This patch implements support for VRAM page table entry compression.
PTE construction is enhanced to identify physically contiguous page
ranges and mark them in the PTE fragment field. L1/L2 TLB support is
enabled for 64KB (SI/CIK) and 256KB (NI) PTE fragments, significantly
improving TLB utilization for VRAM allocations.

Linear store bandwidth is improved from 60GB/s to 125GB/s on Pitcairn.
Unigine Heaven 3.0 sees an average improvement from 24.7 to 27.7 FPS
on default settings at 1920x1200 resolution with vsync disabled.

See main comment in radeon_vm.c for a technical description.

v2 (chk): rebased and simplified.
v3 (chk): add missing hw setup
v4 (chk): rebased on current drm-fixes-3.15
v5 (chk): fix comments and commit text
Signed-off-by: NJay Cornwall <jay@jcornwall.me>
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Just move all fields into radeon_cs_reloc, removing unused/duplicated fields.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>

Not needed any more.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>

No need to make it more complicated than necessary,
just allocate the page tables as normal BO and
flush whenever the address change.

v2: update comments and function name
v3: squash bug fixes, page directory and tables patch
v4: rebased on Mareks changes
Signed-off-by: NChristian König <christian.koenig@amd.com>

Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>

Both are complex enough on their own.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Otherwise we allocate a new VMID on nearly every submit.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Also rename the function to better reflect what it is doing.

agd5f: fix argument size warning
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

If we fail to allocate an indirect buffer (ib) when updating
the ptes, return an error instead of trying to use the ib.
Avoids a null pointer dereference.

Bug:
https://bugzilla.kernel.org/show_bug.cgi?id=58621Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

To workaround bugs and/or certain limits it's sometimes
useful to fall back to waiting on fences.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Clear page tables after allocating them in case
we don't completely fill them later.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

The DMA ring seems to be stable now.

v2: remove pt_ring_index as well
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

We don't pin the BO on allocation, so don't unpin it on free.
Signed-off-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=67016Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

Covers requirements of all current asics.
Reviewed-by: NChristian König <christian.koenig@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

There are cases where we need more than 4k alignment.  No
functional change with this commit.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org

If a buffer is never bound to a virtual memory pagetable than don't try
to unbind it. Only drawback is that we don't update the pagetable when
unbinding the ib pool buffer which is fine because it only happens at
suspend or module unload/shutdown.

Fixes spurious messages about buffers without VM mappings. E.g.:
radeon 0000:01:00.0: bo ffff88020afac400 don't has a mapping in vm ffff88021ca2b900

Cc: stable@kernel.org
Signed-off-by: NJerome Glisse <jglisse@redhat.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

For very large page table updates, we can exceed the
size of the ring.  To avoid this, use an IB to perform
the page table update.

v2(ck): cleanup the IB infrastructure and the use it instead
        of filling the struct ourself.
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NChristian König <christian.koenig@amd.com>

The few places that care should have those checks instead.
This allows destruction of bo backed memory without a reservation.
It's required for being able to rework the delayed destroy path,
as it is no longer guaranteed to hold a reservation before unlocking.

However any previous wait is still guaranteed to complete, and it's
one of the last things to be done before the buffer object is freed.
Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

Signed-off-by: NMaarten Lankhorst <maarten.lankhorst@canonical.com>
Reviewed-by: NThomas Hellstrom <thellstrom@vmware.com>
Signed-off-by: NDave Airlie <airlied@redhat.com>

When allocating more than 2GB of GART the array of pages
gets to big for kzalloc, use vzalloc instead.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Only NI uses 3dw headers, SI uses 4dw headers.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

It's better to handle this in the chipset specific code.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Reported-by: NFengguang Wu <fengguang.wu@intel.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Make it possible to allocate a persistent page table.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

We want to use VMs without the IB pool in the future.

v2: also remove it from radeon_vm_finish.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NMichel Dänzer <michel.daenzer@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Based on Dmitries work, but splitting the code into page
directory and page table handling makes it far more
readable and (hopefully) more reliable.

Allocations of page tables are made from the SA on demand,
that should still work fine since all page tables are of
the same size.

