aboutsummaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd
AgeCommit message (Collapse)AuthorFilesLines
2017-12-12drm/amdgpu: no need to evict VRAM in device_finiMonk Liu1-2/+1
this VRAM evict is not needed and also cost 2seconds to finish because the IRQ is software side disabled before it. Signed-off-by: Monk Liu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: allow get_vm_pde to change flags as wellChristian König11-38/+54
And also provide the level for which we need a PDE. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: batch PDE updates againChristian König1-109/+94
Now instead of one submission for each PDE batch them together over all PDs who need an update. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: remove keeping the addr of the VM PDsChristian König2-10/+5
No more double house keeping. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: remove last_entry_used from the VM codeChristian König2-24/+29
Not needed any more. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: avoid the modulo in amdgpu_vm_get_entryChristian König1-3/+3
We can do this with a simple mask as well. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: use polling mem to set SDMA3 wptr for VFPixel Ding2-8/+20
On Tonga VF, there're 2 sources updating wptr registers for sdma3: 1) polling mem and 2) doorbell. When doorbell and polling mem are both enabled on sdma3, there will be collision hit in occasion between those two sources when ucode and h/w are doing the updating on wptr register in parallel. Issue doesn't happen on CP GFX/Compute since CP drops all doorbell writes when VF is inactive. So enable polling mem and don't use doorbell for SDMA3. Signed-off-by: Pixel Ding <[email protected]> Reviewed-by: Monk Liu <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: update one PDE at a time v2Christian König1-46/+34
Horrible inefficient, but avoids problems when the root PD size becomes to big. v2: remove incr as well. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: stop joining PDEsChristian König1-34/+7
That doesn't hit any more most of the time anyway. Signed-off-by: Christian König <[email protected]> Reviewed-by: Chunming Zhou <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: add amdgpu_evict_vram debugfs fileChristian König1-1/+12
Torture test for MM and VM support, can be used to evict all VRAM while the system is under load. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-12drm/amdgpu: cleanup debugfs handling a bitChristian König3-43/+11
Remove the superflous .debugfs_init callback and register all files in amdgpu_device.c in just one function. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-08drm/amdgpu: Add support for reporting VRAM usageKent Russell5-2/+15
Add functions to report the vram_usage from the amdgpu_device Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Ignore ACPI CRAT for non-APU systemsHarish Kasiviswanathan1-2/+22
Some AMD motherboards without an APU have a broken CRAT table which causes KFD initialization failures or incorrect information about NUMA nodes, CPU cores or system memory. Ignore CRAT tables without GPUs and rely on KFD's code to create a CRAT table for the CPU. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Module option to disable CRAT tableFelix Kuehling3-0/+16
Some systems have broken CRAT tables. Add a module option to ignore a CRAT table. Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Add AQL Queue Memory flag on topologyBen Goz2-0/+5
This is needed for enabling a user-mode workaround for an AQL queue wrapping HW bug on Tonga. Signed-off-by: Ben Goz <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Fixup incorrect info in the CZ CRAT tablePhilip Cox1-3/+6
* Wrong value for max_waves_per_simd * Missing ATC capability bit Signed-off-by: Philip Cox <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Add perf counters to topologyAmber Lin2-2/+127
For hardware blocks whose performance counters are accessed via MMIO registers, KFD provides the support for those privileged blocks. IOMMU is one of those privileged blocks. Most performance counter properties required by Thunk are available at /sys/bus/event_source/devices/amd_iommu. This patch adds properties to topology in KFD sysfs for information not available in /sys/bus/event_source/devices/amd_iommu. They are shown at /sys/devices/virtual/kfd/kfd/topology/nodes/0/perf/iommu/ formatted as /sys/devices/virtual/kfd/kfd/topology/nodes/0/perf/<block>/<property>, i.e. /sys/devices/virtual/kfd/kfd/topology/nodes/0/perf/iommu/max_concurrent. For dGPUs, who don't have IOMMU, nothing appears under /sys/devices/virtual/kfd/kfd/topology/nodes/0/perf. Signed-off-by: Amber Lin <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Add topology support for dGPUsHarish Kasiviswanathan5-51/+748
