[PATCHv3 08/11] arm64: Check for selected granule support
mark.rutland at arm.com
Wed Oct 14 10:24:18 PDT 2015
> @@ -613,10 +614,28 @@ ENDPROC(__secondary_switched)
> * x0 = SCTLR_EL1 value for turning on the MMU.
> * x27 = *virtual* address to jump to upon completion
> - * other registers depend on the function called upon completion
> + * Other registers depend on the function called upon completion.
> + *
> + * Checks if the selected granule size is supported by the CPU.
> + * If it doesn't park the CPU
Nit: "If it isn't, park the CPU."
> +#if defined(CONFIG_ARM64_64K_PAGES)
> +#define id_aa64mmfr0_tgran_shift ID_AA64MMFR0_TGRAN64_SHIFT
> +#define id_aa64mmfr0_tgran_on ID_AA64MMFR0_TGRAN64_ON
> +#define id_aa64mmfr0_tgran_shift ID_AA64MMFR0_TGRAN4_SHIFT
> +#define id_aa64mmfr0_tgran_on ID_AA64MMFR0_TGRAN4_ON
Any reason for not using upper-case names for the macros?
Given they're local you could just call them TGRAN_SHIFT and TRGRAN_ON
to make the asm slightly nicer.
> .section ".idmap.text", "ax"
> + mrs x1, ID_AA64MMFR0_EL1
> + ubfx x2, x1, #id_aa64mmfr0_tgran_shift, 4
> + cmp x2, #id_aa64mmfr0_tgran_on
> + b.ne __no_granule_support
> ldr x5, =vectors
> msr vbar_el1, x5
> msr ttbr0_el1, x25 // load TTBR0
> @@ -634,3 +653,8 @@ __enable_mmu:
> br x27
> + wfe
> + b __no_granule_support
Other than the above, this loogs fine to me.
In future it would be nice if we could somehow signal that these dead
CPUs are trapped in the kernel -- we should have some kind of canary
mechanism for that. That needn't block this patch, though.
More information about the linux-arm-kernel