linux-kernel - Re: [PATCH v4 next 6/9] lib: test_mul_u64_u64_div

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <06s18s2p-97n5-9s2o-0rq0-7o8o2ns9s3qn@syhkavp.arg>
Date: Wed, 29 Oct 2025 14:53:13 -0400 (EDT)
From: Nicolas Pitre <nico@...xnic.net>
To: David Laight <david.laight.linux@...il.com>
cc: Andrew Morton <akpm@...ux-foundation.org>, linux-kernel@...r.kernel.org, 
    u.kleine-koenig@...libre.com, Oleg Nesterov <oleg@...hat.com>, 
    Peter Zijlstra <peterz@...radead.org>, 
    Biju Das <biju.das.jz@...renesas.com>, Borislav Petkov <bp@...en8.de>, 
    Dave Hansen <dave.hansen@...ux.intel.com>, 
    "H. Peter Anvin" <hpa@...or.com>, Ingo Molnar <mingo@...hat.com>, 
    Thomas Gleixner <tglx@...utronix.de>, Li RongQing <lirongqing@...du.com>, 
    Yu Kuai <yukuai3@...wei.com>, Khazhismel Kumykov <khazhy@...omium.org>, 
    Jens Axboe <axboe@...nel.dk>, x86@...nel.org
Subject: Re: [PATCH v4 next 6/9] lib: test_mul_u64_u64_div_u64: Test both
 generic and arch versions

On Wed, 29 Oct 2025, David Laight wrote:

> Change the #if in div64.c so that test_mul_u64_u64_div_u64.c
> can compile and test the generic version (including the 'long multiply')
> on architectures (eg amd64) that define their own copy.
> 
> Test the kernel version and the locally compiled version on all arch.
> Output the time taken (in ns) on the 'test completed' trace.
> 
> For reference, on my zen 5, the optimised version takes ~220ns and the
> generic version ~3350ns.
> Using the native multiply saves ~200ns and adding back the ilog2() 'optimisation'
> test adds ~50ms.
> 
> Signed-off-by: David Laight <david.laight.linux@...il.com>

Reviewed-by: Nicolas Pitre <npitre@...libre.com>

> ---
> 
> Changes for v4:
> - Fix build on non x86 (eg arm32)
> 
>  lib/math/div64.c                    |  8 +++--
>  lib/math/test_mul_u64_u64_div_u64.c | 51 +++++++++++++++++++++++++----
>  2 files changed, 50 insertions(+), 9 deletions(-)
> 
> diff --git a/lib/math/div64.c b/lib/math/div64.c
> index 25295daebde9..f92e7160feb6 100644
> --- a/lib/math/div64.c
> +++ b/lib/math/div64.c
> @@ -177,16 +177,18 @@ EXPORT_SYMBOL(div64_s64);
>   * Iterative div/mod for use when dividend is not expected to be much
>   * bigger than divisor.
>   */
> +#ifndef iter_div_u64_rem
>  u32 iter_div_u64_rem(u64 dividend, u32 divisor, u64 *remainder)
>  {
>  	return __iter_div_u64_rem(dividend, divisor, remainder);
>  }
>  EXPORT_SYMBOL(iter_div_u64_rem);
> +#endif
>  
> -#ifndef mul_u64_add_u64_div_u64
> +#if !defined(mul_u64_add_u64_div_u64) || defined(test_mul_u64_add_u64_div_u64)
>  u64 mul_u64_add_u64_div_u64(u64 a, u64 b, u64 c, u64 d)
>  {
> -#if defined(__SIZEOF_INT128__)
> +#if defined(__SIZEOF_INT128__) && !defined(test_mul_u64_add_u64_div_u64)
>  
>  	/* native 64x64=128 bits multiplication */
>  	u128 prod = (u128)a * b + c;
> @@ -267,5 +269,7 @@ u64 mul_u64_add_u64_div_u64(u64 a, u64 b, u64 c, u64 d)
>  
>  	return res;
>  }
> +#if !defined(test_mul_u64_add_u64_div_u64)
>  EXPORT_SYMBOL(mul_u64_add_u64_div_u64);
>  #endif
> +#endif
> diff --git a/lib/math/test_mul_u64_u64_div_u64.c b/lib/math/test_mul_u64_u64_div_u64.c
> index 4d5e4e5dac67..a3c5e54f37ef 100644
> --- a/lib/math/test_mul_u64_u64_div_u64.c
> +++ b/lib/math/test_mul_u64_u64_div_u64.c
> @@ -73,21 +73,34 @@ done
>  
>   */
>  
> -static int __init test_init(void)
> +static u64 test_mul_u64_add_u64_div_u64(u64 a, u64 b, u64 c, u64 d);
> +
> +static int __init test_run(unsigned int fn_no, const char *fn_name)
>  {
> +	u64 start_time;
>  	int errors = 0;
>  	int tests = 0;
>  	int i;
>  
> -	pr_info("Starting mul_u64_u64_div_u64() test\n");
> +	start_time = ktime_get_ns();
>  
>  	for (i = 0; i < ARRAY_SIZE(test_values); i++) {
>  		u64 a = test_values[i].a;
>  		u64 b = test_values[i].b;
>  		u64 d = test_values[i].d;
>  		u64 expected_result = test_values[i].result;
> -		u64 result = mul_u64_u64_div_u64(a, b, d);
> -		u64 result_up = mul_u64_u64_div_u64_roundup(a, b, d);
> +		u64 result, result_up;
> +
> +		switch (fn_no) {
> +		default:
> +			result = mul_u64_u64_div_u64(a, b, d);
> +			result_up = mul_u64_u64_div_u64_roundup(a, b, d);
> +			break;
> +		case 1:
> +			result = test_mul_u64_add_u64_div_u64(a, b, 0, d);
> +			result_up = test_mul_u64_add_u64_div_u64(a, b, d - 1, d);
> +			break;
> +		}
>  
>  		tests += 2;
>  
> @@ -106,15 +119,39 @@ static int __init test_init(void)
>  		}
>  	}
>  
> -	pr_info("Completed mul_u64_u64_div_u64() test, %d tests, %d errors\n",
> -		tests, errors);
> -	return errors ? -EINVAL : 0;
> +	pr_info("Completed %s() test, %d tests, %d errors, %llu ns\n",
> +		fn_name, tests, errors, ktime_get_ns() - start_time);
> +	return errors;
> +}
> +
> +static int __init test_init(void)
> +{
> +	pr_info("Starting mul_u64_u64_div_u64() test\n");
> +	if (test_run(0, "mul_u64_u64_div_u64"))
> +		return -EINVAL;
> +	if (test_run(1, "test_mul_u64_u64_div_u64"))
> +		return -EINVAL;
> +	return 0;
>  }
>  
>  static void __exit test_exit(void)
>  {
>  }
>  
> +/* Compile the generic mul_u64_add_u64_div_u64() code */
> +#define __div64_32 __div64_32
> +#define div_s64_rem div_s64_rem
> +#define div64_u64_rem div64_u64_rem
> +#define div64_u64 div64_u64
> +#define div64_s64 div64_s64
> +#define iter_div_u64_rem iter_div_u64_rem
> +
> +#undef mul_u64_add_u64_div_u64
> +#define mul_u64_add_u64_div_u64 test_mul_u64_add_u64_div_u64
> +#define test_mul_u64_add_u64_div_u64 test_mul_u64_add_u64_div_u64
> +
> +#include "div64.c"
> +
>  module_init(test_init);
>  module_exit(test_exit);
>  
> -- 
> 2.39.5
> 
>