[pull] amdgpu, amdkfd drm-next-5.3
Hi Dave, Daniel, Fixes for 5.3, mostly for Navi. The following changes since commit 7f963d9f69bf28d639013630da30d7a4c95edd5d: drm/amdgpu/navi10: add uclk activity sensor (2019-07-09 17:43:36 -0500) are available in the Git repository at: git://people.freedesktop.org/~agd5f/linux tags/drm-next-5.3-2019-07-18 for you to fetch changes up to 41a5a2a8531f95d18bb4efddea581ccb469e8ee5: drm/amd/display: init res_pool dccg_ref, dchub_ref with xtalin_freq (2019-07-18 14:12:08 -0500) drm-next-5.3-2019-07-18: amdgpu: - Navi DC fix for secondary adapters - Fix Navi flickering with high res panels - Navi SMU fixes - Vega20 SMU fixes - Fixes for audio hotplug on HG systems - Fix for potential integer overflows on large buffer migrations - debugfs fixes for umr - Various other small fixes amdkfd: - Apply noretry setting consistently - Fix hang in eviction - Properly clean up GWS on uninit UAPI: - clarify a comment on ctx priority Alex Deucher (4): drm/amdgpu/psp: add a mutex to protect access to the psp ring drm/amdgpu: enable IP discovery by default on navi drm/amdgpu: drop dead header drm/amdgpu/pm: remove check for pp funcs in freq sysfs handlers Arnd Bergmann (3): drm/amd/display: Support clang option for stack alignment drm/amd/display: return 'NULL' instead of 'false' from dcn20_acquire_idle_pipe_for_layer drm/amd/amdgpu: hide #warning for missing DC config Emil Velikov (1): drm/amdgpu: extend AMDGPU_CTX_PRIORITY_NORMAL comment Eric Huang (1): drm/amdkfd: fix cp hang in eviction Evan Quan (7): drm/amd/powerplay: increase the SMU msg response waiting time drm/amd/powerplay: fix memory allocation failure check V2 drm/amd/powerplay: avoid access before allocation drm/amd/powerplay: fix deadlock around smu_handle_task V2 drm/amd/powerplay: correct smu_update_table usage drm/amd/powerplay: maintain SMU FW backward compatibility drm/amd/powerplay: update vega20 driver if to fit latest SMU firmware Felix Kuehling (4): drm/amdgpu: Fix potential integer overflows drm/amdkfd: Consistently apply noretry setting drm/amdgpu: Fix unaligned memory copies drm/amdgpu: Fix silent amdgpu_bo_move failures Fuqian Huang (1): drm/amdgpu: remove memset after kzalloc Hawking Zhang (3): drm/amdgpu: switch to macro for psp bootloader command drm/amdgpu: support key database loading for navi10 drm/amdgpu: check kdb_bin_size to exclude kdb loading sequence Joseph Greathouse (1): drm/amdkfd: Remove GWS from process during uninit Kenneth Feng (2): drm/amd/powerplay: bug fix for sysfs drm/amd/powerplay: enable fw ctf,apcc dfll and gfx ss Kent Russell (1): drm/amdgpu: Fix Vega20 Perf counter for pcie_bw Kevin Wang (7): drm/amd/powerplay: fix smu clock type change miss error drm/amd/powerplay: add pstate mclk(uclk) support for navi10 drm/amd/powerplay: add socclk profile dpm support. drm/amd/powerplay: add standard profile dpm support for smu drm/amd/powerplay: avoid double check feature enabled drm/amd/powerplay: fix save dpm level error for smu drm/amd/powerplay: add helper of smu_clk_dpm_is_enabled for smu Nathan Chancellor (1): drm/amd/powerplay: Use proper enums in vega20_print_clk_levels Nicholas Kazlauskas (3): drm/amd/display: Expose audio inst from DC to DM drm/amd/display: Add drm_audio_component support to amdgpu_dm drm/amd/display: Force uclk to max for every state Nicolai Hähnle (1): drm/amdgpu/gfx10: set SH_MEM_CONFIG.INITIAL_INST_PREFETCH Paul Menzel (1): drm/amdgpu: Print out voltage in DM_PPLIB Tom St Denis (3): drm/amd/amdgpu: Add VMID to SRBM debugfs bank selection drm/amd/amdgpu: Add missing select_me_pipe_q() for gfx10 drm/amd/amdgpu: Fix offset for vmid selection in debugfs interface Wang Xiayang (1): drm/amdgpu: replace simple_strtol() by kstrtou32() hersen wu (1): drm/amd/display: init res_pool dccg_ref, dchub_ref with xtalin_freq tiancyin (1): drm/amdgpu/discovery: fix DCE_HWIP mapping error in hw_id_map array drivers/gpu/drm/amd/amdgpu/amdgpu.h| 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c| 9 +- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c| 19 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.h| 4 +- drivers/gpu/drm/amd/amdgpu/amdgpu_pm.c | 13 -- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c| 25 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_psp.h| 12 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c| 51 - drivers/gpu/drm/amd/amdgpu/amdgpu_ucode.c | 6 + drivers/g
[pull] amdgpu, amdkfd drm-next-5.3
Hi Dave, Daniel, Fixes for 5.3. Mostly fixes for Navi10 and a few other odds and ends. Also contains a patch to ease the merge with hmm. Trivial merge fix when the trees are merged: diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c @@ -783,7 +783,7 @@ int amdgpu_ttm_tt_get_user_pages(struct ttm_tt *ttm, struct page **pages) 0 : range->flags[HMM_PFN_WRITE]; range->pfn_flags_mask = 0; range->pfns = pfns; - hmm_range_register(range, mm, start, + hmm_range_register(range, mirror, start, start + ttm->num_pages * PAGE_SIZE, PAGE_SHIFT); The following changes since commit 440e80ce02cde7b810e4eb555768c2d77e7a27c8: drm/amd/display: fix a couple of spelling mistakes (2019-06-27 11:22:57 -0500) are available in the Git repository at: git://people.freedesktop.org/~agd5f/linux tags/drm-next-5.3-2019-07-09 for you to fetch changes up to 7f963d9f69bf28d639013630da30d7a4c95edd5d: drm/amdgpu/navi10: add uclk activity sensor (2019-07-09 17:43:36 -0500) drm-next-5.3-2019-07-09: amdgpu: - GPU reset for navi10 - Powerplay fixes for navi10 - GFX fixes for navi10 - Prepare for hmm_range_register API change - XGMI fixes - clang warning fixes - Fixes for various kconfig scenarios - Misc fixes and cleanups amdkfd: - Add workaround for soft hangs with oversubscribed runlists - Remove duplicated pcie atomics request Alex Deucher (8): drm/amdgpu/gfx9: use reset default for PA_SC_FIFO_SIZE drm/amdgpu/gfx10: use reset default for PA_SC_FIFO_SIZE drm/amdgpu/display: fix interrupt client id for navi drm/amdgpu: properly guard DC support in navi code drm/amdgpu/psp11: simplify the ucode register logic drm/amdgpu: add missing documentation on new module parameters drm/amdgpu: properly guard the generic discovery code drm/amdgpu/navi10: add uclk activity sensor Arnd Bergmann (4): amdgpu: make pmu support optional drm/amd/display: dcn20: include linux/delay.h drm/amd/powerplay: vega20: fix uninitialized variable use drm/amd/display: avoid 64-bit division Evan Quan (2): drm/amdgpu: fix MGPU fan boost enablement for XGMI reset drm/amd/powerplay: use hardware fan control if no powerplay fan table Felix Kuehling (5): drm/amdkfd: Print a warning when the runlist becomes oversubscribed drm/amdgpu: Use FENCE_OWNER_KFD in process_sync_pds_resv drm/amdgpu: Fix tracking of invalid userptrs drm/amdkfd: Add chained_runlist_idle_disable flag to pm4_mes_runlist drm/amdkfd: Disable idle optimization for chained runlist Flora Cui (1): drm/amdgpu: fix scheduler timeout calc Fuqian Huang (1): drm/amdgpu: Use kmemdup rather than duplicating its implementation Jack Xiao (5): drm/amdgpu: add field indicating if has PCIE atomics support drm/amdgpu: enable PCIE atomics ops support drm/amdkfd: remove duplicated PCIE atomics request drm/amdkfd: remove an unused variable drm/amd/powerplay: increase waiting time for smu response Kevin Wang (3): drm/amd/powerplay: add baco smu reset function for smu11 drm/amdgpu: add mode1 (psp) reset for navi asic drm/amd/powerplay: add temperature sensor support for navi10 Lyude Paul (1): drm/amdgpu: Don't skip display settings