Re: [PATCH for 4.21 v6 1/2] xen/riscv: drop CONFIG_RISCV_ISA_RV64G

[Oleksii Kurochko <oleksii.kurochko@xxxxxxxxx>]

On 2/18/25 6:03 PM, Jan Beulich wrote:

> On 12.02.2025 17:50, Oleksii Kurochko wrote:
> > --- a/xen/arch/riscv/Kconfig
> > +++ b/xen/arch/riscv/Kconfig
> > @@ -28,16 +28,6 @@ choice
> >  	help
> >  	  This selects the base ISA extensions that Xen will target.
> >  
> > -config RISCV_ISA_RV64G
> > -	bool "RV64G"
> > -	help
> > -	  Use the RV64I base ISA, plus
> > -	  "M" for multiply/divide,
> > -	  "A" for atomic instructions,
> > -	  “F”/"D" for  {single/double}-precision floating-point instructions,
> > -	  "Zicsr" for control and status register access,
> > -	  "Zifencei" for instruction-fetch fence.
> > -
> >  endchoice
> >       
> Shouldn't the choice be removed altogether then, for now being empty?
>     
Overlooked that, "Base ISA" choice could be removed too then. or just change to:
choice
	prompt "Base ISA"
	default "ima" if RISCV_64
	help
	  This selects the base ISA extensions that Xen will target.

endchoice

> > --- a/xen/arch/riscv/arch.mk
> > +++ b/xen/arch/riscv/arch.mk
> > @@ -6,8 +6,13 @@ $(call cc-options-add,CFLAGS,CC,$(EMBEDDED_EXTRA_CFLAGS))
> >  riscv-abi-$(CONFIG_RISCV_32) := -mabi=ilp32
> >  riscv-abi-$(CONFIG_RISCV_64) := -mabi=lp64
> >  
> > -riscv-march-$(CONFIG_RISCV_ISA_RV64G) := rv64g
> > -riscv-march-$(CONFIG_RISCV_ISA_C)       := $(riscv-march-y)c
> > +riscv-march-$(CONFIG_RISCV_64) := rv64
> > +
> > +riscv-march-y := $(riscv-march-y)ima
> > +
> > +riscv-march-$(CONFIG_RISCV_ISA_C) := $(riscv-march-y)c
> > +
> > +riscv-march-y := $(riscv-march-y)_zicsr_zifencei
> >       
> The repeated use of := makes this longer than necessary, and hence harder to
> read. I understand using += isn't exactly ideal either, because then on the rhs
> no blanks may appear (aiui), being kind of against our style and potentially
> hampering readability. Still maybe:
>
> riscv-march-$(CONFIG_RISCV_64) := rv64
> riscv-march-y+=ima
> riscv-march-$(CONFIG_RISCV_ISA_C)+=c
> riscv-march-y+=_zicsr_zifencei
>
> ?
>     
From my point of view both options hard to read but `+=`, at the moment, look a
little bit better. I will update correspondingly.

Thanks.

~ Oleksii