To sum up what goes on here, we abstract the vma binding, similarly to
the previous object binding. This helps for distinguishing legacy
binding, versus modern binding. To keep the code churn as minimal as
possible, I am leaving in insert_entries(). It serves as the per
platform pte writing basically. bind_vma and insert_entries do share a
lot of similarities, and I did have designs to combine the two, but as
mentioned already... too much churn in an already massive patchset.

What follows are the 3 commits which existed discretely in the original
submissions. Upon rebasing on Broadwell support, it became clear that
separation was not good, and only made for more error prone code. Below
are the 3 commit messages with all their history.

drm/i915: Add bind/unbind object functions to VMA
drm/i915: Use the new vm [un]bind functions
drm/i915: reduce vm->insert_entries() usage

drm/i915: Add bind/unbind object functions to VMA

As we plumb the code with more VM information, it has become more
obvious that the easiest way to deal with bind and unbind is to simply
put the function pointers in the vm, and let those choose the correct
way to handle the page table updates. This change allows many places in
the code to simply be vm->bind, and not have to worry about
distinguishing PPGTT vs GGTT.

Notice that this patch has no impact on functionality. I've decided to
save the actual change until the next patch because I think it's easier
to review that way. I'm happy to squash the two, or let Daniel do it on
merge.

v2:
Make ggtt handle the quirky aliasing ppgtt
Add flags to bind object to support above
Don't ever call bind/unbind directly for PPGTT until we have real, full
PPGTT (use NULLs to assert this)
Make sure we rebind the ggtt if there already is a ggtt binding.  This
happens on set cache levels.
Use VMA for bind/unbind (Daniel, Ben)

v3: Reorganize ggtt_vma_bind to be more concise and easier to read
(Ville). Change logic in unbind to only unbind ggtt when there is a
global mapping, and to remove a redundant check if the aliasing ppgtt
exists.

v4: Make the bind function a bit smarter about the cache levels to avoid
unnecessary multiple remaps. "I accept it is a wart, I think unifying
the pin_vma / bind_vma could be unified later" (Chris)
Removed the git notes, and put version info here. (Daniel)

v5: Update the comment to not suck (Chris)

v6:
Move bind/unbind to the VMA. It makes more sense in the VMA structure
(always has, but I was previously lazy). With this change, it will allow
us to keep a distinct insert_entries.
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>

drm/i915: Use the new vm [un]bind functions

Building on the last patch which created the new function pointers in
the VM for bind/unbind, here we actually put those new function pointers
to use.

Split out as a separate patch to aid in review. I'm fine with squashing
into the previous patch if people request it.

v2: Updated to address the smart ggtt which can do aliasing as needed
Make sure we bind to global gtt when mappable and fenceable. I thought
we could get away without this initialy, but we cannot.

v3: Make the global GTT binding explicitly use the ggtt VM for
bind_vma(). While at it, use the new ggtt_vma helper (Chris)

At this point the original mailing list thread diverges. ie.

v4^:
use target_obj instead of obj for gen6 relocate_entry
vma->bind_vma() can be called safely during pin. So simply do that
instead of the complicated conditionals.
Don't restore PPGTT bound objects on resume path
Bug fix in resume path for globally bound Bos
Properly handle secure dispatch
Rebased on vma bind/unbind conversion
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>

drm/i915: reduce vm->insert_entries() usage

FKA: drm/i915: eliminate vm->insert_entries()

With bind/unbind function pointers in place, we no longer need
insert_entries. We could, and want, to remove clear_range, however it's
not totally easy at this point. Since it's used in a couple of place
still that don't only deal in objects: setup, ppgtt init, and restore
gtt mappings.

v2: Don't actually remove insert_entries, just limit its usage. It will
be useful when we introduce gen8. It will always be called from the vma
bind/unbind.

Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk> (v1)
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

BDW context sizes varies a bit.

v2: Squash in fixup for the hw context size from Ben.

