Drivers, especially i915.ko, can fail during the initial migration of a
dma-buf for CPU access. However, the error code from the driver was not
being propagated back to ioctl and so userspace was blissfully ignorant
of the failure. Rendering corruption ensues.
Whilst fixing the ioctl to return the error code from
dma_buf_start_cpu_access(), also do the same for
dma_buf_end_cpu_access(). For most drivers, dma_buf_end_cpu_access()
cannot fail. i915.ko however, as most drivers would, wants to avoid being
uninterruptible (as would be required to guarrantee no failure when
flushing the buffer to the device). As userspace already has to handle
errors from the SYNC_IOCTL, take advantage of this to be able to restart
the syscall across signals.
This fixes a coherency issue for i915.ko as well as reducing the
uninterruptible hold upon its BKL, the struct_mutex.
Fixes commit c11e391da2
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date: Thu Feb 11 20:04:51 2016 -0200
dma-buf: Add ioctls to allow userspace to flush
Testcase: igt/gem_concurrent_blit/*dmabuf*interruptible
Testcase: igt/prime_mmap_coherency/ioctl-errors
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tiago Vignatti <tiago.vignatti@intel.com>
Cc: Stéphane Marchesin <marcheu@chromium.org>
Cc: David Herrmann <dh.herrmann@gmail.com>
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
CC: linux-media@vger.kernel.org
Cc: dri-devel@lists.freedesktop.org
Cc: linaro-mm-sig@lists.linaro.org
Cc: intel-gfx@lists.freedesktop.org
Cc: devel@driverdev.osuosl.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1458331359-2634-1-git-send-email-chris@chris-wilson.co.uk
Userspace is the one in charge of flush CPU by wrapping mmap with
begin{,end}_cpu_access.
v2: Remove LLC check cause we have dma-buf sync providers now. Also, fix return
before transferring ownership when mmap fails.
v3: Fix return values.
v4: !obj->base.filp is user triggerable, so removed the WARN_ON.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1450820214-12509-6-git-send-email-tiago.vignatti@intel.com
This function is meant to be used with dma-buf mmap, when finishing the CPU
access of the mapped pointer.
The error case should be rare to happen though, requiring the buffer become
active during the sync period and for the end_cpu_access to be interrupted. So
we use a uninterruptible mutex_lock to spit out when it ever happens.
v2: disable interruption to make sure errors are reported.
v3: update to the new end_cpu_access API.
v7: use .write = false cause it doesn't need to know whether it's write.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1450820214-12509-5-git-send-email-tiago.vignatti@intel.com
This patch removes range-based information used for optimizations in
begin_cpu_access and end_cpu_access.
We don't have any user nor implementation using range-based flush. It seems a
consensus that if we ever want something like that again (or even more robust
using 2D, 3D sub-range regions) we can use the upcoming dma-buf sync ioctl for
such.
Cc: Sumit Semwal <sumit.semwal@linaro.org>
Cc: Daniel Vetter <daniel.vetter@intel.com>
Signed-off-by: Tiago Vignatti <tiago.vignatti@intel.com>
Reviewed-by: Stéphane Marchesin <marcheu@chromium.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: http://patchwork.freedesktop.org/patch/msgid/1450820214-12509-3-git-send-email-tiago.vignatti@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>
At present, dma_buf_export() takes a series of parameters, which
makes it difficult to add any new parameters for exporters, if required.
Make it simpler by moving all these parameters into a struct, and pass
the struct * as parameter to dma_buf_export().
While at it, unite dma_buf_export_named() with dma_buf_export(), and
change all callers accordingly.
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Reviewed-by: Daniel Thompson <daniel.thompson@linaro.org>
Acked-by: Mauro Carvalho Chehab <mchehab@osg.samsung.com>
Acked-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Sumit Semwal <sumit.semwal@linaro.org>
This allows reservation objects to be used in dma-buf. it's required
for implementing polling support on the fences that belong to a dma-buf.
