Android 11 AOSP Changes

Changes from 11.0.0_r37 (RQ2A.210505.003) to 11.0.0_r38 (RQ3A.210605.005):

Warning Releases with no significant changes other than version bump in platform/build component are likely to only feature proprietary binary blob (e.g. firmwares) changes.

Newly Added Components (0):

None

Updated Components (103):

  • device/generic/car with 1 change(s)
    • 3c78a69 : Disable adoptable_storage for car emulator

  • device/google/bonito with 1 change(s)
    • 29ba1b1 : Move biometric config to device/google

  • device/google/bramble with 3 change(s)
    • 27f0241 : audio: update speaker firmware and tuning.
    • 74b4ee6 : Temporarily disable ANT_INFO
    • 4a2dc08 : Move config_biometric_sensors to device/google

  • device/google/coral with 7 change(s)
    • 24728b7 : Update Coral SVN to 38
    • d911b53 : Update Coral SVN to 37
    • 0850221 : fsync after block device writes
    • 369dbe6 : Update Coral SVN to 36
    • e972bea : Move config_biometric_sensors to device/google
    • 56686f3 : Update Coral SVN to 35
    • a58f1a4 : Configure Pixel 4XL to show battery-unknown notif

  • device/google/crosshatch with 6 change(s)
    • d3bb4c2 : Update Crosshatch SVN to 51
    • 8c702fb : Update Crosshatch SVN to 50
    • 8db05c3 : Update Crosshatch SVN to 49
    • 2278b89 : Update Crosshatch SVN to 48
    • f27c9f8 : Move config_biometric_sensors to device/google
    • c9bdc80 : Update Crosshatch SVN to 47

  • device/google/cuttlefish with 2 change(s)
    • 77dd584 : Supporting override of BT HAL in cf downstream devices
    • 9bdff85 : Allow userdata image size to be overridden

  • device/google/redbull with 9 change(s)
    • e31e185 : Update Redbull SVN to 27
    • 8eedcc3 : Update Sunfish SVN to 26
    • 899335e : Update Redbull SVN to 25
    • 626479c : config: Override channel utilization estimation
    • b831d07 : Wifi: Enable BTM offload feature
    • fd0c59f : fsync after block device writes
    • e5be87b : Update Redbull SVN to 24
    • 29230f3 : wifi - Move hostapd_cli to userdebug ROM only
    • 6e015b3 : Update Redbull SVN to 23

  • device/google/redbull-sepolicy with 7 change(s)
    • 00d2e42 : camera: correct rls sepolicy
    • 3074803 : Add permissions to set recovery properties and disable nfc
    • 432c71b : Revert "Add policy for Bluetooth uart debug"
    • b3d8cf5 : [IMS-VT] Dump rtp crash information
    • 5d44624 : Suppress the denials related to persist.vendor.service.bdroid.bdaddr
    • 42a8188 : Add policy for Bluetooth uart debug
    • 5bd17ea : Allow to set thermal property

  • device/google/redfin with 4 change(s)
    • cf2ac10 : audio: update speaker firmware and tuning.
    • b31cd78 : Temporarily disable ANT_INFO
    • 2c6c597 : audio: update mixer paths table
    • 79b1790 : Move config_biometric_sensors to device/google

  • device/google/sunfish with 6 change(s)
    • f64a619 : Update Sunfish SVN to 23
    • 8fa166c : Update Sunfish SVN to 22
    • 1ec68a9 : fsync after block device writes
    • f03d3ba : Update Sunfish SVN to 21
    • afd131c : Move config_biometric_sensors to device/google
    • 058e4dc : Update Sunfish SVN to 20

  • device/google/trout with 15 change(s)
    • ec46f03 : Add system properties to configure audio HAL
    • 798bbb9 : Cleanup build warnings for Audio HAL
    • a5d8240 : Add unit tests for GRPC vehicle HAL
    • 97cb3d6 : Adding support for trout derived devices to override the ueventd file
    • ce37f32 : Update the build script
    • 0e1aa20 : Include string.h as strerror depends on it
    • e5656ed : Gracefully shutdown worker threads when destructing GRPC VHAL
    • 47d96b8 : Factor the client/server Library Out
    • 1da488c : Modifying the sensor HAL to support upstream patches
    • 4f346bc : Add a test for iio_utils
    • c8cbe26 : Refactor iio_utils to aid testing
    • f777177 : Add EVS configurations for Trout
    • 9faa825 : Build EVS components for Trout
    • 8c091fa : Create a helper object to wrap polling on multiple file descriptors
    • 0383e5f : Move the threading logic for Sensor HAL into a separate class

  • platform/art with 9 change(s)
    • 7c7cae7 : Fix dm-test invocation
    • c802c37 : dex2oat_vdex_test: add missing dependency on core boot image.
    • c60638a : dex2oat_vdex_test: add missing dependency on core boot image.
    • 2c4ee9b : Do not accept vdex with dex sections from .dm files
    • faf5aa5 : Do not accept vdex with dex sections from .dm files
    • e28d339 : Do not accept vdex with dex sections from .dm files
    • ad72901 : Do not crash on out-of-date oat files.
    • eb21b02 : Do not accept vdex with dex sections from .dm files
    • c762c01 : Added extra safety in veridex

  • platform/build with 15 change(s)
    • 5f402c6 : Version bump to RQ3A.210605.005 [core/build_id.mk]
    • 8b05b10 : Version bump to RQ3A.210605.004 [core/build_id.mk]
    • a1cad6e : Version bump to RQ3A.210605.003 [core/build_id.mk]
    • 21fbe01 : Version bump to RQ3A.210605.002 [core/build_id.mk]
    • e62ef8b : Update Security String to 2021-06-05
    • 6f23fe1 : Update Security String to 2021-05-01
    • 71dd75b : Version bump to RQ3A.210310.001 [core/build_id.mk]
    • 1f4aeb1 : Update Security String to 2021-04-05
    • 927a8e7 : Version bump to RQ3A.210129.002 [core/build_id.mk]
    • 0e3ba7d : Update Security String to 2021-03-01
    • 9c46652 : Update Security String to 2021-03-05
    • 524428a : Switch default execution strategy to racing for developer builds
    • a92f3e0 : Use unified cas uploads / downloads
    • ba9a28b : Source rbesetup.sh from any location inside of a repo directory.
    • e5da4aa : Version bump to RQ3A.210106.001 [core/build_id.mk]

  • platform/build/soong with 3 change(s)
    • 459beab : Minimal license feature.
    • 58d85b8 : Minimal license feature.
    • dc5be90 : Minimal license feature.

