 lines · Simon Glass

Hi Michal,

Quoted 19 lines · Michal Simek

Quoted 17 lines · Simon Glass

Hi Michal,

Quoted 10 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

Quoted 4 lines

Quoted 3 lines

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Quoted 28 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 10 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

Quoted 3 lines

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Thanks,
Michal

_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 33 lines · Michal Simek

Quoted 31 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 10 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Quoted 4 lines

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Would it be appropriate to initialize 'plat->reg_offset' to zero for the non-OF_CONTROL cases?

Hi,

Quoted 39 lines · Derald D. Woods

Quoted 37 lines · Michal Simek

Quoted 23 lines · Adam Ford

Quoted 22 lines · Simon Glass

Hi Michal,

Quoted 19 lines · Michal Simek

Quoted 17 lines · Simon Glass

Hi Michal,

Quoted 10 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

Quoted 8 lines

Quoted 4 lines

Quoted 3 lines

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 2 lines

Would it be appropriate to initialize 'plat->reg_offset' to zero for the non-OF_CONTROL cases?

It should not be needed as driver model promises to zero auto-allocated data.

- Simon

Quoted 28 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 10 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

Quoted 3 lines

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Thanks,
Michal

Quoted 33 lines · Michal Simek

Quoted 31 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 10 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Quoted 4 lines

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 39 lines · Derald D. Woods

Quoted 37 lines · Michal Simek

Quoted 23 lines · Adam Ford

Quoted 22 lines · Simon Glass

Hi Michal,

Quoted 19 lines · Michal Simek

Quoted 17 lines · Simon Glass

Hi Michal,

Quoted 10 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

 drivers/serial/ns16550.c | 6 ++++--
 include/ns16550.h        | 1 +
 2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

Quoted 8 lines

Quoted 4 lines

Quoted 3 lines

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 2 lines

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

You should enable debug console and send the log.
Do you have enough space for malloc?

The patch is quite simple and if your DT has no this property there are not so many options what can be problematic.
I expect when you enable debug console you will get more information.

Thanks,
Michal

Quoted 34 lines · Michal Simek

Quoted 33 lines · Derald D. Woods

Quoted 32 lines · Michal Simek

Quoted 20 lines · Adam Ford

Quoted 19 lines · Simon Glass

Hi Michal,

Quoted 16 lines · Michal Simek

Quoted 15 lines · Simon Glass

Hi Michal,

Quoted 10 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Quoted 7 lines

Quoted 4 lines

Quoted 4 lines

Quoted 1 line

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 5 lines

Quoted 2 lines

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

Quoted 42 lines · Derald D. Woods

Quoted 41 lines · Michal Simek

Quoted 40 lines · Derald D. Woods

Quoted 39 lines · Michal Simek

Quoted 28 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 11 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 6 lines

Quoted 3 lines

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 8 lines

Quoted 5 lines

Quoted 2 lines

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

Quoted 66 lines · Alexander Graf

Quoted 42 lines · Derald D. Woods

Quoted 41 lines · Michal Simek

Quoted 40 lines · Derald D. Woods

Quoted 39 lines · Michal Simek

Quoted 28 lines · Adam Ford

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 11 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 6 lines

Quoted 3 lines

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 8 lines

Quoted 5 lines

Quoted 2 lines

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 
 struct udevice;

I see the following grep results:

$ grep -RI -e "const struct ns16550_platdata" .
./arch/arm/cpu/armv7/am33xx/board.c:static const struct 
ns16550_platdata am33xx_serial[] = {
./arch/arm/cpu/arm926ejs/lpc32xx/devices.c:static const struct 
ns16550_platdata lpc32xx_uart[] = {
./board/timll/devkit8000/devkit8000.c:static const struct 
ns16550_platdata devkit8000_serial = {
./board/ti/beagle/beagle.c:static const struct ns16550_platdata 
beagle_serial = {
./board/logicpd/zoom1/zoom1.c:static const struct ns16550_platdata 
zoom1_serial = {
./board/logicpd/omap3som/omap3logic.c:static const struct 
ns16550_platdata omap3logic_serial = {
./board/quipos/cairo/cairo.c:static const struct ns16550_platdata 
cairo_serial = {
./board/lge/sniper/sniper.c:static const struct ns16550_platdata 
serial_omap_platdata = {
./board/isee/igep00x0/igep00x0.c:static const struct ns16550_platdata 
igep_serial = {
./board/overo/overo.c:static const struct ns16550_platdata 
overo_serial = {

Could the use of 'const' be a part of the problem?

