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: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Christian König <deathsimple@vodafone.de>
Reviewed-by: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Avoid a lot of extra loops through the pm state array.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

The new power tables need to be handled differently when setting
up the profiles.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave 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: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Polarity needs to be set accordingly to connector status (connected
or disconnected). Set it up in hpd_init() so first hotplug works
reliably no matter what is the initial set of connector. hpd_init()
also covers resume so HPD will work correctly after resume as well.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: Jerome Glisse <j.glisse@gmail.com>
Cc: stable@kernel.org
Signed-off-by: Dave 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: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Cayman seems to be particularly sensitive to read cache returning
old data after bind/unbind to GTT. Flush read cache for GTT range
with each fences for all new hw. Should fix several rendering glitches.
Like

V2 flush whole address space
V3 also flush shader read cache

https://bugs.freedesktop.org/show_bug.cgi?id=40221
https://bugs.freedesktop.org/show_bug.cgi?id=38022
https://bugzilla.redhat.com/show_bug.cgi?id=738790Signed-off-by: Jerome Glisse <jglisse@redhat.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

So that we don't get build failures once the implicit module.h
presence is removed.
Signed-off-by: Paul 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: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Covert 4k pages to multiples of 64x64x4 tiles.
This is also more efficient than a scanline based
approach from the MC's perspective.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

With this patch I'm only about 50k larger with DRM debugging
enables (why is that enabled by default?!?), and slightly
smaller without.

[airlied: moved r100.c additions to radeon_ring.c]
Signed-off-by: Andi Kleen <ak@linux.intel.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

The BO blit code inconsistenly handled the page size.  This wasn't
an issue on system with 4k pages since the GPU's page size is 4k as
well.  Switch the driver blit callbacks to take num pages in GPU
page units.

Fixes lemote mipsel systems using AMD rs780/rs880 chipsets.

v2: incorporate suggestions from Michel.
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Michel Dänzer <michel.daenzer@amd.com>
Cc: stable@kernel.org
Signed-off-by: Dave 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: Michel Dänzer <michel.daenzer@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
cc: stable@kernel.org
Signed-off-by: Dave 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: Tormod Volden <debian.tormod@gmail.com>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Signed-off-by: Dave Airlie <airlied@redhat.com>

We should have a read memory barrier between reading the WPTR from
memory and reading ring entries based on that value (ie, we need to
ensure both loads are done in order by the CPU).

It could be argued that the MMIO reads in r600_ack_irq() might be
enough to get that barrier but I prefer keeping an explicit one just
in case.

[airlied: fix evergreen + r/w mixup]
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

When not using MSIs, there is no guarantee that DMA from the device
has been fully flushed to point where it's visible to the CPU when
taking an interrupt. To get this guarantee, we need to perform an
MMIO read from the device, which will flush all outstanding DMAs
from bridges between the device and the system.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>

The writeback ring pointer and IH ring pointer are read using le32_to_cpu
so we do not want the chip to byteswap them on big-endian.

We still want to byteswap the ring itself and the IBs, so we don't touch
that but we remove setting of the byteswap bits in CP_RB_RPTR_ADDR and
IH_CNTL.

In general, for things like that where we control all the accessors easily,
we are better off doing the swap in SW rather than HW. Paradoxally, it does
keep the code closer to x86 and avoid using poorly tested HW features.

I also changed the use of RADEON_ to R600_ in a couple of cases to be more
consistent with the surrounding code.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Dave 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: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave 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: Jerome Glisse <jglisse@redhat.com>
cc: stable@kernel.org
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

0xff01 is not an actual voltage value, but a flag
for the driver.  If the power state as that value,
skip setting the voltage.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Since we were calling the wptr function before checking if the IH was
even enabled, or the GPU wasn't shutdown, we'd get spam in the logs when
the GPU readback 0xffffffff. This reorders things so we return early
in the no IH and GPU shutdown cases.

