PERFORCE change 139294 for review
Andrew Thompson
thompsa at FreeBSD.org
Thu Apr 3 21:11:34 UTC 2008
http://perforce.freebsd.org/chv.cgi?CH=139294
Change 139294 by thompsa at thompsa_burger on 2008/04/03 21:11:27
We may panic if stopping the callout in wpi_newstate as the mutex _must_
be held. Go back to bailing from the callout if we arent in RUN anymore
but hack it for vap. Kinda icky.
Affected files ...
.. //depot/projects/vap/sys/dev/wpi/if_wpi.c#14 edit
Differences ...
==== //depot/projects/vap/sys/dev/wpi/if_wpi.c#14 (text+ko) ====
@@ -1263,8 +1263,6 @@
ieee80211_state_name[vap->iv_state],
ieee80211_state_name[nstate], sc->flags));
- callout_stop(&sc->calib_to);
-
if (nstate == IEEE80211_S_AUTH) {
/* Delay the auth transition until we can update the firmware */
error = wpi_queue_cmd(sc, WPI_AUTH, arg, WPI_QUEUE_NORMAL);
@@ -3161,8 +3159,13 @@
wpi_calib_timeout(void *arg)
{
struct wpi_softc *sc = arg;
+ struct ieee80211com *ic = &sc->sc_ic;
+ struct ieee80211vap *vap = TAILQ_FIRST(&ic->ic_vaps);
int temp;
+ if (vap->iv_state != IEEE80211_S_RUN)
+ return;
+
/* update sensor data */
temp = (int)WPI_READ(sc, WPI_TEMPERATURE);
DPRINTFN(WPI_DEBUG_TEMP,("Temp in calibration is: %d\n", temp));
More information about the p4-projects
mailing list