Bug #1240909 “[performance] Restore support for better-than-trip...” : Bugs : Mir

Daniel van Vugt (vanvugt) on 2013-10-18

summary:	- [feature] Restore support for dynamic switching to double-buffering + [enhancement] Restore support for dynamic switching to double-buffering
tags:	added: enhancement removed: feature

Daniel van Vugt (vanvugt) on 2013-10-18

Changed in mir:
status:	Triaged → In Progress

Daniel van Vugt (vanvugt) on 2013-10-18

Changed in mir:
milestone:	none → 0.1.0

Daniel van Vugt (vanvugt) on 2013-10-22

Changed in mir:
status:	In Progress → Triaged

Daniel van Vugt (vanvugt) on 2013-10-22

Changed in mir:
milestone:	0.1.0 → none

Daniel van Vugt (vanvugt) on 2014-07-09

Changed in mir:
status:	Triaged → In Progress
milestone:	none → 0.5.0

Daniel van Vugt (vanvugt) on 2014-07-16

Changed in mir:
milestone:	0.5.0 → 0.6.0

Daniel van Vugt (vanvugt) on 2014-07-17

tags:

added: performance

Daniel van Vugt (vanvugt) on 2014-07-18

summary:	- [enhancement] Restore support for dynamic switching to double-buffering + [performance] Restore support for dynamic switching to double-buffering
tags:	added: rtm14

Daniel van Vugt (vanvugt) on 2014-07-31

Changed in mir:
milestone:	0.6.0 → 0.7.0

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2014-08-04: Re: [performance] Restore support for dynamic switching to double-buffering

#1

I'm not happy with the reliability of the heuristic right now. If we get it wrong then performance will get worse, not better. Best to not depend on this enhancement to rtm14. It may make it in still, but not sure.

tags:

removed: rtm14

Daniel van Vugt (vanvugt) on 2014-08-28

Changed in mir:
milestone:	0.7.0 → 0.8.0

Daniel van Vugt (vanvugt) on 2014-09-02

Changed in mir:
milestone:	0.8.0 → none
milestone:	none → 0.8.0

Daniel van Vugt (vanvugt) on 2014-09-09

Changed in mir:
milestone:	0.8.0 → 0.9.0

Daniel van Vugt (vanvugt) on 2014-09-10

Changed in mir:
milestone:	0.9.0 → 0.8.0

Daniel van Vugt (vanvugt) on 2014-09-25

Changed in mir:
milestone:	0.8.0 → none
status:	In Progress → Triaged

Daniel van Vugt (vanvugt) on 2014-11-23

Changed in mir:
status:	Triaged → In Progress
milestone:	none → 0.10.0

Daniel van Vugt (vanvugt) on 2015-01-08

Changed in mir:
milestone:	0.10.0 → 0.11.0
milestone:	0.11.0 → none
status:	In Progress → Triaged

Daniel van Vugt (vanvugt) on 2015-02-20

summary:

- [performance] Restore support for dynamic switching to double-buffering
+ [performance] Restore support for double-buffering by default

Daniel van Vugt (vanvugt) on 2015-02-20

Changed in mir:
milestone:	none → 0.12.0
status:	Triaged → In Progress
description:	updated

Revision history for this message

PS Jenkins bot (ps-jenkins) wrote on 2015-02-23: Re: [performance] Restore support for double-buffering by default

#2

Fix committed into lp:mir at revision 2334, scheduled for release in mir, milestone 0.12.0

Changed in mir:
status:	In Progress → Fix Committed

Kevin DuBois (kdub) on 2015-02-24

Changed in mir:
milestone:	0.12.0 → 0.13.0

Daniel van Vugt (vanvugt) on 2015-03-10

Changed in mir (Ubuntu):
importance:	Undecided → Medium
status:	New → Triaged

Cemil Azizoglu (cemil-azizoglu) on 2015-03-18

Changed in mir:
importance:	Medium → High

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-03-24:

#3

"Won't Fix" for 0.12 because in retesting yesterday we've confirmed too many parts of Unity8 suffer with this on arm devices (I know, it was all looking good a month ago). Needs more investigation but most likely we're blocked on Unity8 (especially the dash) getting optimized to use less render time.