  • platform/cts with 286 change(s)
    • cb1e541 : [RESTRICT AUTOMERGE]: STS test fix for CVE-2020-0224
    • c286daa : [RESTRICT AUTOMERGE]: STS test fix for CVE-2018-11261
    • 16e19d9 : Sleep 12 seconds to let potential BAL grace period expire.
    • ab6d039 : Reverting CVE-2017-0630
    • 5a2fbe8 : Revert "[RESTRICT AUTOMERGE] CTS test for Android Security b/72323107"
    • 5fa0de1 : CTS test for Android Security b/130024844
    • 7b12d46 : [RESTRICT AUTOMERGE] CTS test for Android Security b/130024844
    • 354867c : [RESTRICT AUTOMERGE] CTS test for Android Security b/130024844
    • 35e290e : [RESTRICT AUTOMERGE] CTS test for Android Security b/130024844
    • db021bd : Add some sleep time to allow grace period to expire.
    • 9a0b1cc : [RESTRICT AUTOMERGE] Update import statement in CVE_2021_0394.java
    • 98febc2 : @Option to set kptr_restrict in STS
    • 9462b85 : Updated CTS test for Android Security b/172655291
    • 48051dc : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/172655291
    • 9b0d5a8 : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/172655291
    • 857055b : STS test for Android Security CVE-2021-0485
    • 0c96bb9 : WebView: avoid public URL for CTS tests
    • c9ab850 : Revert "Revert "[RESTRICT AUTOMERGE]: CTS test for Android Security b/33139050""
    • 8c09d52 : [RESTRICT AUTOMERGE] CTS test for Android Security CVE-2017-0589
    • 7ef02bb : Fix STS test for CVE-2021-0399
    • f2467b4 : Fix STS test for CVE-2021-0399
    • e7eccad : STS test for Android Security CVE-2021-0486
    • 36725c7 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150706594
    • b3d8759 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150706594
    • 4c2cd98 : Add more logging and locking to StagefrightTest
    • 2244f65 : CTS test for Android Security b/37761553
    • cb0353e : Add test case that loading font files in another package
    • 5f9b54d : [RESTRICT AUTOMERGE]: Crash Parser ignore CANNOT LINK EXECUTABLE
    • a2415f8 : pass an empty file_contexts file to secilc
    • ec13b69 : Remove `@SecurityTest` from CertificateTest.
    • 80ae445 : STS add check exclude for CVE-2018-9537
    • 0485409 : STS add NFC skips for devices without NFC
    • a7d841d : STS add NFC skips for devices without NFC
    • 5fae97c : STS add NFC skips for devices without NFC
    • ef81e53 : STS add NFC skips for devices without NFC
    • cc4e6e2 : Remove Oom Catcher
    • 9b612bb : Remove Oom Catcher
    • 0d8c7a5 : Remove Oom Catcher
    • 8815be2 : STS test for Android Security CVE-2021-0428 [RESTRICT AUTOMERGE]
    • 9616dd3 : CTS test for Android Security b/114237888
    • b1dab76 : CTS test for Android Security b/111660010
    • e98d585 : Don't configure sensor with period higher than max delay
    • 8b26dd6 : STS remove test for CVE-2017-13279
    • a879dd5 : STS remove test for CVE-2017-13279
    • 2fa4be6 : CTS: restriction on overridden min size for PiP
    • 76d0c64 : CTS test for Android Security b/120499324
    • b8f3b4f : CTS test for Android Security b/151644303
    • 3f12af1 : [RESTRICT AUTOMERGE] CTS test for Android Security b/120499324
    • 254dd84 : CTS test for Android Security b/122320256
    • e9e25de : [RESTRICT AUTOMERGE] CTS test for Android Security b/151644303
    • 81819af : [RESTRICT AUTOMERGE] CTS test for Android Security b/120499324
    • 6174a57 : [RESTRICT AUTOMERGE] CTS test for Android Security b/122320256
    • 6b1b2d3 : [RESTRICT AUTOMERGE] CTS test for Android Security b/151644303
    • 741e23f : [RESTRICT AUTOMERGE] CTS test for Android Security b/120499324
    • 71a13d3 : [RESTRICT AUTOMERGE] CTS test for Android Security b/122320256
    • e26caec : [RESTRICT AUTOMERGE] CTS test for Android Security b/151644303
    • 6e570a3 : CTS test for Android Security CVE-2018-13895
    • b6a269a : CTS test for Android Security CVE-2017-18275
    • 0d91b3a : CTS test for Android Security CVE-2019-2245
    • 00003ac : CTS test for Android Security CVE-2018-13925
    • 753a5ec : CTS test for Android Security b/139188579
    • d11edaf : CTS test for Android Security b/122316913
    • 9ffd0be : CTS test for Android Security b/112161557
    • 8517f2c : CTS test for Android Security b/178725766
    • 71ecbb2 : [RESTRICT AUTOMERGE] CTS test for Android Security b/139188579
    • 27113f9 : [RESTRICT AUTOMERGE] CTS test for Android Security b/122316913
    • f86e51c : CTS test for Android Security b/147664838
    • 74618aa : [RESTRICT AUTOMERGE] CTS test for Android Security b/112161557
    • 87d0cdf : [RESTRICT AUTOMERGE] CTS test for Android Security b/139188579
    • 283b26b : [RESTRICT AUTOMERGE] CTS test for Android Security b/122316913
    • 0ecaab8 : [RESTRICT AUTOMERGE] CTS test for Android Security b/147664838
    • 7baa07c : [RESTRICT AUTOMERGE] CTS test for Android Security b/130821293
    • 56e1f8d : [RESTRICT AUTOMERGE] CTS test for Android Security b/112161557
    • d585d75 : [RESTRICT AUTOMERGE] CTS test for Android Security b/139188579
    • 5b44f92 : [RESTRICT AUTOMERGE] CTS test for Android Security b/147664838
    • 9244683 : [RESTRICT AUTOMERGE] CTS test for Android Security b/130821293
    • 4687dd9 : [RESTRICT AUTOMERGE] CTS test for Android Security b/112161557
    • f75e71e : Updated CTS test for b/120426980
    • 1485768 : CTS test for Android Security b/174243830
    • 1508853 : CTS test for Android Security b/120789744
    • a17b598 : [RESTRICT AUTOMERGE] CTS test for Android Security b/120789744
    • 6c8973d : CTS test for Android Security b/114237888
    • 799ba13 : CTS test for Android Security b/111660010
    • 1f431d5 : [RESTRICT AUTOMERGE] CTS test for Android Security b/114237888
    • 1356098 : [RESTRICT AUTOMERGE] CTS test for Android Security b/111660010
    • e15f9b5 : [RESTRICT AUTOMERGE] CTS test for Android Security b/114237888
    • 5c07849 : [RESTRICT AUTOMERGE] CTS test for Android Security b/111660010
    • 4c92a32 : [RESTRICT AUTOMERGE]: STS test fix for CVE-2021-0327
    • e367998 : [RESTRICT AUTOMERGE]: STS test fix for CVE-2021-0327
    • ad4ebe4 : STS test fix for CVE-2021-0305
    • 763ff92 : Make two broadcasts protected
    • 74b7f4c : Fix STS test for CVE-2021-0399
    • 314693f : STS test for Android Security CVE-2021-0305
    • 269d982 : [RESTRICT AUTOMERGE] Exclude @SecurityTest tests from CTS. These are tested in STS.
    • 5968ac6 : Remove upper bound check of getTotal* APIs in TrafficStatsTest
    • 321e204 : Resolve CtsSecurityTest issues introduced with CtsDeviceInfo
    • 6c1586c : STS test for Android Security CVE-2021-0305
    • d5888fc : Revert "[RESTRICT AUTOMERGE] CTS test for Android Security b/137284057"
    • baf326f : STS run pocs with environment variables
    • 676d402 : PocPusher print log when host file not found
    • 3762cf9 : STS Support shell expansion in containsdriver()
    • 4173fd6 : Fixed SecurityException due to Fine Location permission denied.
    • 6330289 : Add StorageEscalation tests
    • c60b68c : Fix testAppLinks_verificationStatus
    • d91d88c : CTS test for Android Security b/124524315
    • 7732ce4 : CTS test for Android Security b/172655291
    • 2c23a5f : [RESTRICT AUTOMERGE] CTS test for Android Security b/124524315
    • 17e10de : [RESTRICT AUTOMERGE] CTS test for Android Security b/122035770
    • 3e999b3 : [RESTRICT AUTOMERGE] CTS test for Android Security b/120664978
    • d218034 : [RESTRICT AUTOMERGE] CTS test for Android Security b/172655291
    • fb21f18 : [RESTRICT AUTOMERGE] CTS test for Android Security b/124524315
    • 36627c5 : [RESTRICT AUTOMERGE] CTS test for Android Security b/122035770
    • 64f3126 : [RESTRICT AUTOMERGE] CTS test for Android Security b/120664978
    • 6cd34c4 : [RESTRICT AUTOMERGE] CTS test for Android Security b/172655291
    • 7b20dd5 : Fix tests that manually override the sts suffix
    • 572c5a0 : Add a consistent suffix to all sts poc binaries
    • 85bda36 : Disable broken test
    • fab5d65 : Flush telephony metrics after CTS SubscriptionManagerTest
    • faeb79b : Defined max timeout for SyncManager CTS tests.
    • 2f3f16b : [RESTRICT AUTOMERGE] Update test certficate for KeyChain storage CtsVerifier test
    • 3fcf673 : [RESTRICT AUTOMERGE] CTS test for Android Security b/168041375
    • 7f65baf : [RESTRICT AUTOMERGE] CTS test for Android Security b/160610106
    • 2d424ea : [RESTRICT AUTOMERGE] CTS test for Android Security b/168041375
    • 3cf1f8d : [RESTRICT AUTOMERGE] CTS test for Android Security b/168041375
    • 6406b5a : [RESTRICT AUTOMERGE] CTS test for Android Security b/168041375
    • 3b286a1 : CTS test for Android Security b/115635871
    • 5677a42 : CTS test for Android Security b/32096880
    • 8a4cec6 : CTS test for Android Security b/64710201
    • 31fdbe4 : [RESTRICT AUTOMERGE] CTS test for Android Security b/160610106
    • 9ea20ba : [sts] add @SecurityTest for testing invalid font size scales
    • 3a153ae : [sts] add @SecurityTest for testing invalid font size scales
    • b2c2324 : [sts] add @SecurityTest for testing invalid font size scales
    • 9da47f5 : [sts] add @SecurityTest for testing invalid font size scales
    • b366d2a : CTS test for Android Security b/150160279
    • 1800442 : [RESTRICT AUTOMERGE] CTS test for Android Security b/115635871
    • 592ee75 : [RESTRICT AUTOMERGE] CTS test for Android Security b/32096880
    • d1a2f9e : [RESTRICT AUTOMERGE] CTS test for Android Security b/64710201
    • 5147615 : [RESTRICT AUTOMERGE] CTS test for Android Security b/115635871
    • 5606dcc : [RESTRICT AUTOMERGE] CTS test for Android Security b/32096880
    • 902034d : [RESTRICT AUTOMERGE] CTS test for Android Security b/64710201
    • 32e19b5 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160279
    • 4ab9598 : [RESTRICT AUTOMERGE] CTS test for Android Security b/147309942
    • e6f9fab : [RESTRICT AUTOMERGE] CTS test for Android Security b/123701862