in hwmgr_resume() Marek Olšák (3): drm/amdgpu: fix transform feedback GDS hang on gfx10 (v2) drm/amdgpu: handle AMDGPU_IB_FLAG_RESET_GDS_MAX_WAVE_ID on gfx10 drm/amdgpu: don't invalidate caches in RELEASE_MEM, only do the writeback Nathan Chancellor (4): drm/amdgpu/mes10.1: Fix header guard drm/amd/powerplay: Use memset to initialize metrics structs drm/amd/powerplay: Zero initialize freq in smu_v11_0_get_current_clk_freq drm/amd/powerplay: Zero initialize current_rpm in vega20_get_fan_speed_percent Philip Yang (1): drm/amdgpu: Prepare for hmm_range_register API change (v2) Yrjan Skrimstad (1): drm/amd/powerplay/smu7_hwmgr: replace blocking delay with non-blocking tiancyin (1): drm/amd/powerplay: update smu11_driver_if_navi10.h xinhui pan (1): drm/amdgpu: Disable ras features on all IPs before gpu reset drivers/gpu/drm/amd/amdgpu/Makefile| 4 +- drivers/gpu/drm/amd/amdgpu/amdgpu.h| 5 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c | 7 ++ drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h | 1 + drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c | 13 ++- drivers/gpu/drm/amd/amdgpu/amdgpu_cs.c | 2 +- drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 38 - drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c| 26 +- drivers/gpu/drm/amd/amdgpu/amdgpu_gds.h| 3 +- drivers/gpu/drm/amd/amdgpu/amdgp
Re: [pull] amdgpu, amdkfd drm-next-5.3
On Tue, Jun 4, 2019 at 4:46 PM Alex Deucher wrote: > > On Mon, Jun 3, 2019 at 11:36 AM Daniel Vetter wrote: > > > > On Thu, May 30, 2019 at 12:09 AM Alex Deucher wrote: > > > > > > Hi Dave, Daniel, > > > > > > New stuff for 5.3: > > > - Add new thermal sensors for vega asics > > > - Various RAS fixes > > > - Add sysfs interface for memory interface utilization > > > - Use HMM rather than mmu notifier for user pages > > > - Expose xgmi topology via kfd > > > - SR-IOV fixes > > > - Fixes for manual driver reload > > > - Add unique identifier for vega asics > > > - Clean up user fence handling with UVD/VCE/VCN blocks > > > - Convert DC to use core bpc attribute rather than a custom one > > > - Add GWS support for KFD > > > - Vega powerplay improvements > > > - Add CRC support for DCE 12 > > > - SR-IOV support for new security policy > > > - Various cleanups > > > > > Chunming Zhou (1): > > > drm/amdgpu: add DRIVER_SYNCOBJ_TIMELINE to amdgpu > > > > This unconditionally enables timeline syncobj support, Which I thought > > we've decided to hold back behind some module_param_named_unsafe or > > experimental Kconfig, at least until KHR ratifies the extensions and > > everyone can publish the mesa patches. This is kinda uapi without > > userspace as-is ... also not on your summary, or I'm blind. > > Sorry, for some reason I had it in my head that we were going to start > enabling this is 5.3. Maybe I was mixing it up with something else. > I need to re-read the thread. I can send an updated pull with a > revert. Yeah I think the idea was to get this in, but somewhere hidden until it's released by khr. I think Kconfig/module option/ifdef or just not yet merging the patch with the driver flag where discussed. Last one I think is what Lionel plans to do for i915. Dave already merged your pull, but it's -next and far from merge window so totally fine if you only take care of this on the next regular feature pull imo. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: [pull] amdgpu, amdkfd drm-next-5.3