Cemil Azizoglu (cemil-azizoglu) on 2015-03-26

no longer affects:

mir/0.12

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-04-30:

#4

Fix reverted from 0.13.0 too. But the ddouble branch is still in progress as an alternate future fix.

Changed in mir:
status:	Fix Committed → In Progress

Daniel van Vugt (vanvugt) on 2015-05-01

Changed in mir:
milestone:	0.13.0 → 0.14.0

Revision history for this message

PS Jenkins bot (ps-jenkins) wrote on 2015-06-17:

#5

Fix committed into lp:mir at revision None, scheduled for release in mir, milestone 0.14.0

Changed in mir:
status:	In Progress → Fix Committed

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-07-16:

#6

Apparently the fix got reverted in the 0.14 branch. It's still present in 0.15 but sounds like it will need fixing.

Changed in mir:
milestone:	0.14.0 → 0.15.0

Daniel van Vugt (vanvugt) on 2015-07-27

Changed in mir:
status:	Fix Committed → In Progress

Alberto Aguirre (albaguirre) on 2015-07-31

Changed in mir:
milestone:	0.15.0 → 0.16.0

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-07-31:

#7

This has actually landed in 0.15 already. But I'm actively trying to disable it.

Revision history for this message

PS Jenkins bot (ps-jenkins) wrote on 2015-07-31:

#8

Fix committed into lp:mir at revision None, scheduled for release in mir, milestone 0.16.0

Changed in mir:
status:	In Progress → Fix Committed

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-07-31:

#9

The branch that landed was a revert, not a fix. See also bug 1476201.

Changed in mir:
status:	Fix Committed → In Progress

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-09-14:

#10

Not in progress right now.

First we need the fix for bug 1476201 to land. But that's still not enough. I find then that our droids tend to exaggerate the frequency and core scale down, making performance stuttery.

So there are at least two prerequisites that need addressing before we can progress on this bug again:
bug 1476201
bug 1488386

Changed in mir:
milestone:	0.16.0 → none
status:	In Progress → Triaged

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2015-09-14:

#11

Bug 1494795 may also be a blocking factor.

Daniel van Vugt (vanvugt) on 2015-09-14

no longer affects:

mir/0.5

Daniel van Vugt (vanvugt) on 2016-02-04

Changed in mir:
status:	Triaged → In Progress
milestone:	none → 0.20.0
summary:	- [performance] Restore support for double-buffering by default + [performance] Restore support for better-than-triple buffering by + default.

Alan Griffiths (alan-griffiths) on 2016-02-12

Changed in mir:
milestone:	0.20.0 → 0.21.0

Revision history for this message

Daniel van Vugt (vanvugt) wrote on 2016-02-18:

#12

I'm done with this for now.

The issue is not that it's technically difficult, but that our primary customer is Unity8. And Unity8 is not presently fast enough on mobile devices to keep up a smooth frame rate, either with constant double buffers or with dynamic queue scaling.

Changed in mir:
milestone:	0.21.0 → none
status:	In Progress → Won't Fix
Changed in mir (Ubuntu):
status:	Triaged → Won't Fix

Daniel van Vugt (vanvugt) on 2017-01-15

Changed in mir (Ubuntu):
status:	Won't Fix → Triaged
Changed in mir:
importance:	High → Medium
milestone:	none → 1.0.0
status:	Won't Fix → In Progress
description:	updated

Revision history for this message

Mir CI Bot (mir-ci-bot) wrote on 2017-01-17:

#13

Fix committed into lp:mir at revision None, scheduled for release in mir, milestone 1.0.0