    • ab37918 : [RESTRICT AUTOMERGE] CTS test for Android Security b/139738828
    • 0844e77 : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/117556220
    • a470784 : [RESTRICT AUTOMERGE] Update pacRunner utility to run proxy-resolution in loop
    • 559569c : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160279
    • 072dc56 : [RESTRICT AUTOMERGE] CTS test for Android Security b/147309942
    • e888e96 : [RESTRICT AUTOMERGE] CTS test for Android Security b/123701862
    • 22fcc4f : [RESTRICT AUTOMERGE] CTS test for Android Security b/139738828
    • 1dd10d9 : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/117556220
    • 2ebe2a9 : [RESTRICT AUTOMERGE] Update pacRunner utility to run proxy-resolution in loop
    • 70ac7e5 : CTS test for Android Security b/147309942
    • ee75844 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160279
    • 4748e67 : CTS test for Android Security b/123701862
    • 7f3450b : CTS test for Android Security b/139738828
    • 8bfdb92 : webview: deflake testRequestImageRef.
    • b0a0dda : Revert "STS test for Android Security CVE-2020-0146"
    • 445294c : Revert "STS test update for CVE-2020-0461"
    • bc3fc89 : STS test for Android Security CVE-2021-0327
    • 195e7d2 : CTS test for Android Security b/114047681
    • 24a7e05 : CTS test for Android Security b/117554809
    • 4fc740a : CTS test for Android Security b/139186193
    • 0361c2c : CTS test for Android Security b/120497583
    • d503f81 : CTS test for Android Security b/123583388
    • 4812675 : [RESTRICT AUTOMERGE] CTS test for Android Security b/114047681
    • 603c404 : [RESTRICT AUTOMERGE] CTS test for Android Security b/117554809
    • 5a55db2 : [RESTRICT AUTOMERGE] CTS test for Android Security b/139186193
    • 239ab4a : [RESTRICT AUTOMERGE] CTS test for Android Security b/120497583
    • 9ebfa12 : [RESTRICT AUTOMERGE] CTS test for Android Security b/123583388
    • 17e1afb : [RESTRICT AUTOMERGE] CTS test for Android Security b/114047681
    • 9fdf971 : [RESTRICT AUTOMERGE] CTS test for Android Security b/117554809
    • a42fdea : [RESTRICT AUTOMERGE] CTS test for Android Security b/139186193
    • e3aef5a : [RESTRICT AUTOMERGE] CTS test for Android Security b/120497583
    • 6cc6ae6 : [RESTRICT AUTOMERGE] CTS test for Android Security b/123583388
    • 6079f7f : STS test for Android Security CVE-2021-0327
    • a01da0e : CTS test for Android Security b/170662121
    • 4000b19 : CTS test for Android Security b/153910727
    • a68482c : CTS test for Android Security b/158575956
    • fda97f5 : CTS test for Android Security b/156280110
    • 766589f : CTS test for Android Security b/168718429
    • e086d00 : [RESTRICT AUTOMERGE] Fix expired test certificate
    • 3f9d2d0 : [RESTRICT AUTOMERGE] Fix leaf certificate for download manager test
    • da6901f : STS test fix for Android Security CVE-2021-0339
    • f7c436a : STS test fix for Android Security CVE-2021-0339
    • 2e4d984 : STS test fix for Android Security CVE-2021-0339
    • c66d139 : Revert "CTS test for Android Security b/159145361" part 2
    • 3fb1f93 : Revert "CTS test for Android Security b/159145361"
    • 6290dd6 : Add mainline module test skipping to CVE-2020-0385
    • 09ad5f3 : Add mainline module test skipping to CVE-2020-0384
    • 67171cf : Add mainline module test skipping to CVE-2020-0381
    • d594883 : CTS test for Android Security b/129768470
    • f1cda3d : Increase max_wait_time_ms for low_end device
    • f816c34 : CTS test for Android Security b/35421151
    • 1b0b6fc : CTS test for Android Security b/125900276
    • 3fb0a57 : CTS test for Android Security b/143155861
    • 0e7e50e : CTS test for Android Security b/143109193
    • a29fe86 : CTS test for Android Security b/143106535
    • a9f8711 : Revert "CTS test for Android security b/169161709"
    • 3bd79ce : Backfill CTS security test owners for oc-mr1 ~ qt
    • 8d4668a : [RESTRICT AUTOMERGE] Updated CTS test for b/120426980
    • 5966b02 : CTS test for Android Security b/117556220
    • ccf7b70 : Update pacRunner utility to run proxy-resolution in loop
    • 9bf395b : CTS test for Android Security b/170732441
    • fceaf08 : [RESTRICT AUTOMERGE] CTS test for Android Security b/129768470
    • 38b99d0 : [RESTRICT AUTOMERGE] CTS test for Android Security b/34749571
    • 73b2362 : [RESTRICT AUTOMERGE] CTS test for Android Security b/30033990
    • fc05674 : [RESTRICT AUTOMERGE] CTS test for Android Security b/170732441
    • df39a05 : [RESTRICT AUTOMERGE] CTS test for Android Security b/129768470
    • 7a0689b : [RESTRICT AUTOMERGE] CTS test for Android Security b/34749571
    • d1dda78 : [RESTRICT AUTOMERGE] CTS test for Android Security b/129768470
    • 9eccbb4 : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/17769851
    • c790f07 : [RESTRICT AUTOMERGE] CTS test for Android Security b/17769851
    • 97a65f9 : Updated CTS test for Android Security b/17769851
    • 0df5683 : CTS test for Android Security b/17769851
    • 9b7251e : Update device identifier tests to support getIccId
    • 6a3b526 : Turn screen on when testing with MediaRouter2TestActivity
    • 16c444b : STS test for Android Security CVE-2021-0339
    • 900b80e : STS test for Android Security CVE-2021-0339
    • 15f332f : Update SubscriptionManager subscription override tests
    • 2aee0ff : STS test for Android Security CVE-2021-0339
    • 1744b72 : Fix failing MediaRouter2(ProviderService)Test
    • cd9d5c7 : [RESTRICT AUTOMERGE] Updated CTS test for b/36389123
    • bf8e922 : [RESTRICT AUTOMERGE] Updated CTS test for b/37239013
    • 953ee95 : [RESTRICT AUTOMERGE] Updated CTS test for b/36576151
    • 3835339 : [RESTRICT AUTOMERGE] Updated CTS test for b/24441553
    • 8656ce5 : [RESTRICT AUTOMERGE] Updated CTS test for b/17769851
    • fae9e7a : [RESTRICT AUTOMERGE] Updated CTS test for b/23247055
    • e71aeda : STS Add extra logging to getDeviceUptime()
    • 711a56f : CTS test for Android Security b/36389123
    • fe77982 : CTS test for Android Security b/150160041
    • ea5ed7d : CTS test for Android Security b/150159669
    • dda7e89 : CTS test for Android Security b/150159906
    • 15ee344 : CTS test for Android Security b/35421151
    • 629580c : [RESTRICT AUTOMERGE] CTS test for Android Security b/36389123
    • 0239cff : CTS test for Android Security b/125900276
    • 6b78f14 : CTS test for Android Security b/143155861
    • ffc5b92 : CTS test for Android Security b/143109193
    • 36d008a : CTS test for Android Security b/143106535
    • 94e19c4 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160041
    • 3f676f9 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159669
    • 7cfd399 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159906
    • fe1326e : [RESTRICT AUTOMERGE] CTS test for Android Security b/35421151
    • 63a8f69 : [RESTRICT AUTOMERGE] CTS test for Android Security b/125900276
    • 68f905a : [RESTRICT AUTOMERGE] CTS test for Android Security b/143155861
    • f281b6c : [RESTRICT AUTOMERGE] CTS test for Android Security b/143109193
    • 6e3ac15 : [RESTRICT AUTOMERGE] CTS test for Android Security b/143106535
    • 7c98197 : [RESTRICT AUTOMERGE] CTS test for Android Security b/137284057
    • f2269d9 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160041
    • cbfc0b6 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159669
    • ea1dbbb : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159906
    • 755afd3 : CTS test for Android Security b/158480899
    • e34c2e7 : [RESTRICT AUTOMERGE] CTS test for Android Security b/125900276
    • 1e2b497 : [RESTRICT AUTOMERGE] CTS test for Android Security b/143155861
    • 558350d : [RESTRICT AUTOMERGE] CTS test for Android Security b/143109193
    • 0179d5e : [RESTRICT AUTOMERGE] CTS test for Android Security b/143106535
    • db9ecc2 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150160041
    • c4130fe : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159669
    • 4a69982 : [RESTRICT AUTOMERGE] CTS test for Android Security b/150159906
    • e39da10 : [RESTRICT AUTOMERGE] CTS test for Android Security b/158480899
    • 86b4757 : CTS test for Android Security b/153912750
    • 04edaa2 : CTS test for Android Security b/151668119
    • 52de74e : CTS test for Android Security b/153910465
    • b188637 : CTS test for Android Security b/153911778
    • d6b9176 : CTS test for Android Security b/144322192
    • c6d1304 : CTS test for Android Security b/153912023
    • 438dde6 : [RESTRICT_AUTOMERGE]: Update Crash Parser with abort messages
    • 13f1040 : [RESTRICT AUTOMERGE]: Fix Crash Parser appending signals/patterns
    • ae71802 : [RESTRICT AUTOMERGE]: STS bp to mk for CVE-2021-0318
    • 803b678 : [RESTRICT AUTOMERGE]: STS bp to mk for CVE-2021-0318
    • 1e6f978 : STS test for Android Security CVE-2021-0340
    • a24ee6b : RESTRICT AUTOMERGE: Add test case testFgsLocationStartFromBGWithBind
    • f611f05 : [cts] fix another font size scale test
    • 7979397 : [cts] fix another font size scale test
    • a92cec5 : [cts] fix another font size scale test
    • dc0c727 : [cts] fix another font size scale test
    • 96447f0 : STS test for Android Security CVE-2021-0314
    • b3485e9 : STS test for Android Security CVE-2021-0314
    • 166cdf4 : STS test for Android Security CVE-2021-0314
    • a4cb128 : [RESTRICT AUTOMERGE] Fix CTS test for Android Security b/72510314
    • d6e7601 : [RESTRICT AUTOMERGE] Updated CTS test for Android Security b/17769851
    • 66a880d : [RESTRICT AUTOMERGE] CTS test for Android Security b/17769851
    • d47143c : STS test for Android Security CVE-2021-0314
    • fed0d0f : [RESTRICT AUTOMERGE] CTS test for Android Security b/117556220
    • 3168a48 : [RESTRICT AUTOMERGE] CTS test for Android Security b/117556220
    • ce65d38 : Add StorageEscalation tests

