- 17 Feb, 2017 40 commits
-
-
git://github.com/skeggsb/linuxDave Airlie authored
- Rework of the secure boot code, in preparation for GP10x secure boot. - Improvements to channel recovery - Initial power budget code - Some preparation for an upcoming MMU rework (probably 4.12) - Misc other fixes. * 'linux-4.11' of git://github.com/skeggsb/linux: (88 commits) drm/nouveau/tmr: provide backtrace when a timeout is hit drm/nouveau/pci/g92: Fix rearm drm/nouveau/drm/therm/fan: add a fallback if no fan control is specified in the vbios drm/nouveau/hwmon: expose power_max and power_crit drm/nouveau/iccsense: Parse max and crit power level drm/nouveau/bios/power_budget: Add basic power budget parsing drm/nouveau/fifo/gk104-: preempt recovery drm/nouveau/fifo/gk104-: trigger mmu fault before attempting engine recovery drm/nouveau/fifo/gk104-: ACK SCHED_ERROR before attempting CTXSW_TIMEOUT recovery drm/nouveau/fifo/gk104-: directly use new recovery code for ctxsw timeout drm/nouveau/fifo/gk104-: directly use new recovery code for mmu faults drm/nouveau/fifo/gk104-: reset all engines a killed channel is still active on drm/nouveau/fifo/gk104-: refactor recovery code drm/nouveau/fifo/gk104-: better detection of chid when parsing engine status drm/nouveau/fifo/gk104-: separate out engine status parsing drm/nouveau/fifo: add an api for initiating channel recovery drm/nouveau/top: add function to translate subdev index to mmu fault id drm/nouveau/gr/gf100-: implement chsw_load() method drm/nouveau/gr: implement chsw_load() method drm/nouveau/core: add engine method to assist in determining chsw direction ...
-
Ben Skeggs authored
Signed-off-by:Ben Skeggs <bskeggs@redhat.com>
-
Karol Herbst authored
704a6c008b7942bb7f30bb43d2a6bcad7f543662 broke pci msi rearm for g92 GPUs. g92 needs the nv46_pci_msi_rearm, where g94+ gpus used nv40_pci_msi_rearm. Reported-by:
Andrew Randrianasulu <randrianasulu@gmail.com> Signed-off-by:
Karol Herbst <karolherbst@gmail.com> Signed-off-by:
-
Martin Peres authored
This seems to be absolutely necessary for a lot of NV40. Reported-by: gsgf on IRC/freenode Signed-off-by:
Martin Peres <martin.peres@free.fr> Signed-off-by:
-
Karol Herbst authored
Signed-off-by:
-
Karol Herbst authored
Signed-off-by:
-
Karol Herbst authored
v2: Set entry to 0xff if not found Add cap entry for ver 0x30 tables Rework to fix memory leak v3: More error checks Simplify check for invalid entries v4: disable for ver 0x10 for now move assignments after the second last return Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Greatly improves the chances of recovering the GPU from a CTXSW_TIMEOUT. Tested with piglit's arb_shader_image_load_store-atomicity, which causes GR to hang in such a way that recovery failed (CTXSW_TIMEOUT continually re-triggers). Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
This will serve as a basis for implementing some improvements to how we recover the GPU from channel errors. Signed-off-by: -
Ben Skeggs authored
The previous commit simply changes the interface, but should result in the same behaviour as previously. This commit has been split out from it as it can result in a different channel being selected. Signed-off-by: -
Ben Skeggs authored
We'll be wanting to reuse this logic in more places. Signed-off-by: -
Ben Skeggs authored
This will be used by callers outside of fifo interrupt handlers. Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
FIFO gives us load/save/switch status, and we need to be able to determine which direction a "switch" is failing during channel recovery. In order to do this, we apparently need to query the engine itself. Signed-off-by: -
Ben Skeggs authored
This prevents *very* long waits while attempting to destroy channels after a fault has occurred. Signed-off-by: -
Ben Skeggs authored
These will be used to improve error recovery behaviour. Signed-off-by: -
Ben Skeggs authored
There are instances (such as non-recoverable GPU page faults) where NVKM decides that a channel's context is no longer viable, and will be removed from the runlist. This commit notifies the owner of the channel when this happens, so it has the opportunity to take some kind of recovery action instead of hanging. Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Martin Peres authored
If the led class registration fails, we free drm->led but do not reset it to NULL, which means that the suspend/resume/fini function will act as if everything went well in init() and will likely crash the kernel. This patch adds the missing drm->led = NULL. Reported-by:
Emmanuel Pescosta <emmanuelpescosta099@gmail.com> Signed-off-by:
-
Karol Herbst authored
Tested on a G92, seems to work. Confirmed by 8 mmiotraces. Signed-off-by:
-
Karol Herbst authored
Signed-off-by:
-
Ben Skeggs authored
Reported-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
Ilia Mirkin authored
The issue was recorded in fd.o bug 27501, not 25701. Signed-off-by:
Ilia Mirkin <imirkin@alum.mit.edu> Signed-off-by:
-
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
This will allow the DRM to share memory objects between clients later down the track. For the moment, the only immediate benefit is less logic required to handle suspend/resume. Signed-off-by: -
Ben Skeggs authored
These were a little insane. Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by: -
Ben Skeggs authored
Signed-off-by:
-
