The command parser scans batch buffers submitted via execbuffer ioctls before
the driver submits them to hardware. At a high level, it looks for several
things:
1) Commands which are explicitly defined as privileged or which should only be
used by the kernel driver. The parser generally rejects such commands, with
the provision that it may allow some from the drm master process.
2) Commands which access registers. To support correct/enhanced userspace
functionality, particularly certain OpenGL extensions, the parser provides a
whitelist of registers which userspace may safely access (for both normal and
drm master processes).
3) Commands which access privileged memory (i.e. GGTT, HWS page, etc). The
parser always rejects such commands.
See the overview comment in the source for more details.
This patch only implements the logic. Subsequent patches will build the tables
that drive the parser.
v2: Don't set the secure bit if the parser succeeds
Fail harder during init
Makefile cleanup
Kerneldoc cleanup
Clarify module param description
Convert ints to bools in a few places
Move client/subclient defs to i915_reg.h
Remove the bits_count field
OTC-Tracker: AXIA-4631
Change-Id: I50b98c71c6655893291c78a2d1b8954577b37a30
Signed-off-by: Brad Volkin <bradley.d.volkin@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
[danvet: Appease checkpatch.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With full ppgtt using acthd is not enough to find guilty
batch buffer. We get multiple false positives as acthd is
per vm.
Instead of scanning which vm was running on a ring,
to find corressponding context, use a different, simpler,
strategy of finding batches that caused gpu hang:
If hangcheck has declared ring to be hung, find first non complete
request on that ring and claim it was guilty.
v2: Rebase
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=73652
Suggested-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net> (v1)
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Ignoring the legacy DRI1 code, and a couple of special cases (to be
discussed later), all access to the ring is mediated through requests.
The first write to a ring will grab a seqno and mark the ring as having
an outstanding_lazy_request. Either through explicitly adding a request
after an execbuffer or through an implicit wait (either by the CPU or by
a semaphore), that sequence of writes will be terminated with a request.
So we can ellide all the intervening writes to the tail register and
send the entire command stream to the GPU at once. This will reduce the
number of *serialising* writes to the tail register by a factor or 3-5
times (depending upon architecture and number of workarounds, context
switches, etc involved). This becomes even more noticeable when the
register write is overloaded with a number of debugging tools. The
astute reader will wonder if it is then possible to overflow the ring
with a single command. It is not. When we start a command sequence to
the ring, we check for available space and issue a wait in case we have
not. The ring wait will in this case be forced to flush the outstanding
register write and then poll the ACTHD for sufficient space to continue.
The exception to the rule where everything is inside a request are a few
initialisation cases where we may want to write GPU commands via the CS
before userspace wakes up and page flips.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Score and action reveals what all the rings were doing
and why hang was declared. Add idle state so that
we can distinguish between waiting and idle ring.
v2: - add idle as a hangcheck action
- consensed hangcheck status to single line (Chris)
- mark active explicitly when we are making progress (Chris)
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It is possible for us to be forced to perform an allocation for the lazy
request whilst running the shrinker. This allocation may fail, leaving
us unable to reclaim any memory leading to premature OOM. A neat
solution to the problem is to preallocate the request at the same time
as acquiring the seqno for the ring transaction. This means that we can
report ENOMEM prior to touching the rings.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Prior to preallocating an request for lazy emission, rename the existing
field to make way (and differentiate the seqno from the request struct).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We now have more devices using ring->private than not, and they all want
the same structure. Worse, I would like to use a scratch page from
outside of intel_ringbuffer.c and so for convenience would like to reuse
ring->private. Embed the object into the struct intel_ringbuffer so that
we can keep the code clean.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The code directly uses the registers and ring->mmio_base.
Signed-off-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The short lowercase names are bound to collide. The default warnings
don't even warn about shadowing.
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With the simplified locking there's no reason any more to keep the
refcounts seperate.
v2: Readd the lost comment that ring->irq_refcount is protected by
dev_priv->irq_lock.
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Now that the rps interrupt locking isn't clearly separated (at elast
conceptually) from all the other interrupt locking having a different
lock stopped making sense: It protects much more than just the rps
workqueue it started out with. But with the addition of VECS the
separation started to blurr and resulted in some more complex locking
for the ring interrupt refcount.
With this we can (again) unifiy the ringbuffer irq refcounts without
causing a massive confusion, but that's for the next patch.
v2: Explain better why the rps.lock once made sense and why no longer,
requested by Ben.
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
For guilty batchbuffer analysis later on when rings are reset,
store what state the ring was on when hang was declared.
This helps to weed out the waiting rings from the active ones.
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If we detect a ring is in a valid wait for another, just let it be.
