[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20200507111727.GB5726@C02TD0UTHF1T.local>
Date: Thu, 7 May 2020 12:17:27 +0100
From: Mark Rutland <mark.rutland@....com>
To: Anshuman Khandual <anshuman.khandual@....com>
Cc: linux-arm-kernel@...ts.infradead.org,
Catalin Marinas <catalin.marinas@....com>,
Will Deacon <will@...nel.org>, Mark Brown <broonie@...nel.org>,
Ard Biesheuvel <ardb@...nel.org>,
Suzuki K Poulose <suzuki.poulose@....com>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH] arm64/cpuinfo: Move HWCAP name arrays alongside their
bit definitions
On Thu, May 07, 2020 at 03:26:30PM +0530, Anshuman Khandual wrote:
> All HWCAP name arrays (i.e hwcap_str, compat_hwcap_str, compat_hwcap2_str)
> that are scanned for /proc/cpuinfo output are detached from their bit fild
> definitions making it difficult to corelate. This is also bit problematic
> because during /proc/cpuinfo dump these arrays get traversed sequentially
> assuming that they reflect and match HWCAP bit sequence, to test various
> features for a given CPU.
Making this relationship explicit is certianly a good idea!
> This moves all HWCAP name arrays near their bit definitions. But first it
> defines all missing COMPAT_HWCAP_XXX that are present in the name string.
> Also defines all COMPAT_KERNEL_HWCAP macros to be used for the HWCAP name
> string indexing.
I think we don't need all the indirection here, and can simplify things
a bit. Comments on that below.
> Cc: Catalin Marinas <catalin.marinas@....com>
> Cc: Will Deacon <will@...nel.org>
> Cc: Mark Brown <broonie@...nel.org>
> Cc: Ard Biesheuvel <ardb@...nel.org>
> Cc: Mark Rutland <mark.rutland@....com>
> Cc: Suzuki K Poulose <suzuki.poulose@....com>
> Cc: linux-arm-kernel@...ts.infradead.org
> Cc: linux-kernel@...r.kernel.org
>
> Signed-off-by: Anshuman Khandual <anshuman.khandual@....com>
> ---
> In longer term we might want to generate these HWCAP macros with scripting,
> and this is a step in that direction. This applies on 5.7-rc4.
Having attempted to build scripting for this in the past, it turns out
to be a bigger job than it seems initially, and I agree it'd make sense
to take a change like this without scripting as a first step.
> arch/arm64/include/asm/hwcap.h | 132 +++++++++++++++++++++++++++++++++
> arch/arm64/kernel/cpuinfo.c | 90 ----------------------
> 2 files changed, 132 insertions(+), 90 deletions(-)
>
> diff --git a/arch/arm64/include/asm/hwcap.h b/arch/arm64/include/asm/hwcap.h
> index 0f00265248b5..cb99b5f75e5d 100644
> --- a/arch/arm64/include/asm/hwcap.h
> +++ b/arch/arm64/include/asm/hwcap.h
> @@ -8,18 +8,27 @@
> #include <uapi/asm/hwcap.h>
> #include <asm/cpufeature.h>
>
> +#define COMPAT_HWCAP_SWP (1 << 0)
> #define COMPAT_HWCAP_HALF (1 << 1)
> #define COMPAT_HWCAP_THUMB (1 << 2)
> +#define COMPAT_HWCAP_26BIT (1 << 3)
> #define COMPAT_HWCAP_FAST_MULT (1 << 4)
> +#define COMPAT_HWCAP_FPA (1 << 5)
> #define COMPAT_HWCAP_VFP (1 << 6)
> #define COMPAT_HWCAP_EDSP (1 << 7)
> +#define COMPAT_HWCAP_JAVA (1 << 8)
> +#define COMPAT_HWCAP_IWMMXT (1 << 9)
> +#define COMPAT_HWCAP_CRUNCH (1 << 10)
