The comment is very old and quite misleading now.

drivers/gpu/drm/i915/i915_gem_context.c:349: warning: No description found for parameter 'dev_priv'
drivers/gpu/drm/i915/i915_gem_context.c:349: warning: No description found for parameter 'file_priv'
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NLionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180208111559.32663-1-chris@chris-wilson.co.uk

Userspace provides a 64b value for the priority, we need to be careful
to preserve the full range before validation to prevent truncation (and
letting an illegal value pass).
Reported-by: NAntonio Argenziano <antonio.argenziano@intel.com>
Fixes: ac14fbd4 ("drm/i915/scheduler: Support user-defined priorities")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Antonio Argenziano <antonio.argenziano@intel.com>
Cc: Michal Winiarski <michal.winiarski@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180208085151.11480-1-chris@chris-wilson.co.ukReviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

If we remove some hardcoded assumptions about the preempt context having
a fixed id, reserved from use by normal user contexts, we may only
allocate the i915_gem_context when required. Then the subsequent
decisions on using preemption reduce to having the preempt context
available.

v2: Include an assert that we don't allocate the preempt context twice.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Michal Winiarski <michal.winiarski@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Acked-by: NDaniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180207210544.26351-3-chris@chris-wilson.co.ukReviewed-by: NMichel Thierry <michel.thierry@intel.com>

Rather than having the high level ioctl interface guess the underlying
implementation details, having the implementation declare what
capabilities it exports. We define an intel_driver_caps, similar to the
intel_device_info, which instead of trying to describe the HW gives
details on what the driver itself supports. This is then populated by
the engine backend for the new scheduler capability field for use
elsewhere.

v2: Use caps.scheduler for validating CONTEXT_PARAM_SET_PRIORITY (Mika)
    One less assumption of engine[RCS] \o/
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tomasz Lis <tomasz.lis@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NTomasz Lis <tomasz.lis@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180207210544.26351-2-chris@chris-wilson.co.ukReviewed-by: NMichel Thierry <michel.thierry@intel.com>

If a fence allocation fails in a blocking context, we will sleep on the
fence as a last resort. We can therefore allow ourselves to fail and
sleep on the fence instead of triggering a system-wide oom. This allows
us to throttle malicious clients that are consuming lots of system
resources by capping the amount of memory used by fences.

Testcase: igt/gem_shrink/execbufX
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171212180652.22061-2-chris@chris-wilson.co.uk

In the upcoming patch we will change the way how to recognize
when GuC is in use. Using helper macros will minimize scope
of that changes. While here, update dev_info message.
Signed-off-by: NMichal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Sagar Arun Kamble <sagar.a.kamble@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20171206135316.32556-3-michal.wajdeczko@intel.com

Even though all rendering should have been flushed at the end of the
previous requests, add an extra flush after switching to the
kernel_context. As the switch to the kernel_context is used when idling
the gpu (e.g. suspend), having an extra layer of paranoia to ensure
everything is flushed to memory seems sensible.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171126214856.23702-1-chris@chris-wilson.co.ukReviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

The legacy i915_switch_context() is only applicable to the legacy
ringbuffer submission method, so move it from the general
i915_gem_context.c to intel_ringbuffer.c (rename pending!).
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171123152631.31385-2-chris@chris-wilson.co.uk

The legacy context switch for ringbuffer submission is multistaged,
where each of those stages may fail. However, we were updating global
state after some stages, and so we had to force the incomplete request
to be submitted because we could not unwind. Save the global state
before performing the switches, and so enable us to unwind back to the
previous global state should any phase fail. We then must cancel the
request instead of submitting it should the construction fail.

v2: s/saved_ctx/from_ctx/; s/ctx/to_ctx/ etc.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171123152631.31385-1-chris@chris-wilson.co.uk

Having disabled the broken semaphores on Sandybridge, there is no need
for a modparam any more, so remove it in favour of a simple
HAS_LEGACY_SEMAPHORES() guard.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NMaarten Lankhorst <maarten.lankhorst@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171120205504.21892-5-chris@chris-wilson.co.uk

Since removing the module parameter to force selection of ringbuffer
emission for gen8, the code is defunct. Remove it.

