| Message ID | 1486985692-26625-1-git-send-email-uri.mashiach@compulab.co.il |
|---|---|
| State | Superseded |
| Delegated to: | Marek Vasut |
| Headers | show |
On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: > The SOC family symbol CONFIG_AM57XX was removed by the commit > 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". > > The symbol is needed by the XHCI OMAP USB driver. > Without the symbol all the AM57x targets symbols should be ored in the > ifdef'ery. > > Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> We need to introduce it into Kconfig files now, thanks!
On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: > On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: > >> The SOC family symbol CONFIG_AM57XX was removed by the commit >> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". >> >> The symbol is needed by the XHCI OMAP USB driver. >> Without the symbol all the AM57x targets symbols should be ored in the >> ifdef'ery. >> >> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> > > We need to introduce it into Kconfig files now, thanks! DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more to do which port to use rather than having anything to do with DRA7 Vs AM57xx. Regards, Nishanth Menon
On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: > On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: > > On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: > > > >> The SOC family symbol CONFIG_AM57XX was removed by the commit > >> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". > >> > >> The symbol is needed by the XHCI OMAP USB driver. > >> Without the symbol all the AM57x targets symbols should be ored in the > >> ifdef'ery. > >> > >> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> > > > > We need to introduce it into Kconfig files now, thanks! > > DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more > to do which port to use rather than having anything to do with DRA7 Vs > AM57xx. Yes, agreed. And we need to solve this, and introduce some Kconfig CONFIG symbol to deal with it. Maybe the better way to look at this is, can we have DRA7xx boards configured in the way Uri wants to be able to configure the AM57xx board he's working with? And vice versa? The answer to that will help drive what the right CONFIG_xxx name is.
On 02/14/2017 01:20 PM, Tom Rini wrote: > On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: >> On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: >>> On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: >>> >>>> The SOC family symbol CONFIG_AM57XX was removed by the commit >>>> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". >>>> >>>> The symbol is needed by the XHCI OMAP USB driver. >>>> Without the symbol all the AM57x targets symbols should be ored in the >>>> ifdef'ery. >>>> >>>> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> >>> >>> We need to introduce it into Kconfig files now, thanks! >> >> DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more >> to do which port to use rather than having anything to do with DRA7 Vs >> AM57xx. > > Yes, agreed. And we need to solve this, and introduce some Kconfig > CONFIG symbol to deal with it. Maybe the better way to look at this is, > can we have DRA7xx boards configured in the way Uri wants to be able to > configure the AM57xx board he's working with? And vice versa? The > answer to that will help drive what the right CONFIG_xxx name is. Should'nt be ideally done via DM model instead? I have no understanding of USB integration done in u-boot to comment better.
On Tue, Feb 14, 2017 at 01:26:24PM -0600, Nishanth Menon wrote: > On 02/14/2017 01:20 PM, Tom Rini wrote: > >On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: > >>On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: > >>>On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: > >>> > >>>>The SOC family symbol CONFIG_AM57XX was removed by the commit > >>>>3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". > >>>> > >>>>The symbol is needed by the XHCI OMAP USB driver. > >>>>Without the symbol all the AM57x targets symbols should be ored in the > >>>>ifdef'ery. > >>>> > >>>>Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> > >>> > >>>We need to introduce it into Kconfig files now, thanks! > >> > >>DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more > >>to do which port to use rather than having anything to do with DRA7 Vs > >>AM57xx. > > > >Yes, agreed. And we need to solve this, and introduce some Kconfig > >CONFIG symbol to deal with it. Maybe the better way to look at this is, > >can we have DRA7xx boards configured in the way Uri wants to be able to > >configure the AM57xx board he's working with? And vice versa? The > >answer to that will help drive what the right CONFIG_xxx name is. > > Should'nt be ideally done via DM model instead? I have no > understanding of USB integration done in u-boot to comment better. That's also something that could be explored.
