UAPI Changes:
- drm/i915: Show per-engine default property values in sysfs
By providing the default values configured into the kernel via sysfs, it
is much more convenient for userspace to restore those sane defaults, or
at least know what are considered good baseline. This is useful, for
example, to cleanup after any failed userspace prior to commencing new
jobs.
Cross-subsystem Changes:
- video/hdmi: Add Unpack only function for DRM infoframe
- Includes pull request gvt-next-2020-05-12
Driver Changes:
- Restore Cherryview back to full-ppgtt (Chris, Mika)
- Document locking guidelines for i915 (Chris, Daniel, Joonas)
- Fix GitLab #1746: Handle idling during i915_gem_evict_something busy loops (Chris)
- Display WA #1105: Require linear fb stride to be multiple of 512 bytes on
gen9/glk (Ville)
- Add Wa_14010685332 for ICP/ICL (Matt R)
- Restrict w/a 1607087056 for EHL/JSL (Swathi)
- Fix interrupt handling for DP AUX transactions on Tigerlake (Imre)
- Revert "drm/i915/tgl: Include ro parts of l3 to invalidate" (Mika)
- Fix HDC pipeline flush hardware bit on Gen12 (Mika)
- Flush L3 when flushing render on Gen12 (Mika)
- Invalidate aux table entries forcibly between BB on Gen12 (Mika)
- Add aux table invalidate for all engines on Gen12 (Mika)
- Force pte cacheline to main memory Gen8+ (Mika)
- Add and enable TGL+ SAGV support (Stanislav)
- Implement vm_ops->access on i915 mmaps for GDB (Chris, Kristian)
- Replace zero-length array with flexible-array (Gustavo)
- Improve batch buffer pool effectiveness to mitigate soft-rc6 hit (Chris)
- Remove wait priority boosting (Chris)
- Keep driver module referenced when PMU is active (Chris)
- Sanitize RPS interrupts upon resume (Chris)
- Extend pcode read timeout to 20 ms (Chris)
- Wait for ACT sent before enabling MST pipe (Ville)
- Extend support to async relocations to SNB (Chris)
- Remove CNL pre-prod workarounds (Ville)
- Don't enable WaIncreaseLatencyIPCEnabled when IPC is disabled (Sultan)
- Record the active CCID from before reset (Chris)
- Mark concurrent submissions with a weak-dependency (Chris)
- Peel dma-fence-chains for await to allow engine-to-engine sync (Lionel)
- Prevent using semaphores to chain up to external fences (Chris)
- Fix GLK watermark calculations (Ville)
- Emit await(batch) before MI_BB_START (Chris)
- Reset execlists registers before HWSP (Chris)
- Drop no-semaphore boosting in favor of fast timeslicing (Chris)
- Fix enabled infoframe states of lspcon (Gwan-gyeong)
- Program DP SDPs on pipe updates (Gwan-gyeong)
- Stop sending DP SDPs on ddi disable (Gwan-gyeong)
- Store CS timestamp frequency in Hz (Ville)
- Remove unused HAS_FWTABLE macro (Pascal)
- Use batchbuffer chaining for relocations to save ring space (Chris)
- Try different engines for relocs if MI ops not supported (Chris, Tvrtko)
- Lazily acquire the device wakeref for freeing objects (Chris)
- Streamline display code arithmetics around rounding etc. (Ville)
- Use bw state for per crtc SAGV evaluation (Stanislav)
- Track active_pipes in bw_state (Stanislav)
- Nuke mode.vrefresh usage (Ville)
- Warn if the FBC is still writing to stolen on removal (Chris)
- Added new PCode commands prepping for QGV rescricting (Stansilav)
- Stop holding onto the pinned_default_state (Chris)
- Propagate error from completed fences (Chris)
- Ignore submit-fences on the same timeline (Chris)
- Pull waiting on an external dma-fence into its routine (Chris)
- Replace the hardcoded I915_FENCE_TIMEOUT with Kconfig (Chris)
- Mark up the racy read of execlists->context_tag (Chris)
- Tidy up the return handling for completed dma-fences (Chris)
- Introduce skl_plane_wm_level accessor (Stanislav)
- Extract SKL SAGV checking (Stanislav)
- Make active_pipes check skl specific (Stanislav)
- Suspend tasklets before resume sanitization (Chris)
- Remove redundant exec_fence (Chris)
- Mark the addition of the initial-breadcrumb in the request (Chris)
- Transfer old virtual breadcrumbs to irq_worker (Chris)
