Re: [Xen-devel] [PATCH for-4.13] xen/arm: Add Skeleton for using configuring early printk using Kconfig

[Julien Grall <julien.grall@xxxxxxx>]

Hi Stefano,

On 10/1/19 8:33 PM, Stefano Stabellini wrote:
> On Fri, 13 Sep 2019, Julien Grall wrote:
> > At the moment, early printk can only be configured on the make command
> > line. It is not very handy because a user has to remove the option
> > everytime it is using another command other than compiling the
> > hypervisor.
> >
> > Furthermore, early printk is one of the few odds one that are not using
> > Kconfig.
> >
> > So this is about time to move it to Kconfig. For now, a skeleton is
> > added with one example based on Cadence UART. Follow-up will continue to
> > convert all the options to Kconfig.
> >
> > Because Kconfig will prefix all the config by CONFIG_, it is necessary
> > to adapt the define within the code.
> >
> > Signed-off-by: Julien Grall <julien.grall@xxxxxxx>
> >
> > ---
> >
> > I have sent it as RFC because this is not complete. I will convert the
> > rest once we agree the approach is correct.
> > ---
> >   xen/Kconfig.debug                  |  2 ++
> >   xen/arch/arm/Kconfig.debug         | 40 ++++++++++++++++++++++++++++++++++++++
> >   xen/arch/arm/Rules.mk              |  5 ++---
> >   xen/arch/arm/arm32/head.S          |  8 ++++----
> >   xen/arch/arm/arm64/debug.S         |  4 ++--
> >   xen/arch/arm/arm64/head.S          |  8 ++++----
> >   xen/arch/x86/Kconfig.debug         |  0
> >   xen/include/asm-arm/early_printk.h |  2 +-
> >   8 files changed, 55 insertions(+), 14 deletions(-)
> >   create mode 100644 xen/arch/arm/Kconfig.debug
> >   create mode 100644 xen/arch/x86/Kconfig.debug
> >
> > diff --git a/xen/Kconfig.debug b/xen/Kconfig.debug
> > index e10e314e25..d0806dba32 100644
> > --- a/xen/Kconfig.debug
> > +++ b/xen/Kconfig.debug
> > @@ -112,6 +112,8 @@ config XMEM_POOL_POISON
> >           Poison free blocks with 0xAA bytes and verify them when a block is
> >           allocated in order to spot use-after-free issues.
> >   
> > +source "arch/$SRCARCH/Kconfig.debug"
> > +
> >   endif # DEBUG || EXPERT
> >   
> >   endmenu
> > diff --git a/xen/arch/arm/Kconfig.debug b/xen/arch/arm/Kconfig.debug
> > new file mode 100644
> > index 0000000000..bc3b622695
> > --- /dev/null
> > +++ b/xen/arch/arm/Kconfig.debug
> > @@ -0,0 +1,40 @@
> > +choice
> > +       prompt "Enable early printk"
> > +
> > +       optional
> > +       config EARLY_PRINTK_ZYNQMP
> > +               bool "Enable early printk on Xilinx ZynQMP"
> > +               select EARLY_UART_CADENCE
> > +               help
> > +                 Say Y here if you want the early printk support on Xilinx
> > +                 ZynQMP platform.
> > +
> > +       config EARLY_PRINTK_CADENCE
> > +               bool "Enable early printk via Cadence UART"
> > +               select EARLY_UART_CADENCE
> Why not select EARLY_PRINTK directly? Is EARLY_UART_CADENCE actually
> needed?
Yes to select the proper EARLY_PRINTK_INC. For other UARTs (such as 
8250), it will be necessary to enable more options.
The other option is to have lengthy if in some part of the Kconfig which 
is not pretty and likely a way to miss some places if we ever decide to 
add more alias (I hope not!).
> > +               help
> > +                 Say Y here if you wish the early printk to direct their
> > +                 output to a Cadence UART. You can use this option to provide