Changed in mir:
status:	In Progress → Fix Committed

Daniel van Vugt (vanvugt) on 2017-01-20

Changed in mir:
milestone:	1.0.0 → 0.26.0

Revision history for this message

Launchpad Janitor (janitor) wrote on 2017-02-01:

#14

Download full text (7.2 KiB)

This bug was fixed in the package mir - 0.26.0+17.04.20170126.3-0ubuntu1

---------------
mir (0.26.0+17.04.20170126.3-0ubuntu1) zesty; urgency=medium

  [ Daniel van Vugt ]
  * New upstream release 0.26.0 (https://launchpad.net/mir/+milestone/0.26.0)
    - ABI summary:
      . mirclient ABI unchanged at 9
      . mirserver ABI bumped to 43
      . mircommon ABI unchanged at 7
      . mirplatform ABI unchanged at 14
      . mirprotobuf ABI unchanged at 3
      . mirplatformgraphics ABI unchanged at 11
      . mirclientplatform ABI unchanged at 5
      . mirinputplatform ABI unchanged at 6
      . mircore ABI unchanged at 1
    - Enhancements:
      . New/improved toolkit APIs: MirInputConfig and related functions,
        MirWindow and related functions, DisplayConfig and related functions,
        MirScreencastSpec and related functions,
      . Support for configuring input configuration
      . Introduce toolkit extension mechanism for platform specific APIs.
      . Toolkit extensions for: mesa_drm_auth, set_gbm_device and
        window_coordinate_translation
      . Screencasting to a specific MirBuffer.
      . Add DisplayConfigurationController::base_configuration() so downstreams
        can get the base configuration. (Weirdly they can already set it.).
      . x11 platform: allow adjustable scale parameter.
      . Added EDID support: client API, server logging and in mirout.
      . mirout: Add newer attributes only available in the new display
        config API: scaling factor, subpixel arrangement and form factor.
      . mirout: Log the orientation and logical size of each output too.
      . Replace the mir::Server-overridable Reports with Observers.
      . Add xkbcommon to mirclient.pc Requires.private.
      . Deprecate legacy toolkit APIs that will be removed in Mir 1.0
      . Introduced 'client-side vsync', which dramatically reduces latency
        from the client to the screen (particularly for nested servers like
        Unity8).
    - Bugs fixed:
      . [performance] Restore support for better-than-triple buffering by
        default. (LP: #1240909)
      . Frame rate is artificially low on Diamondville Intel Atom systems due
        to aggressive power management (LP: #1388490)
      . [testsfail] failure in CI in
        AndroidInputReceiverSetup.slow_raw_input_doesnt_cause_frameskipping
        (LP: #1394369)
      . [trusted prompt sessions] Can't open two prompt sessions at the same
        time (LP: #1494197)
      . Changing scale, formFactor or DPI in display configuration causes
        renderer teardown/recreate unnecessarily (LP: #1556142)
      . [testsfail] ApplicationNotRespondingDetection.failure_to_pong_is_
        noticed (LP: #1570327)
      . CI failure in TestClientInput.receives_one_touch_event_per_frame
        (LP: #1570698)
      . Mir-on-X mouse input is jerky/stuttery compared to Mir-on-KMS
        (LP: #1576600)
      . [regression] Two fingers in mir_proving_server now resizes/moves app
        windows (two finger apps unusable) (LP: #1586311)
      . Pointer/cursor input lag in unity8 session (LP: #1591328)
      . PointerConfinement.test_we_update_our_confined_region_on_a_resize
      ...