To put the difference into perspective, a couple of microbenchmarks
(bdw i7-5557u, 20170324):
                                        ring          execlists
exec continuous nops on all rings:   1.491us            2.223us
exec sequential nops on each ring:  12.508us           53.682us
single nop + sync:                   9.272us           30.291us

vblank_mode=0 glxgears:            ~11000fps           ~9000fps

Since the earlier submission, gen8 ringbuffer submission has fallen
further and further behind in features. So while ringbuffer may hold the
throughput crown, in terms of interactive latency, execlists is much
better. Alas, we have no convenient metrics for such, other than
demonstrating things we can do with execlists but can not using
legacy ringbuffer submission.

We have made a few improvements to lowlevel execlists throughput,
and ringbuffer currently panics on boot! (bdw i7-5557u, 20171026):

                                        ring          execlists
exec continuous nops on all rings:       n/a            1.921us
exec sequential nops on each ring:       n/a           44.621us
single nop + sync:                       n/a           21.953us

vblank_mode=0 glxgears:                  n/a          ~18500fps

References: https://bugs.freedesktop.org/show_bug.cgi?id=87725Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Once-upon-a-time-Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171120205504.21892-2-chris@chris-wilson.co.uk

Execlists and legacy ringbuffer submission are no longer feature
comparable (execlists now offer greater functionality that should
overcome their performance hit) and obsoletes the unsafe module
parameter, i.e. comparing the two modes of execution is no longer
useful, so remove the debug tool.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com> #i915_perf.c
Link: https://patchwork.freedesktop.org/patch/msgid/20171120205504.21892-1-chris@chris-wilson.co.uk

During request construction, after pinning the context we know whether
or not we have to emit a context switch. So move this common operation
from every caller into i915_gem_request_alloc() itself.

v2: Always submit the request if we emitted some commands during request
construction, as typically it also involves changes in global state.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171120102002.22254-2-chris@chris-wilson.co.uk

Take a copy of the HW state after a reset upon module loading by
executing a context switch from a blank context to the kernel context,
thus saving the default hw state over the blank context image.
We can then use the default hw state to initialise any future context,
ensuring that each starts with the default view of hw state.

v2: Unmap our default state from the GTT after stealing it from the
context. This should stop us from accidentally overwriting it via the
GTT (and frees up some precious GTT space).

Testcase: igt/gem_ctx_isolation
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171110142634.10551-7-chris@chris-wilson.co.uk

When we close the VMA, we unbind it from the ppgtt and tear down the
page directory pointing at it. That may trigger us to return WC pages
back to the system, requiring conversion back to WB which itself may
sleep. That makes i915_vma_close() unsuitable for use inside the RCU
read lock, which we need to hold to iterate the radixtree.

The fix is quite simple, we can close all the VMA as we close the ppgtt,
we only need to do that instead of closing them during destruction of
the LUT.

v2: Order between closing the LUT and the ppgtt is important; we use the
vma inside the LUT as a means of retrieving the object, and so we must
clear the LUT before freeing the VMA when closing the ppgtt.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103638
Fixes: 547da76b ("drm/i915: Hold rcu_read_lock when iterating over the radixtree (vma idr)")
Fixes: d1b48c1e ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171109085540.32264-1-chris@chris-wilson.co.ukReviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
(cherry picked from commit 94dec871)
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

When we close the VMA, we unbind it from the ppgtt and tear down the
page directory pointing at it. That may trigger us to return WC pages
back to the system, requiring conversion back to WB which itself may
sleep. That makes i915_vma_close() unsuitable for use inside the RCU
read lock, which we need to hold to iterate the radixtree.

The fix is quite simple, we can close all the VMA as we close the ppgtt,
we only need to do that instead of closing them during destruction of
the LUT.

v2: Order between closing the LUT and the ppgtt is important; we use the
vma inside the LUT as a means of retrieving the object, and so we must
clear the LUT before freeing the VMA when closing the ppgtt.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103638
Fixes: 547da76b ("drm/i915: Hold rcu_read_lock when iterating over the radixtree (vma idr)")
Fixes: d1b48c1e ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Matthew Auld <matthew.william.auld@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171109085540.32264-1-chris@chris-wilson.co.ukReviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>

Kasan spotted

    [IGT] gem_tiled_pread_pwrite: exiting, ret=0
    ==================================================================
    BUG: KASAN: use-after-free in __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
    Read of size 8 at addr ffff8801359da310 by task kworker/3:2/182

