From mboxrd@z Thu Jan 1 00:00:00 1970 From: Henrique de Moraes Holschuh Subject: Re: [PATCH v8 2/7] platform/x86/thinkpad_acpi: Use brightness_set_blocking callback for LEDs Date: Thu, 09 Feb 2017 16:00:16 -0200 Message-ID: <1486663216.1043881.875902856.2EF773DE@webmail.messagingengine.com> References: <20170209154417.19040-1-hdegoede@redhat.com> <20170209154417.19040-3-hdegoede@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20170209154417.19040-3-hdegoede@redhat.com> Sender: platform-driver-x86-owner@vger.kernel.org To: Hans de Goede , Darren Hart , Andy Shevchenko , =?utf-8?Q?Pali=20Roh=C3=A1r?= , Jacek Anaszewski , Pavel Machek Cc: platform-driver-x86@vger.kernel.org, linux-leds@vger.kernel.org List-Id: linux-leds@vger.kernel.org On Thu, Feb 9, 2017, at 13:44, Hans de Goede wrote: > Now a days the LED core can take care of executing brightness_set from > a workqueue if it needs to sleep, make use of this and remove a bunch > of DIY code for this. > > Since this commit removes the workqueue usage for LEDs, the > led_sysfs_blink_set callback may now also sleep, this is fine. > > Signed-off-by: Hans de Goede For the thinkpad-acpi bits: Acked-by: Henrique de Moraes Holschuh BTW, one sure-fire way to test this is to attach a trigger fired from atomic context. If it doesn't defer to a workqueue, the machine *will* crash. Thinkpad-acpi LEDs need to sleep (and will trigger an SMI, etc, etc). -- Henrique Holschuh