This bug was fixed in the package mir - 0.26.0+17.04.20170126.3-0ubuntu1

---------------
mir (0.26.0+17.04.20170126.3-0ubuntu1) zesty; urgency=medium

[ Daniel van Vugt ]
  * New upstream release 0.26.0 (https://launchpad.net/mir/+milestone/0.26.0)
    - ABI summary:
      . mirclient ABI unchanged at 9
      . mirserver ABI bumped to 43
      . mircommon ABI unchanged at 7
      . mirplatform ABI unchanged at 14
      . mirprotobuf ABI unchanged at 3
      . mirplatformgraphics ABI unchanged at 11
      . mirclientplatform ABI unchanged at 5
      . mirinputplatform ABI unchanged at 6
      . mircore ABI unchanged at 1
    - Enhancements:
      . New/improved toolkit APIs: MirInputConfig and related functions,
        MirWindow and related functions, DisplayConfig and related functions,
        MirScreencastSpec and related functions,
      . Support for configuring input configuration
      . Introduce toolkit extension mechanism for platform specific APIs.
      . Toolkit extensions for: mesa_drm_auth, set_gbm_device and
        window_coordinate_translation
      . Screencasting to a specific MirBuffer.
      . Add DisplayConfigurationController::base_configuration() so downstreams
        can get the base configuration. (Weirdly they can already set it.).
      . x11 platform: allow adjustable scale parameter.
      . Added EDID support: client API, server logging and in mirout.
      . mirout: Add newer attributes only available in the new display
        config API: scaling factor, subpixel arrangement and form factor.
      . mirout: Log the orientation and logical size of each output too.
      . Replace the mir::Server-overridable Reports with Observers.
      . Add xkbcommon to mirclient.pc Requires.private.
      . Deprecate legacy toolkit APIs that will be removed in Mir 1.0
      . Introduced 'client-side vsync', which dramatically reduces latency
        from the client to the screen (particularly for nested servers like
        Unity8).
    - Bugs fixed:
      . [performance] Restore support for better-than-triple buffering by
        default. (LP: #1240909)
      . Frame rate is artificially low on Diamondville Intel Atom systems due
        to aggressive power management (LP: #1388490)
      . [testsfail] failure in CI in
        AndroidInputReceiverSetup.slow_raw_input_doesnt_cause_frameskipping
        (LP: #1394369)
      . [trusted prompt sessions] Can't open two prompt sessions at the same
        time (LP: #1494197)
      . Changing scale, formFactor or DPI in display configuration causes
        renderer teardown/recreate unnecessarily (LP: #1556142)
      . [testsfail] ApplicationNotRespondingDetection.failure_to_pong_is_
        noticed (LP: #1570327)
      . CI failure in TestClientInput.receives_one_touch_event_per_frame
        (LP: #1570698)
      . Mir-on-X mouse input is jerky/stuttery compared to Mir-on-KMS
        (LP: #1576600)
      . [regression] Two fingers in mir_proving_server now resizes/moves app
        windows (two finger apps unusable) (LP: #1586311)
      . Pointer/cursor input lag in unity8 session (LP: #1591328)
      . PointerConfinement.test_we_update_our_confined_region_on_a_resize
        (LP: #1606418)
      . [ FAILED ] ClientLatency.throttled_input_rate_yields_lower_latency
        (LP: #1624188)
      . Valgrind failure on mir_acceptance_tests - leak on incoming buffer in
        Requests::free_buffer(int) (LP: #1628794)
      . Add support for returning monitor EDIDs (LP: #1634868)
      . When pointer motion crosses window borders SurfaceInputDispatcher
        injected events increase accumulated relative motion (LP: #1639749)
      . [ FAILED ] ClientLatency.dropping_latency_is_limited_to_one (AKA
        dropping_latency_is_closer_to_zero_than_one) (LP: #1640366)
      . Server::override_the_coordinate_translator() cannot be effectively used
        downstream (LP: #1641166)