  • platform/dalvik with 1 change(s)
    • da7b74d : Update bat files to not use find_java.bat

  • platform/external/boringssl with 1 change(s)
    • ac675a6 : Fix EDIPartyName parsing and GENERAL_NAME_cmp.

  • platform/external/chromium-libpac with 4 change(s)
    • 9295700 : Upgrade V8 to 8.8.278.14
    • 438381d : Upgrade V8 to 8.8.278.14
    • 7fca112 : Upgrade V8 to 8.8.278.14
    • 204ba1d : Upgrade V8 to 8.8.278.14

  • platform/external/dnsmasq with 2 change(s)
    • 303ca27 : backport: dnsmasq: fix heap overflow
    • 2c415e5 : dnsmasq: fix heap overflow

  • platform/external/flac with 1 change(s)
    • 368eb3f : libFlac: Exit at EOS in verify mode

  • platform/external/libavc with 2 change(s)
    • a073696 : Validate input dimensions in process()
    • 10910bf : avcenc: Add bitstream overflow check during emulation prevention

  • platform/external/libhevc with 1 change(s)
    • 45fb34a : encoder: Update chroma modules to avoid reading an extra byte

  • platform/external/mesa3d with 20537 change(s)
    • 4ffb28d : Revert property "default_applicable_licenses"
    • 8971ef9 : [LSC] Add LOCAL_LICENSE_KINDS to external/mesa3d
    • 5c0bad3 : UPSTREAM: egl: android: use num_planes param in createImageFromDmaBufs()
    • 1a8a0f0 : android: softpipe: Set nir headers include path
    • fd5db89 : nvc0: Fix implicit conversion build -Werror on AOSP
    • 1bf5d0a : mesa: i830: Fix -Wxor-used-as-pow build -Werror on AOSP
    • bbedfb4 : nir: Fix implicit conversion build -Werror on AOSP
    • e699a9e : ANDROID: Generate prebuilt-intermediate source files for mesa-20.3.4
    • 08b1ac1 : ANDROID: prebuild: panfrost/Android.bifrost.mk: Use precompiled intermediate sources
    • b5e75a5 : ANDROID: prebuild: compiler/Android.nir.gen.mk: Use precompiled intermediate sources
    • da3fb8c : ANDROID: aosp-gen-prebuilt: Remove obsolete prebuilt targets
    • 5603046 : ANDROID: freedreno/Android.ir3.mk: Fixup usage of lex/yacc
    • fdeee97 : ANDROID: Update Android.gen.bp to mesa-20.3.4
    • 1e6cea7 : [LSC] Add LOCAL_LICENSE_KINDS to external/mesa3d
    • 6d8c686 : VERSION: bump version to 20.3.4
    • d6562bd : docs: add release notes for 20.3.4
    • 62dc4fe : compiler: Use util/bitset.h for system_values_read
    • 74b1a83 : nir: Add a data pointer to the callback in nir_remove_dead_variables
    • c8245ae : radv: fix divide by zero with no tesselation params
    • 918c550 : spirv: Don't remove variables used by resource indexing intrinsics
    • daa0d86 : anv: Skip CCS ambiguate which preceed fast-clears
    • c2a6bbc : intel/gen12: Fix memory corruption issues in fused Gen12 parts.
    • e8b4805 : panfrost: Use normal malloc/free instead of ralloc for surfaces
    • 79eb7c2 : anv: Invalidate the correct AUX-TT entry
    • a8031d0 : anv: fix invalid programming of BLEND_STATE
    • b6310c5 : etnaviv: handle NULL views in set_sampler_views
    • 753f5bb : gallium: Fix leak of shader images on context destruction.
    • 70e586c : panfrost: Stub out set_shader_images().
    • f791884 : swr: Don't report support for shader images.
    • 6e1c91e : aco/optimizer: don't copy-prop logical phis
    • 4257b87 : aco/optimizer: don't propagate subdword temps of different size
    • 773a461 : .pick_status.json: Update to f01ea0aef8a50d2732eb0c64153903e52ed2a757
    • 431541e : mesa/main: remove leftover bumpmap code
    • b28e674 : glx: Provide glvnd wrapper for glXSwapIntervalEXT
    • f155875 : intel/fs: QUAD_SWIZZLE requires packed data
    • 92f23b6 : intel/compiler: Return 1 for immediates in regs_read
    • 5fc1fe8 : .pick_status.json: Update to 9052819ebbff07d82c3eb9adf414144df4868644
    • c5577be : frontends/va/context: don't set max_references with num_render_targets
    • 06fadee : zink: make all xfb caps depend on extension
    • 199fb00 : zink: respect feature-cap for sample-shading
    • 3efcf3b : zink: respect feature-cap for independent blending
    • 613251e : aco: don't consider a phi trivial if same's register doesn't match the def
    • 127e9ef : iris: Consider resolves after changing a resource's aux state
    • c293269 : radv: fix a sync issue with geometry shader primitives query on GFX10+
    • d038bd7 : pan/decode: Free mapped memory objects on BO unreference
    • 5fc12bc : anv: Fix stencil layout in render passes
    • 7d168e8 : .pick_status.json: Update to 3ef89b245e3e1ac4e67fea9c1b13ebeda75769d0
    • a098a70 : st/mesa: fix pbo upload/download for arrays of textures with only 1 layer
    • 2ed63e7 : ci: exclude one CTS test that timeout most of the time for RADV CI
    • 44d9a7f : .pick_status.json: Update to af9977a3d5f3378c297965e21389e36491f47e1b
    • b944f5a : Fix license_type. Contains AMD and GPL
    • 4fab4ed : swr: fix crashes caused by incorrectly reporting SSBO support
    • 79b792e : aco: fix convert_to_SDWA() check in add_subdword_definition()
    • 0d1920b : mesa: fix alpha channel of ETC2_SRGB8 decompression for !bgra
    • 381a5de : zink: handle NULL views in zink_set_sampler_views
    • 6e2e95a : glsl: only expose int64 atomics when extension is enabled
    • 0aa060c : lavapipe: fix missing piece of VK_KHR_get_physical_device_properties2
    • 8eb3a0b : Fall back on clock_gettime when timespec_get() is unavailable
    • affb8ee : device-select-layer: update for vulkan 1.2
    • e875548 : util: Fix pointer to integer conversion error when using libunwind
    • a6e9b36 : util/format: Fix pack/unpack of A1R5G5B5_UINT.
    • 4fb89e0 : .pick_status.json: Update to 3e13c1f8dfef4a4c0fd5e79bbc364f9e5f998856
    • 1d0a7bb : intel/mi_builder: fix self modifying batches
    • d4a0136 : v3d: fix dest offset in TFU setup
    • d5ee841 : iris: only set point sprite overrides if actually using points
    • 851d753 : .pick_status.json: Update to c27347b2e1883a30e023347a36bdcf86cdec4a7c
    • 205e8cd : radeonsi: invalidate compute sgprs in si_rebind_buffer
    • bfb6f66 : anv: add transfer usage for color/depth/stencil attachments
    • cb24282 : lavapipe: Defer lavapipe warning to CreateDevice
    • 3398f1c : frontends/va: Return an error if non-interlaced buffer is not supported
    • 8d20fc6 : .pick_status.json: Update to 184bbef33d1fff3520958c130f2b8e4fce17379c
    • 326d214 : docs: Add sha256sum for 20.3.3
    • 7e4d536 : VERSION: bump to 20.3.3
    • 884c86d : docs: add release notes for 20.3.3
    • 24c7a4d : vulkan/overlay: fix vkGetInstanceProcAddr self-resolving
    • b439d0e : radv: Add Android module info to linker script.
    • 3261d99 : mesa/st: Free the NIR builtins TGSI tokens after passing to the driver.
    • ed87818 : gallium/ntt: Fix leak of the per-instr liveness information.
    • adf09c9 : radv: set invariantgeom for Shadow of the Tomb Raider
    • f8033bd : radv: add RADV_DEBUG=invariantgeom
    • 97365ce : vulkan/overay: fix violation of VUID-VkDeviceCreateInfo-pNext-00373
    • 945bffe : vulkan/device-select: fix vkGetInstanceProcAddr self-resolving
    • c5f73e6 : radv/winsys: Fix offset in range merging.
    • 107e4ee : radv/winsys: Fix inequality for sparse buffer remapping.
    • bf4e0a4 : .pick_status.json: Mark 52b6adfbfb51a3b1a54964cbcb2adb0299d1252b as backported
    • bbb85a7 : util: Avoid pthread_setaffinity_np on NetBSD
    • 76ecdf2 : aco: fix unreachable() for uniform 8/16-bit nir_op_mov from VGPR
    • 6a006c3 : util: Disable memstream for Apple builds
    • e6aa51a : radv/llvm,aco: always split typed vertex buffer loads on GFX6 and GFX10+
    • a29f08b : aco: fix creating the dest vector when 16-bit vertex fetches are splitted
    • eec614e : nir/load_store_vectorize: don't ignore subgroup memory barriers
    • 2c5b592 : .pick_status.json: Update to d9c8422c4168c7433f68b97065283ebf26631d56
    • aef8fc8 : drirc: radeonsi workaround for CS:GO
    • ae1c5dc : radv: disable TC-compat HTILE in GENERAL for Detroit: Become Human
    • 9000c04 : zink: dot leak dummy_buffer
    • 39c386c : nir/algebraic: Fix broken NaN and -0.0 behavior
    • 7223390 : Revert "nir: Replace an odd comparison involving fmin of -b2f"
    • 588cf80 : gallium/util: do not perform n^2 stencil blits
    • cf5fe35 : gallium/ntt: Fix emitting UBO declarations.
    • 071a87a : ac/nir: use llvm.readcyclecounter for LLVM9+
    • e9146fe : util: add AMD CPU family enums and enable L3 cache pinning on Zen3
    • 20a067a : spir-v: Mark floating point comparisons exact
    • f121623 : st/mesa: don't affect original st_CompressedTexSubImage parameters
    • 773d683 : .pick_status.json: Update to 96ceca33c1dd69a1feed13b0e19bfc38e6f7d979
    • da6fb03 : aco: fix incorrect address calculation for load_barycentric_at_sample
    • 2f9e5ca : nir: fix missing nir_lower_pntc_ytransform.c in the makefile
    • 9e1e67b : nir: account for point-coord origin when lowering it
    • d7c09d0 : etnaviv: Fix rework ZSA into a derived state
    • b83d966 : .pick_status.json: Update to b2d000513e4a9b06d3b073913741dae47a00526e
    • 99ff47e : st/mesa: don't do glCopyPixels via blit if depth bounds test is enabled
    • 636a917 : pan/mdg: Fix promoted uniform moves with 64-bit types
    • c7430bb : nir: fix determining if an addition might overflow for phi sources
    • 2b66910 : gallium/draw: Fix intermittent failure to bind new geometry shaders.
    • 2f6f5b3 : .pick_status.json: Update to 9ef2c44ce682b7981bd7a68c65f338f1e33bb001
    • 6d2e76d : docs: Add sha256 sum for 20.3.2
    • b3877e8 : Bump version for 20.3.2 release
    • d9d99af : docs: add release notes for 20.3.2
    • 7ca951d : glsl: default to compat shaders in compat profile
    • e8a6bf6 : st/pbo: fix pbo uploads without PIPE_CAP_TGSI_VS_LAYER_VIEWPORT and skip gs
    • f892a46 : virgl: fix BGRA emulation artifacts during window resize