On Mon, Jun 3, 2019 at 11:36 AM Daniel Vetter wrote: > > On Thu, May 30, 2019 at 12:09 AM Alex Deucher wrote: > > > > Hi Dave, Daniel, > > > > New stuff for 5.3: > > - Add new thermal sensors for vega asics > > - Various RAS fixes > > - Add sysfs interface for memory interface utilization > > - Use HMM rather than mmu notifier for user pages > > - Expose xgmi topology via kfd > > - SR-IOV fixes > > - Fixes for manual driver reload > > - Add unique identifier for vega asics > > - Clean up user fence handling with UVD/VCE/VCN blocks > > - Convert DC to use core bpc attribute rather than a custom one > > - Add GWS support for KFD > > - Vega powerplay improvements > > - Add CRC support for DCE 12 > > - SR-IOV support for new security policy > > - Various cleanups > > > Chunming Zhou (1): > > drm/amdgpu: add DRIVER_SYNCOBJ_TIMELINE to amdgpu > > This unconditionally enables timeline syncobj support, Which I thought > we've decided to hold back behind some module_param_named_unsafe or > experimental Kconfig, at least until KHR ratifies the extensions and > everyone can publish the mesa patches. This is kinda uapi without > userspace as-is ... also not on your summary, or I'm blind. Sorry, for some reason I had it in my head that we were going to start enabling this is 5.3. Maybe I was mixing it up with something else. I need to re-read the thread. I can send an updated pull with a revert. Alex ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: [pull] amdgpu, amdkfd drm-next-5.3
On Thu, May 30, 2019 at 12:09 AM Alex Deucher wrote: > > Hi Dave, Daniel, > > New stuff for 5.3: > - Add new thermal sensors for vega asics > - Various RAS fixes > - Add sysfs interface for memory interface utilization > - Use HMM rather than mmu notifier for user pages > - Expose xgmi topology via kfd > - SR-IOV fixes > - Fixes for manual driver reload > - Add unique identifier for vega asics > - Clean up user fence handling with UVD/VCE/VCN blocks > - Convert DC to use core bpc attribute rather than a custom one > - Add GWS support for KFD > - Vega powerplay improvements > - Add CRC support for DCE 12 > - SR-IOV support for new security policy > - Various cleanups > Chunming Zhou (1): > drm/amdgpu: add DRIVER_SYNCOBJ_TIMELINE to amdgpu This unconditionally enables timeline syncobj support, Which I thought we've decided to hold back behind some module_param_named_unsafe or experimental Kconfig, at least until KHR ratifies the extensions and everyone can publish the mesa patches. This is kinda uapi without userspace as-is ... also not on your summary, or I'm blind. -Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
drop drmP.h usage? [Was: [pull] amdgpu, amdkfd drm-next-5.3]
Hi Alex et all. > - Various cleanups Any chance to persuade one of you guys to sweep through the amd / raedeon tree and drop use of the deprecated header drmP.h? Status at the moment (drm-misc-next): $git grep drmP | cut -d '/' -f 1 | uniq -c | sort -n | tail -n 10 11 meson 12 mediatek 13 vmwgfx 15 rockchip 16 sti 18 sun4i 24 exynos 27 nouveau 103 radeon 110 amd nouveau is already taken care of, the cleanup is in their tree. amd and radeon are the two "worst" drivers in this respect at the moment. Thanks, Sam ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
[pull] amdgpu, amdkfd drm-next-5.3
Hi Dave, Daniel, New stuff for 5.3: - Add new thermal sensors for vega asics - Various RAS fixes - Add sysfs interface for memory interface utilization - Use HMM rather than mmu notifier for user pages - Expose xgmi topology via kfd - SR-IOV fixes - Fixes for manual driver reload - Add unique identifier for vega asics - Clean up user fence handling with UVD/VCE/VCN blocks - Convert DC to use core bpc attribute rather than a custom one - Add GWS support for KFD - Vega powerplay improvements - Add CRC support for DCE 12 - SR-IOV support for new security policy - Various cleanups The following changes since commit a188339ca5a396acc588e5851ed7e19f66b0ebd9: Linux 5.2-rc1 (2019-05-19 15:47:09 -0700) are available in the Git repository at: git://people.freedesktop.org/~agd5f/linux drm-next-5.3 