dri-devel October 2016

dri-devel@lists.freedesktop.org

215 participants
1390 discussions

[Bug 98037] HD6450 KMS failure
by bugzilla-daemon＠freedesktop.org 04 Oct '16

04 Oct '16

1 0

[PATCH] drm: virtio: reinstate drm_virtio_set_busid()
by Laszlo Ersek 04 Oct '16

04 Oct '16

Before commit a325725633c2 ("drm: Lobotomize set_busid nonsense for !pci drivers"), several DRM drivers for platform devices used to expose an explicit "drm_driver.set_busid" callback, invariably backed by drm_platform_set_busid(). Commit a325725633c2 removed drm_platform_set_busid(), along with the referring .set_busid field initializations. This was justified because interchangeable functionality had been implemented in drm_dev_alloc() / drm_dev_init(), which DRM_IOCTL_SET_VERSION would rely on going forward. However, commit a325725633c2 also removed drm_virtio_set_busid(), for which the same consolidation was not appropriate: this .set_busid callback had been implemented with drm_pci_set_busid(), and not drm_platform_set_busid(). The error regressed Xorg/xserver on QEMU's "virtio-vga" card; the drmGetBusid() function from libdrm would no longer return stable PCI identifiers like "pci:0000:00:02.0", but rather unstable platform ones like "virtio0". Reinstate drm_virtio_set_busid() with judicious use of git checkout -p a325725633c2^ -- drivers/gpu/drm/virtio Cc: Daniel Vetter <daniel.vetter(a)ffwll.ch> Cc: Daniel Vetter <daniel.vetter(a)intel.com> Cc: David Airlie <airlied(a)redhat.com> Cc: Emil Velikov <emil.l.velikov(a)gmail.com> Cc: Gerd Hoffmann <kraxel(a)redhat.com> Cc: Gustavo Padovan <gustavo.padovan(a)collabora.co.uk> Cc: Hans de Goede <hdegoede(a)redhat.com> Cc: Joachim Frieben <jfrieben(a)hotmail.com> Cc: stable(a)vger.kernel.org Reported-by: Joachim Frieben <jfrieben(a)hotmail.com> Fixes: a325725633c26aa66ab940f762a6b0778edf76c0 Ref: https://bugzilla.redhat.com/show_bug.cgi?id=1366842 Signed-off-by: Laszlo Ersek <lersek(a)redhat.com> Reviewed-by: Emil Velikov <emil.l.velikov(a)gmail.com> --- drivers/gpu/drm/virtio/virtgpu_drm_bus.c | 10 ++++++++++ drivers/gpu/drm/virtio/virtgpu_drv.c | 1 + drivers/gpu/drm/virtio/virtgpu_drv.h | 1 + 3 files changed, 12 insertions(+) diff --git a/drivers/gpu/drm/virtio/virtgpu_drm_bus.c b/drivers/gpu/drm/virtio/virtgpu_drm_bus.c index 7f0e93f87a55..88a39165edd5 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drm_bus.c +++ b/drivers/gpu/drm/virtio/virtgpu_drm_bus.c @@ -27,6 +27,16 @@ #include "virtgpu_drv.h" +int drm_virtio_set_busid(struct drm_device *dev, struct drm_master *master) +{ + struct pci_dev *pdev = dev->pdev; + + if (pdev) { + return drm_pci_set_busid(dev, master); + } + return 0; +} + static void virtio_pci_kick_out_firmware_fb(struct pci_dev *pci_dev) { struct apertures_struct *ap; diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c index c13f70cfc461..5820b7020ae5 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.c +++ b/drivers/gpu/drm/virtio/virtgpu_drv.c @@ -117,6 +117,7 @@ static const struct file_operations virtio_gpu_driver_fops = { static struct drm_driver driver = { .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_RENDER | DRIVER_ATOMIC, + .set_busid = drm_virtio_set_busid, .load = virtio_gpu_driver_load, .unload = virtio_gpu_driver_unload, .open = virtio_gpu_driver_open, diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index b18ef3111f0c..acf556a35cb2 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -49,6 +49,7 @@ #define DRIVER_PATCHLEVEL 1 /* virtgpu_drm_bus.c */ +int drm_virtio_set_busid(struct drm_device *dev, struct drm_master *master); int drm_virtio_init(struct drm_driver *driver, struct virtio_device *vdev); struct virtio_gpu_object { -- 2.9.2