    CPU: 3 PID: 182 Comm: kworker/3:2 Tainted: G     U          4.14.0-rc6-CI-Custom_3340+ #1
    Hardware name: Intel Corp. Geminilake/GLK RVP1 DDR4 (05), BIOS GELKRVPA.X64.0062.B30.1708222146 08/22/2017
    Workqueue: events __i915_gem_free_work [i915]
    Call Trace:
     dump_stack+0x68/0xa0
     print_address_description+0x78/0x290
     ? __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
     kasan_report+0x23d/0x350
     __asan_report_load8_noabort+0x19/0x20
     __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
     ? i915_gem_object_truncate+0x100/0x100 [i915]
     ? lock_acquire+0x380/0x380
     __i915_gem_object_put_pages+0x30d/0x530 [i915]
     __i915_gem_free_objects+0x551/0xbd0 [i915]
     ? lock_acquire+0x13e/0x380
     __i915_gem_free_work+0x4e/0x70 [i915]
     process_one_work+0x6f6/0x1590
     ? pwq_dec_nr_in_flight+0x2b0/0x2b0
     worker_thread+0xe6/0xe90
     ? pci_mmcfg_check_reserved+0x110/0x110
     kthread+0x309/0x410
     ? process_one_work+0x1590/0x1590
     ? kthread_create_on_node+0xb0/0xb0
     ret_from_fork+0x27/0x40

    Allocated by task 1801:
     save_stack_trace+0x1b/0x20
     kasan_kmalloc+0xee/0x190
     kasan_slab_alloc+0x12/0x20
     kmem_cache_alloc+0xdc/0x2e0
     radix_tree_node_alloc.constprop.12+0x48/0x330
     __radix_tree_create+0x274/0x480
     __radix_tree_insert+0xa2/0x610
     i915_gem_object_get_sg+0x224/0x670 [i915]
     i915_gem_object_get_page+0xb5/0x1c0 [i915]
     i915_gem_pread_ioctl+0x822/0xf60 [i915]
     drm_ioctl_kernel+0x13f/0x1c0
     drm_ioctl+0x6cf/0x980
     do_vfs_ioctl+0x184/0xf30
     SyS_ioctl+0x41/0x70
     entry_SYSCALL_64_fastpath+0x1c/0xb1

    Freed by task 37:
     save_stack_trace+0x1b/0x20
     kasan_slab_free+0xaf/0x190
     kmem_cache_free+0xbf/0x340
     radix_tree_node_rcu_free+0x79/0x90
     rcu_process_callbacks+0x46d/0xf40
     __do_softirq+0x21c/0x8d3

    The buggy address belongs to the object at ffff8801359da0f0
    which belongs to the cache radix_tree_node of size 576
    The buggy address is located 544 bytes inside of
    576-byte region [ffff8801359da0f0, ffff8801359da330)
    The buggy address belongs to the page:
    page:ffffea0004d67600 count:1 mapcount:0 mapping:          (null) index:0x0 compound_mapcount: 0
    flags: 0x8000000000008100(slab|head)
    raw: 8000000000008100 0000000000000000 0000000000000000 0000000100110011
    raw: ffffea0004b52920 ffffea0004b38020 ffff88015b416a80 0000000000000000
    page dumped because: kasan: bad access detected

    Memory state around the buggy address:
     ffff8801359da200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
     ffff8801359da280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
    >ffff8801359da300: fb fb fb fb fb fb fc fc fc fc fc fc fc fc fc fc
			     ^
     ffff8801359da380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
     ffff8801359da400: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
    ==================================================================
    Disabling lock debugging due to kernel taint

which looks like the slab containing the radixtree iter was freed as we
traversed the tree, taking the rcu read lock across the loop should
prevent that (deferring all the frees until the end).
Reported-by: NTomi Sarvela <tomi.p.sarvela@intel.com>
Fixes: d1b48c1e ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171026130032.10677-2-chris@chris-wilson.co.ukReviewed-by: NMatthew Auld <matthew.william.auld@gmail.com>
(cherry picked from commit 547da76b)
Signed-off-by: NRodrigo Vivi <rodrigo.vivi@intel.com>

Kasan spotted

    [IGT] gem_tiled_pread_pwrite: exiting, ret=0
    ==================================================================
    BUG: KASAN: use-after-free in __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
    Read of size 8 at addr ffff8801359da310 by task kworker/3:2/182

