Re: [PATCH v2 1/5] riscv: Checksum header

From: Conor Dooley
Date: Thu Sep 07 2023 - 12:37:11 EST


Hey Charlie,

Just a passing thought that I can't really test since I am meant to be
on holidays...

On Tue, Sep 05, 2023 at 09:46:50PM -0700, Charlie Jenkins wrote:
> Provide checksum algorithms that have been designed to leverage riscv
> instructions such as rotate. In 64-bit, can take advantage of the larger
> register to avoid some overflow checking.
>
> Signed-off-by: Charlie Jenkins <charlie@xxxxxxxxxxxx>
> ---
> arch/riscv/include/asm/checksum.h | 96 +++++++++++++++++++++++++++++++++++++++
> 1 file changed, 96 insertions(+)
>
> diff --git a/arch/riscv/include/asm/checksum.h b/arch/riscv/include/asm/checksum.h
> new file mode 100644
> index 000000000000..573714b9ea15
> --- /dev/null
> +++ b/arch/riscv/include/asm/checksum.h
> @@ -0,0 +1,96 @@
> +/* SPDX-License-Identifier: GPL-2.0 */
> +/*
> + * IP checksum routines
> + *
> + * Copyright (C) 2023 Rivos Inc.
> + */
> +#ifndef __ASM_RISCV_CHECKSUM_H
> +#define __ASM_RISCV_CHECKSUM_H
> +
> +#include <linux/in6.h>
> +#include <linux/uaccess.h>
> +
> +#ifdef CONFIG_32BIT
> +typedef unsigned int csum_t;
> +#else
> +typedef unsigned long csum_t;
> +#endif
> +
> +/*
> + * Fold a partial checksum without adding pseudo headers
> + */
> +static inline __sum16 csum_fold(__wsum sum)
> +{
> + sum += (sum >> 16) | (sum << 16);
> + return (__force __sum16)(~(sum >> 16));
> +}
> +
> +#define csum_fold csum_fold
> +
> +/*
> + * Quickly compute an IP checksum with the assumption that IPv4 headers will
> + * always be in multiples of 32-bits, and have an ihl of at least 5.
> + * @ihl is the number of 32 bit segments and must be greater than or equal to 5.
> + * @iph is assumed to be word aligned.
> + */
> +static inline __sum16 ip_fast_csum(const void *iph, unsigned int ihl)
> +{
> + csum_t csum = 0;
> + int pos = 0;
> +
> + do {
> + csum += ((const unsigned int *)iph)[pos];
> +#ifdef CONFIG_32BIT
> + csum += csum < ((const unsigned int *)iph)[pos];
> +#endif // !CONFIG_32BIT

Some of this ifdeffery really should become IS_ENABLED(), there's
nothing in some of them that can't just get removed by the compiler.

> + } while (++pos < ihl);
> +
> +#ifdef CONFIG_RISCV_ISA_ZBB

This here I can't test since I'm supposed to be AFK, but can this also
be an IS_ENABLED()? I know it is guarding code that the toolchain may
not support, but does it get removed before that matters?

I mainly ask because there's a host of ifdeffery here & the code would
be a lot easier to understand if we could cut it down to a minimum.

> + if (IS_ENABLED(CONFIG_RISCV_ALTERNATIVE)) {
> + csum_t fold_temp;
> +
> + asm_volatile_goto(ALTERNATIVE("j %l[no_zbb]", "nop", 0,
> + RISCV_ISA_EXT_ZBB, 1)
> + :
> + :
> + :
> + : no_zbb);
> +#ifdef CONFIG_32BIT
> + asm(".option push \n\
> + .option arch,+zbb \n\
> + rori %[fold_temp],%[csum],16 \n\
> + add %[csum],%[fold_temp],%[csum] \n\
> + .option pop"
> + : [csum] "+r" (csum), [fold_temp] "=&r" (fold_temp));
> +#else // !CONFIG_32BIT
> + asm(".option push \n\
> + .option arch,+zbb \n\
> + rori %[fold_temp], %[csum], 32 \n\
> + add %[csum], %[fold_temp], %[csum] \n\
> + srli %[csum], %[csum], 32 \n\
> + roriw %[fold_temp], %[csum], 16 \n\
> + addw %[csum], %[fold_temp], %[csum] \n\
> + .option pop"
> + : [csum] "+r" (csum), [fold_temp] "=&r" (fold_temp));
> +#endif // !CONFIG_32BIT

These 3 I think are pretty easy to follow immediately, but...

> + return ~(csum >> 16);
> + }
> + /*
> + * ZBB only saves three instructions on 32-bit and five on 64-bit so not
> + * worth checking if supported without Alternatives.
> + */
> +no_zbb:
> +#endif // CONFIG_RISCV_ISA_ZBB
> +#ifdef CONFIG_32BIT
> +#else // !CONFIG_32BIT
> + csum += (csum >> 32) | (csum << 32);
> + csum >>= 16;
> +#endif // !CONFIG_32BIT

...these ones here could be converted too, afaict.

Thanks,
Conor.

> + return csum_fold((__force __wsum)csum);
> +}
> +
> +#define ip_fast_csum ip_fast_csum
> +
> +#include <asm-generic/checksum.h>
> +
> +#endif // __ASM_RISCV_CHECKSUM_H
>
> --
> 2.42.0
>

Attachment: signature.asc
Description: PGP signature