2 1

[PULL] topic/drm-misc
by Daniel Vetter 04 Oct '16

04 Oct '16

Hi Dave, As promised another pull request. A bit late because flu. - generic pipe crc from Tomeu, required a backmerge to apply - fixes for my fumbled drm_plane.c extraction - display_info cleanup/fixes from Ville - misc stuff all over Cheers, Daniel The following changes since commit c0d5fb4d0d9224ccaad0475c9b58740873970e7e: Merge tag 'drm-qemu-20160921' of git://git.kraxel.org/linux into drm-next (2016-09-28 13:23:07 +1000) are available in the git repository at: git://anongit.freedesktop.org/drm-intel tags/topic/drm-misc-2016-10-03 for you to fetch changes up to 262f75a6288346db44038c63b7a95ee68f8b7bea: drm: Undo damage to page_flip_ioctl (2016-10-03 15:13:38 +0200) ---------------------------------------------------------------- Baoyou Xie (4): drm/rockchip: add missing header dependencies drm/rockchip: mark symbols static where possible drm/mediatek: mark symbols static where possible drm/sti: mark symbols static where possible Chris Wilson (2): drm: Restore lost drm_framebuffer_unreference in drm_mode_page_flip_ioctl drm: Convert prime dma-buf <-> handle to rbtree Daniel Vetter (3): Merge remote-tracking branch 'airlied/drm-next' into topic/drm-misc drm: Document caveats around atomic event handling drm: Undo damage to page_flip_ioctl Dhinakaran Pandiyan (1): drm: Add frame CRC debugfs files only for drivers that have CRTC Emilio López (1): uapi: add missing install of sync_file.h Joe Perches (1): drm: Simplify drm_printk to reduce object size quite a bit Stefan Christ (1): drm/fb-helper: add DRM_FB_HELPER_DEFAULT_OPS for fb_ops Tomeu Vizoso (7): drm/i915/debugfs: Move out pipe CRC code drm: Add API for capturing frame CRCs drm/i915: Use new CRC debugfs API drm/i915: Put "cooked" vlank counters in frame CRC lines drm/i915: Fix build when !CONFIG_DEBUG_FS drm/bridge: analogix_dp: Add analogix_dp_psr_supported drm/rockchip: analogix_dp: Refuse to enable PSR if panel doesn't support it Ville Syrjälä (10): drm/edid: Clear old audio latency values before parsing the new EDID drm/edid: Clear old dvi_dual/max_tmds_clock before parsing the new EDID drm/edid: Make max_tmds_clock kHz instead of MHz drm/edid: Move dvi_dual/max_tmds_clock to drm_display_info drm/edid: Don't pass around drm_display_info needlessly drm/edid: Reduce the number of times we parse the CEA extension block drm/edid: Clear the old cea_rev when there's no CEA extension in the new EDID drm/edid: Move dvi_dual/max_tmds_clock parsing out from drm_edid_to_eld() drm/i915: Replace a bunch of connector->base.display_info with a local variable drm/i915: Account for sink max TMDS clock when checking the port clock Documentation/gpu/drm-uapi.rst | 6 + drivers/gpu/drm/Makefile | 3 +- drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c | 4 +- drivers/gpu/drm/bridge/analogix/analogix_dp_core.c | 8 + drivers/gpu/drm/drm_crtc.c | 29 +- drivers/gpu/drm/drm_debugfs.c | 34 +- drivers/gpu/drm/drm_debugfs_crc.c | 351 +++++++ drivers/gpu/drm/drm_drv.c | 28 +- drivers/gpu/drm/drm_edid.c | 248 ++--- drivers/gpu/drm/drm_internal.h | 16 + drivers/gpu/drm/drm_irq.c | 32 +- drivers/gpu/drm/drm_plane.c | 83 +- drivers/gpu/drm/drm_prime.c | 85 +- drivers/gpu/drm/i915/Makefile | 2 +- drivers/gpu/drm/i915/i915_debugfs.c | 886 +---------------- drivers/gpu/drm/i915/i915_drv.c | 2 +- drivers/gpu/drm/i915/i915_drv.h | 5 +- drivers/gpu/drm/i915/i915_irq.c | 83 +- drivers/gpu/drm/i915/intel_display.c | 15 +- drivers/gpu/drm/i915/intel_drv.h | 13 + drivers/gpu/drm/i915/intel_hdmi.c | 9 +- drivers/gpu/drm/i915/intel_pipe_crc.c | 1014 ++++++++++++++++++++ drivers/gpu/drm/mediatek/mtk_hdmi.c | 11 +- drivers/gpu/drm/radeon/radeon_connectors.c | 4 +- drivers/gpu/drm/rockchip/analogix_dp-rockchip.c | 3 + drivers/gpu/drm/rockchip/rockchip_drm_drv.c | 4 +- drivers/gpu/drm/rockchip/rockchip_drm_fbdev.c | 1 + include/drm/bridge/analogix_dp.h | 1 + include/drm/drmP.h | 35 +- include/drm/drm_connector.h | 15 +- include/drm/drm_crtc.h | 97 +- include/drm/drm_debugfs_crc.h | 73 ++ include/drm/drm_fb_helper.h | 13 + include/uapi/linux/Kbuild | 1 + 34 files changed, 2069 insertions(+), 1145 deletions(-) create mode 100644 drivers/gpu/drm/drm_debugfs_crc.c create mode 100644 drivers/gpu/drm/i915/intel_pipe_crc.c create mode 100644 include/drm/drm_debugfs_crc.h -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch

