Re: [CHECKER] Implementation inconsistencies in 2.6.3

[Matthew Dharm <mdharm-kernel@one-eyed-alien.net>]

[-- Attachment #1: Type: text/plain, Size: 2282 bytes --]

On Tue, Apr 27, 2004 at 02:47:22PM -0700, Ken Ashcraft wrote:
> I'm trying to cross check implementations of the same interface for
> errors.  I assume that if functions are assigned to the same function
> pointer, they are implementations of a common interface and should be
> consistent with each other.  For example, if one implementation checks one
> of its arguments for NULL, the other implementation should also check that
> argument for NULL.
> 
> In this case, I'm looking at which arguments are referenced at all in the
> implementation.  If we have 10 implementations and 9 of them read argument
> 1 and the 10th fails to read argument 1, the 10th implementation may be
> missing some code.  In each of the reports below, I give an example
> implementation that does read the argument.  That is followed by an
> implementation that fails to read the argument.
> 

> ---------------------------------------------------------
> [BUG] (mdharm-usb@one-eyed-alien.net) looks like it should return count
> instead of strlen(buf), but this is in scsiglue.c, so is it special code?
> 
> example:
> /home/kash/linux/2.6.3/linux-2.6.3/drivers/scsi/scsi_sysfs.c:274:store_rescan_field:
> NOTE:READ: Checking arg count [EXAMPLE=device_attribute.store-2]
> 
> /home/kash/linux/2.6.3/linux-2.6.3/drivers/usb/storage/scsiglue.c:321:store_max_sectors:
> ERROR:READ: Not checking arg [COUNTER=device_attribute.store-2]  [fit=1]
> [fit_fn=1] [fn_ex=0] [fn_counter=1] [ex=233] [counter=1] [z =
> 3.20943839741638] [fn-z = -4.35889894354067]
> 
> 	return sprintf(buf, "%u\n", sdev->request_queue->max_sectors);
> }
> 
> /* Input routine for the sysfs max_sectors file */
> 
> Error --->
> static ssize_t store_max_sectors(struct device *dev, const char *buf,
> 		size_t count)
> {
> 	struct scsi_device *sdev = to_scsi_device(dev);

My understanding was that I was supposed to return the number of bytes in
the buffer that I actually used.  I thought 'count' was the maximum size I
could use.

Is that not the case?

Matt

-- 
Matthew Dharm                              Home: mdharm-usb@one-eyed-alien.net 
Maintainer, Linux USB Mass Storage Driver

Oh great modem, why hast thou forsaken me?
					-- Dust Puppy
User Friendly, 3/2/1998

[-- Attachment #2: Type: application/pgp-signature, Size: 189 bytes --]