Also using the fact that allocations from the SA are mostly
continuously (except for end of buffer wraps and under very
high memory pressure) to group updates send to the chipset
specific code into larger chunks.

v3: mostly a rewrite of Dmitries previous patch.
v4: fix some typos and coding style
Signed-off-by: NDmitry Cherkasov <Dmitrii.Cherkasov@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Tested-by: NMichel Dänzer <michel.daenzer@amd.com>
Reviewed-by: NAlex Deucher <alexander.deucher@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Convert #include "..." to #include <path/...> in drivers/gpu/.
Signed-off-by: NDavid Howells <dhowells@redhat.com>
Acked-by: NDave Airlie <airlied@redhat.com>
Acked-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NThomas Gleixner <tglx@linutronix.de>
Acked-by: NPaul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: NDave Jones <davej@redhat.com>

PDE/PTE update code uses CP ring for memory writes.
All page table entries are preallocated for now in alloc_pt().

It is made as whole because it's hard to divide it to several patches
that compile and doesn't break anything being applied separately.

Tested on cayman card.

v2: rebased on top of "refactor set_page chipset interface v3",
    code cleanups

v3: switched offsets calc macros to inline funcs where possible,
    remove pd_addr from radeon_vm, switched RADEON_BLOCK_SIZE define,
    to 9 (and PTE_COUNT to 1 << BLOCK_SIZE)

v4 (ck): move "incr" documentation to previous patch, cleanup and
         document RADEON_VM_* constants, change commit message to
         our usual format, simplify patch allot by removing
         everything current not necessary, disable SI workaround.

v5: (agd5f): Fix typo in tables_size calculation in
             radeon_vm_alloc_pt().  Second line should have been
             '+=' rather than '='.

v6: fix npdes calculation. In scenario when pfns to be mapped overlap
two PDE spans:

   +-----------+-------------+
   | PDE span  | PDE span    |
   +-----------+----+--------+
          |         |
          +---------+
          | pfns    |
          +---------+

the following npdes calculation gives incorrect result:

npdes = (nptes >> RADEON_VM_BLOCK_SIZE) + 1;

For the case above picture it should give npdes = 2, but gives one.

This patch corrects it by rounding last pfn up to 512 border,
first - down to 512 border and then subtracting and dividing by 512.

v7: Make npde calculation clearer, fix ndw calculation.

v8: (agd5f): reserve enough for 2 full VM PTs, add some
             additional comments.

v9: fix typo in npde calculation
Signed-off-by: NDmitry Cherkasov <Dmitrii.Cherkasov@amd.com>
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Cleanup the interface in preparation for hierarchical page tables.

v2: add incr parameter to set_page for simple scattered PTs uptates
    added PDE-specific flags to r600_flags and radeon_drm.h
    removed superfluous value masking with 0xffffffff

v3: removed superfluous bo_va->valid checking
    changed R600_PTE_VALID to R600_ENTRY_VALID to handle PDE too

v4 (ck): fix indention style, rework and fix typos in commit message,
         add documentation for incr parameter, also use incr
         parameter for system pages

v5 (agd5f): use upper_32_bits() and minor white space fixes
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Signed-off-by: NDmitry Cherkassov <Dmitrii.Cherkasov@amd.com>
Signed-off-by: NAlex Deucher <alexander.deucher@amd.com>

Roughly based on how nouveau is handling it. Instead of
adding the bo_va when the address is set add the bo_va
when the handle is opened, but set the address to zero
until userspace tells us where to place it.

This fixes another bunch of problems with glamor.

v2: agd5f: fix build after dropping patch 7/8.
Signed-off-by: NChristian König <deathsimple@vodafone.de>

It doesn't really belong into the object functions,
also rename it to avoid collisions with struct radeon_bo_va.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>

Even GPUs can have a null pointer dereference, so move
the IB pool to another offset to catch those.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>

Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>

The end offset is exclusive not inclusive.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>

Currently doing the update with the CP.

v2: Rebased on Jeromes bugfix. Make validity comparison
    more human readable.
Signed-off-by: NChristian König <deathsimple@vodafone.de>

Makes it easier to move it into the rings.
Signed-off-by: NChristian König <deathsimple@vodafone.de>
Reviewed-by: NJerome Glisse <jglisse@redhat.com>