2 1

[ANNOUNCE] libdrm 2.4.71
by robdclark＠gmail.com 04 Oct '16

04 Oct '16

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Daniel Kurtz (1): modetest: add mediatek to module list Eric Anholt (1): Simplify the RELEASING steps based on current release.sh. Flora Cui (1): amdgpu: expose the AMDGPU_GEM_CREATE_VRAM_CLEARED flag Kristian H. Kristensen (1): Add drmModeAddFB2WithModifiers() which takes format modifiers Leo Liu (1): tests/amdgpu: update vce encRefPic addr mode to tiled Marek Olšák (3): radeon: sync radeon_drm.h with the kernel amdgpu: sync amdgpu_drm.h with the kernel amdgpu: fix valgrind warnings caused by amdgpu_query_firmware_version Michel Dänzer (1): intel: Add new symbols to intel-symbol-check Rob Clark (2): list: fix an issue with android build using clang Bump version for release Ronie Salgado (1): amdgpu: add SI support Stéphane Marchesin (1): modetest: Also print the pixel clock The etnaviv authors (2): libdrm: add etnaviv drm support libdrm: add etnaviv tests Yang Rong (1): intel: Export pooled EU and min no. of eus in a pool. satsahu (1): modetest: Adding amdgpu to module list git tag: libdrm-2.4.71 http://dri.freedesktop.org/libdrm/libdrm-2.4.71.tar.bz2 MD5: 776bccc84618b616fc57a7143836ae7a libdrm-2.4.71.tar.bz2 SHA1: 5c9cd0c45d92bcd4809e42681494163ba9f31f3d libdrm-2.4.71.tar.bz2 SHA256: c66287ddeee5f46ea8f8880b94b80acb3bbc33ba6321d17767eef145046df9b8 libdrm-2.4.71.tar.bz2 PGP: http://dri.freedesktop.org/libdrm/libdrm-2.4.71.tar.bz2.sig http://dri.freedesktop.org/libdrm/libdrm-2.4.71.tar.gz MD5: a9e5bf225873165fd38cad275e7b6a86 libdrm-2.4.71.tar.gz SHA1: 4ff0dcb74af5fb008804b5ed457356435f5e1ec1 libdrm-2.4.71.tar.gz SHA256: c78804adc7f05bccb448188d7839c6769cf9f926dfde88603256942b1c4f6170 libdrm-2.4.71.tar.gz PGP: http://dri.freedesktop.org/libdrm/libdrm-2.4.71.tar.gz.sig -----BEGIN PGP SIGNATURE----- iQIcBAEBAgAGBQJX8v42AAoJEBkcm5BVIrBFi2IP/iQrbS6pl3XHh5oj15orQUuF jm+0lggaLLbyWSXj0a/R5GigjqXWhvakIbBP/G9JRDJJ/tR1U3W9al6SorW7W5Ia pnYbnhmcFFmNmD+mrFUFkyIj+gusiG9kDxvOy3KGfaYSA3WXM35xFBDUevW1LZTe FjLXYSPIfZqyiOvE7uIGoekvaEvvNYRoaoqNkDgSv2Y7ZWOtPrPIyx22ooFycPw5 lDTJwgnnppEelOBSImlX2Ntz2U5Cv+pimsGDvLURlCo2Zba7mR2XihSBMyRCgNuv JVGHS7bYdbv3DvkqHtmgwkquvkVc8Ash7GrVCstTOTFfvWS17N/LJ+QqkTQ5TAkw k/uCMxEP6ToL2crwjI5y/D1Jnx57TQHE8St9FqPEcZSB8EN9a75sdCq7whYfEB/E HOXIuMIlmS915ohxvyt3xBAAIFHNmSxeMQ9qa1MFspbIXDveboL6ERU1vUIU7sEm CaEZJuo3xNqxs7dSgiz7EjnMMeDSnukVvENP1pctT7stJpgDStyu4MbIQILS9Ifk Sohdy/VGfAKn/0olxLhsbB6FkjsL1yB3GkJmAOS4CumIGXesmoKsu8bu+NUr24k0 GxHGhdaav5z1XAMW8QOIbNOIWVx2ybNKFpwlq0BU+OBhFUSRgQUUVH4p5UcEse6C EJj5jNnDZuBzSZnUCo3d =IX18 -----END PGP SIGNATURE-----