      . [regression] FTBFS when cross compiling to yakkety from zesty:
        test_client_extensions.cpp: error: ‘Ne’ was not declared in this scope
        (LP: #1642473)
      . [regression] mir_proving_server crashes with std::exception::what:
        add_options() must be called before the_options() (LP: #1642504)
      . [ FAILED ] NestedServer.nested_platform_connects_and_disconnects
        (LP: #1643432)
      . googletest 1.8.0-2 (on zesty) breaks existing builds [add_library
        cannot create target "gmock" ...] (LP: #1644062)
      . [regression] Those mir_demo_client_* which default to fullscreen now
        ignore the size parameter (-s WIDTHxHEIGHT) (LP: #1644075)
      . Shell doesn't know when base display config has changed (LP: #1644189)
      . miral::ActiveOutputsMonitor (and therefore miral-shell --window-manager
        tiling) are broken by lp:mir (LP: #1645284)
      . mir_proving_server's four finger swipe to switch apps leaks touch
        events to the clients (LP: #1646700)
      . [regression] Cross compiling from zesty to yakkety fails with *** No
        rule to make target 'gmock/libs/googlemock/gtest/libgtest.a' (LP:
        #1646704)
      . [ FAILED ] ThreadedDispatcherDeathTest.destroying_dispatcher_from_a_
        callback_is_an_error (LP: #1647573)
      . [regression] eglapps now all have window title "default" (LP: #1647575)
      . Package mirtest-dev is missing a dependency on mir-renderer-gl-dev
        (LP: #1651391)
      . EDID data is missing in nested servers (LP: #1651633)
      . [regression] Latency for fullscreen interval 0 clients (e.g. games and
        benchmarks) increased with the introduction of nested passthrough
        (LP: #1651638)
      . Race between MirConnection::released() and
        MirConnection::~MirConnection() causes deadlocks, possible crashes and
        memory corruption (LP: #1653658)
      . DisplayConfigurationTest.configure_session_removed_display failure
        (LP: #1653789)
      . [regression] Development headers reference unknown header
        "mir_toolkit/mir_input_device_types.h" so some clients can't build at
        all any more (LP: #1654023)
      . [regression] mir_acceptance_tests' death tests leave behind core files
        (LP: #1654478)
      . Setting the event handler after the surface has been created does
        nothing (LP: #1654612)
      . [ FAILED ] RaiseSurfaces.key_event_with_cookie (LP: #1655293)
      . Black screen with Raspberry Pi 3 VC4 Mesa driver (LP: #1656164)
      . i386 FTBFS:
        src/platforms/eglstream-kms/server/kms_display_configuration.cpp:88:47:
        error: narrowing conversion of ‘..._drmModeConnector::connector_id’
        from ‘uint32_t {aka unsigned int}’ to ‘EGLAttrib {aka int}’ inside { }
        [-Werror=narrowing] (LP: #1656633)
      . android: overlays no longer are activated (LP: #1657755)
      . Mir FTBFS with MIR_USE_PRECOMPILED_HEADERS=OFF: error: ISO C++11
        requires at least one argument for the "..." in a variadic macro
        [-Werror] (LP: #1658604)
      . [regression] Mir cross-compile to vivid/armhf FTBFS with:
        mock_input_device_hub.h:33:18: error: ‘add_observer’ is not a type
        (LP: #1658605)
      . Caught exception at Mir/EGL driver boundary (in setSwapInterval) in
        clients and tests (LP: #1659298)
      . Support display scaling slider in unity8 (LP: #1645372)

-- Cemil Azizoglu <cemil.azizoglu@canonical.com>  Thu, 26 Jan 2017 23:29:33 +0000

Changed in mir (Ubuntu):
status:	Triaged → Fix Released

Daniel van Vugt (vanvugt) on 2017-02-02

Changed in mir:
status:	Fix Committed → Fix Released

Mir

[performance] Restore support for better-than-triple buffering by default.

Bug Description

Related branches

Other bug subscribers

Remote bug watches

Affects		Status	Importance	Assigned to	Milestone
	Mir	Fix Released	Medium	Daniel van Vugt	Mir 0.26.0
	mir (Ubuntu)	Fix Released	Medium	Unassigned