- Derald

Quoted 119 lines · Derald D. Woods

Quoted 82 lines · Alexander Graf

Quoted 58 lines · Derald D. Woods

Quoted 56 lines · Michal Simek

Quoted 53 lines · Derald D. Woods

Quoted 44 lines · Michal Simek

Quoted 36 lines · Adam Ford

Quoted 32 lines · Simon Glass

Hi Michal,

Quoted 27 lines · Michal Simek

Quoted 13 lines · Simon Glass

Hi Michal,

Quoted 8 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Quoted 5 lines

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 6 lines

Quoted 3 lines

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 
 struct udevice;

I see the following grep results:

$ grep -RI -e "const struct ns16550_platdata" .
./arch/arm/cpu/armv7/am33xx/board.c:static const struct ns16550_platdata
am33xx_serial[] = {
./arch/arm/cpu/arm926ejs/lpc32xx/devices.c:static const struct
ns16550_platdata lpc32xx_uart[] = {
./board/timll/devkit8000/devkit8000.c:static const struct ns16550_platdata
devkit8000_serial = {
./board/ti/beagle/beagle.c:static const struct ns16550_platdata
beagle_serial = {
./board/logicpd/zoom1/zoom1.c:static const struct ns16550_platdata
zoom1_serial = {
./board/logicpd/omap3som/omap3logic.c:static const struct ns16550_platdata
omap3logic_serial = {
./board/quipos/cairo/cairo.c:static const struct ns16550_platdata
cairo_serial = {
./board/lge/sniper/sniper.c:static const struct ns16550_platdata
serial_omap_platdata = {
./board/isee/igep00x0/igep00x0.c:static const struct ns16550_platdata
igep_serial = {
./board/overo/overo.c:static const struct ns16550_platdata overo_serial =
{

Quoted 5 lines

Could the use of 'const' be a part of the problem?

- Derald

The structure initializers need rework for the additional member.

- Derald
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 115 lines · Derald D. Woods

Quoted 101 lines · Alexander Graf

Quoted 66 lines · Derald D. Woods

Quoted 42 lines · Michal Simek

Quoted 41 lines · Derald D. Woods

Quoted 40 lines · Michal Simek

Quoted 39 lines · Adam Ford

Quoted 28 lines

Quoted 26 lines · Simon Glass

Hi Michal,

Quoted 22 lines · Michal Simek

Quoted 11 lines · Simon Glass

Hi Michal,

Quoted 7 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel.
It is shifting start of address space by reg-offset.
On Xilinx platform this offset is typically 0x1000.

Signed-off-by: Michal Simek <[email protected]>
---

Quoted 6 lines

Quoted 3 lines

drivers/serial/ns16550.c | 6 ++++--
include/ns16550.h | 1 +
2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

Quoted 8 lines

Quoted 5 lines

Quoted 2 lines

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
   */
  struct ns16550_platdata {
         unsigned long base;
-       int reg_offset;
         int reg_shift;
         int clock;
+       int reg_offset;
  };
 
  struct udevice;

I see the following grep results:

$ grep -RI -e "const struct ns16550_platdata" .
./arch/arm/cpu/armv7/am33xx/board.c:static const struct ns16550_platdata
am33xx_serial[] = {
./arch/arm/cpu/arm926ejs/lpc32xx/devices.c:static const struct
ns16550_platdata lpc32xx_uart[] = {
./board/timll/devkit8000/devkit8000.c:static const struct ns16550_platdata
devkit8000_serial = {
./board/ti/beagle/beagle.c:static const struct ns16550_platdata
beagle_serial = {
./board/logicpd/zoom1/zoom1.c:static const struct ns16550_platdata
zoom1_serial = {
./board/logicpd/omap3som/omap3logic.c:static const struct ns16550_platdata
omap3logic_serial = {
./board/quipos/cairo/cairo.c:static const struct ns16550_platdata
cairo_serial = {
./board/lge/sniper/sniper.c:static const struct ns16550_platdata
serial_omap_platdata = {
./board/isee/igep00x0/igep00x0.c:static const struct ns16550_platdata
igep_serial = {
./board/overo/overo.c:static const struct ns16550_platdata overo_serial =
{

Quoted 5 lines

Could the use of 'const' be a part of the problem?

