aboutsummaryrefslogtreecommitdiff
path: root/drivers/gpu
AgeCommit message (Collapse)AuthorFilesLines
2023-08-31drm/amdgpu: add selftest framework for UMSCHLang Yu1-1/+404
Prepare for VPE and VCN queue submission test. v2: rebase on drm_exec (Alex) Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: enable UMSCH scheduling for VPELang Yu1-0/+4
Add VPE into UMSCH hw resourses, set vmid mask to 0xf00, set hqd mask to 0xfe, then UMSCH can schedule VPE queues. Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add initial support for UMSCHLang Yu6-0/+996
Add basic data structure, dummy ring functions and ip functions for UMSCH. Implement sw_init(ring_init and init_microcodede) and hw_init(load_microcode), UMSCH can boot up now. Implement hw_init(ring_start) and hw_fini(ring_stop), UMSCH is ready for command submission now. Implement set_hw_resources and add/remove_queue, UMSCH is ready for scheduling now. Aggregated doorbell is used to notify UMSCH FW that there is unmapped queue with corresponding priority level (e.g., AGDB[0] for Real time band, etc.) is updating its job. v2: squash together initial patches to avoid breaking the build (Alex) Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UMSCH 4.0 api definitionLang Yu1-0/+434
Add api definition for UMSCH 4.0. v2: adjust coding style. Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UMSCH firmware header definitionLang Yu1-0/+17
Add firmware header definition for UMSCH. Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UMSCH RING TYPE definitionLang Yu1-1/+2
Add RING TYPE definition for Multi Mdeia User Mode Scheduler. Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amd/pm: add fan acoustic limit OD setting support for SMU13Evan Quan7-2/+170
Add SMU13 fan acoustic limit OD setting support. Signed-off-by: Evan Quan <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amd/pm: add fan temperature/pwm curve OD setting support for SMU13Evan Quan7-8/+417
Add SMU13 fan temperature/pwm curve OD setting support. Signed-off-by: Evan Quan <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu/jpeg: initialize number of jpeg ringSaleemkhan Jamadar1-0/+1
Initialize number of jpeg ring for vcn 4.0.5. Signed-off-by: Saleemkhan Jamadar <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: fix amdgpu_cs_p1_user_fenceChristian König1-14/+4
The offset is just 32bits here so this can potentially overflow if somebody specifies a large value. Instead reduce the size to calculate the last possible offset. The error handling path incorrectly drops the reference to the user fence BO resulting in potential reference count underflow. Signed-off-by: Christian König <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: revise the device initialization sequencesEvan Quan1-16/+21
By placing the sysfs interfaces creation after `.late_int`. Since some operations performed during `.late_init` may affect how the sysfs interfaces should be created. Signed-off-by: Evan Quan <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amd/pm: introduce a new set of OD interfacesEvan Quan3-2/+266
There will be multiple interfaces(sysfs files) exposed with each representing a single OD functionality. And all those interface will be arranged in a tree liked hierarchy with the top dir as "gpu_od". Meanwhile all functionalities for the same component will be arranged under the same directory. Signed-off-by: Evan Quan <[email protected]> Reviewed-by: Alex Deucher <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UMSCH IP BLOCK TYPE definitionLang Yu1-0/+2
Add IP BLOCK TYPE definition for Multimedia User Mode Scheduler which is a hardware scheduler for VCN and VPE workload. Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31amdgpu/pm: Optimize emit_clock_levels for arcturus - part 3Darren Powell1-53/+24
split switch statement into two and consolidate the common code for printing most of the types of clock speeds Signed-off-by: Darren Powell <[email protected]> Reviewed-by: Lijo Lazar <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UMSCH 4.0 register headersLang Yu2-0/+1304
Add headers for UMSCH 4.0. v2: updates (Alex) Signed-off-by: Lang Yu <[email protected]> Reviewed-by: Leo Liu <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: enable PG flags for VCNSaleemkhan Jamadar1-0/+2
Enable PG flags for VCN and Jpeg on IP 11_5_0 Signed-off-by: Saleemkhan Jamadar <[email protected]> Reviewed-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu/discovery: add VCN 4.0.5 SupportSaleemkhan Jamadar1-0/+6
Enable VCN 4.0.5 on gc 11_5_0. Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu/soc21: Add video cap query support for VCN_4_0_5Saleemkhan Jamadar1-0/+1
