Commits · 3a8c3400f3e74638bedd0d2410416aa8b794c0fd · Kirill Smelkov / linux

28 Aug, 2015 40 commits
- drm/nouveau/subdev: rename some functions to avoid upcoming conflicts · 3a8c3400
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  3a8c3400
- drm/nouveau/object: rename some functions to avoid upcoming conflicts · aa35888f
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  aa35888f
- drm/nouveau/nvif: device time mthd · d61f4c17
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  d61f4c17
- drm/nouveau/nvif: replace pushbuf with vm in fermi/kepler gpfifo class args · 159045cd
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  159045cd
- drm/nouveau/nvif: return chipset/board names from device info method · 22827fa4
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  22827fa4
- drm/nouveau/nvif: add supported engines query to kepler gpfifo class · d5624179
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  d5624179
- drm/nouveau/nvif: return min/max versions for supported object classes · 41a63406
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  41a63406
- drm/nouveau/nvif: assign internal class identifiers to sw classes · f58ddf95
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  f58ddf95
- drm/nouveau/nvif: use negative oclass identifier for internal classes · 315a8b2e
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  315a8b2e
- drm/nouveau/nvif: return success from sclass even for objects without children · 5c15bf70
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  5c15bf70
- drm/nouveau/nvif: extend nop ioctl to return nvif version identifier · 99d4d36a
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  99d4d36a
- drm/nouveau/nvif: replace path-based object identification · bf81df9b
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  bf81df9b
- drm/nouveau/nvif: simplify and tidy library interfaces · a01ca78c
  Ben Skeggs authored Aug 20, 2015
```
A variety of tweaks to the NVIF library interfaces, mostly ripping out
things that turned out to be not so useful.

- Removed refcounting from nvif_object, callers are expected to not be
  stupid instead.
- nvif_client is directly reachable from anything derived from nvif_object,
  removing the need for heuristics to locate it
- _new() versions of interfaces, that allocate memory for the object
  they construct, have been removed.  The vast majority of callers used
  the embedded _init() interfaces.
- No longer storing constructor arguments (and the data returned from
  nvkm) inside nvif_object, it's more or less unused and just wastes
  memory.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  a01ca78c
- drm/nouveau/client: store default device by handle, not reference · 4e7e62d6
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  4e7e62d6
- drm/nouveau/parent: remove some (now) unnecessary special-case handling · 102b49da
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  102b49da
- drm/nouveau/object: store object type data outside of handle · 493f189d
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  493f189d
- drm/nouveau/device: decouple from engine machinery · a1e88736
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  a1e88736
- drm/nouveau/device: split user device implementation out on its own · a1bfb29a
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  a1bfb29a
- drm/nouveau/device: simplify subdev construction · 0ac9d210
  Ben Skeggs authored Aug 20, 2015
```
Replaces the piece-by-piece (in response to NV_DEVICE ctor args) device
contruction with a once-off all-or-nothing approach, eliminating some
tricky refcounting issues.  The partial device init capability was only
required by some tools, and has been moved to probe time instead.

Temporarily removes a workaround for some boards where we need to fiddle
with AGP registers before executing the DEVINIT scripts.  A later commit
in this series reinstates it.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  0ac9d210
- drm/nouveau/bin: punt client/device argument handling into a common helper · c813d8e0
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  c813d8e0
- drm/nouveau/device: tidy ctor/dtor interfaces · e781dc8f
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  e781dc8f
- drm/nouveau/client: tidy ctor/dtor interfaces · 76ecea5b
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  76ecea5b
- drm/nouveau/platform: remove subclassing of nvkm_device · 47b2505e
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  47b2505e
- drm/nouveau/disp: transition outp/conn away from being based on nvkm_object · f2c906fc
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  f2c906fc
- drm/nouveau/i2c: transition pad/ports away from being based on nvkm_object · 2aa5eac5
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  2aa5eac5
- drm/nouveau/fb: transition nvkm_ram away from being based on nvkm_object · d36a99d2
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  d36a99d2
- drm/nouveau/vga: require nvkm_device pointer in accessor functions · a8dae9fe
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  a8dae9fe
- drm/nouveau/imem: use object accessor function for suspend/resume · 95cf469c
  Ben Skeggs authored Aug 20, 2015
```
Very much not ideal, but this will be improved later.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  95cf469c
- drm/nouveau/core: wrap direct use of object accessor functions · cfdc4c44
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  cfdc4c44
- drm/nouveau/mpeg: switch to gpuobj accessor macros · 572fb13d
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  572fb13d
- drm/nouveau/gr: switch to gpuobj accessor macros · 142ea05f
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  142ea05f
- drm/nouveau/fifo: switch to gpuobj accessor macros · 5444e770
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  5444e770
- drm/nouveau/dma: switch to gpuobj accessor macros · 3f532ef1
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  3f532ef1
- drm/nouveau/cipher: switch to gpuobj accessor macros · 6d06fd68
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  6d06fd68
- drm/nouveau/xtensa: switch to gpuobj accessor macros · edb1dc51
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  edb1dc51
- drm/nouveau/falcon: switch to gpuobj accessor macros · 9cc264a3
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  9cc264a3
- drm/nouveau/mmu: switch to gpuobj accessor macros · cd821077
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  cd821077
- drm/nouveau/bar: switch to gpuobj accessor macros · 19187075
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  19187075
- drm/nouveau/core: switch to gpuobj accessor macros · 45ea5031
  Ben Skeggs authored Aug 20, 2015
```
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  45ea5031
- drm/nouveau/gpuobj: type-safe accessor macros · 384af9ac
  Ben Skeggs authored Aug 20, 2015
```
These require an explicit struct nvkm_gpuobj pointer, unlike the previous
macros which take a void *, and work with any nvkm_object.

New semantics require acquiring/releasing a gpuobj before accessing them,
which will be made use of in later patches to greatly reduce the overhead
of accesses, particularly when a direct mmio mapping of the object is not
available (suspend/resume, out of ioremap() space, and on GK20A).
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
```
  384af9ac