Re: [RFC 2/3] mac80211: Make use of the new F_TX_NOACK and F_TX_NOSEQ Radiotap TX flags

["Stefanik Gábor" <netrolller.3d@gmail.com>]

On Fri, Dec 12, 2008 at 11:45 PM, Johannes Berg
<johannes@sipsolutions.net> wrote:
> On Fri, 2008-12-12 at 23:43 +0100, Stefanik G=E1bor wrote:
>
>> I tried, but for some reason setting IEEE80211_TX_CTL_NO_ACK directl=
y
>> in __ieee80211_parse_tx_radiotap resulted in an oops upon unloading
>> the driver module.
>
> Huh? That doesn't make sense. Do you have more information on that?
>
>> Note that IEEE80211_TX_MON_NO_ACK is not equivalent to
>> !(IEEE80211_TX_CTL_ASSIGN_SEQ), as it also disables mac80211's built
>> in sequence numbering.
>
> Ah. But there's precedence for just doing
>
> if (unlikely(!(info->flags & ASSIGN_SEQ)))
>        return;
>
> in the handler that assigns the seq, and I prefer that over adding a =
new
> flag.
>
> johannes
>

But if we redefine !ASSIGN_SEQ to mean "use the sequence number
supplied by userspace", we will need a new flag to distinguish between
driver-level vs. stack-level sequence number assignments, which is
essentially the same as this patch. There is no way to switch between
3 states (control in driver, stack or userspace) using only one bit.

--=20
Vista: [V]iruses, [I]ntruders, [S]pyware, [T]rojans and [A]dware. :-)
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles=
s" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html