Due to a coherency issue on BXT A steppings we can't guarantee a
coherent view of cached (CPU snooped) GPU mappings, so fail such
requests. User space is supposed to fall back to uncached mappings in
this case.
v2:
- limit the WA to A steppings, on later stepping this HW issue is fixed
v3:
- return error instead of trying to work around the issue in kernel,
since that could confuse user space (Chris)
Testcast: igt/gem_store_dword_batches_loop/cached-mapping
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This fetches the required firmware image from the filesystem,
then loads it into the GuC's memory via a dedicated DMA engine.
This patch is derived from GuC loading work originally done by
Vinit Azad and Ben Widawsky.
v2:
Various improvements per review comments by Chris Wilson
v3:
Removed 'wait' parameter to intel_guc_ucode_load() as firmware
prefetch is no longer supported in the common firmware loader,
per Daniel Vetter's request.
Firmware checker callback fn now returns errno rather than bool.
v4:
Squash uC-independent code into GuC-specifc loader [Daniel Vetter]
Don't keep the driver working (by falling back to execlist mode)
if GuC firmware loading fails [Daniel Vetter]
v5:
Clarify WOPCM-related #defines [Tom O'Rourke]
Delete obsolete code no longer required with current h/w & f/w
[Tom O'Rourke]
Move the call to intel_guc_ucode_init() later, so that it can
allocate GEM objects, and have it fetch the firmware; then
intel_guc_ucode_load() doesn't need to fetch it later.
[Daniel Vetter].
v6:
Update comment describing intel_guc_ucode_load() [Tom O'Rourke]
Issue: VIZ-4884
Signed-off-by: Alex Dai <yu.dai@intel.com>
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently we don't clflush on pin_to_display if the bo is already
UC/WT and is not in the CPU write domain. This causes problems with
pwrite since pwrite doesn't change the write domain, and it avoids
clflushing on UC/WT buffers on LLC platforms unless the buffer is
currently being scanned out.
Fix the problem by marking the cache dirty and adjusting
i915_gem_object_set_cache_level() to clflush when the cache is dirty
even if the cache_level doesn't change.
My last attempt [1] at fixing this via write domain frobbing was shot
down, but now with the cache_dirty flag we can do things in a nicer way.
[1] http://lists.freedesktop.org/archives/intel-gfx/2014-November/055390.html
v2: Drop the I915_CACHE_NONE/WT checks from pwrite
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=86422
Testcase: igt/kms_pwrite_crc
Testcase: igt/gem_pwrite_snooped
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Similar to commit c44ef60e43 ("drm/i915/gtt:
Allow >= 4GB sizes for vm"), i915_gem_obj_offset and i915_gem_obj_ggtt_offset
return an unsigned long, which in only 4-bytes long in 32-bit kernels.
Change return type (and other related offset variables) to u64.
Since Global GTT is always limited to 4GB, this change would not be required
in i915_gem_obj_ggtt_offset, but this is done for consistency.
v2: Remove unnecessary offset variable in do_pin, as we already have
vma->node.start (Chris).
Update GGTT offset too (Tvrtko).
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In a 48b world, users can try to allocate buffers bigger than 4GB; in
these cases it is important that size is a 64b variable.
v2: Drop the warning about bind with size 0, it shouldn't happen anyway.
Signed-off-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
When we shrink our working sets, we want to avoid stealing pages from
objects that likely to be reused in the near future. We first look at
inactive objects before processing active objects - but what about a
recently active object that is about to be used again. That object's
position in the bound_list is ordered by the time of binding, not the
time of last use, so the most recently used inactive object could well
be at the head of the shrink list. To compensate, give the object a bump
to MRU when it becomes inactive (thus transitioning to the end of the
first pass in shrink lists). Conversely, bumping on inactive makes
bumping on active useless, since when we do have to reap from the active
working set, everything is going to become inactive very quickly and the
order pretty much random - just hope for the best at that point, as once
we start stalling on active objects, we can hope that the rebinding
neatly orders vital objects.
