linux-kernel - Re: [PATCH] RDMA/rxe: remove redundant assignment to variable page

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] [day] [month] [year] [list]

Message-ID: <c9f46ef5-b7fe-4592-b458-ecb91652d2ed@linux.dev>
Date: Tue, 14 Oct 2025 07:41:15 -0700
From: Zhu Yanjun <yanjun.zhu@...ux.dev>
To: Colin Ian King <coking@...dia.com>, Zhu Yanjun <zyjzyj2000@...il.com>,
 Jason Gunthorpe <jgg@...pe.ca>, Leon Romanovsky <leon@...nel.org>,
 linux-rdma@...r.kernel.org
Cc: kernel-janitors@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] RDMA/rxe: remove redundant assignment to variable
 page_offset

在 2025/10/14 5:03, Colin Ian King 写道:
> The variable page_offset is being assigned a value at the start of
> a loop and being redundantly zero'd at the end of the loop, there
> is no code that reads the zero'd value. The assignment is redundant
> and can be removed.
> 
> Signed-off-by: Colin Ian King <coking@...dia.com>
> ---
>   drivers/infiniband/sw/rxe/rxe_mr.c  | 1 -
>   drivers/infiniband/sw/rxe/rxe_odp.c | 1 -
>   2 files changed, 2 deletions(-)
> 
> diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c
> index bcb97b3ea58a..b1df05238848 100644
> --- a/drivers/infiniband/sw/rxe/rxe_mr.c
> +++ b/drivers/infiniband/sw/rxe/rxe_mr.c
> @@ -452,7 +452,6 @@ static int rxe_mr_flush_pmem_iova(struct rxe_mr *mr, u64 iova, unsigned int leng
>   
>   		length -= bytes;
>   		iova += bytes;
> -		page_offset = 0;
>   	}
static int rxe_mr_flush_pmem_iova(struct rxe_mr *mr, u64 iova, unsigned 
int length)
{
	unsigned int page_offset;
	...
	while (length > 0) {
		index = rxe_mr_iova_to_index(mr, iova);
		page = xa_load(&mr->page_list, index);
		page_offset = rxe_mr_iova_to_page_offset(mr, iova);
	...
		page_offset = 0;
	}

	return 0;
}>
>   	return 0;
> diff --git a/drivers/infiniband/sw/rxe/rxe_odp.c b/drivers/infiniband/sw/rxe/rxe_odp.c
> index f58e3ec6252f..ae71812bea82 100644
> --- a/drivers/infiniband/sw/rxe/rxe_odp.c
> +++ b/drivers/infiniband/sw/rxe/rxe_odp.c
> @@ -358,7 +358,6 @@ int rxe_odp_flush_pmem_iova(struct rxe_mr *mr, u64 iova,
>   
>   		length -= bytes;
>   		iova += bytes;
> -		page_offset = 0;
>   	}
int rxe_odp_flush_pmem_iova(struct rxe_mr *mr, u64 iova,
			    unsigned int length)
{
	unsigned int page_offset;
...
	while (length > 0) {
		index = rxe_odp_iova_to_index(umem_odp, iova);
		page_offset = rxe_odp_iova_to_page_offset(umem_odp, iova);
...
		page_offset = 0;
	}

	mutex_unlock(&umem_odp->umem_mutex);

	return 0;
}
 From the above, in the functions rxe_mr_flush_pmem_iova and 
rxe_odp_flush_pmem_iova, within the while loop, the variable page_offset 
is assigned a value. At the end of the loop, page_offset is set to 0. 
However, this assignment at the end of the loop is actually unnecessary.

Thanks,
Reviewed-by: Zhu Yanjun <yanjun.zhu@...ux.dev>

Zhu Yanjun

>   
>   	mutex_unlock(&umem_odp->umem_mutex);