Re: [pull] amdgpu/kfd, radeon, ttm drm-next-5.5
On Thu, Oct 10, 2019 at 4:37 PM Koenig, Christian wrote: > Am 10.10.19 um 16:34 schrieb Alex Deucher: > > AOn Thu, Oct 10, 2019 at 5:54 AM Daniel Vetter > > wrote: > >> On Thu, Oct 10, 2019 at 6:17 AM Alex Deucher wrote: > >>> [SNIP] > >>> Christian König (22): > >>>drm/amdgpu: use moving fence instead of exclusive for VM updates > >>>drm/amdgpu: reserve at least 4MB of VRAM for page tables v2 > >>>drm/amdgpu: remove amdgpu_cs_try_evict > >> Patch no handy for a direct reply, so asking here (but this is totally > >> unrelated to the pull): > >> > >> Do you have other stuff than scanout and pagetables that need to be in > >> vram? I was kinda assume this is needed for big vram-only objects to > >> fit, making space by throwing stuff out that could also be put into > >> system memory. But sounds like it was only for making pagetables fit. > > Yes, basically making page tables fit. If you push a bunch of stuff > > to system ram, your page table requirements go up too. See the > > discussion here: > > https://www.spinics.net/lists/amd-gfx/msg38640.html Yeah read that, that's why I asked whether pagetables was the only big thing. > Yeah, typical chicken and egg problem. > > When you evict things to system memory because you don't have enough > VRAM you need more VRAM for page tables so you need to evict even more > things to system memory > > Additional to that we have a few other cases where we really need VRAM > for correct operation (firmware, old MM engines etc...), but nothing > major like page tables. Yeah makes sense. Afaiui we'll have a few more big things in vram only, so I think we'll steal this idea for i915. -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/kfd, radeon, ttm drm-next-5.5
Am 10.10.19 um 16:34 schrieb Alex Deucher: > AOn Thu, Oct 10, 2019 at 5:54 AM Daniel Vetter wrote: >> On Thu, Oct 10, 2019 at 6:17 AM Alex Deucher wrote: >>> [SNIP] >>> Christian König (22): >>>drm/amdgpu: use moving fence instead of exclusive for VM updates >>>drm/amdgpu: reserve at least 4MB of VRAM for page tables v2 >>>drm/amdgpu: remove amdgpu_cs_try_evict >> Patch no handy for a direct reply, so asking here (but this is totally >> unrelated to the pull): >> >> Do you have other stuff than scanout and pagetables that need to be in >> vram? I was kinda assume this is needed for big vram-only objects to >> fit, making space by throwing stuff out that could also be put into >> system memory. But sounds like it was only for making pagetables fit. > Yes, basically making page tables fit. If you push a bunch of stuff > to system ram, your page table requirements go up too. See the > discussion here: > https://www.spinics.net/lists/amd-gfx/msg38640.html Yeah, typical chicken and egg problem. When you evict things to system memory because you don't have enough VRAM you need more VRAM for page tables so you need to evict even more things to system memory Additional to that we have a few other cases where we really need VRAM for correct operation (firmware, old MM engines etc...), but nothing major like page tables. Regards, Christian. > > Alex > >> -Daniel >> >> ___ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
Re: [pull] amdgpu/kfd, radeon, ttm drm-next-5.5
AOn Thu, Oct 10, 2019 at 5:54 AM Daniel Vetter wrote: > > On Thu, Oct 10, 2019 at 6:17 AM Alex Deucher wrote: > > > > Hi Dave, Daniel, > > > > New stuff for 5.5. There's an export of a cgroup function that > > Tejun acked for merging through the drm tree. kfd uses it to handle > > permissions in containers since there is only one /dev/kfd. > > > > The following changes since commit 9a60b2990d6c2b7ab935fe0a5cc274de67d98bed: > > > > Merge branch 'etnaviv/next' of https://git.pengutronix.de/git/lst/linux > > into drm-next (2019-09-06 16:58:10 +1000) > > > > are available in the Git repository at: > > > > git://people.freedesktop.org/~agd5f/linux tags/drm-next-5.5-2019-10-09 > > > > for you to fetch changes up to 1cd4d9eead73c004d08a58536dc726bd172eaaec: > > > > drm/amdkfd: update for drmP.h removal (2019-10-09 12:04:48 -0500) > > > > > > drm-next-5.5-2019-10-09: > > > > amdgpu: > > - Additional