Suggested-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
[danvet: Resolve merge conflict.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
No code changes, just moving all the fence related code into a
separate file (and avoiding a bunch of forward declarations while at
it).
Acked-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
i915_gem_object_create_from_data() is a generic function to save data
from a plain linear buffer in a new pageable gem object that can later
be accessed by the CPU and/or GPU.
We will need this for the microcontroller firmware loading support code.
Derived from i915_gem_object_write(), originally by Alex Dai
v2:
Change of function: now allocates & fills a new object, rather than
writing to an existing object
New name courtesy of Chris Wilson
Explicit domain-setting and other improvements per review comments
by Chris Wilson & Daniel Vetter
v4:
Rebased
Issue: VIZ-4884
Signed-off-by: Alex Dai <yu.dai@intel.com>
Signed-off-by: Dave Gordon <david.s.gordon@intel.com>
Reviewed-by: Tom O'Rourke <Tom.O'Rourke@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Backmerge fixes since it's getting out of hand again with the massive
split due to atomic between -next and 4.2-rc. All the bugfixes in
4.2-rc are addressed already (by converting more towards atomic
instead of minimal duct-tape) so just always pick the version in next
for the conflicts in modeset code.
All the other conflicts are just adjacent lines changed.
Conflicts:
drivers/gpu/drm/i915/i915_drv.h
drivers/gpu/drm/i915/i915_gem_gtt.c
drivers/gpu/drm/i915/intel_display.c
drivers/gpu/drm/i915/intel_drv.h
drivers/gpu/drm/i915/intel_ringbuffer.h
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
After the previous patch this flag will check always clear, as it's
never set for shmem backed and userptr objects, so we can remove it.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Yeah this isn't really fixes but it's a nice cleanup to
clarify the code but not really worth the hassle of backmerging. So
just add to -fixes, we're still early in -rc.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We have 3 types of DMA mappings for GEM objects:
1. physically contiguous for stolen and for objects needing contiguous
memory
2. DMA-buf mappings imported via a DMA-buf attach operation
3. SG DMA mappings for shmem backed and userptr objects
For 1. and 2. the lifetime of the DMA mapping matches the lifetime of the
corresponding backing pages and so in practice we create/release the
mapping in the object's get_pages/put_pages callback.
For 3. the lifetime of the mapping matches that of any existing GPU binding
of the object, so we'll create the mapping when the object is bound to
the first vma and release the mapping when the object is unbound from its
last vma.
Since the object can be bound to multiple vmas, we can end up creating a
new DMA mapping in the 3. case even if the object already had one. This
is not allowed by the DMA API and can lead to leaked mapping data and
IOMMU memory space starvation in certain cases. For example HW IOMMU
drivers (intel_iommu) allocate a new range from their memory space
whenever a mapping is created, silently overriding a pre-existing
mapping.
Fix this by moving the creation/removal of DMA mappings to the object's
get_pages/put_pages callbacks. These callbacks already check for and do
an early return in case of any nested calls. This way objects of the 3.
case also become more like the other object types.
I noticed this issue by enabling DMA debugging, which got disabled after
a while due to its internal mapping tables getting full. It also reported
errors in connection to random other drivers that did a DMA mapping for
an address that was previously mapped by i915 but was never released.
Besides these diagnostic messages and the memory space starvation
problem for IOMMUs, I'm not aware of this causing a real issue.
The fix is based on a patch from Chris.
v2:
- move the DMA mapping create/remove calls to the get_pages/put_pages
callbacks instead of adding new callbacks for these (Chris)
v3:
- also fix the get_page cache logic on the userptr async path (Chris)
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@vger.kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The hang checker needs to inspect whether or not the ring request list is empty
as well as if the given engine has reached or passed the most recently
submitted request. The problem with this is that the hang checker cannot grab
the struct_mutex, which is required in order to safely inspect requests since
requests might be deallocated during inspection. In the past we've had kernel
panics due to this very unsynchronized access in the hang checker.
