linux-kernel - Re: [PATCH][next] scsi: scsi_debug: remove a redundant assignment to variable ret

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: <9151ca6d-7153-4a97-aaa2-7277fc5ffa84@oracle.com>
Date: Wed, 16 Oct 2024 08:16:16 +0100
From: John Garry <john.g.garry@...cle.com>
To: Dan Carpenter <dan.carpenter@...aro.org>,
        Colin Ian King <colin.i.king@...il.com>
Cc: "James E . J . Bottomley" <James.Bottomley@...senpartnership.com>,
        "Martin K . Petersen" <martin.petersen@...cle.com>,
        linux-scsi@...r.kernel.org, kernel-janitors@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH][next] scsi: scsi_debug: remove a redundant assignment to
 variable ret

On 02/10/2024 16:10, Dan Carpenter wrote:
> On Wed, Oct 02, 2024 at 02:50:43PM +0100, Colin Ian King wrote:
>> The variable ret is being assigned a value that is never read, the
>> following break statement exits the loop where ret is being re-assigned
>> a new value. Remove the redundant assignment.
>>
>> Signed-off-by: Colin Ian King<colin.i.king@...il.com>
>> ---
>>   drivers/scsi/scsi_debug.c | 4 +---
>>   1 file changed, 1 insertion(+), 3 deletions(-)
>>
>> diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
>> index d95f417e24c0..7c60f5acc4a3 100644
>> --- a/drivers/scsi/scsi_debug.c
>> +++ b/drivers/scsi/scsi_debug.c
>> @@ -3686,14 +3686,12 @@ static int do_device_access(struct sdeb_store_info *sip, struct scsi_cmnd *scp,
>>   		sdeb_data_sector_lock(sip, do_write);
>>   		ret = sg_copy_buffer(sdb->table.sgl, sdb->table.nents,
> You would think there would be a:
> 
> 	total += ret;
> 
> here.
> 
>>   		   fsp + (block * sdebug_sector_size),
>>   		   sdebug_sector_size, sg_skip, do_write);T
>>   		sdeb_data_sector_unlock(sip, do_write);
>> -		if (ret != sdebug_sector_size) {
>> -			ret += (i * sdebug_sector_size);
>> +		if (ret != sdebug_sector_size)
>>   			break;
>> -		}
>>   		sg_skip += sdebug_sector_size;
>>   		if (++block >= sdebug_store_sectors)
>>   			block = 0;
>>   	}
>>   	ret = num * sdebug_sector_size;
>          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> And that this would be a "return total;"

Right, the function is currently a little messy as there is no variable 
for "total", and we re-assign ret per loop.

So I think that we can either:
a. introduce a variable to hold "total"
b. this change:

diff --git a/drivers/scsi/scsi_debug.c b/drivers/scsi/scsi_debug.c
index af5e3a7f47a9..39218ffc6a31 100644
--- a/drivers/scsi/scsi_debug.c
+++ b/drivers/scsi/scsi_debug.c
@@ -3690,13 +3690,14 @@ static int do_device_access(struct
sdeb_store_info *sip, struct scsi_cmnd *scp,
                sdeb_data_sector_unlock(sip, do_write);
                if (ret != sdebug_sector_size) {
                        ret += (i * sdebug_sector_size);
-                       break;
+                       goto out_unlock;
                }
                sg_skip += sdebug_sector_size;
                if (++block >= sdebug_store_sectors)
                        block = 0;
        }
        ret = num * sdebug_sector_size;
+out_unlock:
        sdeb_data_unlock(sip, atomic);


Maybe a. is better, as b. is maintaining some messiness.

Thanks,
John

> 
> The comment at the start of the function says that it's supposed to return the
> actual number of bytes that were copied.  And you can see how that was the
> intention.
> 
> But what it actually does is it always reports that it copied the complete
> number of bytes.  #Success #Woohoo
> 
> I wouldn't feel comfortable changing it to report partial copies without testing
> it.  Someone needs to look at it more carefully to figure out what the correct
> fix is.