linux-kernel - Re: [PATCH v1 2/3] scsi: ufs: Let command trace only for the cmd != null case

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1d06cc01-a642-e8e0-a251-1b392e4935c7@acm.org>
Date:   Sun, 23 May 2021 18:32:08 -0700
From:   Bart Van Assche <bvanassche@....org>
To:     Bean Huo <huobean@...il.com>, alim.akhtar@...sung.com,
        avri.altman@....com, asutoshd@...eaurora.org, jejb@...ux.ibm.com,
        martin.petersen@...cle.com, stanley.chu@...iatek.com,
        beanhuo@...ron.com, tomas.winkler@...el.com, cang@...eaurora.org
Cc:     linux-scsi@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH v1 2/3] scsi: ufs: Let command trace only for the cmd !=
 null case

On 5/23/21 2:14 PM, Bean Huo wrote:
> +	opcode = cmd->cmnd[0];
> +	if ((opcode == READ_10) || (opcode == WRITE_10)) {
> +		/*
> +		 * Currently we only fully trace read(10) and write(10)
> +		 * commands
> +		 */
> +		if (cmd->request && cmd->request->bio)
> +			lba = cmd->request->bio->bi_iter.bi_sector;

Why does the lba assignment occur inside the if-statement for the
READ_10 and WRITE_10 cases? Has it been considered to move that
assignment before this if-statement?

Does 'lba' represent an offset in units of 512 bytes (sector_t) or an
LBA (logical block address)? In the former case, please rename the
variable 'lba' into 'sector' or 'start_sector'. In the latter case,
please use sectors_to_logical().

Why are READ_16 and WRITE_16 ignored?

Please remove the 'if (cmd->request)' checks since these are not necessary.

> +	} else if (opcode == UNMAP) {
> +		if (cmd->request) {
> +			lba = scsi_get_lba(cmd);
> +			transfer_len = blk_rq_bytes(cmd->request);
>  		}
>  	}

The name of the variable 'transfer_len' is wrong since blk_rq_bytes()
returns the number of bytes affected on the storage medium instead of
the number of bytes transferred from the host to the storage controller.

>  /**
> - * Describes the ufs rpmb wlun.
> - * Used only to send uac.
> + * Describes the ufs rpmb wlun. Used only to send uac.
>   */

Is this change related to the rest of this patch?

Thanks,

Bart.