Signed-off-by: Ben Widawsky <ben@bwidawsk.net> (v1)
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

v2: Fixed the botched locking on init_hw failure in i915_reset (Ville)
Call cleanup_ringbuffer on failed context create in init_hw (Ville)

v3: Add dev argument ti clean_ringbuffer
Reviewed-by: NKenneth Graunke <kenneth@whitecape.org>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

I had this lying around from he original PPGTT series, and thought we
might try to get it in by itself.

With the introduction of context refcounting we never explicitly
ref/unref the backing object. As such, the previous fix was a bit wonky.

Aside from fixing the above, this patch also puts us in good shape for
an upcoming patch which allows a failure to occur in between
context_init and the first do_switch.

CC: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Even though we track object activity and not VMA, because we have the
active_list be based on the VM, it makes the most sense to use VMAs in
the APIs.

NOTE: Daniel intends to eventually rip out active/inactive LRUs, but for
now, leave them be.

v2: Remove leftover hunk from the previous patch which didn't keep
i915_gem_object_move_to_active. That patch had to rely on the ring to
get the dev instead of the obj. (Chris)
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

On both Ivybridge and Haswell, row remapping information is saved and
restored with context. This means, we never actually properly supported
the l3 remapping because our sysfs interface is asynchronous (and not
tied to any context), and the known faulty HW would be reused by the
next context to run.

Not that due to the asynchronous nature of the sysfs entry, there is no
point modifying the registers for the existing context. Instead we set a
flag for all contexts to load the correct remapping information on the
next run. Interested clients can use debugfs to determine whether or not
the row has been remapped.

One could propose at this point that we just do the remapping in the
kernel. I guess since we have to maintain the sysfs interface anyway,
I'm not sure how useful it is, and I do like keeping the policy in
userspace; (it wasn't my original decision to make the
interface the way it is, so I'm not attached).

v2: Force a context switch when we have a remap on the next switch.
(Ville)
Don't let userspace use the interface with disabled contexts.

v3: Don't force a context switch, just let it nop
Improper context slice remap initialization, 1<<1 instead of 1<<i, but I
rewrote it to avoid a second round of confusion.
Error print moved to error path (All Ville)
Added a comment on why the slice remap initialization happens.

CC: Ville Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

I have implemented this patch before without creating a separate list
(I'm having trouble finding the links, but the messages ids are:
<1364942743-6041-2-git-send-email-ben@bwidawsk.net>
<1365118914-15753-9-git-send-email-ben@bwidawsk.net>)

However, the code is much simpler to just use a list and it makes the
code from the next patch a lot more pretty.

As you'll see in the next patch, the reason for this is to be able to
specify when a context needs to get L3 remapping. More details there.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Signed-off-by: NDamien Lespiau <damien.lespiau@intel.com>
Acked-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We use the request to ensure we hold a reference to the context for the
duration that it remains in use by the ring. Each request only holds a
reference to the current context, hence we emit a request after
switching contexts with the final reference to the old context. However,
the extra interrupt caused by that request is not useful (no timing
critical function will wait for the context object), instead the overhead
of servicing the IRQ shows up in some (lightweight) benchmarks. In order
to keep the useful property of using the request to manage the context
lifetime, we want to add a dummy request that is associated with the
interrupt from the subsequent real request following the batch.

The extra interrupt was added as a side-effect of using
i915_add_request() in

commit 112522f6
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Thu May 2 16:48:07 2013 +0300

    drm/i915: put context upon switching

v2: Daniel convinced me that the request here was solely for context
lifetime tracking and that we have the active ref to keep the object
alive whilst the MI_SET_CONTEXT. So the only concern then is which
context should get the blame for MI_SET_CONTEXT failing. The old scheme
added a request for the old context so that any hang upto and including
the switch away would mark the old context as guilty. Now any hang here
implicates the new context. However since we have already gone through a
complete flush with the last context in its last request, and all that
lies in no-man's-land is an invalidate flush and the MI_SET_CONTEXT, we
should be safe in not unduly placing blame on the new context.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <ben@bwidawsk.net>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: NDamien Lespiau <damien.lespiau@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

formerly: "drm/i915: Create VMAs (part 5) - move mm_list"

The mm_list is used for the active/inactive LRUs. Since those LRUs are
per address space, the link should be per VMx .