    CPU: 3 PID: 182 Comm: kworker/3:2 Tainted: G     U          4.14.0-rc6-CI-Custom_3340+ #1
    Hardware name: Intel Corp. Geminilake/GLK RVP1 DDR4 (05), BIOS GELKRVPA.X64.0062.B30.1708222146 08/22/2017
    Workqueue: events __i915_gem_free_work [i915]
    Call Trace:
     dump_stack+0x68/0xa0
     print_address_description+0x78/0x290
     ? __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
     kasan_report+0x23d/0x350
     __asan_report_load8_noabort+0x19/0x20
     __i915_gem_object_reset_page_iter+0x15c/0x170 [i915]
     ? i915_gem_object_truncate+0x100/0x100 [i915]
     ? lock_acquire+0x380/0x380
     __i915_gem_object_put_pages+0x30d/0x530 [i915]
     __i915_gem_free_objects+0x551/0xbd0 [i915]
     ? lock_acquire+0x13e/0x380
     __i915_gem_free_work+0x4e/0x70 [i915]
     process_one_work+0x6f6/0x1590
     ? pwq_dec_nr_in_flight+0x2b0/0x2b0
     worker_thread+0xe6/0xe90
     ? pci_mmcfg_check_reserved+0x110/0x110
     kthread+0x309/0x410
     ? process_one_work+0x1590/0x1590
     ? kthread_create_on_node+0xb0/0xb0
     ret_from_fork+0x27/0x40

    Allocated by task 1801:
     save_stack_trace+0x1b/0x20
     kasan_kmalloc+0xee/0x190
     kasan_slab_alloc+0x12/0x20
     kmem_cache_alloc+0xdc/0x2e0
     radix_tree_node_alloc.constprop.12+0x48/0x330
     __radix_tree_create+0x274/0x480
     __radix_tree_insert+0xa2/0x610
     i915_gem_object_get_sg+0x224/0x670 [i915]
     i915_gem_object_get_page+0xb5/0x1c0 [i915]
     i915_gem_pread_ioctl+0x822/0xf60 [i915]
     drm_ioctl_kernel+0x13f/0x1c0
     drm_ioctl+0x6cf/0x980
     do_vfs_ioctl+0x184/0xf30
     SyS_ioctl+0x41/0x70
     entry_SYSCALL_64_fastpath+0x1c/0xb1

    Freed by task 37:
     save_stack_trace+0x1b/0x20
     kasan_slab_free+0xaf/0x190
     kmem_cache_free+0xbf/0x340
     radix_tree_node_rcu_free+0x79/0x90
     rcu_process_callbacks+0x46d/0xf40
     __do_softirq+0x21c/0x8d3

    The buggy address belongs to the object at ffff8801359da0f0
    which belongs to the cache radix_tree_node of size 576
    The buggy address is located 544 bytes inside of
    576-byte region [ffff8801359da0f0, ffff8801359da330)
    The buggy address belongs to the page:
    page:ffffea0004d67600 count:1 mapcount:0 mapping:          (null) index:0x0 compound_mapcount: 0
    flags: 0x8000000000008100(slab|head)
    raw: 8000000000008100 0000000000000000 0000000000000000 0000000100110011
    raw: ffffea0004b52920 ffffea0004b38020 ffff88015b416a80 0000000000000000
    page dumped because: kasan: bad access detected

    Memory state around the buggy address:
     ffff8801359da200: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
     ffff8801359da280: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
    >ffff8801359da300: fb fb fb fb fb fb fc fc fc fc fc fc fc fc fc fc
			     ^
     ffff8801359da380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
     ffff8801359da400: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
    ==================================================================
    Disabling lock debugging due to kernel taint

which looks like the slab containing the radixtree iter was freed as we
traversed the tree, taking the rcu read lock across the loop should
prevent that (deferring all the frees until the end).
Reported-by: NTomi Sarvela <tomi.p.sarvela@intel.com>
Fixes: d1b48c1e ("drm/i915: Replace execbuf vma ht with an idr")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171026130032.10677-2-chris@chris-wilson.co.ukReviewed-by: NMatthew Auld <matthew.william.auld@gmail.com>

During evict, we wish to idle the GPU if we see that the GGTT is full.
However, our test for idle in i915_gem_evict_something() and in
i915_gem_switch_to_kernel_context() do not match leading to
disappointment - we never believe that we are idle and keep trying to
flush the GGTT ad infinitum.

Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=103438Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171024220855.30155-2-chris@chris-wilson.co.ukReviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

Use a priority stored in the context as the initial value when
submitting a request. This allows us to change the default priority on a
per-context basis, allowing different contexts to be favoured with GPU
time at the expense of lower importance work. The user can adjust the
context's priority via I915_CONTEXT_PARAM_PRIORITY, with more positive
values being higher priority (they will be serviced earlier, after their
dependencies have been resolved). Any prerequisite work for an execbuf
will have its priority raised to match the new request as required.

Normal users can specify any value in the range of -1023 to 0 [default],
i.e. they can reduce the priority of their workloads (and temporarily
boost it back to normal if so desired).

Privileged users can specify any value in the range of -1023 to 1023,
[default is 0], i.e. they can raise their priority above all overs and
so potentially starve the system.

Note that the existing schedulers are not fair, nor load balancing, the
execution is strictly by priority on a first-come, first-served basis,
and the driver may choose to boost some requests above the range
available to users.

This priority was originally based around nice(2), but evolved to allow
clients to adjust their priority within a small range, and allow for a
privileged high priority range.

For example, this can be used to implement EGL_IMG_context_priority
https://www.khronos.org/registry/egl/extensions/IMG/EGL_IMG_context_priority.txt

	EGL_CONTEXT_PRIORITY_LEVEL_IMG determines the priority level of
        the context to be created. This attribute is a hint, as an
        implementation may not support multiple contexts at some
        priority levels and system policy may limit access to high
        priority contexts to appropriate system privilege level. The
        default value for EGL_CONTEXT_PRIORITY_LEVEL_IMG is
        EGL_CONTEXT_PRIORITY_MEDIUM_IMG."

so we can map

	PRIORITY_HIGH -> 1023 [privileged, will failback to 0]
	PRIORITY_MED -> 0 [default]
	PRIORITY_LOW -> -1023

They also map onto the priorities used by VkQueue (and a VkQueue is
essentially a timeline, our i915_gem_context under full-ppgtt).

v2: s/CAP_SYS_ADMIN/CAP_SYS_NICE/
v3: Report min/max user priorities as defines in the uapi, and rebase
internal priorities on the exposed values.

Testcase: igt/gem_exec_schedule
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-9-chris@chris-wilson.co.uk

Add another perma-pinned context for using for preemption at any time.
We cannot just reuse the existing kernel context, as first and foremost
we need to ensure that we can preempt the kernel context itself, so
require a distinct context id. Similar to the kernel context, we may
want to interrupt execution and switch to the preempt context at any
time, and so it needs to be permanently pinned and available.

To compensate for yet another permanent allocation, we shrink the
existing context and the new context by reducing their ringbuffer to the
minimum.

v2: Assert that we never allocate a request from the preemption context.
v3: Limit perma-pin to engines that may preempt.
v4: Onion cleanup for early driver death
v5: Onion ordering in main driver cleanup as well.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMichał Winiarski <michal.winiarski@intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20171003203453.15692-4-chris@chris-wilson.co.uk

Our global struct with params is named exactly the same way
as new preferred name for the drm_i915_private function parameter.
To avoid such name reuse lets use different name for the global.

v5: pure rename
v6: fix

Credits-to: Coccinelle

@@
identifier n;
@@
(
-	i915.n
+	i915_modparams.n
)
Signed-off-by: NMichal Wajdeczko <michal.wajdeczko@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Ville Syrjala <ville.syrjala@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Acked-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Signed-off-by: NJani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170919193846.38060-1-michal.wajdeczko@intel.com

This was the competing idea long ago, but it was only with the rewrite
of the idr as an radixtree and using the radixtree directly ourselves,
along with the realisation that we can store the vma directly in the
radixtree and only need a list for the reverse mapping, that made the
patch performant enough to displace using a hashtable. Though the vma ht
is fast and doesn't require any extra allocation (as we can embed the node
inside the vma), it does require a thread for resizing and serialization
and will have the occasional slow lookup. That is hairy enough to
investigate alternatives and favour them if equivalent in peak performance.
One advantage of allocating an indirection entry is that we can support a
single shared bo between many clients, something that was done on a
first-come first-serve basis for shared GGTT vma previously. To offset
the extra allocations, we create yet another kmem_cache for them.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170816085210.4199-5-chris@chris-wilson.co.uk

When switching between contexts using the aliasing_ppgtt, the VM is
shared. We don't need to reload the PD registers unless they are dirty.