Eventually it will either begin to progress again, or the entire system
will come grinding to a halt and then hangcheck will fire as soon as the
deadlock is detected.
This error was foretold by Ben in
commit 05407ff889
Author: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Date: Thu May 30 09:04:29 2013 +0300
drm/i915: detect hang using per ring hangcheck_score
"If ring B is waiting on ring A via semaphore, and ring A is making
progress, albeit slowly - the hangcheck will fire. The check will
determine that A is moving, however ring B will appear hung because
the ACTHD doesn't move. I honestly can't say if that's actually a
realistic problem to hit it probably implies the timeout value is too
low."
v2: Make sure we don't even incur the KICK cost whilst waiting.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=65394
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is required for tracking render damage for use with FBC and will be
used in subsequent patches.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Keep track of ring seqno progress and if there are no
progress detected, declare hang. Use actual head (acthd)
to distinguish between ring stuck and batchbuffer looping
situation. Stuck ring will be kicked to trigger progress.
This commit adds a hard limit for batchbuffer completion time.
If batchbuffer completion time is more than 4.5 seconds,
the gpu will be declared hung.
Review comment from Ben which nicely clarifies the semantic change:
"Maybe I'm just stating the functional changes of the patch, but in case
they were unintended here is what I see as potential issues:
1. "If ring B is waiting on ring A via semaphore, and ring A is making
progress, albeit slowly - the hangcheck will fire. The check will
determine that A is moving, however ring B will appear hung because
the ACTHD doesn't move. I honestly can't say if that's actually a
realistic problem to hit it probably implies the timeout value is too
low.
2. "There's also another corner case on the kick. If the seqno = 2
(though not stuck), and on the 3rd hangcheck, the ring is stuck, and
we try to kick it... we don't actually try to find out if the kick
helped"
v2: use atchd to detect stuck ring from loop (Ben Widawsky)
v3: Use acthd to check when ring needs kicking.
Declare hang on third time in order to give time for
kick_ring to take effect.
v4: Update commit msg
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: Paste in Ben's review comment.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
v2: Use the correct lock to protect PM interrupt regs, this was
accidentally lost from earlier (Haihao)
Fix return types (Ben)
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
It's overkill on older gens, but it's useful for newer gens.
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
v2: Add set_seqno which didn't exist before rebase (Haihao)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Xiang, Haihao <haihao.xiang@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The video enhancement command streamer is a new ring on HSW which does
what it sounds like it does. This patch provides the most minimal
inception of the ring.
In order to support a new ring, we need to bump the number. The patch
may look trivial to the untrained eye, but bumping the number of rings
is a bit scary. As such the patch is not terribly useful by itself, but
a pretty nice place to find issues during a bisection.
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This replaces the existing MBOX update code with a more generalized
calculation for emitting mbox updates. We also create a sentinel for
doing the updates so we can more abstractly deal with the rings.
When doing MBOX updates the code must be aware of the /other/ rings.
Until now the platforms which supported semaphores had a fixed number of
rings and so it made sense for the code to be very specialized
(hardcoded).
The patch does contain a functional change, but should have no
behavioral changes.
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Semaphores are tied very closely to the rings in the GPU. Trivial patch
adds comments to the existing code so that when we add new rings we can
include comments there as well. It also helps distinguish the ring to
semaphore mailbox interactions by using the ringname in the semaphore
data structures.
This patch should have no functional impact.
v2: The English parts (as opposed to register names) of the comments
were reversed. (Damien)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Instead of relying in acthd, track ring seqno progression
to detect if ring has hung.
v2: put hangcheck stuff inside struct (Chris Wilson)
v3: initialize hangcheck.seqno (Ben Widawsky)
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel 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: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Daniel writes:
- seqno wrap fixes and debug infrastructure from Mika Kuoppala and Chris
Wilson
- some leftover kill-agp on gen6+ patches from Ben
- hotplug improvements from Damien
- clear fb when allocated from stolen, avoids dirt on the fbcon (Chris)
- Stolen mem support from Chris Wilson, one of the many steps to get to
real fastboot support.
- Some DDI code cleanups from Paulo.
- Some refactorings around lvds and dp code.