    • b62bd28 : .pick_status.json: Update to 6c8cc9be12dc5d6c0d2386d6addb69d8f2fb5399
    • b21d692 : .pick_status.json: Mark 967ae12931e0dc12855de903851222b2f0607878 as denominated
    • 009e2aa : zink: handle null ubos
    • b0e8fb5 : ci/deqp: Fix inverted meaning of DEQP_NO_SAVE_RESULTS.
    • 373b6eb : mesa/st: Update FP state when textures change with an ATI_fs bound.
    • 7c74888 : radeon/vcn: fix hevc 10bit profile error
    • 2fd551d : softpipe: count CS invocations for pipeline stats queries.
    • cc7a1ab : .pick_status.json: Update to 9ef23e428bd18c412499471b4e3550a489b72c3f
    • b41a3c8 : turnip: use ir3_compiler_destroy instead of ralloc_free
    • b4cbf37 : panfrost: Fix panfrost_small_padded_vertex_count for 17 vertices
    • 6e75f40 : .pick_status.json: Update to a22b85ce5e825275533ca1be9fbd47667e991356
    • fb68996 : radv: Don't skip layout transitions that only differ in render loop.
    • e2789cd : ir3: Allow tesselation to use all 32 varying slots
    • 633f60e : freedreno/a6xx: Fix assert which checks the count of shader outputs
    • eec213d : mesa: fix layered framebuffer attachment target check
    • 27e6dfb : .pick_status.json: Update to 8db0775f4520e08b3c539f9bd2d1f567639c088c
    • 5e63258 : Fix build issue when USE_3DNOW_ASM is not defined
    • 50b9777 : freedreno/a6xx: Flush depth at the end of bypass rendering, too.
    • d54de72 : spirv: Remove more dead variables
    • a46b833 : nir: Consider pointer initializers in nir_remove_dead_variables
    • d5a9377 : .pick_status.json: Mark f65750d2215242b17309b5aa8845b717bb913e77 as backported
    • d9063a0 : .pick_status.json: Update to b9fccafed616aa2906dad3c77dca0c67af5017b8
    • a5f47bc : spirv: Implement OpArrayLength for OpenGL
    • f99fc9e : zink: fix format-mapping
    • c00c8d8 : mesa/st: Finalize the texture before BlitFramebuffer from it.
    • 8232610 : nir/opt_if: split ALU from Phi more aggressively
    • 79c2aea : .pick_status.json: Update to f65750d2215242b17309b5aa8845b717bb913e77
    • 4ba1dba : aco: add block to worklist in mark_block_wqm()
    • 19b92b7 : tu: pCounterBuffers can be NULL in vkCmd*TransformFeedbackEXT()
    • 81421b7 : tu: Ignore pTessellationState if there is no tesselation shaders
    • abfbaa8 : radv: add missing DB flush after depth/stencil resolve operations
    • 8217078 : aco/ra: fix phi operand renaming
    • 59a9471 : egl: fix EGL_EXT_protected_content/surface mixup
    • cf1af4a : .pick_status.json: Update to 661922f6ac9576fc2077c5d50b53ef7edf9e8a18
    • 840a9a6 : radeonsi: fix small primitive culling with MSAA force-disabled and smoothing
    • 4a87e44 : aco/spill: only prevent rematerializable vars from being DCE'd if they haven't been renamed
    • c9497e2 : aco: fix DCE of rematerializable phi operands
    • 5636ec0 : st/mesa: set drawpixels swizzle before creating sampler view
    • cf0f7d2 : zink: fix 8 bit index handling code
    • e531245 : .pick_status.json: Update to 2d78e28ba702d366becabb8e241b861e6711c76e
    • 3ae62ab : docs: Add sha256 sums for 20.3.1
    • d906229 : VERSION: bump for 20.3.1
    • d4f4d29 : docs: add release notes for 20.3.1
    • adb6c1c : gallium/dri2: Don't forget protected content flag
    • 6c2b0d7 : anv: Add DRM_RDWR flag in anv_gem_handle_to_fd
    • 7c0e171 : .pick_status.json: Update to ada9be1ec9e14fc045086411fbf2d3cb0efbbe2f
    • 7d5c048 : radv: fix access to uninitialized radeon_bo_metadata
    • c040303 : st/pbo: fix pbo uploads without PIPE_CAP_TGSI_VS_LAYER_VIEWPORT
    • 62eea2b : zink: really fix direct image mapping offset (I mean it this time)
    • f9bc767 : nir: Redefine start/end_ip of blocks to fix NIR-to-TGSI liveness bugs.
    • f5246ef : zink: fix direct image mapping offset
    • 07d6fa2 : .pick_status.json: Update to a7fb3954a1318a6b27e1405a9e799dd8f06eaa34
    • 200a2f2 : aco/ra: use get_reg_specified() for p_extract_vector
    • 5e65a36 : aco: Use program-num_waves as maximum in scheduler.
    • ac1145f : android: fix build failure with libbacktrace
    • 18d5fe4 : .pick_status.json: Update to 84c8a35aa2ca4d4de66192933735094ed07b4aaa
    • 4631f38 : radv: do VGT_FLUSH when switching NGG - legacy on Sienna Cichlid
    • 7bf1fbe : radv: fix applying the NGG minimum vertex count requirement
    • 317526e : radv: don't count unusable vertices to the NGG LDS size
    • 7b1ee98 : radv: disable SQTT support for unsupported GPUs
    • 2ae3731 : radv: ignore other blend targets if dual-source blending is enabled
    • 67a3b8e : aco: fix combining max(-min(a, b), c) if a or b uses the neg modifier
    • bd7ccbf : zink: fail if set failed to create
    • bd01bc1 : radeonsi: fix si_get_draw_start_count count value
    • 4691bc5 : ci: .lava-test:amd64 template needs arm_build
    • c763926 : anv: fix calculation of buffer size in case dynamic size is used
    • e0d212c : .pick_status.json: Update to cf3fc79cd0ab55776bbbabe76237c272ab07133e
    • 18fd255 : turnip: no linear_to_srgb for alpha channel for gmem clear value packing
    • 5d24ae6 : iris: Fix resource ptr in resolve_sampler_views
    • 7f485f1 : softpipe: Fix swizzled texture gather of int textures.
    • 4e05bd1 : panfrost: Fix fencing
    • 93ac6a9 : panfrost: Make sure we always add a reader - write dependency when needed
    • c865522 : iris: update depth value for stages after fast clear depth
    • c20c50e : android: spirv: fix '::' typo in gen rules
    • b9dbeb5 : turnip: always set LRZ registers to zero for 3d clear/blit
    • e9407a0 : turnip: move up LRZ invalidate in CmdClearAttachments
    • 4c32042 : turnip: always emit LRZ draw state in DIRTY_DRAW_STATE path
    • 51e8c7e : .pick_status.json: Update to 3f0da800eb4b8184c24707c52d5a519abe948898
    • a8d5e57 : iris: fix memleak for query_buffer_uploader
    • 1c83ec9 : aco: use UINT64_C on 64 bit constant arguments
    • 0c2b29e : radv: disable alphaToOne feature
    • ee83c98 : zink: Cap PIPE_SHADER_CAP_MAX_CONST_BUFFERS to 32
    • c5ffe9e : zink: fix channel ordering in format-mapping
    • 17ef9b5 : gallium: do not reset buffers for unsupported stages
    • d77d3ca : .pick_status.json: Update to f93b7d14d66d8ba70d44772d1a1b6696310b7d17
    • 0cde142 : ci: Drop x86_build_old image
    • b3bfb01 : radv: fix exporting multiviews with NGG
    • 33c2883 : radv: mark GFX10.3 as a non-conformant Vulkan implementation
    • 44bed80 : radeonsi: fix regression on gpus using the radeon winsys.
    • dfadf83 : meson: Fix Clang microsoft-enum-value detection.
    • 03e8cc7 : .pick_status.json: Update to ec3828add38a83b8c09fd5896265abc9d766162e
    • 13e7057 : aco: don't assume src=lower when splitting self-intersecting copies
    • bd95f95 : iris: store copy of the border color in the border color hash table
    • 0380285 : .pick_status.json: Update to 72b68bd2a62d193e26bd6b31123182d7dbae3098
    • ec22eb7 : .pick_status.json: Mark a5e0a2e101bcda0132185a82c3e8c9b4c90ce94c as denominated
    • 9465b0e : .pick_status.json: Update to 872c4bcd27db7b7ca26abe9fc090ae26d502156f
    • a8650e9 : docs: Add sha256 sums for 20.3.0
    • 08169ff : bump version for 20.3.0 final
    • b9d2f63 : docs: add release notes for 20.3.0
    • 404c440 : freedreno: Break out of "should we free the entry" loop once we've freed.
    • 01f9a5e : gallium: Fix leak of currently bound UBOs at CSO context destruction.
    • faeaa81 : gallivm: Fix max const buffer count.
    • 8258c58 : gallium: Fix leak of bound SSBOs at CSO context destruction.
    • d0d3a58 : freedreno: Fix leak of u_transfer_helper.
    • 0ee24d0 : gallium: Fix leak of the merged driconf options.
    • 531b15c : zink: do not require VK_KHR_external_memory
    • cddf1bf : freedreno: Add missing dependency to build
    • 12c40b0 : radv: Fix leak in radv_amdgpu_winsys_destroy()
    • 1578dde : radeonsi: disable WGP mode on gfx10.3 to prevent hangs
    • 2b9da40 : radeonsi: fix a nasty bug in si_pm4.c
    • b7659c5 : radeonsi: fix a memory leak in si_create_dcc_retile_cs
    • 4e39cda : .pick_status.json: Update to d3c67d7e7ec6b9cf10fbea0d08e92751b7b0fbae
    • bcbc20b : nir/algebraic: Avoid creating new fp64 ops when using softfp64
    • 7a4f33b : aco: Fix NGG GS assert failure from the WG scan.
    • f4a059e : etnaviv/drm: fix evil-twin etna_drm_table_lock
    • b52bb0d : st/mesa: fix uninitialized/random clip plane state vars in lower_ucp
    • 770f46c : mesa: call FLUSH_VERTICES before changing sampler uniforms
    • ded8b21 : radv: Deal with unused attachments in mip flush
    • db68b97 : r600/sfn: fix component loading from fixed buffer ID
    • 5355ff3 : radv: Fix a hang on CB change by adding flushes.
    • 25f01a7 : docs: document new zink-flag
    • 3e2a3b4 : nir: fix sampler_lod_parameters_pan indices
    • 1891d30 : freedreno/a6xx: Fix typo in height alignment calculation in a6xx layout
    • bf76f2b : anv: fix descriptor pool leak in VMA object
    • 0998048 : aco/optimizer: Only set scc_needed when it is actually needed.
    • 6e72700 : glsl: avoid an out-of-bound access while setting up a location for variable
    • 0eabed3 : radeonsi/gfx10: flush gfx cs on ngg - legacy transition
    • bfb711b : nir/unsigned_upper_bound: fix buffer overflow in search_phi_bcsel
    • 9c6e0fb : aco: fix v_mul_hi_u32_u24 format
    • 7770f9a : .pick_status.json: Update to 89f6b72f19dbc503386643c6283047bdb1013bef
    • 39bfa34 : bump VERSION for 20.3-rc3
    • 45f5e8d : appveyor: disable for now
    • e1e31f6 : radv: Fix RB+ blending for VK_FORMAT_E5B9G9R9_UFLOAT_PACK32.
    • b2c70c3 : spirv: Call repair SSA for OpTerminateInvocation
    • deccfbc : zink: fall back to util_blitter for scaled resolves
    • 0464d64 : gallium: fix missing bit field in p_state.h
    • e9cdf39 : lavapipe: enable pipeline stats queries
