* Junfeng Yang (yjf@stanford.edu) wrote:
>
> [BUG] not sure. the dereference occurs at a tainted place. call
> copy_from_user (, data, ), then copy_from_user (, *data, )
>
> /home/junfeng/linux-2.5.63/sound/core/seq/instr/ainstr_iw.c:92:snd_seq_iwffff_copy_env_from_stream:
> ERROR:TAINTED deferencing "data" tainted by [dist=0][copy_from_user:parm1]
seems like a bug, although i think it's the first copy_from_user that
would be broken.
Looks like call flow could be something like:
Store user buf ptr in event.data.ext.ptr.
Call snd_seq_iwffff_put with the event.data.ext.ptr (offset by header bits)
Call snd_seq_iwffff_copy_env_from_stream with &event.data.ext.ptr (offset a
little further).
Finally, call copy_from_user on &event.data.ext.ptr, which is copying
into an stype. This looks like it's the bug. I don't think the 32 bit
addr should be copied into the stype, rather the 1st 32bits of the data
(IOW copy_from_user(stype, *data,...)). The second copy_from_user (with
*data) copies in the whole iwffff_env_record_t struct which begins with
an stype, so it really looks as if the first copy is broken. Patch
below. Jaroslav?
thanks,
-chris
-- Linux Security Modules http://lsm.immunix.org http://lsm.bkbits.net===== instr/ainstr_iw.c 1.3 vs edited ===== --- 1.3/sound/core/seq/instr/ainstr_iw.c Mon Feb 10 02:39:27 2003 +++ edited/instr/ainstr_iw.c Fri Mar 21 13:01:32 2003 @@ -78,7 +78,7 @@ while (1) { if (*len < (long)sizeof(__u32)) return -EINVAL; - if (copy_from_user(&stype, data, sizeof(stype))) + if (copy_from_user(&stype, *data, sizeof(stype))) return -EFAULT; if (stype == IWFFFF_STRU_WAVE) return 0; - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Sun Mar 23 2003 - 22:00:39 EST