- Derald

The structure initializers need rework for the additional member.

- Derald
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

I first tried removing the 'const' in the board file as suggested by Derald, but that wasn't successful. I can boot with Alexander's patch, but modifying the order inside the header seems weird to me. I haven't had any time to look this weekend, but I wonder if something in one of the files is modifying the structure and expects the order of the structure to always be a certain order.

adam

Quoted 138 lines · Derald D. Woods

Quoted 138 lines · Derald D. Woods

Quoted 123 lines · Alexander Graf

Quoted 84 lines · Derald D. Woods

Quoted 59 lines · Michal Simek

Quoted 57 lines · Derald D. Woods

Quoted 54 lines · Michal Simek

Quoted 44 lines · Adam Ford

Quoted 36 lines

Quoted 32 lines · Simon Glass

Hi Michal,

Quoted 27 lines · Michal Simek

Quoted 20 lines · Simon Glass

Hi Michal,

On 16 February 2016 at 08:17, Michal Simek
<[email protected]> wrote:

reg-offset is the part of standard 8250 binding in the kernel.
It is shifting start of address space by reg-offset.
On Xilinx platform this offset is typically 0x1000.

Signed-off-by: Michal Simek <[email protected]>
---

drivers/serial/ns16550.c | 6 ++++--
include/ns16550.h | 1 +
2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by: Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Quoted 5 lines

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 6 lines

Quoted 3 lines

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code
is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
   */
  struct ns16550_platdata {
         unsigned long base;
-       int reg_offset;
         int reg_shift;
         int clock;
+       int reg_offset;
  };
 
  struct udevice;

I see the following grep results:

$ grep -RI -e "const struct ns16550_platdata" .
./arch/arm/cpu/armv7/am33xx/board.c:static const struct ns16550_platdata
am33xx_serial[] = {
./arch/arm/cpu/arm926ejs/lpc32xx/devices.c:static const struct
ns16550_platdata lpc32xx_uart[] = {
./board/timll/devkit8000/devkit8000.c:static const struct
ns16550_platdata
devkit8000_serial = {
./board/ti/beagle/beagle.c:static const struct ns16550_platdata
beagle_serial = {
./board/logicpd/zoom1/zoom1.c:static const struct ns16550_platdata
zoom1_serial = {
./board/logicpd/omap3som/omap3logic.c:static const struct
ns16550_platdata
omap3logic_serial = {
./board/quipos/cairo/cairo.c:static const struct ns16550_platdata
cairo_serial = {
./board/lge/sniper/sniper.c:static const struct ns16550_platdata
serial_omap_platdata = {
./board/isee/igep00x0/igep00x0.c:static const struct ns16550_platdata
igep_serial = {
./board/overo/overo.c:static const struct ns16550_platdata overo_serial
=
{

Quoted 5 lines

Could the use of 'const' be a part of the problem?

- Derald

The structure initializers need rework for the additional member.

- Derald
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 11 lines · Adam Ford

I first tried removing the 'const' in the board file as suggested by Derald, but that wasn't successful. I can boot with Alexander's patch, but modifying the order inside the header seems weird to me. I haven't had any time to look this weekend, but I wonder if something in one of the files is modifying the structure and expects the order of the structure to always be a certain order.

adam

According to "doc/driver-model/README.txt" the use of 'U_BOOT_DEVICE' will the problematic going forward.

For now, I would expect something like the following would work, without modifying the structure again:

[board/logicpd/omap3som/omap3logic.c]
...

static struct ns16550_platdata omap3logic_serial = {
    OMAP34XX_UART1,
    0,
    2,
    V_NS16550_CLK
};

...

[board/ti/beagle/beagle.c]
...

static const struct ns16550_platdata beagle_serial = {
    OMAP34XX_UART3,
    0,
    2,
    V_NS16550_CLK
};

...

All static instances of the structure initialization would be incorrect without adding the new 'reg_offset' member after 'base'.

If 'U_BOOT_DEVICE' is not expected to work anymore, then efforts may need to be directed towards the new DM/FDT way.