> +#define COMPAT_HWCAP_THUMBEE (1 << 11)
> #define COMPAT_HWCAP_NEON (1 << 12)
> #define COMPAT_HWCAP_VFPv3 (1 << 13)
> +#define COMPAT_HWCAP_VFPV3D16 (1 << 14)
> #define COMPAT_HWCAP_TLS (1 << 15)
> #define COMPAT_HWCAP_VFPv4 (1 << 16)
> #define COMPAT_HWCAP_IDIVA (1 << 17)
> #define COMPAT_HWCAP_IDIVT (1 << 18)
> #define COMPAT_HWCAP_IDIV (COMPAT_HWCAP_IDIVA|COMPAT_HWCAP_IDIVT)
> +#define COMPAT_HWCAP_VFPD32 (1 << 19)
> #define COMPAT_HWCAP_LPAE (1 << 20)
> #define COMPAT_HWCAP_EVTSTRM (1 << 21)
>
> @@ -95,6 +104,129 @@
> #define KERNEL_HWCAP_DGH __khwcap2_feature(DGH)
> #define KERNEL_HWCAP_RNG __khwcap2_feature(RNG)
>
> +#ifdef CONFIG_COMPAT
> +#define __compat_khwcap_feature(x) const_ilog2(COMPAT_HWCAP_ ## x)
For now I reckon this can be:
#define COMPAT_KERNEL_HWCAP(x) const_ilog2(COMPAT_HWCAP_##x)
> +#define COMPAT_KERNEL_HWCAP_SWP __compat_khwcap_feature(SWP)
> +#define COMPAT_KERNEL_HWCAP_HALF __compat_khwcap_feature(HALF)
> +#define COMPAT_KERNEL_HWCAP_THUMB __compat_khwcap_feature(THUMB)
> +#define COMPAT_KERNEL_HWCAP_26BIT __compat_khwcap_feature(26BIT)
> +#define COMPAT_KERNEL_HWCAP_FAST_MULT __compat_khwcap_feature(FAST_MULT)
> +#define COMPAT_KERNEL_HWCAP_FPA __compat_khwcap_feature(FPA)
> +#define COMPAT_KERNEL_HWCAP_VFP __compat_khwcap_feature(VFP)
> +#define COMPAT_KERNEL_HWCAP_EDSP __compat_khwcap_feature(EDSP)
> +#define COMPAT_KERNEL_HWCAP_JAVA __compat_khwcap_feature(JAVA)
> +#define COMPAT_KERNEL_HWCAP_IWMMXT __compat_khwcap_feature(IWMMXT)
> +#define COMPAT_KERNEL_HWCAP_CRUNCH __compat_khwcap_feature(CRUNCH)
> +#define COMPAT_KERNEL_HWCAP_THUMBEE __compat_khwcap_feature(THUMBEE)
> +#define COMPAT_KERNEL_HWCAP_NEON __compat_khwcap_feature(NEON)
> +#define COMPAT_KERNEL_HWCAP_VFPv3 __compat_khwcap_feature(VFPv3)
> +#define COMPAT_KERNEL_HWCAP_VFPV3D16 __compat_khwcap_feature(VFPV3D16)
> +#define COMPAT_KERNEL_HWCAP_TLS __compat_khwcap_feature(TLS)
> +#define COMPAT_KERNEL_HWCAP_VFPv4 __compat_khwcap_feature(VFPv4)
> +#define COMPAT_KERNEL_HWCAP_IDIVA __compat_khwcap_feature(IDIVA)
> +#define COMPAT_KERNEL_HWCAP_IDIVT __compat_khwcap_feature(IDIVT)
> +#define COMPAT_KERNEL_HWCAP_VFPD32 __compat_khwcap_feature(VFPD32)
> +#define COMPAT_KERNEL_HWCAP_LPAE __compat_khwcap_feature(LPAE)
> +#define COMPAT_KERNEL_HWCAP_EVTSTRM __compat_khwcap_feature(EVTSTRM)
... so we don't need any of the above, as e.g. we can just do:
[COMPAT_KERNEL_HWCAP(SWP)] = "swp",
... at the points of usage.
> +#define __compat_khwcap2_feature(x) const_ilog2(COMPAT_HWCAP2_ ## x)
> +#define COMPAT_KERNEL_HWCAP_AES __compat_khwcap2_feature(AES)
> +#define COMPAT_KERNEL_HWCAP_PMULL __compat_khwcap2_feature(PMULL)
> +#define COMPAT_KERNEL_HWCAP_SHA1 __compat_khwcap2_feature(SHA1)
> +#define COMPAT_KERNEL_HWCAP_SHA2 __compat_khwcap2_feature(SHA2)
> +#define COMPAT_KERNEL_HWCAP_CRC32 __compat_khwcap2_feature(CRC32)
Likewise here with:
#define COMPAT_KERNEL_HWCAP2(x) const_ilog2(COMPAT_HWCAP2_##x)
With that, and the usage in the arrays changed to match:
Acked-by: Mark Rutland <mark.rutland@....com>
Thanks,
Mark.
Powered by blists - more mailing lists