similar to: [Bug 75985] New: HDMI audio device only visible after rescan

Przemysław Kopa reports that since commit b516ea586d71 ("PCI: Enable NVIDIA HDA controllers"), the discrete GPU Nvidia GeForce GT 540M on his 2011 Samsung laptop refuses to runtime suspend, resulting in a power regression and excessive heat. Rivera Valdez witnesses the same issue with a GeForce GT 525M (GF108M) of the same era, as does another Arch Linux user named "R0AR" with

From: Lukas Wunner <lukas at wunner.de> The integrated HDA controller on Nvidia GPUs can be hidden with a bit in the GPU's config space. Information about this scheme was provided by NVIDIA on their forums. Many laptops now ship with this device hidden, meaning that Linux users of affected platforms (where the HDMI connector comes off the NVIDIA GPU) cannot use HDMI audio

From: Lukas Wunner <lukas at wunner.de> The integrated HDA controller on Nvidia GPUs can be hidden with a bit in the GPU's config space. Information about this scheme was provided by NVIDIA on their forums. Many laptops now ship with this device hidden, meaning that Linux users of affected platforms (where the HDMI connector comes off the NVIDIA GPU) cannot use HDMI audio

Since quirk_nvidia_hda() was added there's now two nvidia device functions on any laptops with nvidia GPUs: the HDA controller, and the GPU itself. Unfortunately this has the sideaffect of breaking quirk_reset_lenovo_thinkpad_50_nvgpu() since pci_reset_function() was using pci_parent_bus_reset() to reset the GPU's respective PCI bus, and pci_parent_bus_reset() does not work on busses which

This reverts commit b516ea586d717472178e6ef1c152e85608b0ce32. While this fixes audio for a number of users, this commit has the sideaffect of breaking the BIOS workaround that's required to make the GPU on the nvidia P50 work, by causing the GPU's PCI device function to stop working after it's been set to multifunction mode. Signed-off-by: Lyude Paul <lyude at redhat.com>

Modernize vga_switcheroo by using a device link to enforce a runtime PM dependency from an HDA controller to the GPU it's integrated into, v2. Changes since v1: - Replace patch [1/7] to use pci_save_state() / pci_restore_state() for consistency between runtime PM code path of bound and unbound devices. (Rafael, Bjorn) - Patch [5/7]: Drop an unnecessary initialization. (Bjorn) Rephrase

Modernize vga_switcheroo by using a "device link" to enforce a runtime PM dependency from an HDA controller to the GPU it's integrated into. Remove thereby obsoleted code and fix a bunch of bugs. Device links were introduced in v4.10. Users might see a small power saving if the discrete GPU is in use and its HDA controller is not, because the HDA controller is now allowed to runtime

On Mon, Jul 08, 2019 at 01:17:44PM +0800, Daniel Drake wrote: > From: Lukas Wunner <lukas at wunner.de> > > The integrated HDA controller on Nvidia GPUs can be hidden with a bit in > the GPU's config space. Information about this scheme was provided by > NVIDIA on their forums. > > Many laptops now ship with this device hidden, meaning that Linux users > of

[+cc Rafael, Martin, zigarrre] On Thu, Jun 13, 2019 at 02:35:14PM +0800, Daniel Drake wrote: > From: Lukas Wunner <lukas at wunner.de> > > The integrated HDA controller on Nvidia GPUs can be hidden with a bit in > the GPU's config space. Information about this scheme was provided by > NVIDIA on their forums. > > Many laptops now ship with this device hidden,

On Mon, Feb 19, 2018 at 03:05:53PM +0100, Daniel Vetter wrote: > On Mon, Feb 19, 2018 at 12:58:17PM +0100, Lukas Wunner wrote: > > On Mon, Feb 19, 2018 at 12:34:43PM +0100, Daniel Vetter wrote: > > > Well, userspace expects hotplug events, even when we runtime suspend > > > stuff. Hence waking shit up with polling. Imo ignoring hotplug events is a > > > pretty