Signed-off-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Acked-by: Mauro Carvalho Chehab <m.chehab@samsung.com> #drivers/media/v4l2-core/
Acked-by: Thomas Hellstrom <thellstrom@vmware.com> #drivers/gpu/drm/ttm
Acked-by: Sumit Semwal <sumit.semwal@linaro.org>
Acked-by: Daniel Vetter <daniel@ffwll.ch>
Signed-off-by: Vincent Stehlé <vincent.stehle@laposte.net> #drivers/gpu/drm/armada/
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
By exporting the ability to map user address and inserting PTEs
representing their backing pages into the GTT, we can exploit UMA in order
to utilize normal application data as a texture source or even as a
render target (depending upon the capabilities of the chipset). This has
a number of uses, with zero-copy downloads to the GPU and efficient
readback making the intermixed streaming of CPU and GPU operations
fairly efficient. This ability has many widespread implications from
faster rendering of client-side software rasterisers (chromium),
mitigation of stalls due to read back (firefox) and to faster pipelining
of texture data (such as pixel buffer objects in GL or data blobs in CL).
v2: Compile with CONFIG_MMU_NOTIFIER
v3: We can sleep while performing invalidate-range, which we can utilise
to drop our page references prior to the kernel manipulating the vma
(for either discard or cloning) and so protect normal users.
v4: Only run the invalidate notifier if the range intercepts the bo.
v5: Prevent userspace from attempting to GTT mmap non-page aligned buffers
v6: Recheck after reacquire mutex for lost mmu.
v7: Fix implicit padding of ioctl struct by rounding to next 64bit boundary.
v8: Fix rebasing error after forwarding porting the back port.
v9: Limit the userptr to page aligned entries. We now expect userspace
to handle all the offset-in-page adjustments itself.
v10: Prevent vma from being copied across fork to avoid issues with cow.
v11: Drop vma behaviour changes -- locking is nigh on impossible.
Use a worker to load user pages to avoid lock inversions.
v12: Use get_task_mm()/mmput() for correct refcounting of mm.
v13: Use a worker to release the mmu_notifier to avoid lock inversion
v14: Decouple mmu_notifier from struct_mutex using a custom mmu_notifer
with its own locking and tree of objects for each mm/mmu_notifier.
v15: Prevent overlapping userptr objects, and invalidate all objects
within the mmu_notifier range
v16: Fix a typo for iterating over multiple objects in the range and
rearrange error path to destroy the mmu_notifier locklessly.
Also close a race between invalidate_range and the get_pages_worker.
v17: Close a race between get_pages_worker/invalidate_range and fresh
allocations of the same userptr range - and notice that
struct_mutex was presumed to be held when during creation it wasn't.
v18: Sigh. Fix the refactor of st_set_pages() to allocate enough memory
for the struct sg_table and to clear it before reporting an error.
v19: Always error out on read-only userptr requests as we don't have the
hardware infrastructure to support them at the moment.
v20: Refuse to implement read-only support until we have the required
infrastructure - but reserve the bit in flags for future use.
v21: use_mm() is not required for get_user_pages(). It is only meant to
be used to fix up the kernel thread's current->mm for use with
copy_user().
v22: Use sg_alloc_table_from_pages for that chunky feeling
v23: Export a function for sanity checking dma-buf rather than encode
userptr details elsewhere, and clean up comments based on
suggestions by Bradley.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Cc: "Gong, Zhipeng" <zhipeng.gong@intel.com>
Cc: Akash Goel <akash.goel@intel.com>
Cc: "Volkin, Bradley D" <bradley.d.volkin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Reviewed-by: Brad Volkin <bradley.d.volkin@intel.com>
[danvet: Frob ioctl allocation to pick the next one - will cause a bit
of fuss with create2 apparently, but such are the rules.]
[danvet2: oops, forgot to git add after manual patch application]
[danvet3: Appease sparse.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Since dma_buf_vunmap() procedes blithely on ignorant of whether the
driver failed to actually unmap the backing storage for the dma-buf, we
need to make a best-effort to do so. This involves not allowing
ourselves to be susceptible to signals causing us to leak the storage.
This should have been detectable with the current i-g-t as a misplaced
signal should have left the pages pinned upon freeing the object where
we have a warning in place.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@gmail.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
During the vmap() routine for the dma-buf, we first grab the pages and
then try to allocate a temporary array to pass to the vmap(). However,
the shrinker can and will reap any object that is unbound if the
allocation for the array first fails. This includes the object which we
are attempting to vmap(). The solution is to mark the object's pages as
pinned whilst we try the allocation to prevent the use-after-free
introduced by the potential shrinkage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
As we attempt to kmalloc after calling get_pages, there is a possibility
that the shrinker may reap the pages we just acquired. To prevent this
we need to increment the pages_pin_count early, so rearrange the code
and error paths to make it so.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Need to get my stuff out the door ;-) Highlights:
- pc8+ support from Paulo
- more vma patches from Ben.
- Kconfig option to enable preliminary support by default (Josh
Triplett)
- Optimized cpu cache flush handling and support for write-through caching
of display planes on Iris (Chris)
- rc6 tuning from Stéphane Marchesin for more stability
- VECS seqno wrap/semaphores fix (Ben)
- a pile of smaller cleanups and improvements all over
Note that I've ditched Ben's execbuf vma conversion for 3.12 since not yet
ready. But there's still other vma conversion stuff in here.