    • 10a9d8a : lavapipe: fixup mipmap precsion bits
    • 092a65b : llvmpipe: fix multisample lines.
    • 7e90c4d : llvmpipe: fix multisample point rendering.
    • ad4d87a : llvmpipe/setup: move point stats collection earlier.
    • 904811f : lavapipe: fix wsi acquire fences
    • 6e3894d : lavapipe: fixup device allocate + enable private data
    • c7fdc10 : zink: fix layered resolves
    • 0a39350 : meson: Don't add extra values to shader-cache
    • f998d5b : .pick_status.json: Update to 9fa1cdfe7ffd9e7ebd83055e2008f3e4b8ada549
    • 162a067 : aco: fix combining add/sub to b2i if a new dest needs to be allocated
    • 8464a6e : v3dv/cmd_buffer: missing (uint8_t *) casting when calling memcmp
    • 60bf767 : ac/nir: fix a typo in ac_are_tessfactors_def_in_all_invocs
    • 6adf918 : r600/sfn: lower bool to int32 only after common optimizations
    • 29f1078 : nir: fix gathering patch IO usage with lowered IO
    • 33d7087 : nir: fix gathering TCS cross invocation access with lowered IO
    • 48bcf3b : meson: drop deprecated EGL platform build options
    • 921bd0e : gitlab-ci: drop deprecated platforms that snuck in when nobody was watching
    • f56cbdf : .pick_status.json: Update to a92f597b98bb032b904c7c8a8c3a9fe798b51915
    • 43697f6 : intel/compiler: Rotate instructions ROR and ROL cannot have source modifiers
    • 2d1e152 : .pick_status.json: Update to a59b1b18a95af1f8edb0093baf508e974e3251a2
    • d8e7332 : ac: Don't negate strstr return values in ac_query_gpu_info
    • 84e77da : intel/fs: Fix sampler message headers on Gen11+ when using scratch
    • e169782 : freedreno: Fix warning about uninit size for the size==0 special case.
    • 650e2a1 : radeonsi: fix scan_instruction for bindless inc_wrap/dec_wrap atomics
    • 2ce2c22 : radeonsi: fix NGG streamout regression
    • 5bf5eb4 : ac: fix min/max_good_num_cu_per_sa on gfx10.3 with disabled SEs
    • 32b5779 : ac: fix detection of Pro graphics
    • f396e15 : radeonsi: add options.inline_uniforms to the shader cache key
    • d247a07 : radeonsi: determine correctly if switching from normal launch to fast launch
    • cd3cd16 : radeonsi: only do VGT_FLUSH for fast launch if previous draw was normal launch
    • fe27a78 : radeonsi: do VGT_FLUSH when switching NGG - legacy on Sienna Cichlid
    • 825a884 : radeonsi: fix min_direct_count value
    • 40e3de3 : bump VERSION for 20.3.0-rc2 release
    • 0201fc9 : glcpp: Handle bison-3.6 error message changes
    • 827c0cc : .pick_status.json: Update to bac6cc586fe4c1b24351e0574d3a961eb631f6ae
    • 654376f : anv/image: Fix isl_surf_usage_flags for stencil images
    • c9d10b3 : anv/image: Check DISJOINT in vkGetPhysicalDeviceImageFormatProperties2 (v2)
    • 1534e28 : zink: more accurately track supported blits
    • f076b4d : turnip: Close sync_fd only if it is a valid file descriptor.
    • f6a7322 : clover/spirv: Add missing break for SpvOpExecutionMode case.
    • 5953d04 : frontends/va: Fix *num_entrypoints check.
    • 94f202c : freedreno: Fix leak of shader binary on disk cache hits.
    • d736626 : etnaviv: fix disabling of INT filter for real
    • ce29a10 : mesa/main: add missing include in glformats.h
    • 9d39993 : aco: fix fp16 *0.5 omod
    • dc0580d : aco: disable omod if the sign of zeros should be preserved
    • ded5cd5 : iris: initialize shared screen-vtbl only once
    • 15367d2 : egl/dri2: fix race between image create and egl_image_target_texture
    • ee3ed20 : draw: fix tess eval pipeline statistics.
    • e5327be : etnaviv: nir: do not run opt loop after nir_lower_bool_xxx(..)
    • b8ddfc0 : vdpau: Add missing printf format specifier.
    • 159ded9 : freedreno/ir3: Fix crash in shader compile fail path
    • 4de41de : panfrost: Fix stack shift calculation
    • b37d613 : mesa: Clamp some depth values in glClearBufferfi
    • d47cb41 : mesa: Clamp some depth values in glClearBufferfv
    • 5b83eb0 : mesa: Add and use _mesa_has_depth_float_channel
    • 1416771 : .pick_status.json: Update to 87dc3106b077199b829a082e32ec33d0c6d400ab
    • 004b8b1 : aco: disallow various v_add_u32 opts if modifiers are used
    • 46ab4f9 : aco: fix combine_constant_comparison_ordering() NaN check with 16/64-bit
    • afe279a : aco: don't combine precise max(min()) to med3
    • cdb5bcc : turnip: Fix file descriptor return.
    • 0724abd : gallium/draw: Fix rasterizer_discard for wide points/lines.
    • 0e3bb4a : mesa: Fix vertex_format_to_pipe_format index.
    • 7762b3c : nir: handle float atomics in copy propagation pass
    • fe8c524 : intel/fs: Fix use of undefined value in fixup_nomask_control_flow
    • 51d4d91 : ci: Only install kernel modules for LAVA devices.
    • a442fc2 : llvmpipe: just use draw_regions in draw/line setup.
    • bc3e92a : lavapipe: disable SNORM blending for now
    • 0e4e0a0 : lavapipe: enable alpha to one.
    • 3ee324d : u_blitter: port radv 3D blit coords logic.
    • 4de4f55 : gallium: handle empty cbuf slots in framebuffer samples helper
    • f4d976d : util/set: Fix the _mesa_set_clear function to not leave tombstones.
    • 966b55c : freedreno: Protect gmem_cache ralloc allocations
    • 60ffcfe : st/mesa: fix use-after-free when updating shader info in st_link_nir
    • 46c08b7 : softpipe: correct signature of get_compiler_options
    • 23f4120 : panfrost: Fix -reads_frag_coord assignment
    • f3ff126 : panfrost: Fix Bifrost blend descriptor emission
    • 5c167e8 : pan/bi: Model writemasks correctly
    • 2940fb1 : panfrost: Fix AFBC blits of resources with faked RGTC
    • 5634699 : .pick_status.json: Update to bf5cea7232f9ee2934c212211ebefb6fe766526d
    • f44896c : bump version for 20.3-rc1
    • 3c4e43e : intel: Pointer to SCISSOR_RECT array should be 64B aligned
    • a1a365e : broadcom/compiler: Allow spills of temporaries from TMU reads
    • 1c52713 : nir/algebraic: optimize bitfield_select(a, b, 0) to iand(a, b)
    • d47969e : pan/bi: Add support for load_instance_id
    • 5cd1d8c : pan/bi: Add support for load_vertex_id
    • 255f784 : panfrost: Allow linear ZS resources on Bifrost
    • 4995a4c : pan/bi: Add support for ushr
    • af70987 : pan/bi: Add support for ishr
    • 3257ad2 : pan/bi: Fix ARSHIFT definitions
    • 2a80b2d : pan/bi: Move bitwise op packing out of bi_pack_fma()
    • cc09507 : pan/bi: Get rid of bi_emit_ld_uniform()
    • fd265fa : pan/bi: Lower uniforms to UBO
    • 09da82c : pan/bi: Add support for load_ubo
    • 87e2169 : pan/bi: Fix swizzle handling in bi_copy_src()
    • 2522f50 : pan/bi: Support centroid and sample interpolations
    • ca5a00a : pan/bi: Extract LD_VAR sample field from ins-load_vary.interp_mode
    • 1692088 : panfrost: Expose GLES3 features on Bifrost when PAN_MESA_DEBUG=deqp
    • 23dbf79 : panfrost: Force late pixel kill when depth/stencil is written from the FS
    • 956228d : radeon/vcn : Corrected dpb_size calculation for VP9_2
    • 68092df : intel/nir: Lower 8-bit ops to 16-bit in NIR on Gen11+
    • b98f0d3 : intel/nir: Lower 8-bit scan/reduce ops to 16-bit
    • 3ad2d85 : intel/nir: Refactor lower_bit_size_callback
    • f95665c : nir/lower_bit_size: Add support for lowering subgroup ops
    • 2c4b471 : nir/lower_bit_size: Pass a nir_instr to the callback
    • 15c6e05 : nir/lower_bit_size: Don't cast comparison results
    • 86ef139 : radv: implement VK_EXT_shader_image_atomic_int64
    • 9f43268 : ac/nir: implement 64-bit images
    • 5b81e80 : aco: implement 64-bit images
    • cff1950 : util: add mapping from Vulkan to Gallium R64 integer formats
    • bcc3d53 : gallium: Fix NIR validation when lowering polygon stipple
    • 441feda : gallium/util: do not pass undefined sample-count
    • bae5487 : aco: optimize v_and(a, v_subbrev_co(0, 0, vcc)) - v_cndmask(0, a, vcc)
    • 2bbe01b : spirv: Add support for SPV_EXT_shader_image_atomic_int64
    • 5a3e220 : nir: Allow 64-bit image atomics
    • 79f477c : compiler/types: Add 64-bit image types
    • d44c76b : util,gallium: Add new 64-bit integer formats
    • b725fbd : nir: Validate image atomic formats
    • 72f1c9a : nir: Print formats on image intrinsics as text
    • d22fafa : spirv: Update headers and metadata from latest Khronos commit
    • 2acdcf0 : libgl-gdi: support building without softpipe
    • d7f7d89 : panfrost: Add missing Collabora copyright notices
    • 14bbc24 : pan/mdg: Add missing Collabora copyright notices
    • 92374ae : spirv: correct sematic-typo
    • fc28144 : CI: Disable Panfrost T760
    • 4143572 : radeon/vcn: Bitrate not updated when changing framerate
    • 25066eb : v3dv/device: do nothing when asked physical device pci bus properties
    • ca1969c : v3dv/util: log debug ignored stype only on debug builds
    • 1fa4a37 : v3dv/util: remove several logging functions
    • 1675c83 : radv: Fix exporting/importing multisample images.
    • fb1793b : zink: add some spirv builder functions for barriers
    • 95e15f7 : zink: add a quadop function in spirv_builder
    • 758e932 : zink: use same function for all pipe_context::delete_*_state shader methods
    • d377d84 : zink: make physical device functions use a dynamic function pointers.
    • 91f9bc1 : radv: Fix budget calculations with large BAR.
    • 9948ead : radv: Skip tiny non-visible VRAM heap.
    • b7793e3 : docs: update calendar and link releases notes for 20.2.2
    • e053a02 : dcs: Add sha256 sums for 20.2.2
    • 5e1f518 : docs: add release notes for 20.2.2
    • 39d1576 : st/nine: Remove unnecessary NULL check.
    • 0bc2227 : zink: implement ARB_texture_query_lod
    • bf29daa : ci/deqp: Switch to a new dEQP runner written in Rust.
    • fe61230 : ci/bare-metal: Reset colors at the end of a line of serial output.