- Read the DP SDPs from the video DIP (Gwan-gyeong)
- Program DP SDPs with computed configs (Gwan-gyeong)
- Add state readout for DP VSC and DP HDR Metadata Infoframe SDP
(Gwan-gyeong)
- Add compute routine for DP PSR VSC SDP (Gwan-gyeong)
- Use new DP VSC SDP compute routine on PSR (Gwan-gyeong)
- Restrict qgv points which don't have enough bandwidth. (Stanislav)
- Nuke pointless div by 64bit (Ville)
- Static checker code fixes (Nathan, Mika, Chris)
- Add logging function for DP VSC SDP (Gwan-gyeong)
- Include HDMI DRM infoframe, DP HDR metadata and DP VSC SDP in the
crtc state dump (Gwan-gyeong)
- Make timeslicing explicit engine property (Chris, Tvrtko)
- Selftest and debugging improvements (Chris)
- Align variable names with BSpec (Ville)
- Tidy up gen8+ breadcrumb emission code (Chris)
- Turn intel_digital_port_connected() in a vfunc (Ville)
- Use stashed away hpd isr bits in intel_digital_port_connected() (Ville)
- Extract i915_cs_timestamp_{ns_to_ticks,tick_to_ns}() (Ville)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200515160703.GA19043@jlahtine-desk.ger.corp.intel.com
Now that we have fast timeslicing on semaphores, we no longer need to
prioritise none-semaphore work as we will yield any work blocked on a
semaphore to the next in the queue. Previously with no timeslicing,
blocking on the semaphore caused extremely bad scheduling with multiple
clients utilising multiple rings. Now, there is no impact and we can
remove the complication.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200513173504.28322-1-chris@chris-wilson.co.uk
Driver Changes:
- Fix GitLab #1698: Performance regression with Linux 5.7-rc1 on
Iris Plus 655 and 4K screen (Chris)
- Add Wa_14011059788 for Tigerlake (Matt A)
- Add per ctx batchbuffer wa for timestamp for Gen12 (Mika)
- Use indirect ctx bb to load cmd buffer control value
from context image to avoid corruption (Mika)
- Enable DP Display Audio WA (Uma, Jani)
- Update forcewake firmware ranges for Icelake (Radhakrishna)
- Add missing deinitialization cases of load failure for display (Jose)
- Implement TC cold sequences for Icelake and Tigerlake (Jose)
- Unbreak enable_dpcd_backlight modparam (Lyude)
- Move the late flush_submission in retire to the end (Chris)
- Demote "Reducing compressed framebufer size" message to info (Peter)
- Push MST link retraining to the hotplug work (Ville)
- Hold obj->vma.lock over for_each_ggtt_vma() (Chris)
- Fix timeout handling during TypeC AUX power well enabling for ICL (Imre)
- Fix skl+ non-scaled pfit modes (Ville)
- Prefer soft-rc6 over RPS DOWN_TIMEOUT (Chris)
- Sanitize GT first before poisoning HWSP (Chris)
- Fix up clock RPS frequency readout (Chris)
- Avoid reusing the same logical CCID (Chris)
- Avoid dereferencing a dead context (Chris)
- Always enable busy-stats for execlists (Chris)
- Apply the aggressive downclocking to parking (Chris)
- Restore aggressive post-boost downclocking (Chris)
- Scrub execlists state on resume (Chris)
- Add debugfs attributes for LPSP (Ansuman)
- Improvements to kernel selftests (Chris, Mika)
- Add tiled blits selftest (Zbigniew)
- Fix error handling in __live_lrc_indirect_ctx_bb() (Dan)
- Add pre/post plane updates for SAGV (Stanislav)
- Add ICL PG3 PW ID for EHL (Anshuman)
- Fix Sphinx build duplicate label warning (Jani)
- Error log non-zero audio power refcount after unbind (Jani)
- Remove object_is_locked assertion from unpin_from_display_plane (Chris)
- Use single set of AUX powerwell ops for gen11+ (Matt R)
- Prefer drm_WARN_ON over WARN_ON (Pankaj)
- Poison residual state [HWSP] across resume (Chris, Tvrtko)
- Convert request-before-CS assertion to debug (Chris)
- Carefully order virtual_submission_tasklet (Chris)
- Check carefully for an idle engine in wait-for-idle (Chris)
- Only close vma we open (Chris)
- Trace RPS events (Chris)
- Use the RPM config register to determine clk frequencies (Chris)
