summaryrefslogtreecommitdiffstats
path: root/sys/dev/pci/drm/amd/amdgpu/amdgpu_device.c (follow)
Commit message (Collapse)AuthorAgeFilesLines
* The conversion of the IO BAR access did not completely mirror thepatrick2020-11-191-6/+22
| | | | | | | | | | | | behaviour of Linux' implementation: arm64's bus space operations have no barriers, so while Linux' iowrite32/ioread32 explicitly contain barriers, using bus space read/write is not enough on arm64. Add read barriers after a read to make sure that all reads have completed before the following reads. Add write barriers before a write to make sure all previous writes have completed before that write. This fixes panics on the HoneyComb LX2K with amdgpu(4). ok kettenis@
* Switch from bus_space_read/write to linux io.h interfaces for alljsg2020-10-281-28/+14
| | | | | | pci "memory space" io. Further reduces the diff to linux. ok kettenis@
* don't print pci ids when amdgpu attachesjsg2020-10-151-1/+22
| | | | | Print asic name, rev and number of compute units. Suggested by and ok kettenis@
* drm/amdgpu: move gpu_info parsing after common early initjsg2020-10-151-4/+4
| | | | | | | | From Alex Deucher 6e29c227a4976460ec6d4cc70b998e3a8c30c873 in mainline linux needed to load the correct gpu info firmware on raven2 (Ryzen 3 3200U/Ryzen 3 3250U/Athlon 300U/Athlon 3000G)
* drm/amdgpu: move discovery gfx config fetchingjsg2020-10-151-4/+3
| | | | | From Alex Deucher 6ba57b7a8f94fcf3df6883db362642a0075d185b in mainline linux
* change some drm locks from IPL_TTY to IPL_NONEjsg2020-07-131-2/+2
| | | | | | | | In drm linux spinlocks are mapped to mutex(9). Locks without calls to spin_lock_irqsave(), spin_lock_irq() and the like (which block interrupts) can be changed to IPL_NONE. ok kettenis@
* drm/amdgpu: disable ras query and iject during gpu resetjsg2020-07-091-0/+3
| | | | | | From John Clements 50a8cfb0e71644f2882dbfaf93150184094245bd in linux 5.7.y/5.7.8 61380faa4b4cc577df8a7ff5db5859bac6b351f7 in mainline linux
* Silence amdgpu_device_resize_fb_bar; supporting resizing of PCI BARs iskettenis2020-06-121-24/+2
| | | | | | tricky and using the pre-programmed BAR size should always work. Using bigger BARs might improve performance but it we probably need other changes (such as making the driver mpsafe) to benefit from that.
* update drm to linux 5.7jsg2020-06-081-568/+1516
| | | | | | | | | | adds kernel support for amdgpu: vega20, raven2, renoir, navi10, navi14 inteldrm: icelake, tigerlake Thanks to the OpenBSD Foundation for sponsoring this work, kettenis@ for helping, patrick@ for helping adapt rockchip drm and many developers for testing.
* drm/amdgpu: simplify Raven, Raven2, and Picasso handlingjsg2019-07-041-8/+5
| | | | | From Alex Deucher 741deade2a704a434bd5939118c43d38e9ddac25 in mainline linux
* drm/amd/display/dm: add picasso supportjsg2019-07-041-0/+1
| | | | | From Likun Gao b22ab73314c0dfa91ac1948812c7e8050240c563 in mainline linux
* drm/amdgpu: add soc15 support for picassojsg2019-07-041-1/+6
| | | | | From Likun Gao ad5a67a7ea87e625721a5d0c4e9f12100372f1f6 in mainline linux
* drm/amdgpu: add picasso to asic_type enumjsg2019-07-041-0/+1
| | | | | From Likun Gao be9699e3923000ea32c2f4522e1e4de333d21d47 in mainline linux
* Match radeon_device.c rev 1.2 and return early in amdgpu_device_suspend()jsg2019-05-241-0/+2
| | | | | if "shutdown" is set and the mountroot hook has not run. Otherwise unhibernate causes a uvm fault due to an uninitialised lock.
* match radeondrm and switch DRM_INFO message with device id/asic name tojsg2019-05-231-1/+1
| | | | a printf so it shows without DRMDEBUG
* add amdgpu from linux 4.19.44 for recent AMD Radeon partsjsg2019-05-211-0/+3605
committing now so this can be worked on in tree Thanks to the OpenBSD Foundation for sponsoring this work and kettenis@ for helping.
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.