One solution to this problem is to not inspect the requests directly since
we're only interested in the seqno of the most recently submitted request - not
the request itself. Instead the seqno of the most recently submitted request is
stored separately, which the hang checker then inspects, circumventing the
issue of synchronization from the hang checker entirely.
This fixes a regression introduced in
commit 44cdd6d219
Author: John Harrison <John.C.Harrison@Intel.com>
Date: Mon Nov 24 18:49:40 2014 +0000
drm/i915: Convert 'ring_idle()' to use requests not seqnos
v2 (Chris Wilson):
- Pass current engine seqno to ring_idle() from i915_hangcheck_elapsed() rather
than compute it over again.
- Remove extra whitespace.
Issue: VIZ-5998
Signed-off-by: Tomas Elf <tomas.elf@intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Add regressing commit citation provided by Chris.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This will be useful to PSR and FBC once we start making
dirty fb calls to also flush frontbuffer.
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Pass around requests to carry context deeper in callchain.
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It is found that i915 will not reset gpu under execlist mode when
unload module. that will lead to some issues when unload/load module
with different submission mode. e.g. from execlist mode to ring
buffer mode via loading/unloading i915. Because HW is not in a reset
state and registers are not clean under such condition.
Signed-off-by: Niu,Bing <bing.niu@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Currently only normal views were accounted which under-accounts
the usage as reported in debugfs.
Introduce new helper, i915_gem_obj_total_ggtt_size, and use it
from call sites which want to know how much GGTT space are
objects using.
v2: Single loop in i915_gem_get_aperture_ioctl. (Chris Wilson)
v3: Walk GGTT active/inactive lists in i915_gem_get_aperture_ioctl
for better efficiency. (Chris Wilson, Daniel Vetter)
v4: Make i915_gem_obj_total_ggtt_size private to debugfs. (Chris Wilson)
v5: Change unsigned long to u64. (Chris Wilson)
Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We can't elide the fb tracking invalidate if the buffer is already in
the right domain since that would lead to missed screen updates. I'm
pretty sure I've written this already before but must have gotten lost
unfortunately :(
v2: Chris observed that all internal set_domain users already
correctly do the fb invalidate on their own, hence we can move this
just into the set_domain ioctl instead.
v3: I screwed up setting the invalidate ORIGIN_* correctly (Chris).
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reported-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Pull drm updates from Dave Airlie:
"This is the main drm pull request for v4.2.
I've one other new driver from freescale on my radar, it's been posted
and reviewed, I'd just like to get someone to give it a last look, so
maybe I'll send it or maybe I'll leave it.
There is no major nouveau changes in here, Ben was working on
something big, and we agreed it was a bit late, there wasn't anything
else he considered urgent to merge.
There might be another msm pull for some bits that are waiting on
arm-soc, I'll see how we time it.
This touches some "of" stuff, acks are in place except for the fixes
to the build in various configs,t hat I just applied.
Summary:
New drivers:
- virtio-gpu:
KMS only pieces of driver for virtio-gpu in qemu.
This is just the first part of this driver, enough to run
unaccelerated userspace on. As qemu merges more we'll start
adding the 3D features for the virgl 3d work.
- amdgpu:
a new driver from AMD to driver their newer GPUs. (VI+)
It contains a new cleaner userspace API, and is a clean
break from radeon moving forward, that AMD are going to
concentrate on. It also contains a set of register headers
auto generated from AMD internal database.
core:
- atomic modesetting API completed, enabled by default now.
- Add support for mode_id blob to atomic ioctl to complete interface.
