Re: [PATCH] RISC-V: hwprobe: Add MISALIGNED_PERF key

[Andrew Jones]

On Wed, May 29, 2024 at 11:26:48AM GMT, Evan Green wrote:
> RISCV_HWPROBE_KEY_CPUPERF_0 was mistakenly flagged as a bitmask in
> hwprobe_key_is_bitmask(), when in reality it was an enum value. This
> causes problems when used in conjunction with RISCV_HWPROBE_WHICH_CPUS,
> since SLOW, FAST, and EMULATED have values whose bits overlap with
> each other. If the caller asked for the set of CPUs that was SLOW or
> EMULATED, the returned set would also include CPUs that were FAST.
> 
> Introduce a new hwprobe key, RISCV_HWPROBE_KEY_MISALIGNED_PERF, which
> returns the same values in response to a direct query (with no flags),
> but is properly handled as an enumerated value. As a result, SLOW,
> FAST, and EMULATED are all correctly treated as distinct values under
> the new key when queried with the WHICH_CPUS flag.
> 
> Leave the old key in place to avoid disturbing applications which may
> have already come to rely on the broken behavior.

I appreciate the paranoia, even if I think we could probably get away
with fixing CPUPERF_0.

> 
> Fixes: e178bf146e4b ("RISC-V: hwprobe: Introduce which-cpus flag")
> Signed-off-by: Evan Green <evan@xxxxxxxxxxxx>
> 
> ---
> 
> 
> Note: Yangyu also has a fix out for this issue at [1]. That fix is much
> tidier, but comes with the slight risk that some very broken userspace
> application may break now that FAST cpus are not included for the query
> of which cpus are SLOW or EMULATED. I wanted to get this fix out so that
> we have both as options, and can discuss. These fixes are mutually
> exclusive, don't take both.
> 
> [1] https://lore.kernel.org/linux-riscv/tencent_01F8E0050FB4B11CC170C3639E43F41A1709@xxxxxx/
> 
> ---
>  Documentation/arch/riscv/hwprobe.rst  | 8 ++++++--
>  arch/riscv/include/asm/hwprobe.h      | 2 +-
>  arch/riscv/include/uapi/asm/hwprobe.h | 1 +
>  arch/riscv/kernel/sys_hwprobe.c       | 1 +
>  4 files changed, 9 insertions(+), 3 deletions(-)
> 
> diff --git a/Documentation/arch/riscv/hwprobe.rst b/Documentation/arch/riscv/hwprobe.rst
> index 204cd4433af5..616ee372adaf 100644
> --- a/Documentation/arch/riscv/hwprobe.rst
> +++ b/Documentation/arch/riscv/hwprobe.rst
> @@ -192,8 +192,12 @@ The following keys are defined:
>         supported as defined in the RISC-V ISA manual starting from commit
>         d8ab5c78c207 ("Zihintpause is ratified").
>  
> -* :c:macro:`RISCV_HWPROBE_KEY_CPUPERF_0`: A bitmask that contains performance
> -  information about the selected set of processors.
> +* :c:macro:`RISCV_HWPROBE_KEY_CPUPERF_0`: Deprecated. Returns similar values to
> +     :c:macro:`RISCV_HWPROBE_KEY_MISALIGNED_PERF`, but the key was mistakenly
> +     classified as a bitmask rather than a value.
> +
> +* :c:macro:`RISCV_HWPROBE_KEY_MISALIGNED_PERF`: An enum value describing the
> +  performance of misaligned scalar accesses on the selected set of processors.
>  
>    * :c:macro:`RISCV_HWPROBE_MISALIGNED_UNKNOWN`: The performance of misaligned
>      accesses is unknown.
> diff --git a/arch/riscv/include/asm/hwprobe.h b/arch/riscv/include/asm/hwprobe.h
> index 630507dff5ea..150a9877b0af 100644
> --- a/arch/riscv/include/asm/hwprobe.h
> +++ b/arch/riscv/include/asm/hwprobe.h
> @@ -8,7 +8,7 @@
>  
>  #include <uapi/asm/hwprobe.h>
>  
> -#define RISCV_HWPROBE_MAX_KEY 6
> +#define RISCV_HWPROBE_MAX_KEY 7
>  
>  static inline bool riscv_hwprobe_key_is_valid(__s64 key)
>  {
> diff --git a/arch/riscv/include/uapi/asm/hwprobe.h b/arch/riscv/include/uapi/asm/hwprobe.h
> index dda76a05420b..bc34e33fef23 100644
> --- a/arch/riscv/include/uapi/asm/hwprobe.h
> +++ b/arch/riscv/include/uapi/asm/hwprobe.h
> @@ -68,6 +68,7 @@ struct riscv_hwprobe {
>  #define		RISCV_HWPROBE_MISALIGNED_UNSUPPORTED	(4 << 0)
>  #define		RISCV_HWPROBE_MISALIGNED_MASK		(7 << 0)

Can we also remove the unnecessary ( << 0) shifts for each of the
MISALIGNED_* values? The shifts imply bits of a bitmask (to me).

>  #define RISCV_HWPROBE_KEY_ZICBOZ_BLOCK_SIZE	6
> +#define RISCV_HWPROBE_KEY_MISALIGNED_PERF	7
>  /* Increase RISCV_HWPROBE_MAX_KEY when adding items. */
>  
>  /* Flags */
> diff --git a/arch/riscv/kernel/sys_hwprobe.c b/arch/riscv/kernel/sys_hwprobe.c
> index 969ef3d59dbe..c8b7d57eb55e 100644
> --- a/arch/riscv/kernel/sys_hwprobe.c
> +++ b/arch/riscv/kernel/sys_hwprobe.c
> @@ -208,6 +208,7 @@ static void hwprobe_one_pair(struct riscv_hwprobe *pair,
>  		break;
>  
>  	case RISCV_HWPROBE_KEY_CPUPERF_0:
> +	case RISCV_HWPROBE_KEY_MISALIGNED_PERF:
>  		pair->value = hwprobe_misaligned(cpus);
>  		break;
>  
> -- 
> 2.34.1
>

Otherwise,

Reviewed-by: Andrew Jones <ajones@xxxxxxxxxxxxxxxx>

Thanks,
drew