RAS enablement for vega20 > > - RAS page retirement and bad page storage in EEPROM > > - No GPU reset with unrecoverable RAS errors > > - Reserve vram for page tables rather than trying to evict > > - Fix issues with GPU reset and xgmi hives > > - DC i2c over aux fixes > > - Direct submission for clears, PTE/PDE updates > > - Improvements to help support recoverable GPU page faults > > - Silence harmless SAD block messages > > - Clean up code for creating a bo at a fixed location > > - Initial DC HDCP support > > - Lots of documentation fixes > > - GPU reset for renoir > > - Add IH clockgating support for soc15 asics > > - Powerplay improvements > > - DC MST cleanups > > - Add support for MSI-X > > - Misc cleanups and bug fixes > > > > amdkfd: > > - Query KFD device info by asic type rather than pci ids > > - Add navi14 support > > - Add renoir support > > - Add navi12 support > > - gfx10 trap handler improvements > > - pasid cleanups > > - Check against device cgroup > > > > ttm: > > - Return -EBUSY with pipelining with no_gpu_wait > > > > radeon: > > - Silence harmless SAD block messages > > > > device_cgroup: > > - Export devcgroup_check_permission > > > > > > Aaron Liu (4): > > drm/amd/display: update renoir_ip_offset.h > > drm/amdgpu: disable stutter mode for renoir > > drm/amdgpu: remove program of lbpw for renoir > > Revert "drm/amdgpu: disable stutter mode for renoir" > > > > Adam Zerella (1): > > docs: drm/amdgpu: Resolve build warnings > > > > Alex Deucher (20): > > drm/amdgpu/irq: check if nbio funcs exist > > drm/amdgpu/vm: fix documentation for amdgpu_vm_bo_param > > drm/amdgpu/ras: use GPU PAGE_SIZE/SHIFT for reserving pages > > drm/amdgpu/psp: flush HDP write fifo after submitting cmds to the psp > > drm/amdgpu/psp: invalidate the hdp read cache before reading the psp > > response > > drm/amdgpu: flag navi12 and 14 as experimental for 5.4 > > drm/amdgpu: fix documentation for amdgpu_gem_prime_export > > drm/amdgpu/mn: fix documentation for amdgpu_mn_read_lock > > drm/amdgpu/vm: fix up documentation in amdgpu_vm.c > > drm/amdgpu/ih: fix documentation in amdgpu_irq_dispatch > > drm/amdgpu: fix documentation for amdgpu_pm.c > > drm/amdgpu/ras: fix and update the documentation for RAS > > drm/amdgpu/display: fix 64 bit divide > > drm/amdgpu/display: include slab.h in dcn21_resource.c > > drm/amdgpu/atomfirmware: use proper index for querying vram type (v3) > > drm/amdgpu/atomfirmware: simplify the interface to get vram info > > drm/amdgpu: don't increment vram lost if we are in hibernation > > drm/amdgpu: improve MSI-X handling (v3) > > drm/amdgpu: move amdgpu_device_get_job_timeout_settings > > drm/amdkfd: fix the build when CIK support is disabled > > > > Allen Pais (1): > > drm/amdkfd: fix a potential NULL pointer dereference (v2) > > > > Alvin Lee (1): > > drm/amd/display: Don't allocate payloads if link lost > > > > Andrey Grodzovsky (11): > > drm/amdgpu: Fix bugs in amdgpu_device_gpu_recover in XGMI case. > > drm/amdgpu: Avoid HW GPU reset for RAS. > > dmr/amdgpu: Add system auto reboot to RAS. > > drm/amdgpu: Add smu lock around in pp_smu_i2c_bus_access > > drm/amdgpu: Remove clock gating restore. > > drm/madgpu: Fix EEPROM Checksum calculation. > > drm/amdgpu: Avoid RAS recovery init when no RAS support. > > drm/amdgpu: Add amdgpu_ras_eeprom_reset_table > > drm/amdgpu: Allow to reset to EERPOM table. > > drm/amdgpu: Fix mutex lock from atomic context. > > drm/amdgpu:Fix EEPROM checksum calculation. > > > > Anthony Koo (2): > > drm/amd/display: 3.2.49 > > drm/amd/display: set minimum abm backlight level > > > > Aric Cyr (4): > > drm/amd/display: 3.2.50 > > drm/amd/display: 3.2.51 > > drm/amd/display: 3.2.51.1
Re: [pull] amdgpu/kfd, radeon, ttm drm-next-5.5