- bunch of Displayport MST fixes
- lots of misc fixes.
panel:
- new simple panels
- fix some long-standing build issues with bridge drivers
radeon:
- VCE1 support
- add a GPU reset counter for userspace
- lots of fixes.
amdkfd:
- H/W debugger support module
- static user-mode queues
- support killing all the waves when a process terminates
- use standard DECLARE_BITMAP
i915:
- Add Broxton support
- S3, rotation support for Skylake
- RPS booting tuning
- CPT modeset sequence fixes
- ns2501 dither support
- enable cmd parser on haswell
- cdclk handling fixes
- gen8 dynamic pte allocation
- lots of atomic conversion work
exynos:
- Add atomic modesetting support
- Add iommu support
- Consolidate drm driver initialization
- and MIC, DECON and MIPI-DSI support for exynos5433
omapdrm:
- atomic modesetting support (fixes lots of things in rewrite)
tegra:
- DP aux transaction fixes
- iommu support fix
msm:
- adreno a306 support
- various dsi bits
- various 64-bit fixes
- NV12MT support
rcar-du:
- atomic and misc fixes
sti:
- fix HDMI timing complaince
tilcdc:
- use drm component API to access tda998x driver
- fix module unloading
qxl:
- stability fixes"
* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (872 commits)
drm/nouveau: Pause between setting gpu to D3hot and cutting the power
drm/dp/mst: close deadlock in connector destruction.
drm: Always enable atomic API
drm/vgem: Set unique to "vgem"
of: fix a build error to of_graph_get_endpoint_by_regs function
drm/dp/mst: take lock around looking up the branch device on hpd irq
drm/dp/mst: make sure mst_primary mstb is valid in work function
of: add EXPORT_SYMBOL for of_graph_get_endpoint_by_regs
ARM: dts: rename the clock of MIPI DSI 'pll_clk' to 'sclk_mipi'
drm/atomic: Don't set crtc_state->enable manually
drm/exynos: dsi: do not set TE GPIO direction by input
drm/exynos: dsi: add support for MIC driver as a bridge
drm/exynos: dsi: add support for Exynos5433
drm/exynos: dsi: make use of array for clock access
drm/exynos: dsi: make use of driver data for static values
drm/exynos: dsi: add macros for register access
drm/exynos: dsi: rename pll_clk to sclk_clk
drm/exynos: mic: add MIC driver
of: add helper for getting endpoint node of specific identifiers
drm/exynos: add Exynos5433 decon driver
...
We can have exactly 4GB sized ppgtt with 32bit system.
size_t is inadequate for this.
v2: Convert a lot more places (Daniel)
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Michel Thierry <michel.thierry@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As there is no OLR to check, the check_olr() function is now a no-op and can be
removed.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In _i915_add_request(), the request is associated with a userland client.
Specifically it is linked to the 'file' structure and the current user process
is recorded. One problem here is that the current user process is not
necessarily the same as when the request was submitted to the driver. This is
especially true when the GPU scheduler arrives and decouples driver submission
from hardware submission. Note also that it is only in the case where the add
request comes from an execbuff call that there is a client to associate. Any
other add request call is kernel only so does not need to do it.
This patch moves the client association into a separate function. This is then
called from the execbuffer code path itself at a sensible time. It also removes
the now redundant 'file' pointer from the add request parameter list.
An extra cleanup of the client association is also added to the request clean up
code for the eventuality where the request is killed after association but
before being submitted (e.g. due to out of memory error somewhere). Once the
submission has happened, the request is on the request list and the regular
request list removal will clear the association. Note that this still needs to
happen at this point in time because the request might be kept floating around
much longer (due to someone holding a reference count) and the client should not
be worrying about this request after it has been retired.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The outstanding_lazy_request is no longer used anywhere in the driver.
Everything that was looking at it now has a request explicitly passed in from on
high. Everything that was relying upon it behind the scenes is now explicitly
creating/passing/submitting its own private request. Thus the OLR can be
removed.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that the *_ring_begin() functions no longer call the request allocation
code, it is finally safe for the request allocation code to call *_ring_begin().
This is important to guarantee that the space reserved for the subsequent
i915_add_request() call does actually get reserved.
v2: Renamed functions according to review feedback (Tomas Elf).
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use requests, intel_ring_begin()
can be updated to take a request instead of a ring. This also means that it no
longer needs to lazily allocate a request if no-one happens to have done it
earlier.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the ring->sync_to() implementations to take a request instead of a ring.