- Drop rq->ring->vma peeking from error capture (Chris)
- Check preempt-timeout target before submit_ports (Chris)
- Check HWSP cacheline is valid before acquiring (Chris)
- Use proper fault mask in interrupt postinstall too (Matt R)
- Keep a no-frills swappable copy of the default context state (Chris)
- Add atomic helpers for bandwidth (Stanislav)
- Refactor setting dma info to a common helper from device info (Michael)
- Refactor DDI transcoder code for clairty (Ville)
- Extend PG3 power well ID to ICL (Anshuman)
- Refactor PFIT code for readability and future extensibility (Ville)
- Clarify code split between intel_ddi.c and intel_dp.c (Ville)
- Move out code to return the digital_port of the aux ch (Jose)
- Move rps.enabled/active and use of RPS interrupts to flags (Chris)
- Remove superfluous inlines and dead code (Jani)
- Re-disable -Wframe-address from top-level Makefile (Nick)
- Static checker and spelling fixes (Colin, Nathan)
- Split long lines (Ville)
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200430124904.GA100924@jlahtine-desk.ger.corp.intel.com
Upon waiting a request (when asked), we gave that request a small
priority boost, not enough for it to cause preemption, but enough for it
to be scheduled next before all equals. We also used that bit to give
new clients a small priority boost, similar to FQ_CODEL, such that we
favoured short interactive tasks ahead of long running streams.
However, this is causing lots of complications with timeslicing where we
both want to honour the boost and yet ignore it. Those complications
cause unexpected user behaviour (tasks not being timesliced and run
concurrently as epxected), and the easiest way to resolve that is to
remove the boost. Hopefully, we can find a compromise again if we need
to, but in theory timeslicing itself and future more advanced schedulers
should give us the interactivity boost we seek.
Testcase: igt/gem_exec_schedule/lateslice
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200507152338.7452-3-chris@chris-wilson.co.uk
The older arches did not convert MI_STORE_DATA_IMM to using the GTT, but
left them writing to a physical address. The notes suggest that the
primary reason would be so that the writes were cache coherent, as the
CPU cache uses physical tagging. As such we did not implement the
legacy variant of MI_STORE_DATA_IMM and so left all the relocations
synchronous -- but with a small function to convert from the vma address
into the physical address, we can implement asynchronous relocs on these
older arches, fixing up a few tests that require them.
In order to be able to test the legacy paths, refactor the gpu
relocations so that we can hook them up to a selftest.
v2: Use an array of offsets not enum labels for the selftest
v3: Refactor the common igt_hexdump()
Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/757
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200504140629.28240-1-chris@chris-wilson.co.uk
If at first we don't succeed, try try again.
Not all engines may support the MI ops we need to perform asynchronous
relocation patching, and so we end up falling back to a synchronous
operation that has a liability of blocking. However, Tvrtko pointed out
we don't need to use the same engine to perform the relocations as we
are planning to execute the execbuf on, and so if we switch over to a
working engine, we can perform the relocation asynchronously. The user
execbuf will be queued after the relocations by virtue of fencing.
This patch creates a new context per execbuf requiring asynchronous
relocations on an unusable engines. This is perhaps a bit excessive and
can be ameliorated by a small context cache, but for the moment we only
need it for working around a little used engine on Sandybridge, and only
if relocations are actually required to an active batch buffer.
Now we just need to teach the relocation code to handle physical
addressing for gen2/3, and we should then have universal support!
