Fixes bug on Tegra where we'd strip kind information from system memory
(ie. all) buffers, resulting in misrendering.
Behaviour on dGPU should be unchanged.
Reported-by: Thierry Reding <treding@nvidia.com>
Fixes: d7722134b8 ("drm/nouveau: switch over to new memory and vmm interfaces")
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Tested-by: Thierry Reding <treding@nvidia.com>
If the VMA is being deleted, we don't need to explicity unmap first
anymore. The MMU code will automatically merge the operations into
a single page tree walk.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
This will cause a subtle behaviour change on GPUs that are in mixed-memory
configurations in that VRAM in the degraded section of VRAM will no longer
be used for TTM buffer objects.
That section of VRAM is not meant to be used for displayable/compressed
surfaces, and we have no reliable way with the current interfaces to be
able to make that decision properly.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Another transition step to allow finer-grained patches transitioning to
new MMU backends.
Old backends will continue operate as before (accessing nvkm_mem::tag),
and new backends will get a reference to the tags allocated here.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
v2: Ben Skeggs <bskeggs@redhat.com>
- fill in nouveau_pm.dev to prevent oops
- fix ppc issues (build + OF shadow)
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
This is a HUGE commit, but it's not nearly as bad as it looks - any problems
can be isolated to a particular chipset and engine combination. It was
simply too difficult to port each one at a time, the compat layers are
*already* ridiculous.
Most of the changes here are simply to the glue, the process for each of the
engine modules was to start with a standard skeleton and copy+paste the old
code into the appropriate places, fixing up variable names etc as needed.
v2: Marcin Slusarz <marcin.slusarz@gmail.com>
- fix find/replace bug in license header
v3: Ben Skeggs <bskeggs@redhat.com>
- bump indirect pushbuf size to 8KiB, 4KiB barely enough for userspace and
left no space for kernel's requirements during GEM pushbuf submission.
- fix duplicate assignments noticed by clang
v4: Marcin Slusarz <marcin.slusarz@gmail.com>
- add sparse annotations to nv04_fifo_pause/nv04_fifo_start
- use ioread32_native/iowrite32_native for fifo control registers
v5: Ben Skeggs <bskeggs@redhat.com>
- rebase on v3.6-rc4, modified to keep copy engine fix intact
- nv10/fence: unmap fence bo before destroying
- fixed fermi regression when using nvidia gr fuc
- fixed typo in supported dma_mask checking
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Future work will be headed in the way of separating the policy supplied by
the nouveau drm module from the mechanisms provided by the driver core.
There will be a couple of major classes (subdev, engine) of driver modules
that have clearly defined tasks, and the further directory structure change
is to reflect this.
No code changes here whatsoever, aside from fixing up a couple of include
file pathnames.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Been tested on each major revision that's relevant here, but I'm sure there
are still bugs waiting to be ironed out.
This is a *very* invasive change.
There's a couple of pieces left that I don't like much (eg. other engines
using fifo_priv for the channel count), but that's an artefact of there
being a master channel list still. This is changing, slowly.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
All the places this stuff is actually needed tends to be chipset-specific
anyway, so we're able to just inline the register bashing instead.
The parts of the common code that still directly touch PFIFO temporarily
have conditionals, these will be removed in subsequent commits that will
refactor the fifo modules into engine modules like graph/mpeg etc.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
This adds prime->fd and fd->prime support to nouveau,
it passes the SG object to TTM, and then populates the
GART entries using it.
v2: add stubbed kmap + use new function to fill out pages array
for faulting + add reimport test.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This patch fixes two small issues in timing generation as spotted on
several NVCx cards.
In addition, the header of the file is updated to also contain (some of)
the current developers of this code.
Signed-off-by: Roy Spliet <r.spliet@student.tudelft.nl>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
There's some "extended" GDDR3 chipsets out there with EMRS2 settings that
change the layout of MRS/EMRS1 bitmaps.. Sigh.. Still need to track down
how exactly we're supposed to handle this.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
This will probably result in more lines of code, however, we're going to
have at least 3 slightly different implementations of this very soon and
I'd rather keep the ram reclocking logic separate from the hw specifics.
