Re: [panic] Race in IEEE802.11 layer towards device drivers

From: Hans Petter Selasky <hselasky_at_c2i.net> Date: Wed, 14 Jul 2010 15:11:46 +0200 · This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:40:05 UTC

On Wednesday 14 July 2010 14:31:29 PseudoCylon wrote:
> -if(vap->iv_opmode == IEEE80211_M_HOSTAP){
> -RUN_LOCK(sc);
> +if (vap->iv_opmode == IEEE80211_M_HOSTAP)
>  sc->cmdq_key_set = RUN_CMDQ_GO;
> -RUN_UNLOCK(sc);
> -}
>  

> Why are you removing these locks?

Another question:

	i = RUN_CMDQ_GET(&sc->cmdq_store);
	DPRINTF("cmdq_store=%d\n", i);
	sc->cmdq[i].func = run_update_beacon_cb;
	sc->cmdq[i].arg0 = vap;

Why is this code and similar places not enclosed with mutexes?

--HPS