On Thu, Oct 10, 2019 at 6:17 AM Alex Deucher wrote: > > Hi Dave, Daniel, > > New stuff for 5.5. There's an export of a cgroup function that > Tejun acked for merging through the drm tree. kfd uses it to handle > permissions in containers since there is only one /dev/kfd. > > The following changes since commit 9a60b2990d6c2b7ab935fe0a5cc274de67d98bed: > > Merge branch 'etnaviv/next' of https://git.pengutronix.de/git/lst/linux > into drm-next (2019-09-06 16:58:10 +1000) > > are available in the Git repository at: > > git://people.freedesktop.org/~agd5f/linux tags/drm-next-5.5-2019-10-09 > > for you to fetch changes up to 1cd4d9eead73c004d08a58536dc726bd172eaaec: > > drm/amdkfd: update for drmP.h removal (2019-10-09 12:04:48 -0500) > > > drm-next-5.5-2019-10-09: > > amdgpu: > - Additional RAS enablement for vega20 > - RAS page retirement and bad page storage in EEPROM > - No GPU reset with unrecoverable RAS errors > - Reserve vram for page tables rather than trying to evict > - Fix issues with GPU reset and xgmi hives > - DC i2c over aux fixes > - Direct submission for clears, PTE/PDE updates > - Improvements to help support recoverable GPU page faults > - Silence harmless SAD block messages > - Clean up code for creating a bo at a fixed location > - Initial DC HDCP support > - Lots of documentation fixes > - GPU reset for renoir > - Add IH clockgating support for soc15 asics > - Powerplay improvements > - DC MST cleanups > - Add support for MSI-X > - Misc cleanups and bug fixes > > amdkfd: > - Query KFD device info by asic type rather than pci ids > - Add navi14 support > - Add renoir support > - Add navi12 support > - gfx10 trap handler improvements > - pasid cleanups > - Check against device cgroup > > ttm: > - Return -EBUSY with pipelining with no_gpu_wait > > radeon: > - Silence harmless SAD block messages > > device_cgroup: > - Export devcgroup_check_permission > > > Aaron Liu (4): > drm/amd/display: update renoir_ip_offset.h > drm/amdgpu: disable stutter mode for renoir > drm/amdgpu: remove program of lbpw for renoir > Revert "drm/amdgpu: disable stutter mode for renoir" > > Adam Zerella (1): > docs: drm/amdgpu: Resolve build warnings > > Alex Deucher (20): > drm/amdgpu/irq: check if nbio funcs exist > drm/amdgpu/vm: fix documentation for amdgpu_vm_bo_param > drm/amdgpu/ras: use GPU PAGE_SIZE/SHIFT for reserving pages > drm/amdgpu/psp: flush HDP write fifo after submitting cmds to the psp > drm/amdgpu/psp: invalidate the hdp read cache before reading the psp > response > drm/amdgpu: flag navi12 and 14 as experimental for 5.4 > drm/amdgpu: fix documentation for amdgpu_gem_prime_export > drm/amdgpu/mn: fix documentation for amdgpu_mn_read_lock > drm/amdgpu/vm: fix up documentation in amdgpu_vm.c > drm/amdgpu/ih: fix documentation in amdgpu_irq_dispatch > drm/amdgpu: fix documentation for amdgpu_pm.c > drm/amdgpu/ras: fix and update the documentation for RAS > drm/amdgpu/display: fix 64 bit divide > drm/amdgpu/display: include slab.h in dcn21_resource.c > drm/amdgpu/atomfirmware: use proper index for querying vram type (v3) > drm/amdgpu/atomfirmware: simplify the interface to get vram info > drm/amdgpu: don't increment vram lost if we are in hibernation > drm/amdgpu: improve MSI-X handling (v3) > drm/amdgpu: move amdgpu_device_get_job_timeout_settings > drm/amdkfd: fix the build when CIK support is disabled > > Allen Pais (1): > drm/amdkfd: fix a potential NULL pointer dereference (v2) > > Alvin Lee (1): > drm/amd/display: Don't allocate payloads if link lost > > Andrey Grodzovsky (11): > drm/amdgpu: Fix bugs in amdgpu_device_gpu_recover in XGMI case. > drm/amdgpu: Avoid HW GPU reset for RAS. > dmr/amdgpu: Add system auto reboot to RAS. > drm/amdgpu: Add smu lock around in pp_smu_i2c_bus_access > drm/amdgpu: Remove clock gating restore. > drm/madgpu: Fix EEPROM Checksum calculation. > drm/amdgpu: Avoid RAS recovery init when no RAS support. > drm/amdgpu: Add amdgpu_ras_eeprom_reset_table > drm/amdgpu: Allow to reset to EERPOM table. > drm/amdgpu: Fix mutex lock from atomic context. > drm/amdgpu:Fix EEPROM checksum calculation. > > Anthony Koo (2): > drm/amd/display: 3.2.49 > drm/amd/display: set minimum abm backlight level > > Aric Cyr (4): > drm/amd/display: 3.2.50 > drm/amd/display: 3.2.51 > drm/amd/display: 3.2.51.1 > drm/amd/display: Improve LFC behaviour > > Arnd Bergmann (4): > drm/amd/display: hide an unused variable > drm/amdgpu: make pmu support optional, again > drm/amdgpu: hide another #warning > drm/amdgpu: display_mode_vba_21: remove uint typedef > > Austin Kim (1): >