- some random little bits&pieces
* tag 'drm-intel-next-2012-12-21' of git://people.freedesktop.org/~danvet/drm-intel: (93 commits)
drm/i915: Return the real error code from intel_set_mode()
drm/i915: Make GSM void
drm/i915: Move GSM mapping into dev_priv
drm/i915: Move even more gtt code to i915_gem_gtt
drm/i915: Make next_seqno debugs entry to use i915_gem_set_seqno
drm/i915: Introduce i915_gem_set_seqno()
drm/i915: Always clear semaphore mboxes on seqno wrap
drm/i915: Initialize hardware semaphore state on ring init
drm/i915: Introduce ring set_seqno
drm/i915: Missed conversion to gtt_pte_t
drm/i915: Bug on unsupported swizzled platforms
drm/i915: BUG() if fences are used on unsupported platform
drm/i915: fixup overlay stolen memory leak
drm/i915: clean up PIPECONF bpc #defines
drm/i915: add intel_dp_set_signal_levels
drm/i915: remove leftover display.update_wm assignment
drm/i915: check for the PCH when setting pch_transcoder
drm/i915: Clear the stolen fb before enabling
drm/i915: Access to snooped system memory through the GTT is incoherent
drm/i915: Remove stale comment about intel_dp_detect()
...
Conflicts:
drivers/gpu/drm/i915/intel_display.c
Hardware status page needs to have proper seqno set
as our initial seqno can be arbitrary. If initial seqno is close
to wrap boundary on init and i915_seqno_passed() (31bit space)
refers to hw status page which contains zero, errorneous result
will be returned.
v2: clear mboxes and set hws page directly instead of going
through rings. Suggested by Chris Wilson.
v3: hws needs to be updated for all gens. Noticed by Chris
Wilson.
References: https://bugs.freedesktop.org/show_bug.cgi?id=58230
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>
In preparation for setting per ring initial seqno values
add ring::set_seqno().
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>
Now that Chris Wilson demonstrated that the key for stability on early
gen 2 is to simple _never_ exchange the physical backing storage of
batch buffers I've tried a stab at a kernel solution. Doesn't look too
nefarious imho, now that I don't try to be too clever for my own good
any more.
v2: After discussing the various techniques, we've decided to always blit
batches on the suspect devices, but allow userspace to opt out of the
kernel workaround assume full responsibility for providing coherent
batches. The principal reason is that avoiding the blit does improve
performance in a few key microbenchmarks and also in cairo-trace
replays.
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet:
- Drop the hunk which uses HAS_BROKEN_CS_TLB to implement the ring
wrap w/a. Suggested by Chris Wilson.
- Also add the ACTHD check from Chris Wilson for the error state
dumping, so that we still catch batches when userspace opts out of
the w/a.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
If there are pre-wrap values in semaphore-mbox registers after wrap,
syncing against some after-wrap request will complete immediately.
Fix this by emitting ring commands to set mbox registers to zero
when the wrap happens.
v2: Use __intel_ring_begin to emit ring commands, from
Chris Wilson.
Signed-off-by: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Add a small comment to handle_seqno_wrap.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
From BSpec:
"If the Ring Buffer Head Pointer and the Tail Pointer are on the same
cacheline, the Head Pointer must not be greater than the Tail
Pointer."
The easiest way to enforce this is to reduce the reported ring space.
References:
Gen2 BSpec "1. Programming Environment" / 1.4.4.6 "Ring Buffer Use"
Gen3 BSpec "vol1c Memory Interface Functions" / 2.3.4.5 "Ring Buffer Use"
Gen4+ BSpec "vol1c Memory Interface and Command Stream" / 5.3.4.5 "Ring Buffer Use"
v2: Include the exact BSpec references in the description
v3: s/64/I915_RING_FREE_SPACE, and add the BSpec information to the code
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>
Replace the wait for the ring to be clear with the more common wait for
the ring to be idle. The principle advantage is one less exported
intel_ring_wait function, and the removal of a hardcoded value.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
Signed-off-by: Daniel 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: Chris 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: Daniel Vetter <daniel.vetter@ffwll.ch>
So store into the scratch space of the HWS to make sure the invalidate
occurs.
v2: use GTT address space for store, clean up #defines (Chris)
v3: use correct #define in blt ring flush (Chris)
Signed-off-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Reviewed-by: Antti Koskipää <antti.koskipaa@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
References: https://bugs.launchpad.net/ubuntu/+source/xserver-xorg-video-intel/+bug/1063252
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
With the introduction of per-process GTT space, the hardware designers
thought it wise to also limit the ability to write to MMIO space to only
a "secure" batch buffer. The ability to rewrite registers is the only
way to program the hardware to perform certain operations like scanline
waits (required for tear-free windowed updates). So we either have a
choice of adding an interface to perform those synchronized updates
inside the kernel, or we permit certain processes the ability to write
to the "safe" registers from within its command stream. This patch
exposes the ability to submit a SECURE batch buffer to
DRM_ROOT_ONLY|DRM_MASTER processes.
v2: Haswell split up bit8 into a ppgtt bit (still bit8) and a security
bit (bit 13, accidentally not set). Also add a comment explaining why
secure batches need a global gtt binding.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> (v1)
[danvet: added hsw fixup.]