1 0

[PATCH] drm/virtio: populate .set_busid callback
by Emil Velikov 03 Oct '16

03 Oct '16

Earlier commit was removing all the users of drm_platform_set_busid and removed the virtio hunk (which uses the PCI version of the function) by mistake. This in itself resulted in user space receiving an incorrect value for the busid, which in the case below lead to the failure to detect the (correct) device and ultimately the X server failing to start. Fixes: a325725633c ("drm: Lobotomize set_busid nonsense for !pci drivers") Cc: Daniel Vetter <daniel.vetter(a)ffwll.ch> Cc: Laszlo Ersek <lersek(a)redhat.com> Cc: stable(a)vger.kernel.org Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1366842 Reported-by: Laszlo Ersek <lersek(a)redhat.com> Signed-off-by: Emil Velikov <emil.l.velikov(a)gmail.com> --- Since I'm not 100% sure if we can get into .set_busid() as pci_dev is NULL (for the virtio-vga 'vs' virgio-gpu-pci case) I've left the local wrapper. Note: patch is against mainline (refs/tags/v4.8) but should apply for drm-next/others. --- drivers/gpu/drm/virtio/virtgpu_drm_bus.c | 10 ++++++++++ drivers/gpu/drm/virtio/virtgpu_drv.c | 1 + drivers/gpu/drm/virtio/virtgpu_drv.h | 1 + 3 files changed, 12 insertions(+) diff --git a/drivers/gpu/drm/virtio/virtgpu_drm_bus.c b/drivers/gpu/drm/virtio/virtgpu_drm_bus.c index 7f0e93f87..88a3916 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drm_bus.c +++ b/drivers/gpu/drm/virtio/virtgpu_drm_bus.c @@ -27,6 +27,16 @@ #include "virtgpu_drv.h" +int drm_virtio_set_busid(struct drm_device *dev, struct drm_master *master) +{ + struct pci_dev *pdev = dev->pdev; + + if (pdev) { + return drm_pci_set_busid(dev, master); + } + return 0; +} + static void virtio_pci_kick_out_firmware_fb(struct pci_dev *pci_dev) { struct apertures_struct *ap; diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.c b/drivers/gpu/drm/virtio/virtgpu_drv.c index c13f70c..5820b702 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.c +++ b/drivers/gpu/drm/virtio/virtgpu_drv.c @@ -117,6 +117,7 @@ static const struct file_operations virtio_gpu_driver_fops = { static struct drm_driver driver = { .driver_features = DRIVER_MODESET | DRIVER_GEM | DRIVER_PRIME | DRIVER_RENDER | DRIVER_ATOMIC, + .set_busid = drm_virtio_set_busid, .load = virtio_gpu_driver_load, .unload = virtio_gpu_driver_unload, .open = virtio_gpu_driver_open, diff --git a/drivers/gpu/drm/virtio/virtgpu_drv.h b/drivers/gpu/drm/virtio/virtgpu_drv.h index b18ef31..acf556a 100644 --- a/drivers/gpu/drm/virtio/virtgpu_drv.h +++ b/drivers/gpu/drm/virtio/virtgpu_drv.h @@ -49,6 +49,7 @@ #define DRIVER_PATCHLEVEL 1 /* virtgpu_drm_bus.c */ +int drm_virtio_set_busid(struct drm_device *dev, struct drm_master *master); int drm_virtio_init(struct drm_driver *driver, struct virtio_device *vdev); struct virtio_gpu_object { -- 2.9.3

