From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756156AbbLAMAE (ORCPT ); Tue, 1 Dec 2015 07:00:04 -0500 Received: from mga11.intel.com ([192.55.52.93]:59898 "EHLO mga11.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756109AbbLAMAB (ORCPT ); Tue, 1 Dec 2015 07:00:01 -0500 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.20,368,1444719600"; d="scan'208";a="697824243" From: "Winkler, Tomas" To: "linux@roeck-us.net" , "gregkh@linuxfoundation.org" , "wim@iguana.be" CC: "linux-kernel@vger.kernel.org" , "Usyskin, Alexander" , "linux-watchdog@vger.kernel.org" Subject: Re: [char-misc-next 6/6] mei: wd: re-register device on event Thread-Topic: [char-misc-next 6/6] mei: wd: re-register device on event Thread-Index: AQHRK5Ix19o7xiyWKUOb5yshRb8qWp615p8A Date: Tue, 1 Dec 2015 11:59:38 +0000 Message-ID: <1448971132.30677.20.camel@intel.com> References: <1448541081-11299-1-git-send-email-tomas.winkler@intel.com> <1448541081-11299-7-git-send-email-tomas.winkler@intel.com> <565C8281.9070909@roeck-us.net> In-Reply-To: <565C8281.9070909@roeck-us.net> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.12.87.84] Content-Type: text/plain; charset="utf-8" Content-ID: <08226932A9824D4B97AD01152F456F83@intel.com> MIME-Version: 1.0 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Transfer-Encoding: 8bit X-MIME-Autoconverted: from base64 to 8bit by mail.home.local id tB1C0BdI027330 On Mon, 2015-11-30 at 09:08 -0800, Guenter Roeck wrote: > On 11/26/2015 04:31 AM, Tomas Winkler wrote: > > From: Alexander Usyskin > > > > For Intel SKL platform the ME device can inform the host via > > asynchronous notification that the watchdog feature was activated > > on the > > device. The activation doesn't require reboot. > > In that case the driver register the watchdog device with the > > kernel. > > > > Can the watchdog device also be disabled on the fly ? Yes, it can be disabled also on the fly, but this is not delivered via an event, this is handled via ping return value. > > > Signed-off-by: Alexander Usyskin > > Signed-off-by: Tomas Winkler > > --- > > drivers/watchdog/mei_wdt.c | 26 ++++++++++++++++++++++++-- > > 1 file changed, 24 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/watchdog/mei_wdt.c > > b/drivers/watchdog/mei_wdt.c > > index 47f0dc2e822a..ee3b5d12b1b2 100644 > > --- a/drivers/watchdog/mei_wdt.c > > +++ b/drivers/watchdog/mei_wdt.c > > @@ -279,6 +279,21 @@ out: > > complete(&wdt->response); > > } > > > > +/* > > + * mei_wdt_event_notif - callback for notification > > + * > > + * @cldev: bus device > > + */ > > +static void mei_wdt_event_notif(struct mei_cl_device *cldev) > > +{ > > + struct mei_wdt *wdt = mei_cldev_get_drvdata(cldev); > > + > > + if (wdt->state == MEI_WDT_NOT_REQUIRED) { > > + mei_wdt_register(wdt); > > + wdt->state = MEI_WDT_IDLE; > > + } > > +} > > + > > /** > > * mei_wdt_event - callback for event receive > > * > > @@ -291,6 +306,9 @@ static void mei_wdt_event(struct mei_cl_device > > *cldev, > > { > > if (events & BIT(MEI_CL_EVENT_RX)) > > mei_wdt_event_rx(cldev); > > + > > + if (events & BIT(MEI_CL_EVENT_NOTIF)) > > + mei_wdt_event_notif(cldev); > > } > > > > /** > > @@ -466,9 +484,13 @@ static int mei_wdt_probe(struct mei_cl_device > > *cldev, > > > > wd_info.firmware_version = mei_cldev_ver(cldev); > > > > - ret = mei_cldev_register_event_cb(wdt->cldev, > > BIT(MEI_CL_EVENT_RX), > > + ret = mei_cldev_register_event_cb(wdt->cldev, > > + BIT(MEI_CL_EVENT_RX) | > > + BIT(MEI_CL_EVENT_NOTIF), > > mei_wdt_event, NULL); > > - if (ret) { > > + > > + /* on legacy devices notification is not supported */ > > + if (ret && ret != -EOPNOTSUPP) { > > dev_err(&cldev->dev, "Could not register event > > ret=%d\n", ret); > > goto err_disable; > > } > > > Doesn't that mean, though, that the RX event is not registered > either, > even if it is supported ? If so, should you retry and register only > MEI_CL_EVENT_RX ? The code is correct but we definitely need to think of better API Tomas {.n++%ݶw{.n+{G{ayʇڙ,jfhz_(階ݢj"mG?&~iOzv^m ?I