Suggested-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Testcase: igt/gem_exec_reloc/basic-spin # snb
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200501192945.22215-3-chris@chris-wilson.co.uk
The ring is a precious resource: we anticipate to only use a few hundred
bytes for a request, and only try to reserve that before we start. If we
go beyond our guess in building the request, then instead of waiting at
the start of execbuf before we hold any locks or other resources, we
may trigger a wait inside a critical region. One example is in using gpu
relocations, where currently we emit a new MI_BB_START from the ring
every time we overflow a page of relocation entries. However, instead of
insert the command into the precious ring, we can chain the next page of
relocation entries as MI_BB_START from the end of the previous.
v2: Delay the emit_bb_start until after all the chained vma
synchronisation is complete. Since the buffer pool batches are idle, this
_should_ be a no-op, but one day we may some fancy async GPU bindings
for new vma!
v3: Use pool/batch consitently, once we start thinking in terms of the
batch vma, use batch->obj.
v4: Explain the magic number 4.
Tvrtko spotted that we lose propagation of the error for failing to
submit the relocation request; that's easier to fix up in the next
patch.
Testcase: igt/gem_exec_reloc/basic-many-active
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200501192945.22215-1-chris@chris-wilson.co.uk
Since the introduction of 'soft-rc6', we aim to park the device quickly
and that results in frequent idling of the whole device. Currently upon
idling we free the batch buffer pool, and so this renders the cache
ineffective for many workloads. If we want to have an effective cache of
recently allocated buffers available for reuse, we need to decouple that
cache from the engine powermanagement and make it timer based. As there
is no reason then to keep it within the engine (where it once made
retirement order easier to track), we can move it up the hierarchy to the
owner of the memory allocations.
v2: Hook up to debugfs/drop_caches to clear the cache on demand.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200430111819.10262-2-chris@chris-wilson.co.uk
The history of i915_vma_close() is confusing, as is its use. As the
lifetime of the i915_vma is currently bounded by the object it is
attached to, we needed a means of identify when a vma was no longer in
use by userspace (via the user's fd). This is further complicated by
that only ppgtt vma should be closed at the user's behest, as the ggtt
were always shared.
Now that we attach the vma to a lut on the user's context, the open
count does indicate how many unique and open context/vm are referencing
this vma from the user. As such, we can and should just use the
open_count to track when the vma is still in use by userspace.
It's a poor man's replacement for reference counting.
Closes: https://gitlab.freedesktop.org/drm/intel/issues/1193
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Andi Shyti <andi.shyti@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200422190558.30509-1-chris@chris-wilson.co.uk
While the ggtt vma are protected by their object lifetime, the list