2 4

[Bug 98004] RX470 doesn't reclock memory if a second display has been plugged in at some point
by bugzilla-daemon＠freedesktop.org 03 Oct '16

03 Oct '16

https://bugs.freedesktop.org/show_bug.cgi?id=98004 Bug ID: 98004 Summary: RX470 doesn't reclock memory if a second display has been plugged in at some point Product: DRI Version: unspecified Hardware: x86-64 (AMD64) OS: Linux (All) Status: NEW Severity: minor Priority: medium Component: DRM/AMDgpu Assignee: dri-devel(a)lists.freedesktop.org Reporter: jonathanmoerman(a)gmail.com If a second display has been plugged in the memory clock gets stuck at the highest frequency, even if the display is disabled (using xrandr) or unplugged again. How to reproduce: plug in a second display and either disable it or unplug it again, the memory will continue to be clocked at its maximal frequency (Here: 1750Mhz). Kernels used for testing: Linux 4.4 (amdgpu-pro kernel driver), 4.7, 4.8rc5 The hardware is perfectly capable of clocking back to its lowest frequency if only one display is actually enabled. (It works as expected in Windows) This issue doesn't cause any problems other than using more watts than necessary. I'm not currently using the system with the RX470 GPU, I'll upload a dmesg log in a few minutes. I'm sorry if this isn't the correct place to post this bug, I'm new here. -- You are receiving this mail because: You are the assignee for the bug.

1 4

[PATCH] drm/amdgpu: warn if dp aux is still attached on free
by Grazvydas Ignotas 03 Oct '16

03 Oct '16

If this happens (and it recently did), we free a structure while part of it is still in use, which results in non-obvious crashes. The way it's detached is not trivial (DRM core has to call the connector .destroy callback and things must be torn down in the right order), so better detect it and warn early. Signed-off-by: Grazvydas Ignotas <notasas(a)gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c | 4 +++- drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c | 1 + 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c index ff0b55a..76a7830 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_connectors.c @@ -769,8 +769,10 @@ static void amdgpu_connector_destroy(struct drm_connector *connector) { struct amdgpu_connector *amdgpu_connector = to_amdgpu_connector(connector); - if (amdgpu_connector->ddc_bus->has_aux) + if (amdgpu_connector->ddc_bus->has_aux) { drm_dp_aux_unregister(&amdgpu_connector->ddc_bus->aux); + amdgpu_connector->ddc_bus->has_aux = false; + } amdgpu_connector_free_edid(connector); kfree(amdgpu_connector->con_priv); drm_connector_unregister(connector); diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c index 34bab61..91d3673 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_i2c.c @@ -220,6 +220,7 @@ void amdgpu_i2c_destroy(struct amdgpu_i2c_chan *i2c) { if (!i2c) return; + WARN_ON(i2c->has_aux); i2c_del_adapter(&i2c->adapter); kfree(i2c); } -- 2.7.4

2 1

[PATCH v2] drm: tilcdc: add a da850-specific compatible string
by Bartosz Golaszewski 03 Oct '16