Hi,

Quoted 183 lines · Derald D. Woods

Quoted 138 lines · Derald D. Woods

Quoted 138 lines · Derald D. Woods

Quoted 123 lines · Alexander Graf

Quoted 84 lines · Derald D. Woods

Quoted 59 lines · Michal Simek

Quoted 57 lines · Derald D. Woods

Quoted 54 lines · Michal Simek

Quoted 44 lines · Adam Ford

Quoted 36 lines

Quoted 32 lines · Simon Glass

Hi Michal,

Quoted 27 lines · Michal Simek

On 19.2.2016 21:55, Simon Glass wrote:

Hi Michal,

On 16 February 2016 at 08:17, Michal Simek
<[email protected]> wrote:

reg-offset is the part of standard 8250 binding in the kernel.
It is shifting start of address space by reg-offset.
On Xilinx platform this offset is typically 0x1000.

Signed-off-by: Michal Simek <[email protected]>
---

drivers/serial/ns16550.c | 6 ++++--
include/ns16550.h | 1 +
2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by: Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus
offset and
there is no reason to add one more option to Kconfig.
But let me know if you think that this is incorrect flow.

Quoted 5 lines

This patch seems to break my OMAP3 board. Does anyone know if I need
to set a certain offset for OMAP3 to make this work (and where is the
right place for it) ?

Quoted 5 lines

Are you using DT init? Check your DT description if there is reg-offset
property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 6 lines

Quoted 3 lines

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code
is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
   */
  struct ns16550_platdata {
         unsigned long base;
-       int reg_offset;
         int reg_shift;
         int clock;
+       int reg_offset;
  };
 
  struct udevice;

I see the following grep results:

$ grep -RI -e "const struct ns16550_platdata" .
./arch/arm/cpu/armv7/am33xx/board.c:static const struct ns16550_platdata
am33xx_serial[] = {
./arch/arm/cpu/arm926ejs/lpc32xx/devices.c:static const struct
ns16550_platdata lpc32xx_uart[] = {
./board/timll/devkit8000/devkit8000.c:static const struct
ns16550_platdata
devkit8000_serial = {
./board/ti/beagle/beagle.c:static const struct ns16550_platdata
beagle_serial = {
./board/logicpd/zoom1/zoom1.c:static const struct ns16550_platdata
zoom1_serial = {
./board/logicpd/omap3som/omap3logic.c:static const struct
ns16550_platdata
omap3logic_serial = {
./board/quipos/cairo/cairo.c:static const struct ns16550_platdata
cairo_serial = {
./board/lge/sniper/sniper.c:static const struct ns16550_platdata
serial_omap_platdata = {
./board/isee/igep00x0/igep00x0.c:static const struct ns16550_platdata
igep_serial = {
./board/overo/overo.c:static const struct ns16550_platdata overo_serial
=
{

Quoted 5 lines

Could the use of 'const' be a part of the problem?

- Derald

The structure initializers need rework for the additional member.

- Derald
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 11 lines · Adam Ford

I first tried removing the 'const' in the board file as suggested by Derald, but that wasn't successful. I can boot with Alexander's patch, but modifying the order inside the header seems weird to me. I haven't had any time to look this weekend, but I wonder if something in one of the files is modifying the structure and expects the order of the structure to always be a certain order.

adam

According to "doc/driver-model/README.txt" the use of 'U_BOOT_DEVICE' will the problematic going forward.

For now, I would expect something like the following would work, without modifying the structure again:

[board/logicpd/omap3som/omap3logic.c]
...

static struct ns16550_platdata omap3logic_serial = {
    OMAP34XX_UART1,
    0,
    2,
    V_NS16550_CLK
};

...

[board/ti/beagle/beagle.c]
...

static const struct ns16550_platdata beagle_serial = {
    OMAP34XX_UART3,
    0,
    2,
    V_NS16550_CLK
};

...

All static instances of the structure initialization would be incorrect without adding the new 'reg_offset' member after 'base'.

Quoted 2 lines

If 'U_BOOT_DEVICE' is not expected to work anymore, then efforts may need to be directed towards the new DM/FDT way.

Yes indeed.

For platform data I'd recommend adding member names.

