Commits · fc36076441bae141893bd79899d19aa1b5fdf524 · nexedi / linux

18 May, 2018 40 commits
- drm/nouveau/gr/gf100-: virtualise tpc_mask + apply fixes from traces · fc360764
  Ben Skeggs authored May 08, 2018
```
We weren't placing higher TPC IDs in the right place on some configurations.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  fc360764
- drm/nouveau/gr/gf100-: virtualise r419f78 + apply fixes from traces · aa5e38dc
  Ben Skeggs authored May 08, 2018
```
Removed from GK110[B]/GK208 as RM traces show it not being touched.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  aa5e38dc
- drm/nouveau/gr/gf100-: virtualise gpc_tpc_nr · 60c0264a
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  60c0264a
- drm/nouveau/gr/gf100-: virtualise r406500 · e7163b19
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  e7163b19
- drm/nouveau/gr/gf100-: virtualise dist_skip_table + improve algorithm · 60770fa2
  Ben Skeggs authored May 08, 2018
```
The algorithm for GM200 and newer matches RM for all the boards I have, but
I don't have enough data to try and figure something out for earlier boards,
so these will still write zeroes to the table as we did before.

The code in NVGPU isn't helpful here, it appears to handle specific cases.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  60770fa2
- drm/nouveau/gr/gf100-gf119: modify max_ways_evict where required · c4a2b638
  Ben Skeggs authored May 08, 2018
```
I don't think this is done after Fermi, NVGPU used to do it but removed
the code, and I've not seen RM traces touching it either.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  c4a2b638
- drm/nouveau/gr/gf100-: virtualise alpha_beta_tables + improve algorithms · 43952c6f
  Ben Skeggs authored May 08, 2018
```
I haven't yet been able to find a fully programatic way of calculating the
same mapping as NVIDIA for GF100-GF119, so the algorithm partially depends
on data tables for specific configurations.

I couldn't find traces for every possibility, so the algorithm will switch
to a mapping similar to what GK104-GM10x use if it encounters one.  We did
the wrong thing before anyway, so shouldn't matter too much.

The algorithm used in the GK104 implementation was ported from NVGPU.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  43952c6f
- drm/nouveau/gr/gf100-: virtualise rop_mapping · ff209c23
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  ff209c23
- drm/nouveau/gr/gf100-: virtualise r4060a8 + apply fixes from traces · 9d8a80df
  Ben Skeggs authored May 08, 2018
```
Also fixes some GPUs where we write too many registers.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  9d8a80df
- drm/nouveau/gr/gf100-: virtualise tpc_per_gpc · e51f75d5
  Ben Skeggs authored May 08, 2018
```
GM20B now also shares the same code, as NVGPU shows it doesn't need
special treatment.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  e51f75d5
- drm/nouveau/gr/gf100-: virtualise sm_id/tpc_nr · fc740f54
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  fc740f54
- drm/nouveau/gr/gf100-: virtualise patch_ltc, noting missing init · ea4a2bb5
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  ea4a2bb5
- drm/nouveau/gr/gf100-: support firmware-provided sw_ctx everywhere · aedc49fd
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  aedc49fd
- drm/nouveau/gr/gf100-: delete duplicated init code · 525230cb
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  525230cb
- drm/nouveau/gr/gf100-: virtualise init_400054 · 04547482
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  04547482
- drm/nouveau/gr/gf100-: apply be exception fixes from traces · 6df6d2b9
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  6df6d2b9
- drm/nouveau/gr/gf100-: virtualise init_shader_exceptions · 4615e9b4
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  4615e9b4
- drm/nouveau/gr/gf100-: virtualise init_504430 · ab4d49a3
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  ab4d49a3
- drm/nouveau/gr/gf100-: virtualise init_tex_hww_esr · f3ef80c0
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  f3ef80c0
- drm/nouveau/gr/gf100-: virtualise init_ppc_exceptions · 70d21482
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  70d21482
- drm/nouveau/gr/gf100-: virtualise init_419c9c + apply fixes from traces · 778f18c6
  Ben Skeggs authored May 08, 2018
```
Deliberately removed from non-GP100, as RM doesn't touch it.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  778f18c6
- drm/nouveau/gr/gf100-: virtualise init_419eb4 + apply fixes from traces · 0a84a513
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  0a84a513
- drm/nouveau/gr/gf100-: virtualise init_419cc0 + apply fixes from traces · 0feab025
  Ben Skeggs authored May 08, 2018
```
Pulled some init out of main per-GPC/TPC loops to match RM.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  0feab025
- drm/nouveau/gr/gf100-: virtualise init_sked_hww_esr · 0a5b9730
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  0a5b9730
- drm/nouveau/gr/gf100-: virtualise init_40601c · 2b297b0d
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  2b297b0d
- drm/nouveau/gr/gf100-: virtualise init_ds_hww_esr_2 · 3ac72e98
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  3ac72e98
- drm/nouveau/gr/gf100-: virtualise init_fecs_exceptions + apply fixes from traces · 2585a1b1
  Ben Skeggs authored May 08, 2018
```
The value for GF100 has changed here, but it matches RM now.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  2585a1b1
- drm/nouveau/gr/gf100: write 0x400124 during init · 7c76ebb6
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  7c76ebb6
- drm/nouveau/gr/gf100-: virtualise init_swdx_pes_mask · dff30dbd
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  dff30dbd
- drm/nouveau/gr/gf100-: implement another chunk of bios-provided init · 0f78acc8
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  0f78acc8
- drm/nouveau/gr/gf100-: virtualise init_rop_active_fbps · 429412e2
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  429412e2
- drm/nouveau/gr/gf100-: virtualise init_num_active_ltcs · bfd27f39
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  bfd27f39
- drm/nouveau/gr/gf100-: virtualise init_zcull · 02917aa3
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  02917aa3
- drm/nouveau/gr/gf100-: virtualise init_vsc_stream_master · 2fe5ff63
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  2fe5ff63
- drm/nouveau/gr/gf100-: virtualise init_bios · a37279e9
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  a37279e9
- drm/nouveau/gr/gf100-: support clkgate_pack everywhere · cd9662f8
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  cd9662f8
- drm/nouveau/gr/gf100-: virtualise r405a14 · 8b058ca5
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  8b058ca5
- drm/nouveau/gr/gf100-: support firmware-provided sw_nonctx everywhere · 6f63a5fb
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  6f63a5fb
- drm/nouveau/gr/gf100-: virtualise init_gpc_mmu + apply fixes from traces · 1246f1dc
  Ben Skeggs authored May 08, 2018
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  1246f1dc
- drm/nouveau/fifo/gp100-: force individual channels into a channel group · 334cc26d
  Ben Skeggs authored May 08, 2018
```
RM does this for some reason, and is enforced in HW on Volta.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  334cc26d