    • ff67417 : ci/bare-metal: Apply autopep8 to the bare-metal scripts.
    • 9f1cd99 : turnip: Fix image size for 3D vkGetImageSubresourceLayout.
    • a0b42da : spirv: fix GLSLstd450Modf/GLSLstd450Frexp when the destination is vector
    • 8159775 : zink: For MoltenVk added vkFlushMappedMemoryRanges() to vkMapMemory() to fix empty mapped memory.
    • f0bbd8f : zink: have_triangle_fans support.
    • 2aca374 : zink: add support to device info for macro guards and just VkPhysicalDevice*Features with out the have_.
    • f0a88db : nir/lcssa: consider loops with no back-edge invariant
    • feb9462 : zink: Added inbuilt debug logging from the VK_LAYER_LUNARG_standard_validation layer.
    • 10d44fe : zink: fix pNext chain for resource memory allocation
    • 0b53ca0 : zink: return fail if create_instance fails
    • 6a96729 : docs/features.txt: VK_EXT_separate_stencil_usage not exposed on RADV
    • 6ec7e0d : gallium: document convention for get_handle calls on multi-planar resources
    • b29fe26 : etnaviv: rework ZSA into a derived state
    • 3c7fc95 : etnaviv: expose shader discard usage in etna_shader_variant
    • bff60d6 : etnaviv: update headers from rnndb
    • 83c72ca : etnaviv: flush depth cache when changing depth config
    • 01e2ded : etnaviv: emit RA_EARLY_DEPTH on dirty ZSA
    • d428580 : mesa/st: use a lock to protect access to variants when updating them
    • 77d6fda : nir/algebraic: distribute imul(iadd(a, b), c) when b and c are constants
    • 2fd5079 : clover: implements clSetContextDestructorCallback
    • 04e1b17 : clover: add empty cl 3.0 dispatch entries.
    • 558325c : clover/spirv: hook up spir-v environment for 3.0
    • 7230f39 : clover/llvm: add 3.0 versioning.
    • a52c613 : clover: access 3.0 and deprecated 2.2 API
    • 9790fdf : vtn/opencl: add ctz support
    • 725df7c : CL: update CL headers to 3.0
    • 041e314 : docs: Add MESA_pack_invert and ANGLE_pack_reverse_row_order
    • fd10409 : mesa: Implement GL_ANGLE_pack_reverse_row_order
    • 4506e4d : mesa: Enable GL_MESA_pack_invert unconditionally
    • 923577c : v3dv: Remove unsigned comparison to zero.
    • 03683b9 : nir: Handle ray-tracing intrinsics and storage classes in copy-prop etc.
    • 5a28893 : spirv,nir: Add ray-tracing intrinsics
    • 21b1b91 : nir,spirv: Add support for the ShaderCallKHR scope
    • 6b8fd65 : spirv: Implement the new ray-tracing storage classes
    • 84a8ca1 : nir: Add new variable modes for ray-tracing
    • aa4ea9c : nir: Add intrinsics for object to/from world RT sysvals
    • 07635a3 : nir/builder: Add a select_from_ssa_def_array helper
    • 46cd91b : spirv,nir: Add support for ray-tracing built-ins
    • ed907e5 : spirv: Add support for OpTypeAccelerationStructureKHR
    • 2df055a : spirv: Pass the deref type to storage_class_to_mode for non-forward pointers
    • 2bef026 : spirv: Add a guard for OpTypeForwardPointer storage classes
    • aabe37b : spirv: Remove a redundant vtn_fail_if
    • 3bcebe1 : spirv: Add Ray Tracing execution models
    • d8dbdf2 : spirv: Add basic plumbing for ray-tracing capabilities
    • f651829 : compiler: Add new Vulkan shader stages
    • 1882a02 : tu: Make sure spirv_to_nir knows we support imageStorageWithoutFormat.
    • 0b0f152 : nir/clip_disable: handle 2x vec4 case
    • 5e43ba3 : nir/clip_disable: try for better no-op
    • 1d23a88 : nir/clip_disable: write 0s instead of undefs for disabled clip planes
    • c84ae15 : iris: Move blit scissoring earlier.
    • 0f96a9a : anv: restrict number of subgroups per group
    • b2c7193 : turnip: enable VK_EXT_image_drm_format_modifier
    • f624692 : turnip: don't always fallback to linear for mutable formats
    • 8c4426f : turnip: remove unnecessary/redundant tu_image fields
    • c64cd69 : turnip: remove useless tu_image asserts
    • dfaa8b9 : turnip: LAYOUT_PREINITIALIZED is not different for optimal tiling
    • 43c1648 : turnip: don't implement CreateImage as two separate functions
    • 1a1099c : aco: Fix format string used when raising validation errors
    • 61d2bad : nir/deref: Fix a typo
    • 60c5729 : ci: Distribute ADMGPU driver to LAVA as a module
    • bb41acf : ci: Update dEQP skips and fails for Bifrost on G52
    • 80cbb32 : ci: Update kernel for LAVA to 5.10-rc2 plus patches
    • 092186d : util/threaded_context: use driver's buffer alignment for staging transfers
    • 60b9c00 : v3d: Add GL_ARB_vertex_array_bgra support
    • 9648bfb : v3dv: mark the right bit to swap R/B vertex attributes
    • 1e72374 : v3d/compiler: extend swapping R/B support to all vertex attributes
    • 44925a8 : intel/tools: add missing new lines to few remaining fail_if users
    • c323d7c : intel/tools: refactor logging to be easier to follow by static analyzers
    • f006127 : intel/tools: handle some failures
    • cd9907e : anv: remove dead code from anv_create_cmd_buffer
    • d13b7d6 : intel/tools: allow --color option to be used without arg
    • 0dad1a7 : v3dv: expose more features
    • 0b30336 : broadcom/compiler: Handle non-SSA destinations for tex instructions
    • e331fd7 : vc4: use intmax_t for formatted output of timespec members
    • fe52efa : loader: Print dlerror() output in the failure message
    • 5d5f3e3 : intel/fs: Implement nir_intrinsic_{load,store}_shared_block_intel
    • 9fe158e : intel/fs: Implement nir_intrinsic_{load,store}_ssbo_block_intel
    • d372abe : intel/fs: Add surface OWORD BLOCK opcodes
    • 296137d : intel/fs: Implement nir_intrinsic_{load,store}_global_block_intel
    • d3d2b73 : intel/fs: Add A64 OWORD BLOCK opcodes
    • eb03f29 : spirv: Implement SpvCapabilitySubgroupBufferBlockIOINTEL
    • dd39e31 : nir: Add nir_intrinsic_{load,store}_deref_block_intel
    • b86ce27 : spirv: Implement SpvCapabilitySubgroupShuffleINTEL from SPV_INTEL_subgroups
    • 5194cbc : iris: Flush dmabufs during context flushes
    • c488fb6 : iris: Fix fast-clears of swizzled alpha formats
    • 60336ca : iris: Fix SINT assert in convert_fast_clear_color
    • 4cb9b5d : iris: Fix fast-clears of swizzled LA formats
    • 382451f : iris: fix source/destination layers for 3D blits
    • 61ce544 : st/nir: Drop setting interp mode on system values in builtins.
    • 9143c08 : st/nir: Fix the st-pbo.use_gs case.
    • f8c1d79 : pan/bi: Correctly calculate render target index
    • 9a6dad1 : pan/bi: Lower depth/stencil stores
    • 6433fed : pan/bi: Emit +ZS_EMIT as needed
    • 0c889e7 : pan/bi: Stub handling for nir_intrinsic_store_combined_output_pan
    • f67b49d : pan/bi: Factor out bi_emit_blend
    • 928f2bc : pan/bi: Factor out bi_emit_atest
    • f14beeb : pan/bi: Infer z/stencil flags from sources passed
    • 2b1db36 : pan/bi: Add +ZS_EMIT instruction to IR
    • cd66aa7 : panfrost: Deduplicate shader properties
    • 4a2ecc7 : panfrost: Pass through src_type
    • 42319c5 : pan/mdg: Move writeout lowering to common panfrost
    • dfaa4c5 : pan/mdg: Deduplicate nir_find_variable_with_driver_location
    • a05921b : nir: Add SRC_TYPE to store_combined_output_pan
    • fa869fb : v3dv: add a v3dv_bo_init helper
    • 456beb4 : aco/ra: Fix counting of subdword variables in get_reg_create_vector
    • 7868281 : aco: implement 8/16-bit instructions which can be trivially widened
    • ef95ba8 : aco: implement some 16-bit arithmetic instead of lowering
    • 8ed7cad : radv: rework nir_lower_bit_size callback and run DA on GFX8+
    • b2ce7da : radv: do nir_lower_bit_size after algebraic optimizations
    • c771149 : radv: move a few passes to after load/store vectorization
    • 475077c : nir/lower_bit_size: optimize upcast of b2i8/b2i16
    • 4e5c855 : nir: add shader_info::bit_sizes_used
    • 7d3df69 : va: support VA_RT_FORMAT_PROTECTED
    • 81be8b3 : va/picture: make sure destination buffer is protected if needed
    • 70191f3 : frontends/va: Added protected playback support for VP9
    • f20a20b : radeon/vcn: program drm message buffer
    • ebbf40f : radeon/vcn: delay dec-ctx and dec-dpb allocation
    • cf324df : radeon: add si_vid_create_tmz_buffer helper
    • 024be71 : radeon/vcn: add defines for drm message buffer
    • b9753fd : radeon: add decryption params definition header
    • deb7dc8 : frontends/va: handle protected slice data buffer
    • 5f0816e : vl: add flag and definition for protected playback
    • 58e7088 : nir/find_array_copies: Don't assume all children exist
    • e1736c7 : radesonsi: Remove unnecessary shader-selector NULL check.
    • 5fb77a5 : lavapipe: request correct sample mask behaviour
    • 47701d4 : llvmpipe: respect the sample mask in non-multisample flag
    • 3dc6da1 : gallium: add a non-multisample sample mask out behaviour flag.
    • 4ff4d4e : nir/opt_intrinsic: Optimize bcsel(b, shuffle(x, i), shuffle(x, j))
    • 2f5b56a : nir/opt_intrinsics: Refactor a bit
    • 3b28186 : nir/constant_folding: Fold subgroup shuffle intrinsics
    • e59d635 : nir: Move constant folding of vote to opt_constant_folding
    • 9492ab2 : nir/constant_folding: Use the standard variable naming convention
    • 9d2ccbf : nir/constant_folding: Use a switch in try_fold_intrinsic
    • d9c0f36 : nir/opt_intrinsics: Report progress for the gl_SampleMask optimization
    • b900632 : nir: use nir_alu_src_is_trivial_ssa() in nir_ssa_for_alu_src()
    • 233a820 : nir: skip bcsel with non-trivial swizzle in opt_simplify_bcsel_of_phi()
    • 1df2fc9 : nir: add nir_alu_src_is_trivial_ssa()
    • b9f9528 : nir/lower_io: Add a new 62bit_generic address format
    • b84f74f : nir/lower_io: Support generic pointer access
    • a451f03 : nir/lower_io: Add support for lowering deref_mode_is
    • 5794311 : nir/lower_io: Add support for 32/64bit_global for shared
    • c50332f : nir/lower_io: Add a mode parameter to addr_format_is_*