Reported-and-tested-by: ManDay on #radeon
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

debug statement for GUI idle interrupt is wrong and incorrectly
reports CP EOP interrupt; trivial issue, but confusing for
someone trying to distinguish interrupt sources while debugging
... fixed
Signed-off-by: Ilija Hadzic <ihadzic@research.bell-labs.com>
Reviewed-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

- add gpu init support
- add blit support
- add ucode loader
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

agd5f: fix commit message.
Signed-off-by: Cedric Cano <ccano@interfaceconcept.com>
Reviewed-by: Michel Dänzer <michel@daenzer.net>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

This is needed for setting voltages other than vddc.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

As per Konrad's original patch, the dummy page used
by the gart code and allocated in radeon_gart_init()
was not freed properly in radeon_gart_fini().

At the same time r6xx and newer allocated and freed the
dummy page on their own.  So to do Konrad's patch one
better, just remove the allocation and freeing of the
dummy page in the r6xx, 7xx, evergreen, and ni code and
allocate and free in the gart_init/fini() functions for
all asics.

Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Acked-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

So we used to use lpfn directly to restrict VRAM when we couldn't
access the unmappable area, however this was removed in
93225b0d as it also restricted
the gtt placements. However it was only later noticed that this
broke on some hw.

This removes the active_vram_size, and just explicitly sets it
when it changes, TTM/drm_mm will always use the real_vram_size,
and the active vram size will change the TTM size used for lpfn
setting.

We should re-work the fpfn/lpfn to per-placement at some point
I suspect, but that is too late for this kernel.

Hopefully this addresses:
https://bugs.freedesktop.org/show_bug.cgi?id=35254

v2: fix reported useful VRAM size to userspace to be correct.
Signed-off-by: Dave Airlie <airlied@redhat.com>

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

Unconditionally initialize the drm gem object - it's not
worth the trouble not to for the few kernel objects.

This patch only changes the place of the drm gem object,
access is still done via pointers.

v2: Uncoditionally align the size in radeon_bo_create. At
least the r600/evergreen blit code didn't to this, angering
the paranoid gem code.
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
Signed-off-by: Jiri Kosina <jkosina@suse.cz>

agd5f: minor cleanups
Signed-off-by: Cédric Cano <ccano@interfaceconcept.com>
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Some fixes to the thermal sensor code:
- handle negative numbers
- properly handle temp calculation on different asics
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Don't reset if the engine isn't busy.  This matches the behavior of
previous asics.  Reseting a non-hung block can lead to a hang.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=33272Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@gmail.com>

Use the common one for all asics.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Switching to pcie gen2 causes problems on some
boards.  Add a module option to turn it on/off.

There are gen2 compatability issues with some
motherboards it seems.

Fixes:
https://bugs.freedesktop.org/show_bug.cgi?id=33027Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

Supported on rv6xx/r7xx/evergreen.  Cards come up in gen1 mode.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

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

With cmwq, there's no reason for radeon to use a dedicated workqueue.
Drop dev_priv->wq and use system_wq instead.

Because radeon_driver_irq_uninstall_kms() may be called from
unsleepable context, the work items can't be flushed from there.
Instead, init and flush from radeon_irq_kms_init/fini().

While at it, simplify canceling/flushing of rdev->pm.dynpm_idle_work.
Always initialize and sync cancel instead of being unnecessarily smart
about it.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Alex Deucher <alexdeucher@gmail.com>
Cc: dri-devel@lists.freedesktop.org
Signed-off-by: Dave Airlie <airlied@redhat.com>

We were using the lockup struct from the wrong union.
Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Cc: Jerome Glisse <jglisse@redhat.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>

It should be required for all 7xx asics, but seems to cause
problems on some AGP 7xx chips.

Fixes:
https://bugzilla.kernel.org/show_bug.cgi?id=19002Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
Reported-and-Tested-by: Duncan <1i5t5.duncan@cox.net>
Cc: stable@kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>