Added the video capability query support for VCN version 4_0_5 Signed-off-by: Saleemkhan Jamadar <[email protected]> Reviewed-by: Veerabadhran Gopalakrishnan <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu:enable CG and PG flags for VCNSaleemkhan Jamadar1-2/+4
Enable CG and PG flags for VCN on IP 11_5_0 Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VCN_4_0_5 firmware supportSaleemkhan Jamadar1-0/+2
Add VCN_4_0_5 firmware support Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu/jpeg: add jpeg support for VCN4_0_5Saleemkhan Jamadar3-1/+658
Add jpeg support for VCN4_0_5 v2 - update license year (Leo Liu) Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VCN4 ip block supportSaleemkhan Jamadar3-0/+1815
Add VCN 4.0.5 initialization and decoder/encoder ring functions. v2 - update license year (Leo Liu) Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add vcn 4_0_5 header filesSaleemkhan Jamadar2-0/+10411
Add VCN 4.0.5 registers v2 - Add license header (Alexander Deucher) v3 - updates (Alex) Signed-off-by: Saleemkhan Jamadar <[email protected]> Acked-by: Leo Liu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: fix VPE front door loading issueLang Yu6-0/+63
Implement proper front door loading for vpe 6.1. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE FW version query supportLang Yu1-0/+12
Add support to query VPE FW version. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: enable VPE for VPE 6.1.0Lang Yu1-0/+19
Enable Video Processing Engine on SoCs that contain VPE 6.1.0. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add user space CS support for VPELang Yu3-0/+9
Enable command submission to VPE from user space. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add PSP loading support for VPELang Yu2-0/+28
Add PSP loading support for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE 6.1.0 supportLang Yu6-0/+1019
Add skeleton driver code. (Ray) Add initial support for Video Processing Engine. (Lang) Signed-off-by: Huang Rui <[email protected]> Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add nbio 7.11 callback for VPELang Yu1-0/+24
Add nbio callback to configure doorbell settings. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add nbio callback for VPELang Yu1-0/+2
Add nbio callback to configure doorbell settings. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add PSP FW TYPE for VPELang Yu2-0/+8
Add PSP FW TYPE for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add UCODE ID for VPELang Yu1-0/+2
Add UCODE ID for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add support for VPE firmware name decodingLang Yu1-0/+3
Add decoding VPE firmware name support. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add doorbell index for VPELang Yu2-2/+6
Add doorbell index for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add irq src id definitions for VPELang Yu1-0/+40
The irq src id is used to route interrupts to the corresponding handlers. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add IH CLIENT ID for VPELang Yu1-0/+2
Add Interrupt Handler Client ID for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add HWID for VPELang Yu2-0/+3
Add HWID for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE firmware interfaceLang Yu1-0/+217
Add initial firmware interface. (Ray) Add more opcodes and rename to vpe_v6_1. (Lang) v2: Update copyright date (Alex) Signed-off-by: Huang Rui <[email protected]> Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE firmware header definitionLang Yu1-0/+13
Add firmware header definition for Video Processing Engine. Signed-off-by: Lang Yu <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE HW IP BLOCK definitionHuang Rui1-0/+1
Add HW IP BLOCK for Video Processing Engine. Signed-off-by: Huang Rui <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE IP BLOCK definitionHuang Rui1-0/+1
Add IP BLOCK for Video Processing Engine. Signed-off-by: Huang Rui <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/amdgpu: add VPE RING TYPE definitionHuang Rui1-0/+2
Add RING TYPE for Video Processing Engine. Signed-off-by: Huang Rui <[email protected]> Signed-off-by: Alex Deucher <[email protected]>
2023-08-31drm/i915/gvt: Drop final dependencies on KVM internal detailsSean Christopherson2-2/+3