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Avoid the forcewake overhead when simply retiring requests, as often the
last seen seqno is good enough to satisfy the retirment process and will
be promptly re-run in any case. Only ensure that we force the coherent
seqno read when we are explicitly waiting upon a completion event to be
sure that none go missing, and also for when we are reporting seqno
values in case of error or debugging.
This greatly reduces the load for userspace using the busy-ioctl to
track active buffers, for instance halving the CPU used by X in pushing
the pixels from a software render (flash). The effect will be even more
magnified with userptr and so providing a zero-copy upload path in that
instance, or in similar instances where X is simply compositing DRI
buffers.
v2: Reverse the polarity of the tachyon stream. Daniel suggested that
'force' was too generic for the parameter name and that 'lazy_coherency'
better encapsulated the semantics of it being an optimization and its
purpose. Also notice that gen6_get_seqno() is only used by gen6/7
chipsets and so the test for IS_GEN6 || IS_GEN7 is redundant in that
function.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By moving the function to intel_ringbuffer and currying the appropriate
parameter, hopefully we make the callsites easier to read and
understand.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is now handled by a global flag to ensure we emit a flush before
the next serialisation point (if we failed to queue one previously).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This is just the minimal patch to disable all this code so that we can
do decent amounts of QA before we rip it all out.
The complicating thing is that we need to flush the gpu caches after
the batchbuffer is emitted. Which is past the point of no return where
execbuffer can't fail any more (otherwise we risk submitting the same
batch multiple times).
Hence we need to add a flag to track whether any caches associated
with that ring are dirty. And emit the flush in add_request if that's
the case.
Note that this has a quite a few behaviour changes:
- Caches get flushed/invalidated unconditionally.
- Invalidation now happens after potential inter-ring sync.
I've bantered around a bit with Chris on irc whether this fixes
anything, and it might or might not. The only thing clear is that with
these changes it's much easier to reason about correctness.
Also rip out a lone get_next_request_seqno in the execbuffer
retire_commands function. I've dug around and I couldn't figure out
why that is still there, with the outstanding lazy request stuff it
shouldn't be necessary.
v2: Chris Wilson complained that I also invalidate the read caches
when flushing after a batchbuffer. Now optimized.
v3: Added some comments to explain the new flushing behaviour.
Cc: Eric Anholt <eric@anholt.net>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
From http://intellinuxgraphics.org/documentation/SNB/IHD_OS_Vol1_Part3.pdf
[DevSNB] If Flush TLB invalidation Mode is enabled it's the driver's
responsibility to invalidate the TLBs at least once after the previous
context switch after any GTT mappings changed (including new GTT
entries). This can be done by a pipelined PIPE_CONTROL with TLB inv bit
set immediately before MI_SET_CONTEXT.
On GEN7 the invalidation mode is explicitly set, but this appears to be
lacking for GEN6. Since I don't know the history on this, I've decided
to dynamically read the value at ring init time, and use that value
throughout.
v2: better comment (daniel)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Implement the context switch code as well as the interfaces to do the
context switch. This patch also doesn't match 1:1 with the RFC patches.
The main difference is that from Daniel's responses the last context
object is now stored instead of the last context. This aids in allows us
to free the context data structure, and context object independently.
There is room for optimization: this code will pin the context object
until the next context is active. The optimal way to do it is to
actually pin the object, move it to the active list, do the context
switch, and then unpin it. This allows the eviction code to actually
evict the context object if needed.
The context switch code is missing workarounds, they will be implemented
in future patches.
v2: actually do obj->dirty=1 in switch (daniel)
Modified comment around above
Remove flags to context switch (daniel)
Move mi_set_context code to i915_gem_context.c (daniel)
Remove seqno , use lazy request instead (daniel)
v3: use i915_gem_request_next_seqno instead of
outstanding_lazy_request (Daniel)
remove id's from trace events (Daniel)
Put the context BO in the instruction domain (Daniel)
Don't unref the BO is context switch fails (Chris)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
Invent an abstraction for a hw context which is passed around through
the core functions. The main bit a hw context holds is the buffer object
which backs the context. The rest of the members are just helper
functions. Specifically the ring member, which could likely go away if
we decide to never implement whatever other hw context support exists.
Of note here is the introduction of the 64k alignment constraint for the
BO. If contexts become heavily used, we should consider tweaking this
down to 4k. Until the contexts are merged and tested a bit though, I
think 64k is a nice start (based on docs).
