Mason <slash.tmp@xxxxxxx> writes: > On 04/11/2015 16:47, Måns Rullgård wrote: > >> Mason writes: >> >>> On 27/10/2015 13:57, Mans Rullgard wrote: >>> >>>> Some SoCs have a Palmchip UART with a non-standard register layout. >>>> This allows the debug console to work with these. >>>> >>>> Signed-off-by: Mans Rullgard <mans@xxxxxxxxx> >>>> --- >>>> arch/arm/Kconfig.debug | 8 ++++++++ >>>> arch/arm/include/debug/8250.S | 12 ++++++++++++ >>>> 2 files changed, 20 insertions(+) >>>> >>>> diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug >>>> index 0cfd7f9..9039fff 100644 >>>> --- a/arch/arm/Kconfig.debug >>>> +++ b/arch/arm/Kconfig.debug >>>> @@ -1597,6 +1597,14 @@ config DEBUG_UART_8250_WORD >>>> DEBUG_BCM_KONA_UART || DEBUG_RK32_UART2 || \ >>>> DEBUG_BRCMSTB_UART >>>> >>>> +config DEBUG_UART_8250_PALMCHIP >>>> + bool "8250 UART is Palmchip variant" >>>> + depends on DEBUG_LL_UART_8250 || DEBUG_UART_8250 >>> >>> I am 100% clueless about the UART sequence in Linux. >>> >>> There's earlyprintk, the boot console (polled?), the "full-blown" >>> console (with IRQs)... >>> >>> If I want earlyprintk, I have to enable DEBUG_LL (?) so I'll have >>> DEBUG_LL_UART_8250, right? There's no way to pick DEBUG_UART_8250 >>> in menuconfig, is there? >>> >>> So I picked the following options: >>> >>> CONFIG_DEBUG_LL=y >>> # CONFIG_DEBUG_ICEDCC is not set >>> # CONFIG_DEBUG_SEMIHOSTING is not set >>> CONFIG_DEBUG_LL_UART_8250=y >>> # CONFIG_DEBUG_LL_UART_PL01X is not set >>> CONFIG_DEBUG_LL_INCLUDE="debug/8250.S" >>> # CONFIG_DEBUG_UART_8250 is not set >>> # CONFIG_DEBUG_UART_BCM63XX is not set >>> CONFIG_DEBUG_UART_PHYS=0x10700 >>> CONFIG_DEBUG_UART_VIRT=0xf0010700 >>> CONFIG_DEBUG_UART_8250_SHIFT=2 >>> CONFIG_DEBUG_UART_8250_WORD=y >>> CONFIG_DEBUG_UART_8250_PALMCHIP=y >>> # CONFIG_DEBUG_UART_8250_FLOW_CONTROL is not set >>> CONFIG_DEBUG_UNCOMPRESS=y >>> CONFIG_UNCOMPRESS_INCLUDE="debug/uncompress.h" >>> CONFIG_EARLY_PRINTK=y >> >> Those are the settings I use. >> >>> I suppose I have to provide DEBUG_UART_PHYS because it needs >>> the address before it can parse the DT? But how am I supposed >>> to know the virtual address? Isn't that from iomap(UART_PHYS)? >> >> The decompression code and early setup don't have that luxury. > > But how am I supposed to compute the virtual address? > Can I set DEBUG_UART_VIRT to 0xdeadbeef? :-) Add 0xf0000000. > Also is it expected that setting DEBUG_LL forces make to recompile > every source file in the source tree? Apparently. :) -- Måns Rullgård mans@xxxxxxxxx -- To unsubscribe from this list: send the line "unsubscribe linux-serial" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