> > +                 the parameters for the Cadence UART rather than selecting
> > +                 one of the platform specific options above if you know the
> > +                 parameters for the port.
> > +
> > +                 This option is preferred over the platform specific options;
> > +                 the platform specific options are deprecated and will soon
> > +                 be removed.
> > +endchoice
> > +
> > +config EARLY_PRINTK
> > +       bool
> > +
> > +config EARLY_UART_CADENCE
> > +       bool
> > +       select EARLY_PRINTK
> > +
> > +config EARLY_UART_BASE_ADDRESS
> > +       hex "Physical base address of debug UART" if EARLY_PRINTK
> > +       default 0xff000000 if EARLY_PRINTK_ZYNQMP
> This only works for EARLY_PRINTK_ZYNQMP and not for
> EARLY_PRINTK_CADENCE. I imagine we need a default line for each of the
> possible options above.
If you don't set any default value, then it will be requested when 
compiling Xen. The risk of setting a default value for "generic" config 
is the user may not notice that it needs to fill it up.
So he/she may end up to compile Xen with the wrong address and will get 
not output later on at best. This sort of things is quite difficult to 
debug until you know what you are doing.
> > +config EARLY_PRINTK_INC
> > +       string
> > +       default "debug-cadence.inc" if EARLY_UART_CADENCE
> > diff --git a/xen/arch/arm/Rules.mk b/xen/arch/arm/Rules.mk
> > index 3d9a0ed357..084f1725a8 100644
> > --- a/xen/arch/arm/Rules.mk
> > +++ b/xen/arch/arm/Rules.mk
> > @@ -46,7 +46,6 @@ EARLY_PRINTK_thunderx       := pl011,0x87e024000000
> >   EARLY_PRINTK_vexpress       := pl011,0x1c090000
> >   EARLY_PRINTK_xgene-mcdivitt := 8250,0x1c021000,2
> >   EARLY_PRINTK_xgene-storm    := 8250,0x1c020000,2
> > -EARLY_PRINTK_zynqmp         := cadence,0xff000000
> >   
> >   ifneq ($(EARLY_PRINTK_$(CONFIG_EARLY_PRINTK)),)
> >   EARLY_PRINTK_CFG := $(subst $(comma), ,$(EARLY_PRINTK_$(CONFIG_EARLY_PRINTK)))
> > @@ -82,9 +81,9 @@ endif
> >   
> >   CFLAGS-$(EARLY_PRINTK) += -DCONFIG_EARLY_PRINTK
> >   CFLAGS-$(EARLY_PRINTK_INIT_UART) += -DEARLY_PRINTK_INIT_UART
> > -CFLAGS-$(EARLY_PRINTK) += -DEARLY_PRINTK_INC=\"debug-$(EARLY_PRINTK_INC).inc\"
> > +CFLAGS-$(EARLY_PRINTK) += 
> > -DCONFIG_EARLY_PRINTK_INC=\"debug-$(EARLY_PRINTK_INC).inc\"
> >   CFLAGS-$(EARLY_PRINTK) += -DEARLY_PRINTK_BAUD=$(EARLY_PRINTK_BAUD)
> > -CFLAGS-$(EARLY_PRINTK) += -DEARLY_UART_BASE_ADDRESS=$(EARLY_UART_BASE_ADDRESS)
> > +CFLAGS-$(EARLY_PRINTK) += 
> > -DCONFIG_EARLY_UART_BASE_ADDRESS=$(EARLY_UART_BASE_ADDRESS)
> I don't have an opinion on the naming, the following is a suggestion to
> make the patch easier to write.
I got bored enough on the plan to try to do more cleanup at the same 
time (see below). ;)
>  It looks like we could get away without
> any of the renaming below if we choose to export
> EARLY_UART_BASE_ADDRESS and keep using EARLY_UART_BASE_ADDRESS directly
> in the code below.
I want to get rid of anything not starting with CONFIG_ because this is 
misleading. So they either got drop in this patch (and follow-ups) or 
they get renamed before hand.
I chose the former because some options will not be a straight prefixing 
of CONFIG_. For instance, EARLY_UART_REG_SHIFT is a specific to 8250, so 
I would like to rename it CONFIG_EARLY_UART_8250_SHIFT.
I can try to do the renaming before hand if you prefer.

Cheers,

--
Julien Grall

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel