Return-path: Received: from panicking.kicks-ass.org ([95.141.38.207]:52408 "EHLO panicking.kicks-ass.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1755801Ab1DHJzB (ORCPT ); Fri, 8 Apr 2011 05:55:01 -0400 Subject: Re: [PATCH RFC 2/3] libertas: add set_power callback for iface subdrivers From: Alberto Panizzo To: Dan Williams Cc: Vasily Khoruzhick , linux-wireless@vger.kernel.org, "John W. Linville" , libertas-dev@lists.infradead.org In-Reply-To: <1302215732.17586.25.camel@dcbw.foobar.com> References: <1302207960-28536-1-git-send-email-anarsoul@gmail.com> <1302207960-28536-3-git-send-email-anarsoul@gmail.com> <1302215732.17586.25.camel@dcbw.foobar.com> Content-Type: text/plain; charset="UTF-8" Date: Fri, 08 Apr 2011 11:54:48 +0200 Message-ID: <1302256488.2649.70.camel@realization> Mime-Version: 1.0 Sender: linux-wireless-owner@vger.kernel.org List-ID: On Thu, 2011-04-07 at 17:35 -0500, Dan Williams wrote: > On Thu, 2011-04-07 at 23:25 +0300, Vasily Khoruzhick wrote: > > Add set_power callback, so driver can disable device when > > interface is down or before going into suspend > > > > Signed-off-by: Vasily Khoruzhick > > --- > > drivers/net/wireless/libertas/cmd.c | 18 ++++- > > drivers/net/wireless/libertas/dev.h | 3 +- > > drivers/net/wireless/libertas/main.c | 177 ++++++++++++++++++++++------------ > > 3 files changed, 135 insertions(+), 63 deletions(-) > > > > diff --git a/drivers/net/wireless/libertas/cmd.c b/drivers/net/wireless/libertas/cmd.c > > index 7e8a658..d0e7df6 100644 > > --- a/drivers/net/wireless/libertas/cmd.c > > +++ b/drivers/net/wireless/libertas/cmd.c > > @@ -84,6 +84,7 @@ static u8 is_command_allowed_in_ps(u16 cmd) > > int lbs_update_hw_spec(struct lbs_private *priv) > > { > > struct cmd_ds_get_hw_spec cmd; > > + struct cmd_ds_802_11_mac_address hw_addr_cmd; > > int ret = -1; > > u32 i; > > > > @@ -151,8 +152,20 @@ int lbs_update_hw_spec(struct lbs_private *priv) > > memcpy(priv->mesh_dev->dev_addr, > > priv->current_addr, ETH_ALEN); > > priv->copied_hwaddr = 1; > > + } else { > > + /* Copy addr back to hw */ > > + memcpy(priv->current_addr, priv->dev->dev_addr, ETH_ALEN); > > + hw_addr_cmd.hdr.size = cpu_to_le16(sizeof(hw_addr_cmd)); > > + hw_addr_cmd.action = cpu_to_le16(CMD_ACT_SET); > > + memcpy(hw_addr_cmd.macadd, priv->current_addr, ETH_ALEN); > > + > > + ret = lbs_cmd_with_response(priv, CMD_802_11_MAC_ADDRESS, > > + &hw_addr_cmd); > > + if (ret) > > + lbs_deb_net("set MAC address failed\n"); > > Might want to make a note here that this part is only used if the MAC > was set while the card was powered down. > > > } > > > > + > > out: > > lbs_deb_leave(LBS_DEB_CMD); > > return ret; > > @@ -1111,6 +1124,7 @@ out: > > > > void lbs_set_mac_control(struct lbs_private *priv) > > { > > + int ret; > > struct cmd_ds_mac_control cmd; > > > > lbs_deb_enter(LBS_DEB_CMD); > > @@ -1119,7 +1133,9 @@ void lbs_set_mac_control(struct lbs_private *priv) > > cmd.action = cpu_to_le16(priv->mac_control); > > cmd.reserved = 0; > > > > - lbs_cmd_async(priv, CMD_MAC_CONTROL, &cmd.hdr, sizeof(cmd)); > > + ret = lbs_cmd_with_response(priv, CMD_MAC_CONTROL, &cmd); > > + if (ret) > > + lbs_deb_net("set MAC control failed\n"); > > > > lbs_deb_leave(LBS_DEB_CMD); > > } > > diff --git a/drivers/net/wireless/libertas/dev.h b/drivers/net/wireless/libertas/dev.h > > index bc461eb..9a873bc 100644 > > --- a/drivers/net/wireless/libertas/dev.h > > +++ b/drivers/net/wireless/libertas/dev.h > > @@ -90,12 +90,13 @@ struct lbs_private { > > void *card; > > u8 fw_ready; > > u8 surpriseremoved; > > - u8 setup_fw_on_resume; > > + int power_on_cnt; > > int (*hw_host_to_card) (struct lbs_private *priv, u8 type, u8 *payload, u16 nb); > > void (*reset_card) (struct lbs_private *priv); > > int (*enter_deep_sleep) (struct lbs_private *priv); > > int (*exit_deep_sleep) (struct lbs_private *priv); > > int (*reset_deep_sleep_wakeup) (struct lbs_private *priv); > > + void (*set_power) (struct lbs_private *priv, int enable); > > > > /* Adapter info (from EEPROM) */ > > u32 fwrelease; > > diff --git a/drivers/net/wireless/libertas/main.c b/drivers/net/wireless/libertas/main.c > > index ca8149c..7922ead 100644 > > --- a/drivers/net/wireless/libertas/main.c > > +++ b/drivers/net/wireless/libertas/main.c > > @@ -89,6 +89,73 @@ u8 lbs_data_rate_to_fw_index(u32 rate) > > return 0; > > } > > > > +/** > > + * @brief This function gets the HW spec from the firmware and sets > > + * some basic parameters. > > + * > > + * @param priv A pointer to struct lbs_private structure > > + * @return 0 or -1 > > + */ > > +static int lbs_setup_firmware(struct lbs_private *priv) > > +{ > > + int ret = -1; > > + s16 curlevel = 0, minlevel = 0, maxlevel = 0; > > + > > + lbs_deb_enter(LBS_DEB_FW); > > + > > + /* Read MAC address from firmware */ > > + memset(priv->current_addr, 0xff, ETH_ALEN); > > + ret = lbs_update_hw_spec(priv); > > + if (ret) > > + goto done; > > + > > + /* Read power levels if available */ > > + ret = lbs_get_tx_power(priv, &curlevel, &minlevel, &maxlevel); > > + if (ret == 0) { > > + priv->txpower_cur = curlevel; > > + priv->txpower_min = minlevel; > > + priv->txpower_max = maxlevel; > > + } > > + > > + /* Send cmd to FW to enable 11D function */ > > + ret = lbs_set_snmp_mib(priv, SNMP_MIB_OID_11D_ENABLE, 1); > > + > > + lbs_set_mac_control(priv); > > +done: > > + lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret); > > + return ret; > > +} > > + > > +/** > > + * @brief This function enables device > > + * > > + * @param priv A pointer to struct lbs_private structure > > + */ > > +static void lbs_power_on(struct lbs_private *priv) > > +{ > > + if (priv->set_power) { > > + priv->power_on_cnt++; > > + if (priv->power_on_cnt == 1) { > > + priv->set_power(priv, 1); > > + lbs_setup_firmware(priv); > > + lbs_init_mesh(priv); > > + } > > + } > > +} > > + > > +/** > > + * @brief This function disables device > > + * > > + * @param priv A pointer to struct lbs_private structure > > + */ > > +static void lbs_power_off(struct lbs_private *priv) > > +{ > > + if (priv->set_power) { > > + priv->power_on_cnt--; > > + if (priv->power_on_cnt == 0) > > + priv->set_power(priv, 0); > > + } > > +} > > > > /** > > * @brief This function opens the ethX interface > > @@ -103,6 +170,8 @@ static int lbs_dev_open(struct net_device *dev) > > > > lbs_deb_enter(LBS_DEB_NET); > > > > + lbs_power_on(priv); > > + > > spin_lock_irq(&priv->driver_lock); > > priv->stopping = false; > > > > @@ -136,13 +205,14 @@ static int lbs_eth_stop(struct net_device *dev) > > netif_stop_queue(dev); > > spin_unlock_irq(&priv->driver_lock); > > > > - schedule_work(&priv->mcast_work); > > Any reason this is getting removed? > > > cancel_delayed_work_sync(&priv->scan_work); > > if (priv->scan_req) { > > cfg80211_scan_done(priv->scan_req, false); > > priv->scan_req = NULL; > > } > > > > + lbs_power_off(priv); > > + > > So the idea here is that when the device is opened, the card powers up, > but when it's closed, the card powers down? If that's the case, we can > make mesh work with this too if you also call > lbs_power_on()/lbs_power_off() from the mesh.c functions that open/close > the mesh device, since power on/off is refcounted, essentially. > > > lbs_deb_leave(LBS_DEB_NET); > > return 0; > > } > > @@ -201,15 +271,16 @@ int lbs_set_mac_address(struct net_device *dev, void *addr) > > > > /* In case it was called from the mesh device */ > > dev = priv->dev; > > + if (priv->power_on_cnt) { > > + cmd.hdr.size = cpu_to_le16(sizeof(cmd)); > > + cmd.action = cpu_to_le16(CMD_ACT_SET); > > + memcpy(cmd.macadd, phwaddr->sa_data, ETH_ALEN); > > > > - cmd.hdr.size = cpu_to_le16(sizeof(cmd)); > > - cmd.action = cpu_to_le16(CMD_ACT_SET); > > - memcpy(cmd.macadd, phwaddr->sa_data, ETH_ALEN); > > - > > - ret = lbs_cmd_with_response(priv, CMD_802_11_MAC_ADDRESS, &cmd); > > - if (ret) { > > - lbs_deb_net("set MAC address failed\n"); > > - goto done; > > + ret = lbs_cmd_with_response(priv, CMD_802_11_MAC_ADDRESS, &cmd); > > + if (ret) { > > + lbs_deb_net("set MAC address failed\n"); > > + goto done; > > + } > > } > > > > memcpy(priv->current_addr, phwaddr->sa_data, ETH_ALEN); > > @@ -539,59 +610,27 @@ static int lbs_thread(void *data) > > return 0; > > } > > > > -/** > > - * @brief This function gets the HW spec from the firmware and sets > > - * some basic parameters. > > - * > > - * @param priv A pointer to struct lbs_private structure > > - * @return 0 or -1 > > - */ > > -static int lbs_setup_firmware(struct lbs_private *priv) > > -{ > > - int ret = -1; > > - s16 curlevel = 0, minlevel = 0, maxlevel = 0; > > - > > - lbs_deb_enter(LBS_DEB_FW); > > - > > - /* Read MAC address from firmware */ > > - memset(priv->current_addr, 0xff, ETH_ALEN); > > - ret = lbs_update_hw_spec(priv); > > - if (ret) > > - goto done; > > - > > - /* Read power levels if available */ > > - ret = lbs_get_tx_power(priv, &curlevel, &minlevel, &maxlevel); > > - if (ret == 0) { > > - priv->txpower_cur = curlevel; > > - priv->txpower_min = minlevel; > > - priv->txpower_max = maxlevel; > > - } > > - > > - /* Send cmd to FW to enable 11D function */ > > - ret = lbs_set_snmp_mib(priv, SNMP_MIB_OID_11D_ENABLE, 1); > > - > > - lbs_set_mac_control(priv); > > -done: > > - lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret); > > - return ret; > > -} > > - > > int lbs_suspend(struct lbs_private *priv) > > { > > - int ret; > > + int ret = 0; > > > > lbs_deb_enter(LBS_DEB_FW); > > > > - if (priv->is_deep_sleep) { > > - ret = lbs_set_deep_sleep(priv, 0); > > - if (ret) { > > - lbs_pr_err("deep sleep cancellation failed: %d\n", ret); > > - return ret; > > + if (priv->set_power) > > + lbs_power_off(priv); > > + else { > > + if (priv->is_deep_sleep) { > > + ret = lbs_set_deep_sleep(priv, 0); > > + if (ret) { > > + lbs_pr_err( > > + "deep sleep cancellation failed: %d\n", ret); > > + return ret; > > + } > > + priv->deep_sleep_required = 1; > > } > > - priv->deep_sleep_required = 1; > > - } > > > > - ret = lbs_set_host_sleep(priv, 1); > > + ret = lbs_set_host_sleep(priv, 1); > > + } > > > > netif_device_detach(priv->dev); > > if (priv->mesh_dev) > > @@ -604,26 +643,26 @@ EXPORT_SYMBOL_GPL(lbs_suspend); > > > > int lbs_resume(struct lbs_private *priv) > > { > > - int ret; > > + int ret = 0; > > > > lbs_deb_enter(LBS_DEB_FW); > > > > - ret = lbs_set_host_sleep(priv, 0); > > + if (priv->set_power) > > + lbs_power_on(priv); > > + else > > + ret = lbs_set_host_sleep(priv, 0); > > > > netif_device_attach(priv->dev); > > if (priv->mesh_dev) > > netif_device_attach(priv->mesh_dev); > > > > - if (priv->deep_sleep_required) { > > + if (!priv->set_power && priv->deep_sleep_required) { > > priv->deep_sleep_required = 0; > > ret = lbs_set_deep_sleep(priv, 1); > > if (ret) > > lbs_pr_err("deep sleep activation failed: %d\n", ret); > > } > > > > - if (priv->setup_fw_on_resume) > > - ret = lbs_setup_firmware(priv); > > - > > lbs_deb_leave_args(LBS_DEB_FW, "ret %d", ret); > > return ret; > > } > > @@ -917,6 +956,9 @@ void lbs_remove_card(struct lbs_private *priv) > > priv->surpriseremoved = 1; > > kthread_stop(priv->main_thread); > > > > + /* Disable card power if it's still on */ > > + lbs_power_off(priv); > > + > > lbs_free_adapter(priv); > > lbs_cfg_free(priv); > > free_netdev(dev); > > @@ -944,6 +986,16 @@ int lbs_start_card(struct lbs_private *priv) > > > > lbs_deb_enter(LBS_DEB_MAIN); > > > > + /* We're not using lbs_power_on here because we don't want > > + * to setup firmware twice. > > + * TODO: replace following code with lbs_power_on() when set_power > > + * callback become mandatory. > > + */ > > + if (priv->set_power) { > > + priv->power_on_cnt++; > > + priv->set_power(priv, 1); > > + } > > + > > /* poke the firmware */ > > ret = lbs_setup_firmware(priv); > > if (ret) > > @@ -964,6 +1016,9 @@ int lbs_start_card(struct lbs_private *priv) > > > > ret = 0; > > > > + /* Init is done, so we can disable card power */ > > + lbs_power_off(priv); > > + > > done: > > lbs_deb_leave_args(LBS_DEB_MAIN, "ret %d", ret); > > return ret; > > The thing that worries me here (and it's not your fault) is that there > are 4 different and overlapping ways to do power saving with libertas: > > 1) normal 802.11 power save (CMD_802_11_PS_MODE): independent of > everything else > > 2) deep sleep + host sleep: SDIO-only for now, used by OLPC, but could > be used by all bus types. Bus remains powered unless the host sleeps > too, and it requires bus-specific interaction (GPIO) to wake up the > card. Doesn't require a firmware reload because the card is still > powered. > > 3) bus sleep/power on bus power events: requires a firmware reload when > the bus comes back because the bus and card aren't powered, triggered by > some external suspend/resume mechanism > > 4) bus sleep/power on dev open/close: same as #3 except the trigger is > internal to the driver > > and it's getting pretty complicated in the code... You'll want to do > any of [2, 3, 4] depending on your platform, so that's really the only > difference. I'd think that we'd want to enable the dev open/close power > save stuff by default. > > We want deep sleep when we know that (a) the card will be powered > despite the bus or host being unpowered, and that's a platform decision. > I wonder if there's a better way to integrate all this stuff so it's > less confusing in the code? > > Dan Maybe I can say something on this. We got the libertas driver from 2.6.34 working very well with Power Save mode and Host sleep with an 8686 chip and SDIO interface. This means that: While the wifi chip is connected to an AP: - We can suspend the CPU maintaining the wifi chip in full power. And this is lot reliable. - We can suspend the CPU when the wifi chip in Power Save mode. This is less reliable since the libertas driver do not manage the firmware specifications strictly in this mode. This have to be fixed redesigning the libertas driver in the way of Power save complete awareness: I think that all power save command should be moved to a new command layer between the cmd.c and the bus interface where to implement the documentation logic. There should not be any direct calls to command executions, since there are strict rules in times where commands are allowed and what command are allowed too. We got this working also modifying MMC subsystem code. Because in PS mode the chip do not accept any bus commands, consequently, when resuming, the MMC subsystem does not to do any probes of SDIO bus state if the wifi chip is in PS mode. Well, unfortunately the code is really bound in our use case (fits the Android opportunistic power management) and needs cleanups to get merged. I know that this hurt your ears, but the cleanup is scheduled and I hope this will happen asap. We got tested also the wake filters as the 8686 documentation describe. These filters are working properly but the driver do not manage these in the right way so from the first wake and next suspend the wake conditions got lost. I didn't gone too deep in this because I think this is something that will be managed proficiently in the new power aware layer I told. >From our tests, we found that if the station remains connected to one AP, all the power management can be really useful to save power in a reliable way. But if the station is moving through different rooms with different APs (which are broadcasting the same SSID) If the host is suspended and wake conditions are bound to the wifi chip, the wifi chip does not be able to wake the system on network events because (I think and this have to be proven) the wifi chip is listening at the original AP and does not manage roaming correctly. This even when the wifi chip is maintained full active. Lot of things I know. But even if we will not have enough time to cleanup the code for submission I am really open to be a functionality tester. Best regards, Alberto! > _______________________________________________ > libertas-dev mailing list > libertas-dev@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/libertas-dev