for you to fetch changes up to cf401e2856b27b2deeada498eab864e2a50cf219: drm/amdkfd: fix null pointer dereference on dev (2019-05-29 16:50:09 -0500) Alex Deucher (3): drm/amdgpu/vega20: use mode1 reset for RAS and XGMI drm/amdgpu: use pcie_bandwidth_available rather than open coding it drm/amdgpu/soc15: skip reset on init Amber Lin (1): drm/amdkfd: Add domain number into gpu_id Anthony Koo (1): drm/amd/display: fix multi display seamless boot case Aric Cyr (5): drm/amd/display: 3.2.28 drm/amd/display: 3.2.29 drm/amd/display: 3.2.30 drm/amd/display: Use VCP for extended colorimetry drm/amd/display: 3.2.31 Bhawanpreet Lakha (1): drm/amd/powerplay: Fix maybe-uninitialized in get_ppfeature_status Charlene Liu (5): drm/amd/display: add SW_USE_I2C_REG request. drm/amd/display: color space ycbcr709 support drm/amd/display: reset retimer/redriver below 340Mhz drm/amd/display: define v_total_min and max parameters drm/amd/display: enabling stream after HPD low to high happened Chengming Gui (2): drm/amd/powerplay: Enable "disable dpm" feature to support swSMU debug (v2) drm/amd/powerplay: Fix code error for translating int type to bool type correctly Chris Park (2): drm/amd/display: Support AVI InfoFrame V3 and V4 drm/amd/display: Define Byte 14 on AVI InfoFrame Christian König (2): drm/amdgpu: rename amdgpu_prime.[ch] into amdgpu_dma_buf.[ch] drm/amdgpu: remove static GDS, GWS and OA allocation Chunming Zhou (1): drm/amdgpu: add DRIVER_SYNCOBJ_TIMELINE to amdgpu Colin Ian King (2): drm/amdgpu: fix spelling mistake "retrived" -> "retrieved" drm/amdkfd: fix null pointer dereference on dev Dmytro Laktyushkin (4): drm/amd/display: move signal type out of otg dlg params drm/amd/display: stop external access to internal optc sync params drm/amd/display: fix acquire_first_split_pipe function drm/amd/display: add null checks and set update flags Emily Deng (2): drm/amdgpu: fix unload driver fail drm/amdgpu: Need to set the baco cap before baco reset Eric Yang (2): drm/amd/display: Set dispclk and dprefclock directly drm/amd/display: move back vbios cmd table for set dprefclk Evan Quan (26): drm/amd/powerplay: support hotspot/memory critical limit values drm/amd/powerplay: support temperature emergency max values drm/amd/powerplay: support SMU metrics table on Vega12 drm/amd/powerplay: expose current hotspot and memory temperatures V2 drm/amd/powerplay: support hwmon temperature channel labels V2 drm/amd/powerplay: expose Vega12 current power drm/amd/powerplay: expose Vega12 current gpu activity drm/amd/powerplay: expose Vega20 realtime memory utilization drm/amd/powerplay: expose Vega12 realtime memory utilization drm/amd/powerplay: expose SMU7 asics realtime memory utilization drm/amdgpu: add new sysfs interface for memory realtime utilization drm/amdgpu: enable separate timeout setting for every ring type V4 drm/amd/powerplay: fix Vega10 mclk/socclk voltage link setup drm/amd/powerplay: valid Vega10 DPMTABLE_OD_UPDATE_VDDC settings V2 drm/amd/powerplay: avoid repeat AVFS enablement/disablement drm/amd/powerplay: update Vega10 power state on OD drm/amd/powerplay: force to update all clock tables on OD reset drm/amd/powerplay: update Vega10 ACG Avfs Gb parameters drm/amd/powerplay: drop unnecessary sw smu check drm/amd/powerplay: drop redundant smu call drm/amd/powerplay: support ppfeatures sysfs interface on sw smu routine drm/amd/powerplay: honor hw limit on fetching metrics data drm/amd/powerplay: support uclk activity retrieve on sw smu routine drm/amd/powerplay: support sw smu hotspot and memory temperature retrieval drm/amd/powerplay: fix sw SMU wrong UVD/VCE powergate setting drm/amd/powerplay: enable ppfeaturemask module parameter support on Vega20 Felix Kuehling (3): drm/amdgpu: Rese