linux-kernel - Re: [PATCH v12 2/2] mm: page_alloc: reduce unnecessary binary search in memblock_next_valid

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: <20190723083353.GC4896@rapoport-lnx>
Date:   Tue, 23 Jul 2019 11:33:54 +0300
From:   Mike Rapoport <rppt@...ux.ibm.com>
To:     Hanjun Guo <guohanjun@...wei.com>
Cc:     Ard Biesheuvel <ard.biesheuvel@...aro.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Catalin Marinas <catalin.marinas@....com>,
        Jia He <hejianet@...il.com>, Will Deacon <will@...nel.org>,
        linux-arm-kernel@...ts.infradead.org, linux-mm@...ck.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v12 2/2] mm: page_alloc: reduce unnecessary binary search
 in memblock_next_valid_pfn

On Tue, Jul 23, 2019 at 01:51:13PM +0800, Hanjun Guo wrote:
> From: Jia He <hejianet@...il.com>
> 
> After skipping some invalid pfns in memmap_init_zone(), there is still
> some room for improvement.
> 
> E.g. if pfn and pfn+1 are in the same memblock region, we can simply pfn++
> instead of doing the binary search in memblock_next_valid_pfn.
> 
> Furthermore, if the pfn is in a gap of two memory region, skip to next
> region directly to speedup the binary search.

How much speed up do you see with this improvements relatively to simple
binary search in memblock_next_valid_pfn()?
  
> Signed-off-by: Jia He <hejianet@...il.com>
> Signed-off-by: Hanjun Guo <guohanjun@...wei.com>
> ---
>  mm/memblock.c | 37 +++++++++++++++++++++++++++++++------
>  1 file changed, 31 insertions(+), 6 deletions(-)
> 
> diff --git a/mm/memblock.c b/mm/memblock.c
> index d57ba51bb9cd..95d5916716a0 100644
> --- a/mm/memblock.c
> +++ b/mm/memblock.c
> @@ -1256,28 +1256,53 @@ int __init_memblock memblock_set_node(phys_addr_t base, phys_addr_t size,
>  unsigned long __init_memblock memblock_next_valid_pfn(unsigned long pfn)
>  {
>  	struct memblock_type *type = &memblock.memory;
> +	struct memblock_region *regions = type->regions;
>  	unsigned int right = type->cnt;
>  	unsigned int mid, left = 0;
> +	unsigned long start_pfn, end_pfn, next_start_pfn;
>  	phys_addr_t addr = PFN_PHYS(++pfn);
> +	static int early_region_idx __initdata_memblock = -1;
>  
> +	/* fast path, return pfn+1 if next pfn is in the same region */
> +	if (early_region_idx != -1) {
> +		start_pfn = PFN_DOWN(regions[early_region_idx].base);
> +		end_pfn = PFN_DOWN(regions[early_region_idx].base +
> +				regions[early_region_idx].size);
> +
> +		if (pfn >= start_pfn && pfn < end_pfn)
> +			return pfn;
> +
> +		/* try slow path */
> +		if (++early_region_idx == type->cnt)
> +			goto slow_path;
> +
> +		next_start_pfn = PFN_DOWN(regions[early_region_idx].base);
> +
> +		if (pfn >= end_pfn && pfn <= next_start_pfn)
> +			return next_start_pfn;
> +	}
> +
> +slow_path:
> +	/* slow path, do the binary searching */
>  	do {
>  		mid = (right + left) / 2;
>  
> -		if (addr < type->regions[mid].base)
> +		if (addr < regions[mid].base)
>  			right = mid;
> -		else if (addr >= (type->regions[mid].base +
> -				  type->regions[mid].size))
> +		else if (addr >= (regions[mid].base + regions[mid].size))
>  			left = mid + 1;
>  		else {
> -			/* addr is within the region, so pfn is valid */
> +			early_region_idx = mid;
>  			return pfn;
>  		}
>  	} while (left < right);
>  
>  	if (right == type->cnt)
>  		return -1UL;
> -	else
> -		return PHYS_PFN(type->regions[right].base);
> +
> +	early_region_idx = right;
> +
> +	return PHYS_PFN(regions[early_region_idx].base);
>  }
>  EXPORT_SYMBOL(memblock_next_valid_pfn);
>  #endif /* CONFIG_HAVE_MEMBLOCK_PFN_VALID */
> -- 
> 2.19.1
> 

-- 
Sincerely yours,
Mike.