Generate and parse VCRAT tables for dGPUs in kfd_topology_add_device. Some information that isn't available in the CRAT table is patched into the topology after parsing. HSA_CAP_DOORBELL_TYPE_1_0 is dependent on the ASIC feature CP_HQD_PQ_CONTROL.SLOT_BASED_WPTR, which was not introduced in VI until Carrizo. Report HSA_CAP_DOORBELL_TYPE_PRE_1_0 on Tonga ASICs. v2: Added #include <linux/pci.h> to kfd_crat.c to make it compile Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Ben Goz <[email protected]> Signed-off-by: Amber Lin <[email protected]> Signed-off-by: Jay Cornwall <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Add topology support for CPUsFelix Kuehling5-35/+489
Currently, the KFD topology information is generated by parsing the CRAT (ACPI) table. However, at present CRAT table is available only for AMD APUs. To support CPUs on systems without a CRAT table, the KFD driver will create a Virtual CRAT (VCRAT) table and then the existing code will parse that table to generate topology. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Fix sibling_map[] sizeHarish Kasiviswanathan2-10/+14
Change kfd_cache_properties.sibling_map[256] to kfd_cache_properties.sibling_map[32]. Since, CRAT uses bitmap for sibling_map, it is more efficient to use bitmap in the kfd structure also. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Simplify counting of memory banksFelix Kuehling3-15/+3
Only count memory banks in one place. Ignore redundant num_banks entry in crat_subtype_computeunit. Signed-off-by: Felix Kuehling <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Turn verbose topology messages into pr_debugFelix Kuehling2-9/+10
Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: sync IOLINK defines to thunk specHarish Kasiviswanathan1-5/+16
Current thunk spec v1.07 dated Feb 1, 2016 v2: fix indentation Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Support enumerating non-GPU devicesHarish Kasiviswanathan4-10/+19
Modify kfd_topology_enum_kfd_devices(..) function to support non-GPU nodes. The function returned NULL when it encountered non-GPU (say CPU) nodes. This caused kfd_ioctl_create_event and kfd_init_apertures to fail for Intel + Tonga. kfd_topology_enum_kfd_devices will now parse all the nodes and return valid kfd_dev for nodes with GPU. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Decouple CRAT parsing from device list updateHarish Kasiviswanathan4-79/+132
Currently, CRAT parsing is intertwined with topology_device_list and hence repeated calls to kfd_parse_crat_table() will fail. Decouple kfd_parse_crat_table() and topology_device_list. kfd_parse_crat_table() will parse CRAT and add topology devices to a temporary list temp_topology_device_list and then kfd_topology_update_device_list will move contents from temporary list to master list. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Reorganize CRAT fetching from ACPIHarish Kasiviswanathan3-30/+54
Reorganize and rename kfd_topology_get_crat_acpi function. In this way acpi_get_table(..) needs to be called only once. This will also aid in dGPU topology implementation. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Group up CRAT related functionsFelix Kuehling5-332/+362
Take CRAT related functions out of kfd_topology.c and place them in kfd_crat.c. This is the initial step of supporting more CRAT features, i.e. creating virtual CRAT table for KFD devices without CRAT. v2: Minor cleanup that was missed previously because code moved around Signed-off-by: Amber Lin <[email protected]> Signed-off-by: Kent Russell <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Fix memory leaks in kfd topologyYong Zhao1-0/+10
Kobject created using kobject_create_and_add() can be freed using kobject_put() when there is no referenece any more. However, kobject memory allocated with kzalloc() has to set up a release callback in order to free it when the counter decreases to 0. Otherwise it causes memory leak. Signed-off-by: Yong Zhao <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Topology: Fix location_idHarish Kasiviswanathan1-2/+2
Fix location_id format to match Thunk specification. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Update number of compute unit from KGDFlora Cui1-2/+5
Overwrite the active simd_count from KGD at driver loading time. This is based on assumption that register GC_USER_SHADER_ARRAY_CONFIG won’t get changed. V2: remove the incorrect simd_count reported at loading module. Signed-off-by: Flora Cui <[email protected]> Reviewed by: Yair Shachar< [email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amd: Remove get_vmem_size from KGD-KFD interfaceHarish Kasiviswanathan1-3/+0
Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Remove deprecated get_vmem_sizeHarish Kasiviswanathan4-13/+0
Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdkfd: Stop using get_vmem_size KGD-KFD interfaceHarish Kasiviswanathan1-1/+5
get_vmem_size() is deprecated. Instead use get_local_mem_info(). Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Ben Goz <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdgpu: Implement get_local_mem_infoHarish Kasiviswanathan4-0/+34
Implement new kgd-kfd interface function get_local_mem_info. Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Ben Goz <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Acked-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amd: Add get_local_mem_info to KGD-KFD interfaceHarish Kasiviswanathan1-0/+12