Quoted 55 lines · Derald D. Woods

Quoted 53 lines · Michal Simek

Quoted 50 lines · Derald D. Woods

Quoted 38 lines · Michal Simek

Quoted 36 lines · Adam Ford

Quoted 32 lines · Simon Glass

Hi Michal,

Quoted 27 lines · Michal Simek

Quoted 13 lines · Simon Glass

Hi Michal,

Quoted 8 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 8 lines

Quoted 3 lines

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 6 lines

Quoted 3 lines

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

Hi,

Quoted 85 lines · Alexander Graf

Quoted 55 lines · Derald D. Woods

Quoted 53 lines · Michal Simek

Quoted 50 lines · Derald D. Woods

Quoted 38 lines · Michal Simek

Quoted 36 lines · Adam Ford

Quoted 32 lines · Simon Glass

Hi Michal,

Quoted 27 lines · Michal Simek

Quoted 13 lines · Simon Glass

Hi Michal,

Quoted 8 lines · Michal Simek

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>

---

Quoted 7 lines

Quoted 3 lines

  drivers/serial/ns16550.c | 6 ++++--
  include/ns16550.h        | 1 +
  2 files changed, 5 insertions(+), 2 deletions(-)

Reviewed-by Simon Glass <[email protected]>

Do you support the debug UART feature on your boards?

yes. I do support it but there you can put just address plus offset and
there is no reason to add one more option to Kconfig. But let me know if you think that this is incorrect flow.

Quoted 8 lines

Quoted 3 lines

This patch seems to break my OMAP3 board. Does anyone know if I need to set a certain offset for OMAP3 to make this work (and where is the right place for it) ?

Are you using DT init? Check your DT description if there is reg-offset property. I expect if your board worked before and you remove this property it will start to work again.

I am seeing the same problem with my BeagleBoard Rev. C4. There is something common, to more than one board, happening with this commit.

Quoted 6 lines

Quoted 3 lines

You should enable debug console and send the log.
Do you have enough space for malloc?

I will have little time this weekend to go further. Some things will need to be un-configured to have enough space. I am around 7 KiB over with DEBUG enabled.

I'm not quite sure what exactly is going wrong here - maybe some asm code is accessing the fields without proper offset generation?

Either way, the patch below seems to fix the issue for me (on beaglebone):

include/ns16550.h +1 -1

index 5eeacd6..1311f4c 100644
@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 

include/ns16550.h line 62

Quoted 2 lines

include/ns16550.h +0 -0

 struct udevice;

I picked up Alexander's suggestion: Still, we should fix everything up to use member names.

Applied to u-boot-dm/next, thanks!

- Simon
_______________________________________________
U-Boot mailing list
U-Boot at lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Quoted 14 lines

include/ns16550.h +1 -1

index 5eeacd6..1311f4c 100644
@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 

include/ns16550.h line 62

Quoted 2 lines

include/ns16550.h +0 -0

 struct udevice;

I picked up Alexander's suggestion: Still, we should fix everything up to use member names.

Applied to u-boot-dm/next, thanks!

Hi Simon,

Quoted 20 lines

Quoted 6 lines

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 
 struct udevice;

I picked up Alexander's suggestion: Still, we should fix everything up to use member names.

Quoted 1 line

Applied to u-boot-dm/next, thanks!

I can't see this patch applied to your next branch. Can you please check it?

Hi Michal,

Quoted 23 lines · Michal Simek

Hi Simon,

Quoted 20 lines

Quoted 6 lines

include/ns16550.h +0 -0

index 5eeacd6..1311f4c 100644

include/ns16550.h line 62

Quoted 10 lines

include/ns16550.h +1 -1

@@ -54,9 +54,9 @@
  */
 struct ns16550_platdata {
        unsigned long base;
-       int reg_offset;
        int reg_shift;
        int clock;
+       int reg_offset;
 };
 
 struct udevice;

I picked up Alexander's suggestion: Still, we should fix everything up to use member names.

Quoted 4 lines

Quoted 1 line

Applied to u-boot-dm/next, thanks!

I can't see this patch applied to your next branch. Can you please check it?

It's in mainline now - I have have updated the u-boot-dm/next branch already.

reg-offset is the part of standard 8250 binding in the kernel. It is shifting start of address space by reg-offset. On Xilinx platform this offset is typically 0x1000.

Signed-off-by Michal Simek <[email protected]>