continues until it hits a non-ggtt vma, and that vma is not protected
and may be freed as we inspect it. Hence, we require the obj->vma.lock
to protect the list as we iterate.
An example of forgetting to hold the obj->vma.lock is
[1642834.464973] general protection fault, probably for non-canonical address 0xdead000000000122: 0000 [#1] SMP PTI
[1642834.464977] CPU: 3 PID: 1954 Comm: Xorg Not tainted 5.6.0-300.fc32.x86_64 #1
[1642834.464979] Hardware name: LENOVO 20ARS25701/20ARS25701, BIOS GJET94WW (2.44 ) 09/14/2017
[1642834.465021] RIP: 0010:i915_gem_object_set_tiling+0x2c0/0x3e0 [i915]
[1642834.465024] Code: 8b 84 24 18 01 00 00 f6 c4 80 74 59 49 8b 94 24 a0 00 00 00 49 8b 84 24 e0 00 00 00 49 8b 74 24 10 48 8b 92 30 01 00 00 89 c7 <80> ba 0a 06 00 00 03 0f 87 86 00 00 00 ba 00 00 08 00 b9 00 00 10
[1642834.465025] RSP: 0018:ffffa98780c77d60 EFLAGS: 00010282
[1642834.465028] RAX: ffff8d232bfb2578 RBX: 0000000000000002 RCX: ffff8d25873a0000
[1642834.465029] RDX: dead000000000122 RSI: fffff0af8ac6e408 RDI: 000000002bfb2578
[1642834.465030] RBP: ffff8d25873a0000 R08: ffff8d252bfb5638 R09: 0000000000000000
[1642834.465031] R10: 0000000000000000 R11: ffff8d252bfb5640 R12: ffffa987801cb8f8
[1642834.465032] R13: 0000000000001000 R14: ffff8d233e972e50 R15: ffff8d233e972d00
[1642834.465034] FS: 00007f6a3d327f00(0000) GS:ffff8d25926c0000(0000) knlGS:0000000000000000
[1642834.465036] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[1642834.465037] CR2: 00007f6a2064d000 CR3: 00000002fb57c001 CR4: 00000000001606e0
[1642834.465038] Call Trace:
[1642834.465083] i915_gem_set_tiling_ioctl+0x122/0x230 [i915]
[1642834.465121] ? i915_gem_object_set_tiling+0x3e0/0x3e0 [i915]
[1642834.465151] drm_ioctl_kernel+0x86/0xd0 [drm]
[1642834.465156] ? avc_has_perm+0x3b/0x160
[1642834.465178] drm_ioctl+0x206/0x390 [drm]
[1642834.465216] ? i915_gem_object_set_tiling+0x3e0/0x3e0 [i915]
[1642834.465221] ? selinux_file_ioctl+0x122/0x1c0
[1642834.465226] ? __do_munmap+0x24b/0x4d0
[1642834.465231] ksys_ioctl+0x82/0xc0
[1642834.465235] __x64_sys_ioctl+0x16/0x20
[1642834.465238] do_syscall_64+0x5b/0xf0
[1642834.465243] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[1642834.465245] RIP: 0033:0x7f6a3d7b047b
[1642834.465247] Code: 0f 1e fa 48 8b 05 1d aa 0c 00 64 c7 00 26 00 00 00 48 c7 c0 ff ff ff ff c3 66 0f 1f 44 00 00 f3 0f 1e fa b8 10 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d ed a9 0c 00 f7 d8 64 89 01 48
[1642834.465249] RSP: 002b:00007ffe71adba28 EFLAGS: 00000246 ORIG_RAX: 0000000000000010
[1642834.465251] RAX: ffffffffffffffda RBX: 000055f99048fa40 RCX: 00007f6a3d7b047b
[1642834.465253] RDX: 00007ffe71adba30 RSI: 00000000c0106461 RDI: 000000000000000e
[1642834.465254] RBP: 0000000000000002 R08: 000055f98f3f1798 R09: 0000000000000002
[1642834.465255] R10: 0000000000001000 R11: 0000000000000246 R12: 0000000000000080
[1642834.465257] R13: 000055f98f3f1690 R14: 00000000c0106461 R15: 00007ffe71adba30
Now to take the spinlock during the list iteration, we need to break it
down into two phases. In the first phase under the lock, we cannot sleep
and so must defer the actual work to a second list, protected by the
ggtt->mutex.
We also need to hold the spinlock during creation of a new vma to
serialise with updates of the tiling on the object.
