Virtual address space are per drm client (opener of /dev/drm).
Client are in charge of virtual address space, they need to
map bo into it by calling DRM_RADEON_GEM_VA ioctl.

First 16M of virtual address space is reserved by the kernel.

Once using 2 level page table we should be able to have a small
vram memory footprint for each pt (there would be one pt for all
gart, one for all vram and then one first level for each virtual
address space).

Plan include using the sub allocator for a common vm page table
area and using memcpy to copy vm page table in & out. Or use
a gart object and copy things in & out using dma.

v2: agd5f fixes:
- Add vram base offset for vram pages.  The GPU physical address of a
vram page is FB_OFFSET + page offset.  FB_OFFSET is 0 on discrete
cards and the physical bus address of the stolen memory on
integrated chips.
- VM_CONTEXT1_PROTECTION_FAULT_DEFAULT_ADDR covers all vmid's >= 1

v3: agd5f:
- integrate with the semaphore/multi-ring stuff

v4:
- rebase on top ttm dma & multi-ring stuff
- userspace is now in charge of the address space
- no more specific cs vm ioctl, instead cs ioctl has a new
  chunk

v5:
- properly handle mem == NULL case from move_notify callback
- fix the vm cleanup path

v6:
- fix update of page table to only happen on valid mem placement

v7:
- add tlb flush for each vm context
- add flags to define mapping property (readable, writeable, snooped)
- make ring id implicit from ib->fence->ring, up to each asic callback
  to then do ring specific scheduling if vm ib scheduling function

v8:
- add query for ib limit and kernel reserved virtual space
- rename vm->size to max_pfn (maximum number of page)
- update gem_va ioctl to also allow unmap operation
- bump kernel version to allow userspace to query for vm support

v9:
- rebuild page table only when bind and incrementaly depending
  on bo referenced by cs and that have been moved
- allow virtual address space to grow
- use sa allocator for vram page table
- return invalid when querying vm limit on non cayman GPU
- dump vm fault register on lockup

v10: agd5f:
- Move the vm schedule_ib callback to a standalone function, remove
  the callback and use the existing ib_execute callback for VM IBs.

v11:
- rebase on top of lastest Linus

v12: agd5f:
- remove spurious backslash
- set IB vm_id to 0 in radeon_ib_get()

v13: agd5f:
- fix handling of RADEON_CHUNK_ID_FLAGS

v14:
- fix va destruction
- fix suspend resume
- forbid bo to have several different va in same vm

v15:
- rebase

v16:
- cleanup left over of vm init/fini

v17: agd5f:
- cs checker

v18: agd5f:
- reworks the CS ioctl to better support multiple rings and
VM.  Rather than adding a new chunk id for VM, just re-use the
IB chunk id and add a new flags for VM mode.  Also define additional
dwords for the flags chunk id to define the what ring we want to use
(gfx, compute, uvd, etc.) and the priority.

v19:
- fix cs fini in weird case of no ib
- semi working flush fix for ni
- rebase on top of sa allocator changes

v20: agd5f:
- further CS ioctl cleanups from Christian's comments

v21: agd5f:
- integrate CS checker improvements

v22: agd5f:
- final cleanups for release, only allow VM CS on cayman
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Somewhat specializaed sub-allocator designed to perform sub-allocation
for command buffer not only for current cs ioctl but for future command
submission ioctl as well. Patch also convert current ib pool to use
the sub allocator. Idea is that ib poll buffer can be share with other
command buffer submission not having 64K granularity.

v2 Harmonize pool handling and add suspend/resume callback to pin/unpin
sa bo (tested on rv280, rv370, r420, rv515, rv610, rv710, redwood, cayman,
rs480, rs690, rs880)
v3 Simplify allocator
v4 Fix radeon_ib_get error path to properly free fence
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

cayman is wb only and doesn't have a VC.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Some rptr/wptrs fields have different offsets and not all rings are pm4
so add a new nop field.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Add a start fence driver helper function which will be call
once for each ring and will compute cpu/gpu addr for fence
depending on wether to use wb buffer or scratch reg.

This patch replace initialize fence driver separately which
was broken in regard of GPU lockup. The fence list for created,
emited, signaled must be initialize once and only from the
asic init callback not from the startup call back which is
call from the gpu reset.

v2: With this in place we no longer need to know the number of
    rings in fence_driver_init, also writing to the scratch reg
    before knowing its offset is a bad idea.

v3: rebase on top of change to previous patch in the serie
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

That naming seems to make more sense, since we not
only want to run PM4 rings with it.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Disable the additional compute rings on cayman
until their setup is fully implemented.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Replace cp, cp1 and cp2 members with just an array
of radeon_cp structs.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Every ring seems to have the concept of read and
write pointers. Make the register offset variable
so we can use the functions for different types of rings.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Give all asic and radeon_ring_* functions a
radeon_cp parameter, so they know the ring to work with.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

They are used to sync between rings, while fences
sync between a ring and the cpu.

v2 Fix radeon_semaphore_driver_fini when no semaphore were
allocated.

v3 Initialize list early on to avoid issue in case or early
error
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

For supporting multiple CP ring buffers, async DMA
engines and UVD.  We still need a way to synchronize
between engines.

v2 initialize unused fence driver ring to avoid issue in
   suspend/unload
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

After GPU lockup VRAM gart table is unpinned and thus its pointer
becomes unvalid. This patch move the unpin code to a common helper
function and set pointer to NULL so that page update code can check
if it should update GPU page table or not. That way bo still bound
to GART can be unbound (pci_unmap_page for all there page) properly
while there is no need to update the GPU page table.