On 14/02/17 22:12, Tom Rini wrote: > On Tue, Feb 14, 2017 at 01:26:24PM -0600, Nishanth Menon wrote: >> On 02/14/2017 01:20 PM, Tom Rini wrote: >>> On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: >>>> On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: >>>>> On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: >>>>> >>>>>> The SOC family symbol CONFIG_AM57XX was removed by the commit >>>>>> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". >>>>>> >>>>>> The symbol is needed by the XHCI OMAP USB driver. >>>>>> Without the symbol all the AM57x targets symbols should be ored in the >>>>>> ifdef'ery. >>>>>> >>>>>> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> >>>>> >>>>> We need to introduce it into Kconfig files now, thanks! >>>> >>>> DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more >>>> to do which port to use rather than having anything to do with DRA7 Vs >>>> AM57xx. >>> >>> Yes, agreed. And we need to solve this, and introduce some Kconfig >>> CONFIG symbol to deal with it. Maybe the better way to look at this is, >>> can we have DRA7xx boards configured in the way Uri wants to be able to >>> configure the AM57xx board he's working with? And vice versa? The >>> answer to that will help drive what the right CONFIG_xxx name is. >> >> Should'nt be ideally done via DM model instead? I have no >> understanding of USB integration done in u-boot to comment better. > > That's also something that could be explored. > Yes, ideally this should be taken care of by DM. Till that migration is complete I'm fine with fixing this temporarily. OMAP5, DRA7, AM57 and AM43 use different XHCI_BASE address.
On 15/02/17 10:24, Roger Quadros wrote: > On 14/02/17 22:12, Tom Rini wrote: >> On Tue, Feb 14, 2017 at 01:26:24PM -0600, Nishanth Menon wrote: >>> On 02/14/2017 01:20 PM, Tom Rini wrote: >>>> On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: >>>>> On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: >>>>>> On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: >>>>>> >>>>>>> The SOC family symbol CONFIG_AM57XX was removed by the commit >>>>>>> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". >>>>>>> >>>>>>> The symbol is needed by the XHCI OMAP USB driver. >>>>>>> Without the symbol all the AM57x targets symbols should be ored in the >>>>>>> ifdef'ery. >>>>>>> >>>>>>> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> >>>>>> >>>>>> We need to introduce it into Kconfig files now, thanks! >>>>> >>>>> DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more >>>>> to do which port to use rather than having anything to do with DRA7 Vs >>>>> AM57xx. >>>> >>>> Yes, agreed. And we need to solve this, and introduce some Kconfig >>>> CONFIG symbol to deal with it. Maybe the better way to look at this is, >>>> can we have DRA7xx boards configured in the way Uri wants to be able to >>>> configure the AM57xx board he's working with? And vice versa? The >>>> answer to that will help drive what the right CONFIG_xxx name is. >>> >>> Should'nt be ideally done via DM model instead? I have no >>> understanding of USB integration done in u-boot to comment better. >> >> That's also something that could be explored. >> > > Yes, ideally this should be taken care of by DM. > Till that migration is complete I'm fine with fixing this temporarily. > > OMAP5, DRA7, AM57 and AM43 use different XHCI_BASE address. > Just to clarify further on this DRA7 and AM57 are identical SoCs but the EVMs are using different instances of the USB as host so the difference in XHCI_BASE. DRA7-evm uses USB1 instance whereas AM57x-EVM uses USB0 instance as host.
On 02/15/2017 01:30 PM, Roger Quadros wrote: > On 15/02/17 10:24, Roger Quadros wrote: >> On 14/02/17 22:12, Tom Rini wrote: >>> On Tue, Feb 14, 2017 at 01:26:24PM -0600, Nishanth Menon wrote: >>>> On 02/14/2017 01:20 PM, Tom Rini wrote: >>>>> On Tue, Feb 14, 2017 at 01:18:20PM -0600, menon.nishanth@gmail.com wrote: >>>>>> On Mon, Feb 13, 2017 at 7:27 AM, Tom Rini <trini@konsulko.com> wrote: >>>>>>> On Mon, Feb 13, 2017 at 01:34:52PM +0200, Uri Mashiach wrote: >>>>>>> >>>>>>>> The SOC family symbol CONFIG_AM57XX was removed by the commit >>>>>>>> 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". >>>>>>>> >>>>>>>> The symbol is needed by the XHCI OMAP USB driver. >>>>>>>> Without the symbol all the AM57x targets symbols should be ored in the >>>>>>>> ifdef'ery. >>>>>>>> >>>>>>>> Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> >>>>>>> >>>>>>> We need to introduce it into Kconfig files now, thanks! >>>>>> >>>>>> DRA7 and AM57xx are literally the same. the #ifdeffery in USB is more >>>>>> to do which port to use rather than having anything to do with DRA7 Vs >>>>>> AM57xx. >>>>> >>>>> Yes, agreed. And we need to solve this, and introduce some Kconfig >>>>> CONFIG symbol to deal with it. Maybe the better way to look at this is, >>>>> can we have DRA7xx boards configured in the way Uri wants to be able to >>>>> configure the AM57xx board he's working with? And vice versa? The >>>>> answer to that will help drive what the right CONFIG_xxx name is. >>>> >>>> Should'nt be ideally done via DM model instead? I have no >>>> understanding of USB integration done in u-boot to comment better. >>> >>> That's also something that could be explored. >>> >> >> Yes, ideally this should be taken care of by DM. >> Till that migration is complete I'm fine with fixing this temporarily. >> >> OMAP5, DRA7, AM57 and AM43 use different XHCI_BASE address. >> > Just to clarify further on this > DRA7 and AM57 are identical SoCs but the EVMs are using different > instances of the USB as host so the difference in XHCI_BASE. > > DRA7-evm uses USB1 instance whereas AM57x-EVM uses USB0 instance as host. > Alternative solution: * Definition of registers base groups for USB0 and USB1. * Target based definition of XHCI_USB0 or XHCI_USB1. * Selection of register base group based on the definition of XHCI_USB0 or XHCI_USB1.