Because we'll only ever have 1 VMA before this point, it's not incorrect
to defer this change until this point in the patch series, and doing it
here makes the change much easier to understand.

Shamelessly manipulated out of Daniel:
"active/inactive stuff is used by eviction when we run out of address
space, so needs to be per-vma and per-address space. Bound/unbound otoh
is used by the shrinker which only cares about the amount of memory used
and not one bit about in which address space this memory is all used in.
Of course to actual kick out an object we need to unbind it from every
address space, but for that we have the per-object list of vmas."

v2: only bump GGTT LRU in i915_gem_object_set_to_gtt_domain (Chris)

v3: Moved earlier in the series

v4: Add dropped message from v3
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
[danvet: Frob patch to apply and use vma->node.size directly as
discused with Ben. Also drop a needles BUG_ON before move_to_inactive,
the function itself has the same check.]
[danvet 2nd: Rebase on top of the lost "drm/i915: Cleanup more of VMA
in destroy", specifically unlink the vma from the mm_list in
vma_unbind (to keep it symmetric with bind_to_vm) instead of
vma_destroy.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

MLC_LLC was never validated for Sandybridge and was superseded by a new
level of cacheing for the GPU in Ivybridge. Update our names to be
consistent with usage, and in the process stop setting the unwanted bit
on Sandybridge.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NVille Syrjälä <ville.syrjala@linux.intel.com>
[danvet: s/BUG/WARN_ON(1) bikeshed.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

To verbalize it, one can say, "pin an object into the given address
space." The semantics of pinning remain the same otherwise.

Certain objects will always have to be bound into the global GTT.
Therefore, global GTT is a special case, and keep a special interface
around for it (i915_gem_obj_ggtt_pin).

v2: s/i915_gem_ggtt_pin/i915_gem_obj_ggtt_pin
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The default context is always supported (as it contains the global
hangcheck stats) and the contexts for hangcheck are not limited
to any ring.

References: https://bugs.freedesktop.org/show_bug.cgi?id=65845Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Soon we want to gut a lot of our existing assumptions how many address
spaces an object can live in, and in doing so, embed the drm_mm_node in
the object (and later the VMA).

It's possible in the future we'll want to add more getter/setter
methods, but for now this is enough to enable the VMAs.

v2: Reworked commit message (Ben)
Added comments to the main functions (Ben)
sed -i "s/i915_gem_obj_set_color/i915_gem_obj_ggtt_set_color/" drivers/gpu/drm/i915/*.[ch]
sed -i "s/i915_gem_obj_bound/i915_gem_obj_ggtt_bound/" drivers/gpu/drm/i915/*.[ch]
sed -i "s/i915_gem_obj_size/i915_gem_obj_ggtt_size/" drivers/gpu/drm/i915/*.[ch]
sed -i "s/i915_gem_obj_offset/i915_gem_obj_ggtt_offset/" drivers/gpu/drm/i915/*.[ch]
(Daniel)

v3: Rebased on new reserve_node patch
Changed DRM_DEBUG_KMS to actually work (will need fixing later)
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

With updates to the spec, we can actually see the context layout, and
how many dwords are allocated. That table suggests we need 70720 bytes
per HW context. Rounded up, this is 18 pages. Looking at what lives
after the current 4 pages we use, I can't see too much important (mostly
it's d3d related), but there are a couple of things which look scary. I
am hopeful this can explain some of our odd HSW failures.

v2: Make the context only 17 pages. The power context space isn't used
ever, and execlists aren't used in our driver, making the actual total
66944 bytes.

v3: Add a comment to the code. (Jesse & Paulo)
Reported-by: N"Azad, Vinit" <vinit.azad@intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: NJesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Only execbuffer needed all the parameters on i915_add_request().
By putting __i915_add_request behind macro, all current callsites
become cleaner. Following patch will introduce a new parameter
for __i915_add_request. With this patch, only the relevant callsite
will reflect the change making commit smaller and easier to understand.

v2: _i915_add_request as function name (Chris Wilson)

v3: change name __i915_add_request and fix ordering of params (Ben Widawsky)
Signed-off-by: NMika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Acked-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

To get context hang statistics for specified context,
add i915_gem_context_get_hang_stats().