V2 move the test for null gart out of the loop, small optimization
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Bail if we hit the timeout limit.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

The vram scratch was originally only used on some 7xx asics
to work around a hw bug.  Allocate the scratch page on all 6xx+
radeons and set the MC_VM_SYSTEM_APERTURE_DEFAULT_ADDR to point
to it.  We shouldn't ever hit it since we limit the system
aperture to vram or vram and AGP, but better safe than sorry.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

So that we don't get build failures once the implicit module.h
presence is removed.
Signed-off-by: NPaul Gortmaker <paul.gortmaker@windriver.com>

r600-NI shared the same blit suspend code.  Clean it up
and make it a shared function.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Reviewed-by: NIlija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

factor out most of evergreen blit code and use the refactored code
from r600 that is now common for both r600 and evergreen
Signed-off-by: NIlija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Looks like the same pcie gen2 speed initialization for
Evergreen also works on Cayman and seems to come up fine,
so enable it if the module parameter says so
Signed-off-by: NIlija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Most asics just use the hw default value which requires
no explicit programming.  For those that need a different
value, the vbios will program it properly.  As such,
there's no need to program these registers explicitly
in the driver.  Changing MC_SHARED_CHREMAP requires a reload
of all data in vram otherwise its contents will be scambled.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=40103

v2: drop now unused channel_remap functions.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Reviewed-by: NMichel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

Apparently this doesn't always work reliably, e.g. at resume time.

Just initialize to 0, so the ring is considered empty.

Tested with hibernation on Sumo and Cayman cards.

Should fix https://bugs.launchpad.net/ubuntu/+source/linux/+bug/820746/ .
Signed-off-by: NMichel Dänzer <michel.daenzer@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

This was previously done for r300 only. Use %016llX instead of %08X for
printing the table address.

Also fix typos in gart warning messages.
Signed-off-by: NTormod Volden <debian.tormod@gmail.com>
Reviewed-by: NMichel Dänzer <michel@daenzer.net>
Signed-off-by: NDave Airlie <airlied@redhat.com>

If the bios or OS sets the pci max read request size to 0 or an
invalid value (6,7), it can result in a hang or slowdown.  Check
and set it to something sane if it's invalid.

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=42162

v2: use pci reg defines from include/linux/pci_regs.h
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@kernel.org
Reviewed-by: NMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

For some reason SPI block is in broken state after module
unloading. This lead to broken rendering after reloading
module. Fix this by reseting SPI block in CP resume function

Signed-off-by: Jerome Glisse <jglisse@redhat.com
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

The 3D driver need to get the pipe to backend
map to certain things.  Add a query to get the
info.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

ib pool weren't free for various newer asic on module unload.
This doesn't cause much arm but still could be candidate for
stable.
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
cc: stable@kernel.org
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Allows us to use the 3D engine for memory management
and allows us to use vram beyond the BAR aperture.

v2: fix copy paste typo
Reported-by: NNils Wallménius <nils.wallmenius@gmail.com>
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

cayman is reporting the wrong tile config value to userspace, this
causes piglit mipmap generation tests to fail.
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

This was leading to a bogus value being programmed to the backend
routing register.
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

On cayman we need to set the bit to cause HDP flushes to invalidate the
HDP cache also.
Reviewed-by: NAlex Deucher <alexdeucher@gmail.com>
cc: stable@kernel.org
Signed-off-by: NDave Airlie <airlied@redhat.com>

The TCC disable setup was incorrect.  This
prevents the GPU from hanging when draw commands
are issued.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

We are accessing totally the wrong struct in this case, and putting
uninitialised values into the GPU, which it doesn't like unsurprisingly.
Signed-off-by: NDave Airlie <airlied@redhat.com>

These should be handled by the clear_state setup, but set them
directly as well just to be sure.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Cayman is different enough from evergreen to warrant it's own functions.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Cayman asics have 3 ring buffers:
ring 0 supports both gfx and compute
rings 1 and 2 are compute only

At the moment we only support ring 0.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This patch sets up the gart in legacy mode.  We
probably want to switch to full VM mode at some point.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

This may some work to get accel going.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

The MC ucode is no longer loaded by the vbios
tables as on previous asics.  It now must be loaded
by the driver.
Signed-off-by: NAlex Deucher <alexdeucher@gmail.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>