Reported-by: Dave Airlie <airlied@redhat.com>
Fixes: 2850748ef8 ("drm/i915: Pull i915_vma_pin under the vm->mutex")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Dave Airlie <airlied@redhat.com>
Cc: <stable@vger.kernel.org> # v5.5+
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200422072805.17340-1-chris@chris-wilson.co.uk
drm-misc-next for 5.8:
UAPI Changes:
- drm: error out with EBUSY when device has existing master
- drm: rework SET_MASTER and DROP_MASTER perm handling
Cross-subsystem Changes:
- mm: export two symbols from slub/slob
- fbdev: savage: fix -Wextra build warning
- video: omap2: Use scnprintf() for avoiding potential buffer overflow
Core Changes:
- Remove drm_pci.h
- drm_pci_{alloc/free)() are now legacy
- Introduce managed DRM resourcesA
- Allow drivers to subclass struct drm_framebuffer
- Introduce struct drm_afbc_framebuffer and helpers
- fbdev: remove return value from generic fbdev setup
- Introduce simple-encoder helper
- vram-helpers: set fence on plane
- dp_mst: ACT timeout improvements
- dp_mst: Remove drm_dp_mst_has_audio()
- TTM: ttm_trace_dma_{map/unmap}() cleanups
- dma-buf: add flag for PCIP2P support
- EDID: Various improvements
- Encoder: cleanup semantics of possible_clones and possible_crtcs
- VBLANK documentation updates
- Writeback documentation updates
Driver Changes:
- Convert several drivers to i2c_new_client_device()
- Drop explicit drm_mode_config_cleanup() calls from drivers
- Auto-release device structures with drmm_add_final_kfree()
- Init bfdev console after registering DRM device
- Make various .debugfs functions return 0 unconditionally; ignore errors
- video: Use scnprintf() to avoid buffer overflows
- Convert drivers to simple encoders
- drm/amdgpu: note that we can handle peer2peer DMA-buf
- drm/amdgpu: add support for exporting VRAM using DMA-buf v3
- drm/kirin: Revert change to register connectors
- drm/lima: Add optional devfreq and cooling device support
- drm/lima: Various improvements wrt. task handling
- drm/panel: nt39016: Support multiple modes and 50Hz
- drm/panel: Support Leadtek LTK050H3146W
- drm/rockchip: Add support for afbc
- drm/virtio: Various cleanups
- drm/hisilicon/hibmc: Enforce 128-byte stride alignment
- drm/qxl: Fix notify port address of cursor ring buffer
- drm/sun4i: Improvements to format handling
- drm/bridge: dw-hdmi: Various improvements
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Thomas Zimmermann <tzimmermann@suse.de>
Link: https://patchwork.freedesktop.org/patch/msgid/20200414090738.GA16827@linux-uq9g
igt_ppgtt_pin_update() invokes i915_gem_context_get_vm_rcu(), which
returns a reference of the i915_address_space object to "vm" with
increased refcount.
When igt_ppgtt_pin_update() returns, "vm" becomes invalid, so the
refcount should be decreased to keep refcount balanced.
The reference counting issue happens in two exception handling paths of
igt_ppgtt_pin_update(). When i915_gem_object_create_internal() returns
IS_ERR, the refcnt increased by i915_gem_context_get_vm_rcu() is not
decreased, causing a refcnt leak.
Fix this issue by jumping to "out_vm" label when
i915_gem_object_create_internal() returns IS_ERR.
Fixes: a4e7ccdac3 ("drm/i915: Move context management under GEM")
Signed-off-by: Xiyu Yang <xiyuyang19@fudan.edu.cn>
Signed-off-by: Xin Tan <tanxin.ctf@gmail.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/1587361342-83494-1-git-send-email-xiyuyang19@fudan.edu.cn
If we want to percolate information back from the HW, up through the GEM
context, we need to wait until the intel_context is scheduled out for
the last time. This is handled by the retirement of the intel_context's
barrier, i.e. by listening to the pulse after the notional unpin. So
wait until the intel_context is finally retired before releasing the
engine, so that we can inspect the final context state and pass it on.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200406155840.1728-3-chris@chris-wilson.co.uk
Now that we can peek at GEM->engines[] and obtain a reference to them
using RCU, do so for instances where we can safely iterate the
potentially old copy of the engines. For setting, we can do this when we
know the engine properties are copied over before swapping, so we know
the new engines already have the global property and we update the old
before they are discarded. For reading, we only need to be safe; as we
do so on behalf of the user, their races are their own problem.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200402124218.6375-1-chris@chris-wilson.co.uk
We cached the number of vma bound to the object in order to speed up
shrinker decisions. This has been superseded by being more proactive in
removing objects we cannot shrink from the shrinker lists, and so we can
drop the clumsy attempt at atomically counting the bind count and
comparing it to the number of pinned mappings of the object. This will
only get more clumsier with asynchronous binding and unbinding.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Matthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200401223924.16667-1-chris@chris-wilson.co.uk
A recent commit in clang added -Wtautological-compare to -Wall, which is
enabled for i915 after -Wtautological-compare is disabled for the rest
of the kernel so we see the following warning on x86_64:
../drivers/gpu/drm/i915/gem/i915_gem_execbuffer.c:1433:22: warning:
result of comparison of constant 576460752303423487 with expression of
type 'unsigned int' is always false
[-Wtautological-constant-out-of-range-compare]
if (unlikely(remain > N_RELOC(ULONG_MAX)))
~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~
../include/linux/compiler.h:78:42: note: expanded from macro 'unlikely'
# define unlikely(x) __builtin_expect(!!(x), 0)
^
1 warning generated.