Add get_local_mem_info which provides more information about local memory than get_vmem_size: - public and private framebuffer size - memory clock Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Ben Goz <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amdgpu: add amdgpu interface to query cu infoFlora Cui7-0/+39
Signed-off-by: Flora Cui <[email protected]> Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/amd: add new interface to query cu infoFlora Cui1-0/+19
Signed-off-by: Flora Cui <[email protected]> Signed-off-by: Harish Kasiviswanathan <[email protected]> Signed-off-by: Felix Kuehling <[email protected]> Reviewed-by: Oded Gabbay <[email protected]> Signed-off-by: Oded Gabbay <[email protected]>
2017-12-08drm/admgpu: Reduce the usage of soc15ip.hShaoyun Liu16-17/+0
Remove the header where it's not used. Acked-by: Christian Konig <[email protected]> Signed-off-by: Shaoyun Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-08drm/amdgpu: Change SOC15_REG_OFFSET to use dynamic register offsetShaoyun Liu8-18/+46
Acked-by: Christian Konig <[email protected]> Signed-off-by: Shaoyun Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-08drm/amdgpu: Avoid use SOC15_REG_OFFSET in static const arrayShaoyun Liu13-283/+354
Handle dynamic offsets correctly in static arrays. Acked-by: Christian Konig <[email protected]> Signed-off-by: Shaoyun Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-08drm/amdgpu: Use dynamic IP offset for register access on SOC15Shaoyun Liu2-95/+65
Update the register access macros and functions to take into account the new dynamic IP base offsets. Acked-by: Christian Konig <[email protected]> Signed-off-by: Shaoyun Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-08drm/amdgpu: Dynamic initialize IP base offsetShaoyun Liu5-1/+96
The base offsets of the IP blocks may change across asics even though the relative register offsets are the same for an IP. Handle this dynamically. Acked-by: Christian Konig <[email protected]> Signed-off-by: Shaoyun Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-07drm/sched: move fence slab handling to module init/exitLucas Stach1-8/+0
This is the only part of the scheduler which must not be called from different drivers. Move it to module init/exit so it is done a single time when loading the scheduler. Reviewed-by: Chunming Zhou <[email protected]> Tested-by: Dieter Nützel <[email protected]> Acked-by: Alex Deucher <[email protected]> Signed-off-by: Lucas Stach <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-07drm: move amd_gpu_scheduler into common locationLucas Stach31-1427/+119
This moves and renames the AMDGPU scheduler to a common location in DRM in order to facilitate re-use by other drivers. This is mostly a straight forward rename with no code changes. One notable exception is the function to_drm_sched_fence(), which is no longer a inline header function to avoid the need to export the drm_sched_fence_ops_scheduled and drm_sched_fence_ops_finished structures. Reviewed-by: Chunming Zhou <[email protected]> Tested-by: Dieter Nützel <[email protected]> Acked-by: Alex Deucher <[email protected]> Signed-off-by: Lucas Stach <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-07drm/amdgpu: add license to files where it was missingAlex Deucher15-8/+254
These files were missing it before. Acked-by: Harry Wentland <[email protected]> Acked-by: Felix Kuehling <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-07drm/amdgpu: add license to MakefilesAlex Deucher26-6/+547
Was missing license text. Acked-by: Harry Wentland <[email protected]> Acked-by: Felix Kuehling <[email protected]> Acked-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-07drm/amdgpu: Fix amdgpu_sync_add_later to preserve explicit flag.Andrey Grodzovsky1-3/+6
Signed-off-by: Andrey Grodzovsky <[email protected]> Reviewed-by: Christian König <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-06drm/amdgpu: Use drm_fb_helper_lastclose() and _poll_changed()Noralf Trønnes5-42/+3
This driver can use drm_fb_helper_lastclose() in its .lastclose function. It can also use drm_fb_helper_output_poll_changed() as its .output_poll_changed callback. Remove the unused driver implementations. Cc: Alex Deucher <[email protected]> Cc: "Christian König" <[email protected]> Signed-off-by: Noralf Trønnes <[email protected]> Acked-by: Daniel Vetter <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-06drm/amd/display: Use drm_fb_helper_poll_changed()Noralf Trønnes1-1/+1
This driver can use drm_fb_helper_output_poll_changed() as its .output_poll_changed callback. Cc: Alex Deucher <[email protected]> Cc: "Christian König" <[email protected]> Signed-off-by: Noralf Trønnes <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2017-12-06drm/amdgpu: fix amdgpu_sync_resv v2Christian König1-5/+2
Fixes a bug introduced by AMDGPU_GEM_CREATE_EXPLICIT_SYNC. We still need to wait for pipelined moves in the shared fences list. v2: fix typo Signed-off-by: Christian König <[email protected]> Reviewed-by: Andres Rodriguez <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.