Re: ath5k oops (recent regression, I think)

[Johannes Berg <johannes@sipsolutions.net>]

[-- Attachment #1: Type: text/plain, Size: 1243 bytes --]

Hi,

I'm absolutely unable to read the stack trace you pasted, I just don't
know how to read x86 stack traces as they're usually badly mangled.

> seems ath5k likes to write some rate registers before vif is set up. i used 
> the following as a stopgap fix. johannes, do you have any advice how to 
> properly fix that?

> +       if (sc->vif == NULL) {
> +               printk("*** sc->vif NULL\n");
> +               return;
> +       }
> +

Superficially, this fix looks correct if this function
(ath5k_hw_write_rate_duration) is called before a virtual interface is
brought up or down. I didn't think that would happen so I didn't protect
in mac80211 against it, ieee80211_generic_frame_duration() would have
returned 0 with the original behaviour.

Hence, if you want to restore the original behaviour, do something like

tx_time = 0;
if (sc->vif)
	tx_time = ieee80211_generic_frame_duration(....)

instead at the spot where ieee80211_generic_frame_duration() is used.

I don't think that's correct though and I have no idea why ath5k needs a
frame duration before it has an interface assigned (since it will then
never send a frame), but that's something for the ath5k people to figure
out.

johannes

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