Re: [PATCH 1/2] kexec: fix KEXEC_FILE dependencies

[Baoquan He]

Just add people and mailing list to CC since I didn't find this mail in
my box, just drag it via 'b4 am'.

On 10/23/23 at 01:01pm, Arnd Bergmann wrote:
......
> ---
>  arch/powerpc/Kconfig | 4 ++--
>  arch/riscv/Kconfig   | 4 +---
>  arch/s390/Kconfig    | 4 ++--
>  arch/x86/Kconfig     | 4 ++--
>  kernel/Kconfig.kexec | 1 +
>  5 files changed, 8 insertions(+), 9 deletions(-)
> 
> diff --git a/arch/powerpc/Kconfig b/arch/powerpc/Kconfig
> index d5d5388973ac7..4640cee33f123 100644
> --- a/arch/powerpc/Kconfig
> +++ b/arch/powerpc/Kconfig
> @@ -607,10 +607,10 @@ config ARCH_SUPPORTS_KEXEC
>  	def_bool PPC_BOOK3S || PPC_E500 || (44x && !SMP)
>  
>  config ARCH_SUPPORTS_KEXEC_FILE
> -	def_bool PPC64 && CRYPTO=y && CRYPTO_SHA256=y
> +	def_bool PPC64
>  
>  config ARCH_SUPPORTS_KEXEC_PURGATORY
> -	def_bool KEXEC_FILE
> +	def_bool y
>  
>  config ARCH_SELECTS_KEXEC_FILE
>  	def_bool y
> diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
> index 25474f8c12b79..f571bad2d22d0 100644
> --- a/arch/riscv/Kconfig
> +++ b/arch/riscv/Kconfig
> @@ -687,9 +687,7 @@ config ARCH_SELECTS_KEXEC_FILE
>  	select KEXEC_ELF
>  
>  config ARCH_SUPPORTS_KEXEC_PURGATORY
> -	def_bool KEXEC_FILE
> -	depends on CRYPTO=y
> -	depends on CRYPTO_SHA256=y
> +	def_bool y
>  
>  config ARCH_SUPPORTS_CRASH_DUMP
>  	def_bool y
> diff --git a/arch/s390/Kconfig b/arch/s390/Kconfig
> index b0d67ac8695f9..ec77106af4137 100644
> --- a/arch/s390/Kconfig
> +++ b/arch/s390/Kconfig
> @@ -253,13 +253,13 @@ config ARCH_SUPPORTS_KEXEC
>  	def_bool y
>  
>  config ARCH_SUPPORTS_KEXEC_FILE
> -	def_bool CRYPTO && CRYPTO_SHA256 && CRYPTO_SHA256_S390
> +	def_bool y
>  
>  config ARCH_SUPPORTS_KEXEC_SIG
>  	def_bool MODULE_SIG_FORMAT
>  
>  config ARCH_SUPPORTS_KEXEC_PURGATORY
> -	def_bool KEXEC_FILE
> +	def_bool y
>  
>  config ARCH_SUPPORTS_CRASH_DUMP
>  	def_bool y
> diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig
> index 94efde80ebf35..f9975b15ccd57 100644
> --- a/arch/x86/Kconfig
> +++ b/arch/x86/Kconfig
> @@ -2073,7 +2073,7 @@ config ARCH_SUPPORTS_KEXEC
>  	def_bool y
>  
>  config ARCH_SUPPORTS_KEXEC_FILE
> -	def_bool X86_64 && CRYPTO && CRYPTO_SHA256
> +	def_bool X86_64
>  
>  config ARCH_SELECTS_KEXEC_FILE
>  	def_bool y
> @@ -2081,7 +2081,7 @@ config ARCH_SELECTS_KEXEC_FILE
>  	select HAVE_IMA_KEXEC if IMA
>  
>  config ARCH_SUPPORTS_KEXEC_PURGATORY
> -	def_bool KEXEC_FILE
> +	def_bool y
>  
>  config ARCH_SUPPORTS_KEXEC_SIG
>  	def_bool y
> diff --git a/kernel/Kconfig.kexec b/kernel/Kconfig.kexec
> index 7aff28ded2f48..bfc636d64ff2b 100644
> --- a/kernel/Kconfig.kexec
> +++ b/kernel/Kconfig.kexec
> @@ -36,6 +36,7 @@ config KEXEC
>  config KEXEC_FILE
>  	bool "Enable kexec file based system call"
>  	depends on ARCH_SUPPORTS_KEXEC_FILE
> +	depends on CRYPTO_SHA256=y || !ARCH_SUPPORTS_KEXEC_PURGATORY

I am not sure if the logic is correct. In theory, kexec_file code
utilizes purgatory to verify the checksum digested during kernel loading
when try to jump to the kernel. That means kexec_file depends on
purgatory, but not contrary?

With these changes, we can achieve the goal to avoid building issue,
whereas the code logic becomes confusing. E.g people could disable
CONFIG_KEXEC_FILE, but still get purgatory code built in which is
totally useless.

Not sure if I think too much over this.

>  	select KEXEC_CORE
>  	help
>  	  This is new version of kexec system call. This system call is
> -- 
> 2.39.2