This fixes an issus that made "iwconfig eth1 power on" non-working.
When we get a "PS sleep" event, we have to confirm this to the firmware.
The confirm happens with a command, but this command is special: the
firmware won't send us a response. if_cs_host_to_card() is setting
priv->dnld_sent anyway, so this variable stayed at DNLD_DATA_SENT and
was never cleared back.
Now I put the special knowledge that the CMD_802_11_PS_MODE with
CMD_SUBCMD_SLEEP_CONFIRMED doesn't need to need a response by directly
clearing the dnld_sent state in lbs_send_confirmsleep().
Signed-off-by: Holger Schurig <hs4233@mail.mn-solutions.de>
Acked-by: Dan Williams <dcbw@redhat.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
 
        spin_lock_irqsave(&priv->driver_lock, flags);
 
+       /* We don't get a response on the sleep-confirmation */
+       priv->dnld_sent = DNLD_RES_RECEIVED;
+
        /* If nothing to do, go back to sleep (?) */
        if (!__kfifo_len(priv->event_fifo) && !priv->resp_len[priv->resp_idx])
                priv->psstate = PS_STATE_SLEEP;
 
        lbs_deb_enter(LBS_DEB_HOST);
 
+       spin_lock_irqsave(&priv->driver_lock, flags);
        if (priv->dnld_sent) {
                allowed = 0;
                lbs_deb_host("dnld_sent was set\n");
        }
 
-       spin_lock_irqsave(&priv->driver_lock, flags);
        /* In-progress command? */
        if (priv->cur_cmd) {
                allowed = 0;
 
                lbs_deb_thread("4: currenttxskb %p, dnld_sent %d\n",
                       priv->currenttxskb, priv->dnld_sent);
 
-               spin_lock_irq(&priv->driver_lock);
                /* Process any pending command response */
+               spin_lock_irq(&priv->driver_lock);
                resp_idx = priv->resp_idx;
                if (priv->resp_len[resp_idx]) {
                        spin_unlock_irq(&priv->driver_lock);