03 Oct '16

Due to some potential tweaks for the da850 LCDC (for example: the required memory bandwith settings) we need a separate compatible for the IP present on the da850 boards. Suggested-by: Sekhar Nori <nsekhar(a)ti.com> Signed-off-by: Bartosz Golaszewski <bgolaszewski(a)baylibre.com> --- v1 -> v2: - added the new compatible to the bindings documentation Documentation/devicetree/bindings/display/tilcdc/tilcdc.txt | 4 ++-- drivers/gpu/drm/tilcdc/tilcdc_drv.c | 1 + 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/Documentation/devicetree/bindings/display/tilcdc/tilcdc.txt b/Documentation/devicetree/bindings/display/tilcdc/tilcdc.txt index a83abd7..33b6e8a 100644 --- a/Documentation/devicetree/bindings/display/tilcdc/tilcdc.txt +++ b/Documentation/devicetree/bindings/display/tilcdc/tilcdc.txt @@ -1,7 +1,7 @@ Device-Tree bindings for tilcdc DRM driver Required properties: - - compatible: value should be "ti,am33xx-tilcdc". + - compatible: value should be "ti,am33xx-tilcdc" or "ti,da850-tilcdc". - interrupts: the interrupt number - reg: base address and size of the LCDC device @@ -51,7 +51,7 @@ Optional nodes: Example: fb: fb@4830e000 { - compatible = "ti,am33xx-tilcdc"; + compatible = "ti,am33xx-tilcdc", "ti,da850-tilcdc"; reg = <0x4830e000 0x1000>; interrupt-parent = <&intc>; interrupts = <36>; diff --git a/drivers/gpu/drm/tilcdc/tilcdc_drv.c b/drivers/gpu/drm/tilcdc/tilcdc_drv.c index a694977..231f2c7 100644 --- a/drivers/gpu/drm/tilcdc/tilcdc_drv.c +++ b/drivers/gpu/drm/tilcdc/tilcdc_drv.c @@ -723,6 +723,7 @@ static int tilcdc_pdev_remove(struct platform_device *pdev) static struct of_device_id tilcdc_of_match[] = { { .compatible = "ti,am33xx-tilcdc", }, + { .compatible = "ti,da850-tilcdc", }, { }, }; MODULE_DEVICE_TABLE(of, tilcdc_of_match); -- 2.9.3

2 2

[PATCH] drm/amdgpu/dce11: add missing drm_mode_config_cleanup call
by Grazvydas Ignotas 03 Oct '16

03 Oct '16

All other amdgpu/dce_v* files have this call, it's only mysteriously missing from dce_v11_0.c since the file was added and causes leaks. Fixes: aaa36a976bbb ("drm/amdgpu: Add initial VI support") Signed-off-by: Grazvydas Ignotas <notasas(a)gmail.com> --- drivers/gpu/drm/amd/amdgpu/dce_v11_0.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c index 0bed302..2696428 100644 --- a/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c +++ b/drivers/gpu/drm/amd/amdgpu/dce_v11_0.c @@ -3075,6 +3075,7 @@ static int dce_v11_0_sw_fini(void *handle) dce_v11_0_afmt_fini(adev); + drm_mode_config_cleanup(adev->ddev); adev->mode_info.mode_config_initialized = false; return 0; -- 2.7.4

2 1

[PATCH] drm/amdgpu: also track late init state
by Grazvydas Ignotas 03 Oct '16

03 Oct '16

Successful sw_init() and hw_init() states are tracked, but not late_init(). Various error paths may result in amdgpu_fini() being called before .late init is done, so late_init needs to be tracked to avoid unexpected or multiple .late_fini() calls. Signed-off-by: Grazvydas Ignotas <notasas(a)gmail.com> --- drivers/gpu/drm/amd/amdgpu/amdgpu.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 4 ++++ 2 files changed, 5 insertions(+) diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h index b9cc003..4e510092 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h @@ -1943,6 +1943,7 @@ struct amdgpu_ip_block_status { bool valid; bool sw; bool hw; + bool late_initialized; bool hang; }; diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c index 9933ab7..0c57898 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c @@ -1450,6 +1450,7 @@ static int amdgpu_late_init(struct amdgpu_device *adev) DRM_ERROR("late_init of IP block <%s> failed %d\n", adev->ip_blocks[i].funcs->name, r); return r; } + adev->ip_block_status[i].late_initialized = true; } } @@ -1495,8 +1496,11 @@ static int amdgpu_fini(struct amdgpu_device *adev) } for (i = adev->num_ip_blocks - 1; i >= 0; i--) { + if (!adev->ip_block_status[i].late_initialized) + continue; if (adev->ip_blocks[i].funcs->late_fini) adev->ip_blocks[i].funcs->late_fini((void *)adev); + adev->ip_block_status[i].late_initialized = false; } return 0; -- 2.7.4

2 1

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

dri-devel October 2016