warning in ieee80211 tkip code

warning in ieee80211 tkip code

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 821 bytes --]

net/ieee80211/ieee80211_crypt_tkip.c: In function 'michael_mic_hdr':
net/ieee80211/ieee80211_crypt_tkip.c:587: warning: overflow in implicit
constant conversion

The reason is pretty obvious:
hdr[12] = qoshdr->qos_ctl & cpu_to_le16(IEEE80211_QCTL_TID);

where IEEE80211_QCTL_TID is 0xF, so that's made into 0xF00 on BE and
then truncated to 0x00. Not very useful.

This was introduced in this commit:

commit 65b6a2775102cd81e57158ef4b1cb89641f76cfd
Author: Zhu Yi <yi.zhu@intel.com>
Date:   Mon Aug 21 11:32:31 2006 +0800

with the change

-               hdr[12] = le16_to_cpu(qoshdr->qos_ctl) & IEEE80211_QCTL_TID;
+               hdr[12] = qoshdr->qos_ctl & cpu_to_le16(IEEE80211_QCTL_TID);

The code before the change looks much saner since you're accessing the
header as bytes...

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 190 bytes --]

Re: warning in ieee80211 tkip code

[Johannes Berg]

[-- Attachment #1: Type: text/plain, Size: 690 bytes --]

On Mon, 2007-08-13 at 08:35 +0200, Johannes Berg wrote:
> net/ieee80211/ieee80211_crypt_tkip.c: In function 'michael_mic_hdr':
> net/ieee80211/ieee80211_crypt_tkip.c:587: warning: overflow in implicit
> constant conversion
> 
> The reason is pretty obvious:
> hdr[12] = qoshdr->qos_ctl & cpu_to_le16(IEEE80211_QCTL_TID);
> 
> where IEEE80211_QCTL_TID is 0xF, so that's made into 0xF00 on BE and
> then truncated to 0x00. Not very useful.
> 
> This was introduced in this commit:
> 
> commit 65b6a2775102cd81e57158ef4b1cb89641f76cfd
> Author: Zhu Yi <yi.zhu@intel.com>
> Date:   Mon Aug 21 11:32:31 2006 +0800

Ping? Any comments on this? It's been two months.

johannes

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 828 bytes --]