Re: [PATCH v4 06/11] qla2xxx: Increase the size of struct qla_fcp_prio_cfg to FCP_PRIO_CFG_SIZE

From: Arun Easi <aeasi@marvell.com>
To: Bart Van Assche <bvanassche@acm.org>
Cc: "Martin K . Petersen" <martin.petersen@oracle.com>,
	"James E . J . Bottomley" <jejb@linux.vnet.ibm.com>,
	<linux-scsi@vger.kernel.org>,
	"Nilesh Javali" <njavali@marvell.com>,
	Himanshu Madhani <himanshu.madhani@oracle.com>,
	Quinn Tran <qutran@marvell.com>, Martin Wilck <mwilck@suse.com>,
	"Daniel Wagner" <dwagner@suse.de>,
	Roman Bolshakov <r.bolshakov@yadro.com>
Subject: Re: [PATCH v4 06/11] qla2xxx: Increase the size of struct qla_fcp_prio_cfg to FCP_PRIO_CFG_SIZE
Date: Mon, 11 May 2020 00:36:00 -0700	[thread overview]
Message-ID: <alpine.LRH.2.21.9999.2005110035150.23618@irv1user01.caveonetworks.com> (raw)
In-Reply-To: <20200427030310.19687-7-bvanassche@acm.org>

On Sun, 26 Apr 2020, 8:03pm, Bart Van Assche wrote:

> This patch fixes the following Coverity complaint without changing any
> functionality:
> 
> CID 337793 (#1 of 1): Wrong size argument (SIZEOF_MISMATCH)
> suspicious_sizeof: Passing argument ha->fcp_prio_cfg of type
> struct qla_fcp_prio_cfg * and argument 32768UL to function memset is
> suspicious because a multiple of sizeof (struct qla_fcp_prio_cfg) /*48*/
> is expected.
> 
> memset(ha->fcp_prio_cfg, 0, FCP_PRIO_CFG_SIZE);
> 
> Cc: Nilesh Javali <njavali@marvell.com>
> Cc: Himanshu Madhani <himanshu.madhani@oracle.com>
> Cc: Quinn Tran <qutran@marvell.com>
> Cc: Martin Wilck <mwilck@suse.com>
> Cc: Daniel Wagner <dwagner@suse.de>
> Cc: Roman Bolshakov <r.bolshakov@yadro.com>
> Signed-off-by: Bart Van Assche <bvanassche@acm.org>
> ---
>  drivers/scsi/qla2xxx/qla_fw.h | 3 ++-
>  drivers/scsi/qla2xxx/qla_os.c | 1 +
>  2 files changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/scsi/qla2xxx/qla_fw.h b/drivers/scsi/qla2xxx/qla_fw.h
> index b364a497e33d..4fa34374f34f 100644
> --- a/drivers/scsi/qla2xxx/qla_fw.h
> +++ b/drivers/scsi/qla2xxx/qla_fw.h
> @@ -2217,8 +2217,9 @@ struct qla_fcp_prio_cfg {
>  #define FCP_PRIO_ATTR_PERSIST   0x2
>  	uint8_t  reserved;      /* Reserved for future use          */
>  #define FCP_PRIO_CFG_HDR_SIZE   0x10
> -	struct qla_fcp_prio_entry entry[1];     /* fcp priority entries  */
> +	struct qla_fcp_prio_entry entry[1023]; /* fcp priority entries  */
>  #define FCP_PRIO_CFG_ENTRY_SIZE 0x20
> +	uint8_t  reserved2[16];
>  };
>  
>  #define FCP_PRIO_CFG_SIZE       (32*1024) /* fcp prio data per port*/
> diff --git a/drivers/scsi/qla2xxx/qla_os.c b/drivers/scsi/qla2xxx/qla_os.c
> index 2dd9c2a39cd5..30c2750c5745 100644
> --- a/drivers/scsi/qla2xxx/qla_os.c
> +++ b/drivers/scsi/qla2xxx/qla_os.c
> @@ -7877,6 +7877,7 @@ qla2x00_module_init(void)
>  	BUILD_BUG_ON(sizeof(struct qla82xx_uri_data_desc) != 28);
>  	BUILD_BUG_ON(sizeof(struct qla82xx_uri_table_desc) != 32);
>  	BUILD_BUG_ON(sizeof(struct qla83xx_fw_dump) != 51196);
> +	BUILD_BUG_ON(sizeof(struct qla_fcp_prio_cfg) != FCP_PRIO_CFG_SIZE);
>  	BUILD_BUG_ON(sizeof(struct qla_fdt_layout) != 128);
>  	BUILD_BUG_ON(sizeof(struct qla_flt_header) != 8);
>  	BUILD_BUG_ON(sizeof(struct qla_flt_region) != 16);
> 

The changes themselves look ok, but..

Could the warning be avoided by memset of FCP_PRIO_CFG_HDR_SIZE
before first read_optrom(), and another memset of
"FCP_PRIO_CFG_SIZE - FCP_PRIO_CFG_HDR_SIZE" before second
read_optrom() call?

The reason I ask is that, the kind of "1" element array
declaration in a struct is a common way of mapping a header
followed by N records of some nature. It is a bit sad if we are
moving away from that style and hard computing the structure by hand.