We were not very carefully checking to see if an older request on the
engine was an earlier switch-to-kernel-context before deciding to emit a
new switch. The end result would be that we could get into a permanent
loop of trying to emit a new request to perform the switch simply to
flush the existing switch.
What we need is a means of tracking the completion of each timeline
versus the kernel context, that is to detect if a more recent request
has been submitted that would result in a switch away from the kernel
context. To realise this, we need only to look in our syncmap on the
kernel context and check that we have synchronized against all active
rings.
v2: Since all ringbuffer clients currently share the same timeline, we do
have to use the gem_context to distinguish clients.
As a bonus, include all the tracing used to debug the death inside
suspend.
v3: Test, test, test. Construct a selftest to exercise and assert the
expected behaviour that multiple switch-to-contexts do not emit
redundant requests.
Reported-by: Mika Kuoppala <mika.kuoppala@intel.com>
Fixes: a89d1f921c ("drm/i915: Split i915_gem_timeline into individual timelines")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Reviewed-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180524081135.15278-1-chris@chris-wilson.co.uk
To no surprise (since we've flip-flopped over the use of PIN_HIGH a few
times), doing a search by address over a pathologically fragmented
address space is exceeding slow. To protect ourselves from nearly
unbounded latency (think searching a million holes while under
struct_mutex), limit the search for the highest available hole and
fallback to best-fit if it fails.
In the pathologically fragmented case, such as igt/gem_ctx_thrash, the
effect is dramatic, bringing the runtime down from hours to seconds
(depending on how many other slow searches you hit, e.g. alloc_iova()
and alloc_vmap_area() both degrade to a slow rbtree walk after their
small cache is exhausted). For the real world, the number of search
steps is unlikely to be significant as we should only need to search
once per new context.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180521082131.13744-3-chris@chris-wilson.co.uk
PSR hardware and hence the driver code for VLV and CHV deviates a lot from
their DDI counterparts. While the feature has been disabled for a long time
now, retaining support for these platforms is a maintenance burden. There
have been multiple refactoring commits to just keep the existing code for
these platforms in line with the rest. There are known issues that need to
be fixed to enable PSR on these platforms, and there is no PSR capable
platform in CI to ensure the code does not break again if we get around to
fixing the existing issues. On account of all these reasons, let's nuke
this code for now and bring it back if a need arises in the future.
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Acked-by: Jani Nikula <jani.nikula@intel.com>
Acked-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180511230059.19387-1-dhinakaran.pandiyan@intel.com
For psr block #9, the vbt description has moved to options [0-3] for
TP1,TP2,TP3 Wakeup time from decimal value without any change to vbt
structure. Since spec does not mention from which VBT version this
change was added to vbt.bsf file, we cannot depend on bdb->version check
to change for all the platforms.
There is RCR inplace for GOP team to provide the version number
to make generic change. Since Kabylake with bdb version 209 is having this
change, limiting this change to gen9_bc and version 209+ to unblock google.
Tested on skl(bdb version 203,without options) and
kabylake(bdb version 209,212) having new options.
bspec 20131
v2: (Jani and Rodrigo)
move the 165 version check to intel_bios.c
v3: Jani
Move the abstraction to intel_bios.
v4: Jani
Rename tp*_wakeup_time to have "us" suffix.
For values outside range[0-3],default to max 2500us.
Old decimal value was wake up time in multiples of 100us.
v5: Jani and Rodrigo
Handle option 2 in default condition.
Print oustide range value.
For negetive values default to 2500us.
v6: Jani
Handle default first and then fall through for case 2.
v7: Rodrigo
Apply this change for IS_GEN9_BC and vbt version > 209
v8: Puthik
Add new function vbt_psr_to_us.
v9: Jani
Change to v7 version as it's more readable.
DK
add comment /*fall through*/ after case2.
Cc: Rodrigo Vivi <rodrigo.vivi@intel.com>
Cc: Puthikorn Voravootivat <puthik@chromium.org>
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Maulik V Vaghela <maulik.v.vaghela@intel.com>
Signed-off-by: Vathsala Nagaraju <vathsala.nagaraju@intel.com>
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
Signed-off-by: Jani Nikula <jani.nikula@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526981243-2745-1-git-send-email-vathsala.nagaraju@intel.com
L3Bank could be fused off in hardware for debug purpose, and it
is possible that subslice is enabled while its corresponding L3Bank pairs
are disabled. In such case, if MCR packet control register(0xFDC) is
programed to point to a disabled bank pair, a MMIO read into L3Bank range
will return 0 instead of correct values.
However, this is not going to be the case in any production silicon.
Therefore, we only check at initialization and issue a warning should
this really happen.
References: HSDES#1405586840
v2:
- use fls instead of find_last_bit (Chris)
- use is_power_of_2() instead of counting bit set (Chris)
v3:
- rebase on latest tip
v5:
- Added references (Mika)
- Move local variable into scope where they are used (Ursulin)
- use a new local variable to reduce long line of code (Ursulin)
v6:
- Some coding style and use more local variables for clearer
logic (Ursulin)
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Yunwei Zhang <yunwei.zhang@intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526683285-24861-1-git-send-email-yunwei.zhang@intel.com
WaProgramMgsrForCorrectSliceSpecificMmioReads dictate that before any MMIO
read into Slice/Subslice specific registers, MCR packet control
register(0xFDC) needs to be programmed to point to any enabled
slice/subslice pair. Otherwise, incorrect value will be returned.