* tag 'drm-intel-next-2013-08-23' of git://people.freedesktop.org/~danvet/drm-intel: (62 commits)
drm/i915: Print seqnos as unsigned in debugfs
drm/i915: Fix context size calculation on SNB/IVB/VLV
drm/i915: Use POSTING_READ in lcpll code
drm/i915: enable Package C8+ by default
drm/i915: add i915.pc8_timeout function
drm/i915: add i915_pc8_status debugfs file
drm/i915: allow package C8+ states on Haswell (disabled)
drm/i915: fix SDEIMR assertion when disabling LCPLL
drm/i915: grab force_wake when restoring LCPLL
drm/i915: drop WaMbcDriverBootEnable workaround
drm/i915: Cleaning up the relocate entry function
drm/i915: merge HSW and SNB PM irq handlers
drm/i915: fix how we mask PMIMR when adding work to the queue
drm/i915: don't queue PM events we won't process
drm/i915: don't disable/reenable IVB error interrupts when not needed
drm/i915: add dev_priv->pm_irq_mask
drm/i915: don't update GEN6_PMIMR when it's not needed
drm/i915: wrap GEN6_PMIMR changes
drm/i915: wrap GTIMR changes
drm/i915: add the FCLK case to intel_ddi_get_cdclk_freq
...
Makes it more obviously correct what tricks we play by reusing the drm
prime release helper.
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This fixes a WARN in i915_gem_free_object when the
obj->pages_pin_count isn't 0.
v2: Add locking to unmap, noticed by Chris Wilson. Note that even
though we call unmap with our own dev->struct_mutex held that won't
result in an immediate deadlock since we never go through the dma_buf
interfaces for our own, reimported buffers. But it's still easy to
blow up and anger lockdep, but that's already the case with our ->map
implementation. Fixing this for real will involve per dma-buf ww mutex
locking by the callers. And lots of fun. So go with the duct-tape
approach for now.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reported-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Cc: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Tested-by: Armin K. <krejzi@email.com> (v1)
Acked-by: Maarten Lankhorst <maarten.lankhorst@canonical.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Note that this is slightly tricky since both drivers store their
native objects in dma_buf->priv. But both also embed the base
drm_gem_object at the first position, so the implicit cast is ok.
To use the release helper we need to export it, too.
Cc: Inki Dae <inki.dae@samsung.com>
Cc: Intel Graphics Development <intel-gfx@lists.freedesktop.org>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
drm_gem_object_init() and drm_gem_private_object_init() do exactly the
same (except for shmem alloc) so make the first use the latter to reduce
code duplication.
Also drop the return code from drm_gem_private_object_init(). It seems
unlikely that we will extend it any time soon so no reason to keep it
around. This simplifies code paths in drivers, too.
Last but not least, fix gma500 to call drm_gem_object_release() before
freeing objects that were allocated via drm_gem_private_object_init().
That isn't actually necessary for now, but might be in the future.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Patrik Jakobsson <patrik.r.jakobsson@gmail.com>
Acked-by: Rob Clark <robdclark@gmail.com>
Signed-off-by: Dave Airlie <airlied@gmail.com>
In commit be8a42ae60 we inroduced a refcount problem, where on the
drm_gem_prime_fd_to_handle() error path we'll call dma_buf_put() for
self imported dma buffers.
Fix this by taking a reference on the dma buffer in the .gem_import
hook instead of assuming the caller had taken one. Besides fixing the
bug this is also more logical.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
The i915 driver uses sg lists for memory without backing 'struct page'
pages, similarly to other IO memory regions, setting only the DMA
address for these. It does this, so that it can program the HW MMU
tables in a uniform way both for sg lists with and without backing pages.
Without a valid page pointer we can't call nth_page to get the current
page in __sg_page_iter_next, so add a helper that relevant users can
call separately. Also add a helper to get the DMA address of the current
page (idea from Daniel).
Convert all places in i915, to use the new API.
Signed-off-by: Imre Deak <imre.deak@intel.com>
Reviewed-by: Damien Lespiau <damien.lespiau@intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
So far the assumption was that each dma scatter list entry contains only
a single page. This might not hold in the future, when we'll introduce
compact scatter lists, so prepare for this everywhere in the i915 code
where we walk such a list.
We'll fix the place _creating_ these lists separately in the next patch
to help the reviewing/bisectability.