diff --git a/include/configs/am57xx_evm.h b/include/configs/am57xx_evm.h index 3d8b996..2896628 100644 --- a/include/configs/am57xx_evm.h +++ b/include/configs/am57xx_evm.h @@ -15,6 +15,7 @@ #include <environment/ti/dfu.h> #define CONFIG_DRA7XX +#define CONFIG_AM57XX #ifdef CONFIG_SPL_BUILD #define CONFIG_IODELAY_RECALIBRATION diff --git a/include/configs/cl-som-am57x.h b/include/configs/cl-som-am57x.h index e1f724b..4d1ceff 100644 --- a/include/configs/cl-som-am57x.h +++ b/include/configs/cl-som-am57x.h @@ -12,6 +12,7 @@ #define __CONFIG_CL_SOM_AM57X_H #define CONFIG_DRA7XX +#define CONFIG_AM57XX #define CONFIG_NR_DRAM_BANKS 2 diff --git a/include/linux/usb/xhci-omap.h b/include/linux/usb/xhci-omap.h index 9de80d7..ba21d60 100644 --- a/include/linux/usb/xhci-omap.h +++ b/include/linux/usb/xhci-omap.h @@ -14,7 +14,7 @@ #define OMAP_XHCI_BASE 0x488d0000 #define OMAP_OCP1_SCP_BASE 0x4A081000 #define OMAP_OTG_WRAPPER_BASE 0x488c0000 -#elif defined CONFIG_TARGET_AM57XX_EVM +#elif defined CONFIG_AM57XX #define OMAP_XHCI_BASE 0x48890000 #define OMAP_OCP1_SCP_BASE 0x4A084c00 #define OMAP_OTG_WRAPPER_BASE 0x48880000 diff --git a/scripts/config_whitelist.txt b/scripts/config_whitelist.txt index d21589b..7c7f372 100644 --- a/scripts/config_whitelist.txt +++ b/scripts/config_whitelist.txt @@ -85,6 +85,7 @@ CONFIG_AM335X_USB0_MODE CONFIG_AM335X_USB1 CONFIG_AM335X_USB1_MODE CONFIG_AM437X_USB2PHY2_HOST +CONFIG_AM57XX CONFIG_AMBAPP_IOAREA CONFIG_AMCC_DEF_ENV CONFIG_AMCC_DEF_ENV_NOR_UPD
The SOC family symbol CONFIG_AM57XX was removed by the commit 3891a54: "ARM: DRA7x/AM57xx: Get rid of CONFIG_AM57XX". The symbol is needed by the XHCI OMAP USB driver. Without the symbol all the AM57x targets symbols should be ored in the ifdef'ery. Signed-off-by: Uri Mashiach <uri.mashiach@compulab.co.il> --- V1 -> V2: Replace the commit "fix XHCI registers base address". include/configs/am57xx_evm.h | 1 + include/configs/cl-som-am57x.h | 1 + include/linux/usb/xhci-omap.h | 2 +- scripts/config_whitelist.txt | 1 + 4 files changed, 4 insertions(+), 1 deletion(-)