On Fri, Feb 4, 2011 at 6:16 AM, Jeff Layton <jlayton@xxxxxxxxxx> wrote:
> On Thu, 3 Feb 2011 23:01:45 -0600
> Steve French <smfrench@xxxxxxxxx> wrote:
>
>> Looking at this comment in the patch:
>>
>> + * FIXME: why 48 bytes?
>> + */
>> + length = checkSMB(smb_buffer, smb_buffer->Mid, total_read);
>> + if (length != 0)
>> cifs_dump_mem("Bad SMB: ", smb_buffer,
>> total_read < 48 ? total_read : 48);
>>
>>
>> got me thinking, if you prefer a longer length that is fine. To
>> avoid huge log entries, while still dumping the most useful info, the
>> original length dumped was limited. 48 bytes was chosen because it
>> printed 3 lines of 16 to log and thus would include the most important
>> parts of the SMB (past wct). It is large enough to fit a close
>> request and response, although about 10 bytes to small for getting all
>> of a SMB ReadX request. It would probably be fine to make that 64
>> bytes if you prefer. If we made it smaller (e.g. 32 instead of 48)
>> IIRC we wouldn't get wct.
>
> The length doesn't much matter to me. What does matter is that this
> would do a cifs_dump_mem for 48 bytes even when the client received
> less than 48 bytes. Granted, the buffers are larger than that so I
> don't think there's any danger of a buffer overrun or anything, but the
> stuff at the end could just be garbage or zeroed out.
Yes, agreed. I liked that change. It improves
readability of debug info.
--
Thanks,
Steve
--
To unsubscribe from this list: send the line "unsubscribe linux-cifs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
