drm/nouveau/gr/gf117-: make ppc_nr[gpc] accurate

We're going to be pulling in a chunk of code from NVGPU to fixup our
SMID mappings on Volta and above, which depends on ppc_nr[gpc]
reflecting the actual number of PPCs present, not the maximum number.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

drm/nouveau/gr/gf100-: make global attrib_cb actually global

This was thought to be per-channel initially - it's not. The backing
pages for the VMM mappings are shared for all channels.

- switches to more straight-forward patch interfaces
- prepares for sub-context support
- this is saving a *sizeable* amount of vram

v2:
- whitespace

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Reviewed-by: Lyude Paul <lyude@redhat.com>

drm/nouveau/gr/gf100-: calculate and use sm mapping table

There's a number of places that require this data, so let's separate out
the calculations to ensure they remain consistent.

This is incorrect for GM200 and newer, but will produce the same results
as we did before.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: delete duplicated grctx init code

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: update 419a3c where required

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise r418e94

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise tpc_mask + apply fixes from traces

We weren't placing higher TPC IDs in the right place on some configurations.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise gpc_tpc_nr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise r406500

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise dist_skip_table + improve algorithm

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>

drm/nouveau/gr/gf100-: virtualise rop_mapping

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise tpc_per_gpc

GM20B now also shares the same code, as NVGPU shows it doesn't need
special treatment.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: virtualise sm_id/tpc_nr

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gk104-: move rop_active_fbps init to nonctx

Matches newer RM.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200-: fix bad hardcoding of a max-tpcs-per-gpc value

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200: make generate_main() static

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gf100-: remove hardcoded idle_timeout values

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200: switch over to using sw_method_init from firmware

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200: switch over to using sw_bundle_init from firmware

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200: switch over to using sw_ctx from firmware

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>

drm/nouveau/gr/gm200: s/gm204/gm200/

Most of the per-chipset differences will go away when we fully switch
to using the register lists provided by the firmware files, which will
leave all the remaining code "belonging" to GM200.

This is a preemptive rename from GM204 to GM200.

Signed-off-by: Ben Skeggs <bskeggs@redhat.com>