Martin Peres reported an issue that looks like corruption between
Haswell context switches, bisecting to commit f9326be5 ("drm/i915:
Rearrange switch_context to load the aliasing ppgtt on first use").
Switching between the same mm (the aliasing_ppgtt is used for all
contexts in this case) should be a nop, but appears to trigger some
side-effects in the context switch. However, as we know the switch
is redundant in this case, we can skip it and continue to ignore the
issue until somebody feels strong enough to investigate full-ppgtt on
gen7 again!

Except.. Martin was using full-ppgtt which is not supported as it
doesn't work correctly yet. So whilst the bisect did yield valuable
information about the failures, the fix should not have any user impact
under default settings, with the exception of a slightly lower
throughput on xcs as the VM would always be reloaded.

v2: Also remember to set the legacy_active_context following the switch
on xcs (commit e8a9c58f ("drm/i915: Unify active context tracking
between legacy/execlists/guc"))

Fixes: f9326be5 ("drm/i915: Rearrange switch_context to load the aliasing ppgtt on first use")
Fixes: e8a9c58f ("drm/i915: Unify active context tracking between legacy/execlists/guc")
Reported-by: NMartin Peres <martin.peres@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Martin Peres <martin.peres@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170812152724.6883-1-chris@chris-wilson.co.uk
(cherry picked from commit 12124bea)
Signed-off-by: NJani Nikula <jani.nikula@intel.com>

When switching between contexts using the aliasing_ppgtt, the VM is
shared. We don't need to reload the PD registers unless they are dirty.

Martin Peres reported an issue that looks like corruption between
Haswell context switches, bisecting to commit f9326be5 ("drm/i915:
Rearrange switch_context to load the aliasing ppgtt on first use").
Switching between the same mm (the aliasing_ppgtt is used for all
contexts in this case) should be a nop, but appears to trigger some
side-effects in the context switch. However, as we know the switch
is redundant in this case, we can skip it and continue to ignore the
issue until somebody feels strong enough to investigate full-ppgtt on
gen7 again!

Except.. Martin was using full-ppgtt which is not supported as it
doesn't work correctly yet. So whilst the bisect did yield valuable
information about the failures, the fix should not have any user impact
under default settings, with the exception of a slightly lower
throughput on xcs as the VM would always be reloaded.

v2: Also remember to set the legacy_active_context following the switch
on xcs (commit e8a9c58f ("drm/i915: Unify active context tracking
between legacy/execlists/guc"))

Fixes: f9326be5 ("drm/i915: Rearrange switch_context to load the aliasing ppgtt on first use")
Fixes: e8a9c58f ("drm/i915: Unify active context tracking between legacy/execlists/guc")
Reported-by: NMartin Peres <martin.peres@linux.intel.com>
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Martin Peres <martin.peres@linux.intel.com>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170812152724.6883-1-chris@chris-wilson.co.uk

Since we make call i915_gem_context_mark_guilty() concurrently when
resetting different engines in parallel, we need to make sure that our
updates are safe for the unlocked access.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Reviewed-by: NMika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170721123238.16428-12-chris@chris-wilson.co.ukSigned-off-by: NDaniel Vetter <daniel.vetter@ffwll.ch>

Currently, we move all unreferenced contexts to an RCU free list and
then onto a worker for eventual reaping. To compensate against this
growing into a long list with frequent allocations starving the system
of available memory, before we allocate a new context we reap all the
stale contexts. This puts all the cost of destroying the context into
the next allocator, which is presumably more sensitive to syscall
latency and unfair. We can limit the number of contexts being freed by
the new allocator to both keep the list trimmed and to allow the
allocator to be reasonably fast.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170705142634.18554-4-chris@chris-wilson.co.ukReviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>

Before we create a new context, we try and reap all the stale contexts
(i.e. those that are freed but waiting for a worker to come and return
their allocations to the system). Before we do this, we retire all
requests so that we clear any inflight no longer used contexts (who are
only being kept alived by those inflght requests). However, any context
that is finally unreferenced by this retirement is put onto an RCU list
and not available for immediately reaping, we stall for no immediate
benefit.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170705142634.18554-3-chris@chris-wilson.co.ukReviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>

We need to reap the stale contexts for all new contexts, be they created
by user in i915_gem_context_ioctl or from opening a new file in
i915_gem_context_open. Both paths may be called very frequently
accumulating many stale contexts before any worker has a chance to run
and free their memory.