Since we don't yet switch contexts, there is really not much complexity
here. Creation/destruction works pretty much as one would expect. An idr
is used to generate the context id numbers which are unique per file
descriptor.
v2: add DRM_DEBUG_DRIVERS to distinguish ENOMEM failures (ben)
convert a BUG_ON to WARN_ON, default destruction is still fatal (ben)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
In many places we wish to iterate over the rings associated with the
GPU, so refactor them to use a common macro.
Along the way, there are a few code removals that should be side-effect
free and some rearrangement which should only have a cosmetic impact,
such as error-state.
Note that this slightly changes the semantics in the hangcheck code:
We now always cycle through all enabled rings instead of
short-circuiting the logic.
v2: Pull in a couple of suggestions from Ben and Daniel for
intel_ring_initialized() and not removing the warning (just moving them
to a new home, closer to the error).
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: Added note to commit message about the small behaviour
change, suggested by Ben Widawsky.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Two things:
- ring->virtual start is an __iomem pointer, treat it accordingly.
- dev_priv->status_page.page_addr is now always a cpu addr, no pointer
casting needed for that.
Take the opportunity to remove the unnecessary drm indirection when
setting up the ringbuffer iomapping.
v2: Add a compiler barrier before reading the hw status page.
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Wohoo!
Now we only need to move all the gem/kms stuff that accidentally
landed in i915_dma.c out of it, and this will be our legacy dri1
grave-yard.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The waiting_seqno is not terribly useful, and as such we can remove it
so that we'll be able to extract lockless code.
v2: Keep the information for error_state (Chris)
Check if ring is initialized in hangcheck (Chris)
Capture the waiting ring (Chris)
Signed-off-by: Ben Widawsky <ben@bwidawsk.net>
[danvet: add some bikeshed to clarify a comment.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We were attempting to use a per-ring spinlock whilst modifying global
IRQ flags. A recipe for rare missed interrupts.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Acked-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We only ever enable/disable one interrupt (namely user_interrupts and
pipe_notify), so we don't need to track the interrupt masking state.
Also rename irq_enable to irq_enable_mask, now that it won't collide -
beforehand both a irq_mask and irq_enable_mask would have looked a bit
strange.
Reviewed-by: Eric Anholt <eric@anholt.net>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We can now open-code the get/put irq functions as they were just
abstracting single register definitions.
It would be nice to merge this in with the IRQ handling code... but that
is too much work for me at present. In addition I could probably
collapse this in to a lot of the Ironlake stuff, but I don't think it's
worth the potential regressions.
This patch itself should not effect functionality.
CC: Jesse Barnes <jbarnes@virtuousgeek.org>
Signed-off-by: Ben Widawsky <benjamin.widawsky@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
By recording the location of every request in the ringbuffer, we know
that in order to retire the request the GPU must have finished reading
it and so the GPU head is now beyond the tail of the request. We can
therefore provide a conservative estimate of where the GPU is reading
from in order to avoid having to read back the ring buffer registers
when polling for space upon starting a new write into the ringbuffer.
A secondary effect is that this allows us to convert
intel_ring_buffer_wait() to use i915_wait_request() and so consolidate
upon the single function to handle the complicated task of waiting upon
the GPU. A necessary precaution is that we need to make that wait
uninterruptible to match the existing conditions as all the callers of
intel_ring_begin() have not been audited to handle ERESTARTSYS
correctly.
By using a conservative estimate for the head, and always processing all
outstanding requests first, we prevent a race condition between using
the estimate and direct reads of I915_RING_HEAD which could result in
the value of the head going backwards, and the tail overflowing once
again. We are also careful to mark any request that we skip over in
order to free space in ring as consumed which provides a
self-consistency check.
Given sufficient abuse, such as a set of unthrottled GPU bound
cairo-traces, avoiding the use of I915_RING_HEAD gives a 10-20% boost on
Sandy Bridge (i5-2520m):
firefox-paintball 18927ms -> 15646ms: 1.21x speedup
firefox-fishtank 12563ms -> 11278ms: 1.11x speedup
which is a mild consolation for the performance those traces achieved from
exploiting the buggy autoreported head.
v2: Add a few more comments and make request->tail a conservative
estimate as suggested by Daniel Vetter.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: resolve conflicts with retirement defering and the lack of
the autoreport head removal (that will go in through -fixes).]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
... and add a helpr function for the places where we want a flag.
This way we can use ring->id to index into arrays.
v2: Resurrect the missing beautification-space Chris Wilson noted.
I'm moving this space around because I'll reuse ring_str in the next
patch.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Reviewed-by: Eugeni Dodonov <eugeni.dodonov@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