It is not wrong in the case where ULONG_MAX > UINT_MAX but it does not
account for the case where this file is built for 32-bit x86, where
ULONG_MAX == UINT_MAX and this check is still relevant.
Cast remain to unsigned long, which keeps the generated code the same
(verified with clang-11 on x86_64 and GCC 9.2.0 on x86 and x86_64) and
the warning is silenced so we can catch more potential issues in the
future.
Closes: https://github.com/ClangBuiltLinux/linux/issues/778
Suggested-by: Michel Dänzer <michel@daenzer.net>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200214054706.33870-1-natechancellor@gmail.com
If the caller allows and we do not have to wait for any signals,
immediately execute the work within the caller's process. By doing so we
avoid the overhead of scheduling a new task, and the latency in
executing it, at the cost of pulling that work back into the immediate
context. (Sometimes we still prefer to offload the task to another cpu,
especially if we plan on executing many such tasks in parallel for this
client.)
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200325120227.8044-2-chris@chris-wilson.co.uk
It looks like some callers expect a non-volatile object, that they do not
want the contents of the pages lost if they happen to not be looking at it.
The shrinker however sees that we mark the pages as DONTNEED and
believes that it can freely reap them. However, since the huge object
use plain pages, they cannot be swapped out as they have no backing
storge, and the only way we can shrink them is by discarding the
contents. In light of the callers wanting to keep the contents around,
both IS_SHRINKABLE and marking the pages as volatile are incorrect.
If we drop the IS_SHRINKABLE flag we avoid the immediate issue of the
shrinker accidentally removing valuable content. We will have to
remember that a huge object is not suitable for exercising the shrinker
interaction -- although we can introduce a shrinkable one if we require.
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20200323130821.47914-1-matthew.auld@intel.com
As we store the handle lookup inside a radix tree, we do not need the
gem_context->mutex except until we need to insert our lookup into the
common radix tree. This takes a small bit of rearranging to ensure that
the lut we insert into the tree is ready prior to actually inserting it
(as soon as it is exposed via the radixtree, it is visible to any other
submission).
v2: For brownie points, remove the goto spaghetti.
v3: Tighten up the closed-handle checks.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200323092841.22240-8-chris@chris-wilson.co.uk
On Gen11 powergating half the execution units is a functional
requirement when using the VME samplers. Not fullfilling this
requirement can lead to hangs.
This unfortunately plays fairly poorly with the NOA requirements. NOA
requires a stable power configuration to maintain its configuration.
As a result using OA (and NOA feeding into it) so far has required us
to use a power configuration that can work for all contexts. The only
power configuration fullfilling this is powergating half the execution
units.
This makes performance analysis for 3D workloads somewhat pointless.
Failing to find a solution that would work for everybody, this change
introduces a new i915-perf stream open parameter that punts the
decision off to userspace. If this parameter is omitted, the existing
Gen11 behavior remains (half EU array powergating).
This change takes the initiative to move all perf related sseu
configuration into i915_perf.c
v2: Make parameter priviliged if different from default
v3: Fix context modifying its sseu config while i915-perf is enabled
v4: Always consider global sseu a privileged operation (Tvrtko)
Override req_sseu point in intel_sseu_make_rpcs() (Tvrtko)
Remove unrelated changes (Tvrtko)
v5: Some typos (Tvrtko)
Process sseu param in read_properties_unlocked() (Tvrtko)
v6: Actually commit the bits from v5...
Fixup some checkpath warnings
v7: Only compare engine uabi field (Chris)
Signed-off-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200317132222.2638719-3-lionel.g.landwerlin@intel.com