Open code gpa_to_gfn() in kvmgt_page_track_write() and drop KVMGT's dependency on kvm_host.h, i.e. include only kvm_page_track.h. KVMGT assumes "gfn == gpa >> PAGE_SHIFT" all over the place, including a few lines below in the same function with the same gpa, i.e. there's no reason to use KVM's helper for this one case. No functional change intended. Reviewed-by: Yan Zhao <[email protected]> Tested-by: Yongwei Ma <[email protected]> Reviewed-by: Zhi Wang <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31KVM: x86/mmu: Handle KVM bookkeeping in page-track APIs, not callersSean Christopherson1-10/+7
Get/put references to KVM when a page-track notifier is (un)registered instead of relying on the caller to do so. Forcing the caller to do the bookkeeping is unnecessary and adds one more thing for users to get wrong, e.g. see commit 9ed1fdee9ee3 ("drm/i915/gvt: Get reference to KVM iff attachment to VM is successful"). Reviewed-by: Yan Zhao <[email protected]> Tested-by: Yongwei Ma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31KVM: x86/mmu: Drop @slot param from exported/external page-track APIsSean Christopherson1-29/+8
Refactor KVM's exported/external page-track, a.k.a. write-track, APIs to take only the gfn and do the required memslot lookup in KVM proper. Forcing users of the APIs to get the memslot unnecessarily bleeds KVM internals into KVMGT and complicates usage of the APIs. No functional change intended. Reviewed-by: Yan Zhao <[email protected]> Tested-by: Yongwei Ma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31KVM: x86/mmu: Rename page-track APIs to reflect the new realitySean Christopherson1-2/+2
Rename the page-track APIs to capture that they're all about tracking writes, now that the facade of supporting multiple modes is gone. Opportunstically replace "slot" with "gfn" in anticipation of removing the @slot param from the external APIs. No functional change intended. Tested-by: Yongwei Ma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31KVM: x86/mmu: Drop infrastructure for multiple page-track modesSean Christopherson1-2/+2
Drop "support" for multiple page-track modes, as there is no evidence that array-based and refcounted metadata is the optimal solution for other modes, nor is there any evidence that other use cases, e.g. for access-tracking, will be a good fit for the page-track machinery in general. E.g. one potential use case of access-tracking would be to prevent guest access to poisoned memory (from the guest's perspective). In that case, the number of poisoned pages is likely to be a very small percentage of the guest memory, and there is no need to reference count the number of access-tracking users, i.e. expanding gfn_track[] for a new mode would be grossly inefficient. And for poisoned memory, host userspace would also likely want to trap accesses, e.g. to inject #MC into the guest, and that isn't currently supported by the page-track framework. A better alternative for that poisoned page use case is likely a variation of the proposed per-gfn attributes overlay (linked), which would allow efficiently tracking the sparse set of poisoned pages, and by default would exit to userspace on access. Link: https://lore.kernel.org/all/[email protected] Cc: Ben Gardon <[email protected]> Tested-by: Yongwei Ma <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31drm/i915/gvt: switch from ->track_flush_slot() to ->track_remove_region()Yan Zhao1-12/+9
Switch from the poorly named and flawed ->track_flush_slot() to the newly introduced ->track_remove_region(). From KVMGT's perspective, the two hooks are functionally equivalent, the only difference being that ->track_remove_region() is called only when KVM is 100% certain the memory region will be removed, i.e. is invoked slightly later in KVM's memslot modification flow. Cc: Zhenyu Wang <[email protected]> Suggested-by: Sean Christopherson <[email protected]> Signed-off-by: Yan Zhao <[email protected]> [sean: handle name change, massage changelog, rebase] Tested-by: Yan Zhao <[email protected]> Tested-by: Yongwei Ma <[email protected]> Reviewed-by: Zhi Wang <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
2023-08-31drm/i915/gvt: Don't bother removing write-protection on to-be-deleted slotSean Christopherson1-7/+1
When handling a slot "flush", don't call back into KVM to drop write protection for gfns in the slot. Now that KVM rejects attempts to move memory slots while KVMGT is attached, the only time a slot is "flushed" is when it's being removed, i.e. the memslot and all its write-tracking metadata is about to be deleted. Reviewed-by: Yan Zhao <[email protected]> Tested-by: Yongwei Ma <[email protected]> Reviewed-by: Zhi Wang <[email protected]> Link: https://lore.kernel.org/r/[email protected] Signed-off-by: Sean Christopherson <[email protected]> Signed-off-by: Paolo Bonzini <[email protected]>
Powered by cgit, Gruvboxed by Blaster4385.