For arb-robustness, every context needs to have its own
hang statistics tracking. Added function will return
the user specified context statistics or in case of
default context, statistics from drm_i915_file_private.

v2: handle default context inside get_reset_state

v3: return struct pointer instead of passing it in as param
    (Chris Wilson)
Signed-off-by: NMika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Acked-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Add some debug messages to help figure out what goes wrong on context
initialization.

Later in the PPGTT series, I ended up having a lot of failures after
reset. In many cases it was extra difficult to debug because I hadn't
even realized that contexts failed to reinitialize after reset (again an
artifact of some later patches).

This fairly benign patch does help debug some potential issues which
arise later.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We did not mention the workaround name when implementing those. This
should help us track what we already implement.
Signed-off-by: NDamien Lespiau <damien.lespiau@intel.com>
Reviewed-by: NImre Deak <imre.deak@intel.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Because our context refcounting doesn't grab a ref at lookup time, it is
unsafe to do so without the lock.

NOTE: We don't have an easy way to put the assertion in the lookup
function which is where this really belongs. Context switching is good
enough because it actually asserts even more correctness by protecting
the default_context.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Reviewed-by: NJesse Barnes <jbarnes@virtuousgeek.org>
[danvet: s/BUG/WARN/]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

In order to be notified of when the context and all of its associated
objects is idle (for if the context maps to a ppgtt) we need a callback
from the retire handler. We can arrange this by using the kref_get/put
of the context for request tracking and by inserting a request to
demarque the switch away from the old context.

[Ben: fixed minor error to patch compile, AND s/last_context/from/]
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Before module unload is called, gpu_idle() will switch
to default context. This will increment ref count of base
object as the default context is 'running' on module unload
time. Unreference the drm object so that when context
is freed, base object is freed as well.

v2: added comment to explain the refcounts (Ben Widawsky)
Signed-off-by: NMika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Enabling PPGTT and also the need to track which context was guilty of
gpu hang (arb robustness enabling) have put pressure for struct i915_hw_context
to be more than just a placeholder for hw context state.

In order to track object lifetime properly in a multi peer usage, add reference
counting for i915_hw_context.

v2: track i915_hw_context pointers instead of using ctx_ids
(from Chris Wilson)

v3 (Ben): Get rid of do_release() and handle refcounting more compactly.
(recommended by Chis)

v4: kref_* put inside static inlines (Daniel Vetter)
remove code duplication on freeing context (Chris Wilson)

v5: idr_remove and ctx->file_priv = NULL in destroy ioctl (Chris)
This actually will cause a problem if one destroys a context and later
refers to the idea of the context (multiple contexts may have the same
id, but only 1 will exist in the idr).

v6: Strip out the request related stuff. Reworded commit message.
Got rid of do_destroy and introduced i915_gem_context_release_handle,
suggested by Chris Wilson.

v7: idr_remove can't be called inside idr_for_each (Chris Wilson)

Signed-off-by: Ben Widawsky <ben@bwidawsk.net> (v5)
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com> (v7)
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
[danvet: Squash sob lines, the patch ping-ponged between Ben and Mika
a bit ...]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Enabling context support increases SwapBuffers latency by about 20%
(measured on an i7-3720qm). We can offset that loss slightly by enabling
faster caching for the contexts. As they are not backed by any
particular cache (such as the sampler or render caches) our only option
is to select the generic mid-level cache. This reduces the latency of
the swap by about 5%.

Oddly this effect can be observed running smokin-guns on IVB at
1280x1024:
Using BLT copies for swaps: 151.67 fps
Using Render copies for swaps (unpatched):  141.70 fps
With contexts disabled: 150.23 fps
With contexts in L3$: 150.77 fps
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <ben@bwidawsk.net>
Cc: Kenneth Graunke <kenneth@whitecape.org>
Reviewed-by: NKenneth Graunke <kenneth@whitecape.org>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Convert to the much saner new idr interface.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NDaniel Vetter <daniel.vetter@ffwll.ch>
Acked-by: NDavid Airlie <airlied@linux.ie>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

It's not that the assertion is incorrect, but rather that we can call
do_destroy early in loading, and we will falsely BUG().