DDR2/DDR3/GDDR3 implemented thus far, others will be added as necessary.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Statically generating the PFB register and MR values for each timing set
turns out to be insufficient. There's at least one (so far) known piece
of information which effects MR values which is stored in the perflvl
entry on some chipsets (and in another table on later ones), which is
disconnected from the timing table entries.
After this change we will generate a timing set based on an input clock
frequency instead, and have this data stored in the performance level
data.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Roy Spliet:
- Implement according to specs
- Simplify
- Make array for mc latency registers
Martin Peres:
- squash and split all the commits from Roy
- rework following Ben Skeggs comments
- add a form of timings validation
- store the initial timings for later use
Ben Skeggs
- merge slightly modified tidy-up patch with this one
- remove perflvl-dropping logic for the moment
Signed-off-by: Roy Spliet <r.spliet@student.tudelft.nl>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
- Rename several VBIOS entries to closer match the real world
- Add the missing 0x100238 and 0x100240 register values
- Parse bit 14 of the VBIOS timing table
- "Magic value" -> tCWL, fixing some minor bugs in the process
- Also name a few more by their name rather than their number.
- Some values seem to be dependent on the memory type. Fix
Edits by Martin Peres <martin.peres@labri.fr>:
- this is a squash commit
- reworked for fixing some style issues
Signed-off-by: Roy Spliet <r.spliet@student.tudelft.nl>
Signed-off-by: Martin Peres <martin.peres@labri.fr>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
M version 2 appears to have a table with some form of memory type info
available.
NVIDIA appear to ignore the table information except for this DDR2/DDR3
case (which has the same value in 0x100714). My guess is this is due to
some of the supported memory types not being represented in the table.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
If the card is capable of more than 32-bit, then use the default
TTM page pool code which allocates from anywhere in the memory.
Note: If the 'ttm.no_dma' parameter is set, the override is ignored
and the default TTM pool is used.
V2 use pci_set_consistent_dma_mask
V3 Rebase on top of no memory account changes (where/when is my
delorean when i need it ?)
CC: Ben Skeggs <bskeggs@redhat.com>
CC: Francisco Jerez <currojerez@riseup.net>
CC: Dave Airlie <airlied@redhat.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
NV30: Create framework for memtm
NV50: Improve reg creation,
NV50: Use P.version instead of card codename/stepping,
NVC0: Initial memtiming code for Fermi,
Renamed regs for consistency,
Overall redesign to improve readability,
Avoid kfree on null-pointer
Signed-off-by: Roy Spliet <r.spliet@student.tudelft.nl>
drm_pci_device_is_pcie duplicates the funcationality of pci_is_pcie.
Convert callers of the former to the latter. This has the side benefit
of removing an unnecessary search in the PCI configuration space due to
using a saved PCIe capability offset.
[airlied: update for new callsite]
Signed-off-by: Jon Mason <jdmason@kudzu.us>
Signed-off-by: Dave Airlie <airlied@redhat.com>
'drm/nouveau: rework vram init/fini ordering a little' changed
the order of instmem.init() and nouveau_mem_vram_init() which
resulted in using ramin_rsvd_vram before it was calculated and
failing to init any accel on pre-NV50 cards.
Since it's only used on <NV50 just calculate it where it's needed
and leave it as default 0 for NV50.
Signed-off-by: Younes Manton <younes.m@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
It's a regression from "drm/nouveau: create temp vmas for both src and
dst of bo moves".
Signed-off-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Commit "drm/nouveau: add some debug output if nouveau_mm busy at destroy time"
revealed an issue where vram mm takedown would actually fail due to there
still being nodes present, causing nouveau to leak a small amount of memory
on module unload.
This splits TTM/nouveau_mm a bit more cleanly and ensures nouveau_mm fini
isn't done until all gpuobjs are also destroyed.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Greatly simplifies a number of things, particularly once per-client GPU
address spaces are involved.
May add this back later once I know what things'll look like.
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
Ben Skeggs