On Thu, Jun 13, 2019 at 2:35 AM Daniel Drake <drake at endlessm.com> wrote: > > From: Lukas Wunner <lukas at wunner.de> > > The integrated HDA controller on Nvidia GPUs can be hidden with a bit in > the GPU's config space. Information about this scheme was provided by > NVIDIA on their forums. > > Many laptops now ship with this device hidden, meaning that

On Wed, 2019-07-31 at 23:26 +0200, Karol Herbst wrote: > On Wed, Jul 31, 2019 at 11:18 PM Lukas Wunner <lukas at wunner.de> wrote: > > On Wed, Jul 31, 2019 at 04:19:27PM -0400, Lyude Paul wrote: > > > While this fixes audio for a number of users, this commit has the > > > sideaffect of breaking the BIOS workaround that's required to make the > > > GPU

On Wed, Jul 31, 2019 at 04:19:27PM -0400, Lyude Paul wrote: > While this fixes audio for a number of users, this commit has the > sideaffect of breaking the BIOS workaround that's required to make the > GPU on the nvidia P50 work, by causing the GPU's PCI device function to > stop working after it's been set to multifunction mode. This is missing a reference to the commit

On Tue, Mar 06, 2018 at 11:29:40AM +0100, Daniel Vetter wrote: > On Sat, Mar 03, 2018 at 10:53:24AM +0100, Lukas Wunner wrote: > > Modernize vga_switcheroo by using a device link to enforce a runtime PM > > dependency from an HDA controller to the GPU it's integrated into, v2. > > > > https://github.com/l1k/linux/commits/switcheroo_devlink_v2 > > This all

On Mon, Feb 19, 2018 at 12:34:43PM +0100, Daniel Vetter wrote: > On Sun, Feb 11, 2018 at 10:38:28AM +0100, Lukas Wunner wrote: > > Fix a deadlock on hybrid graphics laptops that's been present since 2013: > > > > DRM drivers poll connectors in 10 sec intervals. The poll worker is > > stopped on ->runtime_suspend with cancel_delayed_work_sync(). However >

Users kept complaining about those messages and it's a little spammy on prime systems so turn it into a debug print. Cc: Bjorn Helgaas <bhelgaas at google.com> Cc: Lukas Wunner <lukas at wunner.de> Cc: linux-pci at vger.kernel.org Cc: nouveau at lists.freedesktop.org Fixes: b516ea586d71 ("PCI: Enable NVIDIA HDA controllers") Signed-off-by: Karol Herbst <kherbst at

Hi Peter, > [12:42] Lekensteyn: Should the video card always be powered up when a > register is read from the HDMI audio controller? Or would it be > better to leave the video card suspended and just fail the HDA > register reads? It should probably be powered up. > [12:43] Lekensteyn: I'm trying to figure out how vga_switcheroo should handle this, maybe l1k has an

On Tue, May 24, 2016 at 06:03:27PM +0200, Lukas Wunner wrote: > nouveau_drm_load() calls pm_runtime_put() if nouveau_runtime_pm != 0, > but nouveau_drm_unload() calls pm_runtime_get_sync() unconditionally. > We therefore leak a runtime pm ref whenever nouveau is loaded with > runpm=0 and then unloaded. The GPU will subsequently never runtime > suspend even if nouveau is loaded again

Fix a deadlock on hybrid graphics laptops that's been present since 2013: DRM drivers poll connectors in 10 sec intervals. The poll worker is stopped on ->runtime_suspend with cancel_delayed_work_sync(). However the poll worker invokes the DRM drivers' ->detect callbacks, which call pm_runtime_get_sync(). If the poll worker starts after runtime suspend has begun,

On Sun, Feb 18, 2018 at 09:38:32AM +0100, Lukas Wunner wrote: > PCI devices not bound to a driver are supposed to stay in D0 during > runtime suspend. Doesn't "runtime suspend" mean an individual device is suspended while the rest of the system remains active? If so, maybe it would be more direct to say "PCI devices not bound to a driver cannot be runtime