Since contexts have been in for a while now, and in the internal APIs
are pretty stable, it should be fairly safe to remove this.

v2: Remove unused dev_priv, and dev
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

This was a rebase error from when the patches originally landed. Since
the context size is unsigned, there is also no use in checking if it's
less than 0.

The existing code is not really wrong, but it's not simple as it should
be.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Based on the work by Mika Kuoppala, we realised that we need to handle
seqno wraparound prior to committing our changes to the ring. The most
obvious point then is to grab the seqno inside intel_ring_begin(), and
then to reuse that seqno for all ring operations until the next request.
As intel_ring_begin() can fail, the callers must already be prepared to
handle such failure and so we can safely add further checks.

This patch looks like it should be split up into the interface
changes and the tweaks to move seqno wrapping from the execbuffer into
the core seqno increment. However, I found no easy way to break it into
incremental steps without introducing further broken behaviour.

v2: Mika found a silly mistake and a subtle error in the existing code;
inside i915_gem_retire_requests() we were resetting the sync_seqno of
the target ring based on the seqno from this ring - which are only
related by the order of their allocation, not retirement. Hence we were
applying the optimisation that the rings were synchronised too early,
fortunately the only real casualty there is the handling of seqno
wrapping.

v3: Do not forget to reset the sync_seqno upon module reinitialisation,
ala resume.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=863861
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com> [v2]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Whoops. This was fixed previously, but not sure how it got lost. It's
not needed for -fixes or stable because at the moment
drm_i915_file_private is way bigger than i915_hw_context (by 120 bytes
on my 64b build).
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

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>

A side-effect of commit 7d54a904
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Fri Aug 10 10:18:10 2012 +0100

    drm/i915: Apply post-sync write for pipe control invalidates

was that only a request to emit invalidate flush would result in the
TLB being invalidated (since it requires synchronisation and so incurs a
performance penalty). However, the stated w/a for hardware contexts is
that the TLBs must be invalidated prior to a MI_SET_CONTEXT, yet the w/a
itself did not request the TLBs to be invalidated...

Note this w/a does not prevent the hard system hang I experience when
using hw contexts (with rc6 enabled) on SNB GT1.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Avoid stalling and waiting for the GPU by checking to see if there is
sufficient inactive space in the aperture for us to bind the buffer
prior to writing through the GTT. If there is inadequate space we will
have to stall waiting for the GPU, and incur overheads moving objects
about. Instead, only incur the clflush overhead on the target object by
writing through shmem.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

the following warning was produced,

drivers/gpu/drm/i915/i915_gem_context.c: In function ‘i915_switch_context’:
drivers/gpu/drm/i915/i915_gem_context.c:454:6: warning: unused variable ‘ret’ [-Wunused-variable]

fix up by removing it
Signed-off-by: NDevendra Naga <devendra.aaru@gmail.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Basic context support on HSW is no different than previous generations.
The size of the context object changes, but that's about it.
Signed-off-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

When bug hunting, I found the interface to do_switch() overly
complicated and I believe festered the earlier bug. This aims to make
the code a little clearer.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

We need to check that "ctx" is a valid pointer before dereferencing it.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Otherwise we end up trying to unpin a freed object and BUG.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

The issue is that we stale data in the CPU caches, when we come to
swap-out the object, the CPU may short-circuit the reads from those
cacheline and so corrupt the context object.

Secondary, leaving the context object as being marked in the CPU write
domain whilst on the GPU active list is a bad idea and will throw
warnings later.

Note: Thanks to calling set_to_gtt_domain with write = false and not
setting any gpu write domain when putting a context object onto the
active list (when we switch away from it) the set_to_gtt_domain call
won't block.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NBen Widawsky <ben@bwidawsk.net>
[danvet: Added a note to the commit message and a comment in the code
to explain the clever non-blocking trick.]
Signed-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>