However, that means each subsequent MMIO read will be forwarded to a
specific slice/subslice combination as read is unicast. This is OK since
slice/subslice specific register values are consistent in almost all cases
across slice/subslice. There are rare occasions such as INSTDONE that this
value will be dependent on slice/subslice combo, in such cases, we need to
program 0xFDC and recover this after. This is already covered by
read_subslice_reg.
Also, 0xFDC will lose its information after TDR/engine reset/power state
change.
References: HSD#1405586840, BSID#0575
v2:
- use fls() instead of find_last_bit() (Chris)
- added INTEL_SSEU to extract sseu from device info. (Chris)
v3:
- rebase on latest tip
v5:
- Added references (Mika)
- Change the ordered of passing arguments and etc. (Ursulin)
v7:
- Moved WA explanation Comments(Oscar)
- Rebased.
v8:
- Renamed sanitize_mcr to calculate_s_ss_select. (Oscar)
- calculate s/ss selector instead of whole mcr. (Oscar)
v9:
- Updated function name (Oscar)
- Remove redundant variables (Oscar)
v10:
- Separate pre-GEN10 and GEN11 mask. (Oscar)
Cc: Oscar Mateo <oscar.mateo@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
Signed-off-by: Yunwei Zhang <yunwei.zhang@intel.com>
Reviewed-by: Oscar Mateo <oscar.mateo@intel.com>
Signed-off-by: Mika Kuoppala <mika.kuoppala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1526683197-24656-1-git-send-email-yunwei.zhang@intel.com
'Pipe CSC enable' bit is more than just deprecated in ICL+, it was
disabled in commit 077ef1f09c ("drm/i915/icl: Don't set pipe
CSC/Gamma in PLANE_COLOR_CTL") for primary and sprite planes as it was
causing those planes to be rendered as always black but it was not
disabled in cursor plane, also causing it to be rendered as black.
As mentioned in the commit referenced above, this is a workaround
too and the CSC and gamma per plane values needs to be setup before
enable CSC and gamma again.
BSpec: 4278 and 7635
Cc: James Ausmus <james.ausmus@intel.com>
Cc: Paulo Zanoni <paulo.r.zanoni@intel.com>
Reviewed-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Signed-off-by: José Roberto de Souza <jose.souza@intel.com>
Signed-off-by: Paulo Zanoni <paulo.r.zanoni@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180518201547.15793-1-jose.souza@intel.com
Clean up the DP pipe select bits. To make the whole situation a bit
less ugly we'll start to share the same code between .get_hw_state(),
the port state asserts, and the VLV power sequencer code.
v2: Return PIPE_A for cpt/ppt when the port isn't selected by
any transcoder. Returning INVALID_PIPE explodes *somewhere*
on some machines (can't immediately see where though). This
now matches the old behaviour.
v3: Order the defines shift,mask,value (Jani)
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180518152931.13104-4-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
for_each_encoder_on_crtc() is legacy and shouldn't be used by atomic
drivers. Let's throw out intel_trans_dp_port_sel() and replace it
with intel_get_crtc_new_encoder() which looks the atomic state instead.
Since we now have to call intel_get_crtc_new_encoder() during the commit
phase we'll need to plumb in the top level atomic state. The
crtc_state->state pointers are no longer valid at that point.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180518152931.13104-3-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
nospec quite reasonably asserts that it will never be used with an index
larger than unsigned long (that being the largest possibly index into an
C array). However, our ubi uses the convention of u64 for any large
integer, running afoul of the assertion on 32b. Reduce our index to an
unsigned long, checking for type overflow first.
drivers/gpu/drm/i915/i915_query.c: In function 'i915_query_ioctl':
include/linux/compiler.h:339:38: error: call to '__compiletime_assert_119' declared with attribute error: BUILD_BUG_ON failed: sizeof(_s) > sizeof(long)
Reported-by: kbuild-all@01.org
Fixes: 84b510e22d ("drm/i915/query: Protect tainted function pointer lookup")
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Reviewed-by: Lionel Landwerlin <lionel.g.landwerlin@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180522121018.15199-1-chris@chris-wilson.co.uk
When testing reset, we wait for 1s on the main thread for the hang to
start. Meanwhile, we continue submitting requests on all the background
threads, and we may have more threads than cores and so potentially
starve the waiter from being woken within the timeout. As the hang
timeout and the active timeouts are the same, it is hard to distinguish
which caused the timeout. Bump the active thread timeouts to 5s,
compared to the 1s timeout for the hang, so that we preferentially
report the hang timing out, while hopefully ensuring that we do at least
wake up the hang thread first before declaring the background active
timeout.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Link: https://patchwork.freedesktop.org/patch/msgid/20180517142442.16979-1-chris@chris-wilson.co.uk
Reviewed-by: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
We want to be able to reset the GPU from inside a timer callback
(hardirq context). One step requires us to copy the default context
state over to the guilty context, which means we need to plan in advance
to have that object accessible from within an atomic context. The atomic
context prevents us from pinning the object or from peeking into the
shmemfs backing store (all may sleep), so we choose to pin the
default_state into memory when the engine becomes active. This
compromise allows us to swap out the default state when idle, when
required.
References: 5692251c25 ("drm/i915/lrc: Scrub the GPU state of the guilty hanging request")
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/20180518090212.5349-2-chris@chris-wilson.co.uk
Use intel_ddi_dp_voltage_max() for HSW/BDW too instead of letting these
fall through the if ladder in a weird way. This function will look at
the actual buf trans tables we have for HSW/BDW to determine the max
vswing level.
It looks to me like the current code leads HSW port A down the IVB port
A path, HSW port B+ and BDW fall through to the very end. Both cases do
result in the correct max vswing level 2, but it's very hard to see that
from the code.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180517170309.28630-1-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