Also updated the tracer to include the request id.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
[danvet: Rebase since I didn't merge the patch which added ->uniq.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the ring->emit_request() implementation to take a request instead of a
ringbuf/request pair. Also removed its use of the OLR for obtaining the
request's seqno.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the various ring->add_request() implementations to take a request
instead of a ring. This removes their reliance on the OLR to obtain the seqno
value that the request should be tagged with.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Updated the *_ring_flush_all_caches() functions to take requests instead of
rings or ringbuf/context pairs.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Converted i915_gem_l3_remap() to take a request structure instead of a ring.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that everything above has been converted to use request structures, it is
possible to update the lower level move_to_active() functions to be request
based as well.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that all callers of i915_add_request() have a request pointer to hand, it is
possible to update the add request function to take a request pointer rather
than pulling it out of the OLR.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The plan is to pass requests around as the basic submission tracking structure
rather than rings and contexts. This patch updates the i915_gem_object_sync()
code path.
v2: Much more complex patch to share a single request between the sync and the
page flip. The _sync() function now supports lazy allocation of the request
structure. That is, if one is passed in then that will be used. If one is not,
then a request will be allocated and passed back out. Note that the _sync() code
does not necessarily require a request. Thus one will only be created until
certain situations. The reason the lazy allocation must be done within the
_sync() code itself is because the decision to need one or not is not really
something that code above can second guess (except in the case where one is
definitely not required because no ring is passed in).
The call chains above _sync() now support passing a request through which most
callers passing in NULL and assuming that no request will be required (because
they also pass in NULL for the ring and therefore can't be generating any ring
code).
The exeception is intel_crtc_page_flip() which now supports having a request
returned from _sync(). If one is, then that request is shared by the page flip
(if the page flip is of a type to need a request). If _sync() does not generate
a request but the page flip does need one, then the page flip path will create
its own request.
v3: Updated comment description to be clearer about 'to_req' parameter (Tomas
Elf review request). Rebased onto newer tree that significantly changed the
synchronisation code.
v4: Updated comments from review feedback (Tomas Elf)
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that the request is guaranteed to specify the context, it is possible to
update the context switch code to use requests rather than ring and context
pairs. This patch updates i915_switch_context() accordingly.
Also removed the warning that the request's context must match the last context
switch's context. As the context switch now gets the context object from the
request structure, there is no longer any scope for the two to become out of
step.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The final step in removing the OLR from i915_gem_init_hw() is to pass the newly
allocated request structure in to each step rather than passing a ring
structure. This patch updates both i915_ppgtt_init_ring() and
i915_gem_context_enable() to take request pointers.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that a single per ring loop is being done for all the different
intialisation steps in i915_gem_init_hw(), it is possible to add proper request
management as well. The last remaining issue is that the context enable call
eventually ends up within *_render_state_init() and this does its own private
_i915_add_request() call.
This patch adds explicit request creation and submission to the top level loop
and removes the add_request() from deep within the sub-functions.
v2: Updated for removal of batch_obj from add_request call in previous patch.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The start of day context initialisation code in i915_gem_context_enable() loops
over each ring and calls the legacy switch context or the execlist init context
code as appropriate.
This patch moves the ring looping out of that function in to the top level
caller i915_gem_init_hw(). This means the a single pass can be made over all
rings doing the PPGTT, L3 remap and context initialisation of each ring
altogether.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The i915_gem_init_hw() function calls a bunch of smaller initialisation
functions. Multiple of which have generic sections and per ring sections. This
means multiple passes are done over the rings. Each pass writes data to the ring
which floats around in that ring's OLR until some random point in the future
when an add_request() is done by some random other piece of code.
This patch breaks i915_ppgtt_init_hw() in two with the per ring initialisation
now being done in i915_ppgtt_init_ring(). The ring looping is now done at the
top level in i915_gem_init_hw().
v2: Fix dumb loop variable re-use.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Added explicit request creation and submission to the GPU idle code path.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In order to explcitly track all GPU work (and completely remove the outstanding
lazy request), it is necessary to add extra i915_add_request() calls to various
places. Some of these do not need the implicit cache flush done as part of the
standard batch buffer submission process.