Reference: http://www.spinics.net/lists/dri-devel/msg33917.html
Signed-off-by: Imre Deak <imre.deak@intel.com>
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
Increasing ref counts of both dma-buf and gem for imported dma-buf come from gem
makes memory leak. release function of dma-buf cannot be called because f_count
of dma-buf increased by importing gem and gem ref count cannot be decrease
because of exported dma-buf.
So I add dma_buf_put() for imported gem come from its own gem into each drivers
having prime_import and prime_export capabilities. With this, only gem ref
count is increased if importing gem exported from gem of same driver.
Signed-off-by: Seung-Woo Kim <sw0312.kim@samsung.com>
Signed-off-by: Kyungmin.park <kyungmin.park@samsung.com>
Cc: Inki Dae <inki.dae@samsung.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Rob Clark <rob.clark@linaro.org>
Cc: Alex Deucher <alexander.deucher@amd.com>
Cc: stable <stable@vger.kernel.org>
Signed-off-by: Dave Airlie <airlied@redhat.com>
As pointed out by Seung-Woo Kim this should have been
passing flags like nouveau/radeon have.
Cc: stable@vger.kernel.org
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
The primary purpose of this was to debug some use-after-free memory
corruption that was causing an OOPS inside drm/i915. As it turned out
the corruption was being caused elsewhere and i915.ko as a major user of
many objects was being hit hardest.
Indeed as we do frequent the generic kmalloc caches, dedicating one to
ourselves (or at least naming one for us depending upon the core) aids
debugging our own slab usage.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Jesse Barnes <jbarnes@virtuousgeek.org>
Reviewed-by: Ben Widawsky <ben@bwidawsk.net>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Pull drm merge (part 1) from Dave Airlie:
"So first of all my tree and uapi stuff has a conflict mess, its my
fault as the nouveau stuff didn't hit -next as were trying to rebase
regressions out of it before we merged.
Highlights:
- SH mobile modesetting driver and associated helpers
- some DRM core documentation
- i915 modesetting rework, haswell hdmi, haswell and vlv fixes, write
combined pte writing, ilk rc6 support,
- nouveau: major driver rework into a hw core driver, makes features
like SLI a lot saner to implement,
- psb: add eDP/DP support for Cedarview
- radeon: 2 layer page tables, async VM pte updates, better PLL
selection for > 2 screens, better ACPI interactions
The rest is general grab bag of fixes.
So why part 1? well I have the exynos pull req which came in a bit
late but was waiting for me to do something they shouldn't have and it
looks fairly safe, and David Howells has some more header cleanups
he'd like me to pull, that seem like a good idea, but I'd like to get
this merge out of the way so -next dosen't get blocked."
Tons of conflicts mostly due to silly include line changes, but mostly
mindless. A few other small semantic conflicts too, noted from Dave's
pre-merged branch.
* 'drm-next' of git://people.freedesktop.org/~airlied/linux: (447 commits)
drm/nv98/crypt: fix fuc build with latest envyas
drm/nouveau/devinit: fixup various issues with subdev ctor/init ordering
drm/nv41/vm: fix and enable use of "real" pciegart
drm/nv44/vm: fix and enable use of "real" pciegart
drm/nv04/dmaobj: fixup vm target handling in preparation for nv4x pcie
drm/nouveau: store supported dma mask in vmmgr
drm/nvc0/ibus: initial implementation of subdev
drm/nouveau/therm: add support for fan-control modes
drm/nouveau/hwmon: rename pwm0* to pmw1* to follow hwmon's rules
drm/nouveau/therm: calculate the pwm divisor on nv50+
drm/nouveau/fan: rewrite the fan tachometer driver to get more precision, faster
drm/nouveau/therm: move thermal-related functions to the therm subdev
drm/nouveau/bios: parse the pwm divisor from the perf table
drm/nouveau/therm: use the EXTDEV table to detect i2c monitoring devices
drm/nouveau/therm: rework thermal table parsing
drm/nouveau/gpio: expose the PWM/TOGGLE parameter found in the gpio vbios table
drm/nouveau: fix pm initialization order
drm/nouveau/bios: check that fixed tvdac gpio data is valid before using it
drm/nouveau: log channel debug/error messages from client object rather than drm client
drm/nouveau: have drm debugging macros build on top of core macros
...
Convert #include "..." to #include <path/...> in drivers/gpu/.
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Dave Airlie <airlied@redhat.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Acked-by: Dave Jones <davej@redhat.com>
By providing a callback for when we need to bind the pages, and then
release them again later, we can shorten the amount of time we hold the
foreign pages mapped and pinned, and importantly the dmabuf objects then
behave as any other normal object with respect to the shrinker and
memory management.
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>
Rather than have multiple data structures for describing our page layout
in conjunction with the array of pages, we can migrate all users over to
a scatterlist.