Fixes: 1acfc104 ("drm/i915: Enable rcu-only context lookups")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170705142634.18554-2-chris@chris-wilson.co.uk

Avoid any pointer dereference in inspecting a potential PTR_ERR by
checking for the error pointer before checking for an invalid context.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170705142634.18554-1-chris@chris-wilson.co.uk

When iterating the list of contexts to free, we need to use a safe
iterator as we are freeing the link as we go. Pass an extra thick brown
paper bag.

Fixes: 5f09a9c8 ("drm/i915: Allow contexts to be unreferenced locklessly")
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Matthew Auld <matthew.auld@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170630230517.1938-1-chris@chris-wilson.co.ukReviewed-by: NMatthew Auld <matthew.auld@intel.com>

Whilst the contents of the context is still protected by the big
struct_mutex, this is not much of an improvement. It is just one tiny
step towards reducing our BKL.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620110547.15947-3-chris@chris-wilson.co.uk

If we move the actual cleanup of the context to a worker, we can allow
the final free to be called from any context and avoid undue latency in
the caller.

v2: Negotiate handling the delayed contexts free by flushing the
workqueue before calling i915_gem_context_fini() and performing the final
free of the kernel context directly
v3: Flush deferred frees before new context allocations
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620110547.15947-2-chris@chris-wilson.co.uk

Create a substruct to hold all the global context state under
drm_i915_private.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170620110547.15947-1-chris@chris-wilson.co.uk

If we take a reference to the object/vma when it is first used in an
execbuf, we can keep that reference until the object's file-local handle
is closed. Thereby saving a frequent ref/unref pair.
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

The advent of full-ppgtt lead to an extra indirection between the object
and its binding. That extra indirection has a noticeable impact on how
fast we can convert from the user handles to our internal vma for
execbuffer. In order to bypass the extra indirection, we use a
resizable hashtable to jump from the object to the per-ctx vma.
rhashtable was considered but we don't need the online resizing feature
and the extra complexity proved to undermine its usefulness. Instead, we
simply reallocate the hastable on demand in a background task and
serialize it before iterating.

In non-full-ppgtt modes, multiple files and multiple contexts can share
the same vma. This leads to having multiple possible handle->vma links,
so we only use the first to establish the fast path. The majority of
buffers are not shared and so we should still be able to realise
speedups with multiple clients.

v2: Prettier names, more magic.
v3: Many style tweaks, most notably hiding the misuse of execobj[].rsvd2
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>

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

This was dropped in 2013 because it was supposed to have been included
with Ben's full-ppgtt patch set. It never landed and the bug remains.

References: https://bugs.freedesktop.org/show_bug.cgi?id=65845
Link: http://patchwork.freedesktop.org/patch/msgid/1372175222-27622-1-git-send-email-mika.kuoppala@intel.comSigned-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170616132849.29597-1-chris@chris-wilson.co.uk

Explicitly assign the default priority, and give it a name. After much
discussion, we have chosen to call it I915_PRIORITY_NORMAL!
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: NMika Kuoppala <mika.kuoppala@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170517121007.27224-7-chris@chris-wilson.co.uk

Pre-calculate engine context size based on engine class and device
generation and store it in the engine instance.

v2:
- Squash and get rid of hw_context_size (Chris)

v3:
- Move after MMIO init for probing on Gen7 and 8 (Chris)
- Retained rounding (Tvrtko)
v4:
- Rebase for deferred legacy context allocation
Signed-off-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniele Ceraolo Spurio <daniele.ceraolospurio@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Zhenyu Wang <zhenyuw@linux.intel.com>
Cc: intel-gvt-dev@lists.freedesktop.org
Acked-by: NTvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: NChris Wilson <chris@chris-wilson.co.uk>

Almost from the outset for execlists, we used deferred allocation of the
logical context and rings. Then we ported the infrastructure for pinning
contexts back to legacy, and so now we are able to also implement
deferred allocation for context objects prior to first use on the legacy
submission.

v2: We still need to differentiate between legacy engines, Joonas is
fixing that but I want this first ;) (Joonas)
Signed-off-by: NChris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170427104651.22394-1-chris@chris-wilson.co.ukReviewed-by: NJoonas Lahtinen <joonas.lahtinen@linux.intel.com>