This patch adds a flag to _add_request() to specify whether the flush is
required or not.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The alloc_request() function does not actually return the newly allocated
request. Instead, it must be pulled from ring->outstanding_lazy_request. This
patch fixes this so that code can create a request and start using it knowing
exactly which request it actually owns.
v2: Updated for new i915_gem_request_alloc() scheme.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In execlist mode, the context object pointer is written in to the request
structure (and reference counted) at the point of request creation. In legacy
mode, this only happens inside i915_add_request().
This patch updates the legacy code path to match the execlist version. This
allows all the intermediate code between request creation and request submission
to get at the context object given only a request structure. Thus negating the
need to pass context pointers here, there and everywhere.
v2: Moved the context reference so it does not need to be undone if the
get_seqno() fails.
v3: Fixed execlist mode always hitting a warning about invalid last_contexts
(which don't exist in execlist mode).
v4: Updated for new i915_gem_request_alloc() scheme.
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The i915_add_request() function is called to keep track of work that has been
written to the ring buffer. It adds epilogue commands to track progress (seqno
updates and such), moves the request structure onto the right list and other
such house keeping tasks. However, the work itself has already been written to
the ring and will get executed whether or not the add request call succeeds. So
no matter what goes wrong, there isn't a whole lot of point in failing the call.
At the moment, this is fine(ish). If the add request does bail early on and not
do the housekeeping, the request will still float around in the
ring->outstanding_lazy_request field and be picked up next time. It means
multiple pieces of work will be tagged as the same request and driver can't
actually wait for the first piece of work until something else has been
submitted. But it all sort of hangs together.
This patch series is all about removing the OLR and guaranteeing that each piece
of work gets its own personal request. That means that there is no more
'hoovering up of forgotten requests'. If the request does not get tracked then
it will be leaked. Thus the add request call _must_ not fail. The previous patch
should have already ensured that it _will_ not fail by removing the potential
for running out of ring space. This patch enforces the rule by actually removing
the early exit paths and the return code.
Note that if something does manage to fail and the epilogue commands don't get
written to the ring, the driver will still hang together. The request will be
added to the tracking lists. And as in the old case, any subsequent work will
generate a new seqno which will suffice for marking the old one as complete.
v2: Improved WARNings (Tomas Elf review request).
For: VIZ-5115
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Reviewed-by: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It is a bad idea for i915_add_request() to fail. The work will already have been
send to the ring and will be processed, but there will not be any tracking or
management of that work.
The only way the add request call can fail is if it can't write its epilogue
commands to the ring (cache flushing, seqno updates, interrupt signalling). The
reasons for that are mostly down to running out of ring buffer space and the
problems associated with trying to get some more. This patch prevents that
situation from happening in the first place.
When a request is created, it marks sufficient space as reserved for the
epilogue commands. Thus guaranteeing that by the time the epilogue is written,
there will be plenty of space for it. Note that a ring_begin() call is required
to actually reserve the space (and do any potential waiting). However, that is
not currently done at request creation time. This is because the ring_begin()
code can allocate a request. Hence calling begin() from the request allocation
code would lead to infinite recursion! Later patches in this series remove the
need for begin() to do the allocate. At that point, it becomes safe for the
allocate to call begin() and really reserve the space.
Until then, there is a potential for insufficient space to be available at the
point of calling i915_add_request(). However, that would only be in the case
where the request was created and immediately submitted without ever calling
ring_begin() and adding any work to that request. Which should never happen. And
even if it does, and if that request happens to fall down the tiny window of
opportunity for failing due to being out of ring space then does it really
matter because the request wasn't doing anything in the first place?
v2: Updated the 'reserved space too small' warning to include the offending
sizes. Added a 'cancel' operation to clean up when a request is abandoned. Added
re-initialisation of tracking state after a buffer wrap to keep the sanity
checks accurate.
v3: Incremented the reserved size to accommodate Ironlake (after finally
managing to run on an ILK system). Also fixed missing wrap code in LRC mode.
v4: Added extra comment and removed duplicate WARN (feedback from Tomas).