One major advantage, other than unifying the page tracking structures,
this offers is that we replace the vmalloc'ed array (which can be up to
a megabyte in size) with a chain of individual pages which helps reduce
memory pressure.
The disadvantage is that we then do not have a simple array to iterate,
or to access randomly. The common case for this is in the relocation
processing, which will typically fit within a single scatterlist page
and so be almost the same cost as the simple array. For iterating over
the array, the extra function call could be optimised away, but in
reality is an insignificant cost of either binding the pages, or
performing the pwrite/pread.
v2: Fix drm_clflush_sg() to not invoke wbinvd as well! And fix the
trivial compile error from rebasing.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
We need to refcount our pages in order to prevent reaping them at
inopportune times, such as when they currently vmapped or exported to
another driver. However, we also wish to keep the lazy deallocation of
our pages so we need to take a pin/unpinned approach rather than a
simple refcount.
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>
In order to specialise functions depending upon the type of object, we
can attach vfuncs to each object via a new ->ops pointer.
For instance, this will be used in future patches to only bind pages from
a dma-buf for the duration that the object is used by the GPU - and so
prevent them from pinning those pages for the entire of the object.
v2: Bonus comments.
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>
When dealing with a working set larger than the GATT, or even the
mappable aperture when touching through the GTT, we end up with evicting
objects only to rebind them at a new offset again later. Moving an
object into and out of the GTT requires clflushing the pages, thus
causing a double-clflush penalty for rebinding.
To avoid having to clflush on rebinding, we can track the pages as they
are evicted from the GTT and only relinquish those pages on memory
pressure.
As usual, if it were not for the handling of out-of-memory condition and
having to manually shrink our own bo caches, it would be a net reduction
of code. Alas.
Note: The patch also contains a few changes to the last-hope
evict_everything logic in i916_gem_execbuffer.c - we no longer try to
only evict the purgeable stuff in a first try (since that's superflous
and only helps in OOM corner-cases, not fragmented-gtt trashing
situations).
Also, the extraction of the get_pages retry loop from bind_to_gtt (and
other callsites) to get_pages should imo have been a separate patch.
v2: Ditch the newly added put_pages (for unbound objects only) in
i915_gem_reset. A quick irc discussion hasn't revealed any important
reason for this, so if we need this, I'd like to have a git blame'able
explanation for it.
v3: Undo the s/drm_malloc_ab/kmalloc/ in get_pages that Chris noticed.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
[danvet: Split out code movements and rant a bit in the commit message
with a few Notes. Done v2]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
In order for udl vmap to work properly, we need to push the object
into the CPU domain before we start copying the data to the USB device.
This along with the udl change avoids userspace explicit mapping to
be used.
v2: add a flag for userspace to query to know if Intel kernel driver can
deal with the vmap flushing properly. In theory udl would need a flag also,
but I intend to push the patches very close to each other and other drivers
should do the right thing from the start.
I've added a test to my intel-gpu-tools prime branch, however testing
this is a bit messy since the only way to get udl to vmap is to rendering
something. I've tested this with real code as well to make sure it works.
Signed-off-by: Dave Airlie <airlied@redhat.com>
[danvet: resolved conflict, which required reallocating the PARAM
number to 21.]
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
This adds handle->fd and fd->handle support to i915, this is to allow
for offloading of rendering in one direction and outputs in the other.
v2 from Daniel Vetter:
- fixup conflicts with the prepare/finish gtt prep work.
- implement ppgtt binding support.
Note that we have squat i-g-t testcoverage for any of the lifetime and
access rules dma_buf/prime support brings along. And there are quite a
few intricate situations here.
Also note that the integration with the existing code is a bit
hackish, especially around get_gtt_pages and put_gtt_pages. It imo
would be easier with the prep code from Chris Wilson's unbound series,
but that is for 3.6.
Also note that I didn't bother to put the new prepare/finish gtt hooks
to good use by moving the dma_buf_map/unmap_attachment calls in there
(like we've originally planned for).
Last but not least this patch is only compile-tested, but I've changed
very little compared to Dave Airlie's version. So there's a decent
chance v2 on drm-next works as well as v1 on 3.4-rc.
v3: Right when I've hit sent I've noticed that I've screwed up one
obj->sg_list (for dmar support) and obj->sg_table (for prime support)
disdinction. We should be able to merge these 2 paths, but that's
material for another patch.
v4: fix the error reporting bugs pointed out by ickle.
v5: fix another error, and stop non-gtt mmaps on shared objects
stop pread/pwrite on imported objects, add fake kmap
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-Off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Chris Wilson