For: VIZ-5115
CC: Tomas Elf <tomas.elf@intel.com>
Signed-off-by: John Harrison <John.C.Harrison@Intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Backmerge drm-next because the conflict between Ander's atomic fixes
for 4.2 and Maartens future work are getting to unwielding to handle.
Conflicts:
drivers/gpu/drm/i915/intel_display.c
drivers/gpu/drm/i915/intel_ringbuffer.h
Just always take ours, same as git merge -X ours, but done by hand
because I didn't trust git: It's confusing that it doesn't show any
conflicts in the merge diff at all.
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
This patch doesn't have any functional change, but organize fruntbuffer
invalidate and busy by removing unecesarry signature argument for ring.
It was unsed on mark_fb_busy and only used on fb_obj_invalidate for the
same ORIGIN_CS usage. So let's clean it a bit
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Apparently we can have requests even if though the active list is empty,
so do the request retirement regardless of whether there's anything
on the active list.
The way it happened here is that during suspend intel_ring_idle()
notices the olr hanging around and then proceeds to get rid of it by
adding a request. However since there was nothing on the active lists
i915_gem_retire_requests() didn't clean those up, and so the idle work
never runs, and we leave the GPU "busy" during suspend resulting in a
WARN later.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
With the introduction of multiple views of an obj in the same vm, each
vma was taught to cache its copy of the pages (so that different views
could have different page arrangements). However, this missed decoupling
those vma->ggtt_view.pages when the vma released its reference on the
obj->pages. As we don't always free the vma, this leads to a possible
scenario (e.g. execbuffer interrupted by the shrinker) where the vma
points to a stale obj->pages, and explodes.
Fixes regression from commit fe14d5f4e5
Author: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Date: Wed Dec 10 17:27:58 2014 +0000
drm/i915: Infrastructure for supporting different GGTT views per object
Tvrtko says, if someone else will be confused how this can happen, key
is the reservation execbuffer path. That puts the VMA on the exec_list
which prevents i915_vma_unbind and i915_gem_vma_destroy from fully
destroying the VMA. So the VMA is left existing as an empty object in
the list - unbound and disassociated with the backing store. Kind of a
cached memory object. And then re-using it needs to clear the cached
pages pointer which is fixed above.
Bugzilla: https://bugzilla.redhat.com/show_bug.cgi?id=1227892
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
[Jani: Added Tvrtko's explanation to commit message.]
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Apparently we can have requests even if though the active list is empty,
so do the request retirement regardless of whether there's anything
on the active list.
The way it happened here is that during suspend intel_ring_idle()
notices the olr hanging around and then proceeds to get rid of it by
adding a request. However since there was nothing on the active lists
i915_gem_retire_requests() didn't clean those up, and so the idle work
never runs, and we leave the GPU "busy" during suspend resulting in a
WARN later.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: stable@vger.kernel.org
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
In commit 1854d5ca0d
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Tue Apr 7 16:20:32 2015 +0100
drm/i915: Deminish contribution of wait-boosting from clients
we removed an atomic timer based check for allowing waitboosting and
moved it below the mutex taken during RPS. However, that mutex can be
held for long periods of time on Vallyview/Cherryview as communication
with the PCU is slow. As clients may frequently wait for results (e.g.
such as tranform feedback) we introduced contention between the client
and the RPS worker. We can take advantage of the RPS worker, by
switching the wait boost decision to use spin locks and defer the
actual reclocking to the worker.
Fixes a regression of up to 45% on Baytrail and Baswell!
v2 (Daniel):
- Use max_freq_softlimit instead of the not-yet-merged boost
frequency.
- Don't inject a fake irq into the boost work, instead treat
client_boost as just another legit waker.
v3: Drop the now unused mask (Chris).
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90112
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