    • 7007d06 : nir/lower_io: Add a mode parameter to build_addr_iadd
    • ff838ab : nir/opt_deref: Add an optimization for deref_mode_is
    • df51518 : nir/opt_deref: Add a deref mode specialization optimization
    • a8e53a7 : spirv: Add generic pointer support
    • d6415b5 : nir: Add support for generic pointers
    • 9d377c0 : nir: Make nir_deref_instr::mode a bitfield
    • 7d5f3b5 : nir/split_*_vars: Prepare for generic pointers
    • d50a4db : nir/find_array_copies: Prepare for generic pointers
    • ced9b6f : nir: Use nir_deref_mode_may_be in deref optimizations
    • 379d535 : nir/vec3_to_vec4: Use nir_deref_must_be
    • 8a2cda1 : nir/vars_to_ssa: Use nir_deref_must_be
    • 0f94ff8 : nir: Only force loop unrolling if we know it's a in/out/temp
    • fff78fc : nir/phis_to_scalar,gcm: Use nir_deref_mode_may_be
    • 9f3e3df : nir/lower_io: Use nir_deref_mode_* helpers
    • 9ae87a6 : nir/lower_array_deref_of_vec: Use nir_deref_mode_must_be
    • 3cc58e6 : nir: Add and use some deref mode helpers
    • 74886ca : nir/opt_find_array_copies: Allow copies from mem_constant
    • 89abfbe : nir: Disallow writes to system values and mem_constant
    • bb5d502 : nir: Use var-data.mode instead of deref-mode in a few cases
    • 5664713 : nir: Handle incomplete derefs in split_struct_vars
    • 6b72004 : nir/phis_to_scalar: Use a deny-list for load_deref modes
    • 3f0a29f : nir/builder: Add a nir_ieq_imm helper
    • 9d36151 : lavapipe: don't advertise linear filtering on integer textures.
    • 3d81cf4 : lavapipe: use clear_buffer callback
    • c3d8a69 : llvmpipe: add clear_buffer callback. (v2)
    • 1e3fbee : lavapipe: stop crashes with 3D z blits
    • c871ac0 : lavapipe: fix 3d compressed texture copies.
    • 42b5cfd : gallivm/nir: fix vulkan vertex inputs
    • 621c4f8 : gallivm/nir: handle dvec3/4 inputs properly.
    • a5aab63 : lavapipe: fix dEQP-VK.info.device_properties
    • 6dabb80 : lavapipe: constify state pointers into command buffers.
    • 33a2f27 : lavapipe: don't write to pending clear aspects in cmd buffer
    • ad38492 : gallivm: fix f16 quantize.
    • 7ae81d6 : r600: amend space check for chips older than EVERGREEN
    • 902b318 : zink: break up dynamic access lowering
    • e12e674 : zink: add pass for lowering dynamic ubo/ssbo block indexing to constants
    • 71a1979 : mesa/st: Fix a use-after-free of the NIR shader stage.
    • 6b85a88 : mesa/spirv: Lower variable initializers for global variables
    • c2cb2dd : zink: Added support for MacOS MoltenVK APIs.
    • a0f889b : zink: Basic framework to check for optional instance layers and instance extensions.
    • bfa9fd8 : radv,radv/winsys: Move RADV_MAX_IBS_PER_SUBMIT
    • f6e5baf : docs/features: add some extensions we missed
    • a2056b0 : docs/features: VK_KHR_mir_surface is disabled, remove it
    • 3d58ab7 : docs/features: Minor update extensions support
    • 89c4bba : nir/algebraic: better propagate constants up fadd chains
    • 16c756e : spirv: reverse order in matrix multiplication
    • 24a18b1 : nir: scalarize fdot in reverse
    • 6dc6b8a : panfrost: Set .array_size on Bifrost
    • eaf0be8 : pan/bi: Don't emit TEXS for array textures
    • 3a86e1e : pan/bi: Handle 3D/array coordinates
    • 56bda0f : pan/bi: Track tex data register swizzles
    • 1463824 : panfrost: Add bi_emit_array_index helper
    • 914eebb : panfrost: Drop unused swizzles
    • b6c7142 : panfrost: Advertise Bifrost support
    • 828c7e7 : panfrost: Disable point sprites on Bifrost
    • 2ed0036 : pan/bi: Lower +CUBEFACE2
    • 3186401 : pan/bi: Suppress disassembly for internal shaders
    • 2f00f82 : pan/bi: Lower cube map coordinates
    • f76558b : pan/bi: Hook up cube instructions packing
    • 0ed8eee : pan/bi: Split special class in two
    • 20cc638 : pan/bi: Move special instruction packing to a separate helper
    • 07f1df5 : pan/bi: s/t0/t1/ in bi_disasm_dest_add()
    • 4713191 : panfrost: Implement v7 texture payloads
    • 4e0826d : panfrost: Add array size to XML
    • 557633b : panfrost: Suppress Bifrost prefetching
    • c0b8f73 : panfrost: Leave push_constants pointer to NULL if there's no uniform
    • 29fc115 : mesa: do not throw _mesa_problem when invalid enum is used
    • e02e1cc : mesa/st: call memobj_destroy only if there is memory imported
    • 57c152a : aco: select v_mul_{hi}_u32_u24 for 24-bit multiplications
    • 3a72021 : aco: store NIR range analysis data to the isel context
    • 81ff389 : zink: clamped maxPerStageDescriptorUniformBuffers limits to INT_MAX when stored as uint32_t.
    • ac65d3b : radv: fix shader caching with NaN fixup workaround
    • 36f6249 : radv: fix shader caching with discard-demote workaround
    • 19f3911 : radv: add some missing radv_{start,stop}_feedback
    • 200bcd7 : android: freedreno: Add freedreno_dev_info.[ch] to Makefile.sources
    • 09b2bd1 : broadcom/compiler: remove v3d_fs_key depth_enabled field.
    • 21ffacf : intel/compiler: remove branch weight heuristic
    • 06764e0 : intel/compiler: use C++ template instead of preprocessor
    • 788f6dc : Revert "gallium/dri: fix dri2_from_planar for multiplanar images"
    • fe3e571 : tu: Support rasterizerDiscardEnable and RasterizationStreamSelect
    • 841f736 : tu: Support geometryStreams
    • 6b8d30e : util/bitset: Add a range iterator helper
    • 563789c : ir3: Support geometry streams
    • 48cfaec : freedreno/a6xx: Update SO registers for streams
    • 92022f2 : zink: add VK_STRUCTURE_TYPE_WSI_MEMORY_ALLOCATE_INFO_MESA for WSI allocations
    • 98ebffc : etnaviv: move etna_destroy_shader(..) to generic location
    • 2dc73d2 : etnaviv: move etna_dump_shader(..) to generic location
    • 3bef6df : etnaviv: convert from tgsi semantic/index to varying-slot
    • edbdd97 : nir: make tgsi_varying_semantic_to_slot(..) public
    • e351559 : zink: require Vulkan timestamp queries for time query caps
    • b07b943 : tgsi: Initialize tgsi_declaration_dimension padding.
    • 7278f90 : gallium: Remove duplicate resource variable.
    • 78a420c : nir/validate: Explain why we don't use nir_foreach_block
    • e8a332b : zink: always reset query pools on next query begin
    • 60db5af : zink: always use query-type for starting/stopping xfb queries
    • e1e3484 : zink: rework query overflow handling
    • 7fc806b : zink: fixup gs/xfb tracking for primitives generated queries
    • 28ad3c1 : zink: store batch id onto query object at time of start
    • b073398 : zink: also create an xfb query for every primitives generated query
    • 0b2475f : zink: more correctly handle PIPE_QUERY_PRIMITIVES_GENERATED queries
    • a878557 : zink: deduplicate some query result code
    • b6a0309 : lavapipe: use resource get param.
    • 2dcc9c7 : llvmpipe: add resource get param support.
    • ae17e1f : gallium: add a layer stride pipe resource parameter.
    • e972472 : gallium: add a level parameter to resource parameter get
    • 0e3f2e6 : zink: call the reset callback not only during a status check
    • efadeb3 : i965: remove prototypes of not-existing functions
    • e3f6a9e : intel: remove dead code
    • b5e2c58 : anv: always annotate memory returned from anv_gem_mmap
    • 990343b : turnip: rework android gralloc path so it doesn't call tu_image_create
    • b2a21fe : os: Fix open result check.
    • 1eda842 : iris/bufmgr: Handle NULL bufmgr in iris_bufmgr_get_for_fd
    • eb61f89 : i965/bufmgr: Handle NULL bufmgr in brw_bufmgr_get_for_fd
    • aa2f6bd : freedreno: Use freedreno_dev_info
    • a1d2b21 : tu: Use freedreno_dev_info
    • 4a0bdf4 : freedreno: Introduce common device info struct
    • ea81889 : nir/large_constants: only search for constant duplicates
    • ce5e2e2 : nir: Stabilize compact_components sort
    • aaa94d9 : docs/features: add Vulkan 1.2
    • 8c4d15d : docs/features: update unpromoted Vulkan extensions
    • 8077f3f : nir/lower_returns: Append missing phis' sources after "break" insertion
    • 95b9fc4 : zink: implement pipe_device_reset_callback
    • 7d6609e : v3dv: fix occlusion query inheritance in secondary command buffers
    • 12f87b6 : v3dv: add support for timestamp queries
    • 520f3e2 : radeonsi: fix RADEON_FLUSH flags conflicts
    • 18b7caf : driconf: add disable_protected_content_check option
    • 9b0ffa9 : egl/dri2: implement createImageFromDmaBufs3
    • d9582ff : dri: introduce createImageFromDmaBufs3
    • 2aff88a : egl: handle EGL_PROTECTED_CONTENT_EXT for eglImage
    • 52268a3 : radeonsi: enable PIPE_CAP_DEVICE_PROTECTED_CONTENT
    • bd18277 : egl: implement EGL_EXT_protected_surface support
    • 9698a22 : radeonsi: honor PIPE_BIND_PROTECTED
    • a653504 : gallium: introduce PIPE_BIND_PROTECTED
    • 7748e50 : gallium: add new cap PIPE_CAP_DEVICE_PROTECTED_CONTENT
    • b54c054 : v3dv: expose VK_KHR_maintenance1
    • 53c6dc5 : v3dv: implement vkTrimCommandPool
    • 0dedee7 : v3dv: update assertion to match VK_KHR_maintenance1 semantics
    • e4988e6 : v3dv: fix base slice selection for copies involving 3D images
    • 0a9b807 : v3dv: add image view debug checks for VK_KHR_maintenance1
    • 03f260c : radv,aco: optimize computing the sample mask for per-sample shading
    • c63bcda : radv,aco: adjust the sample mask only if per-sample shading is enabled
    • ea83fd9 : glsl: drop NMS OpenGL workarounds
    • d5dd779 : panfrost: Add missing alpha-first special formats
    • 3a67806 : panfrost: Fix BGR233 component order
    • b5bc093 : panfrost: Fix RGB5A1 formats
    • f2af23b : panfrost: Use macro for panfrost_get_default_swizzle
    • c5d8619 : panfrost: Add missing 1/2/4/64-bit formats to XML
    • d473622 : panfrost: Rename VARYING_POS to SNAP4
    • 1b5cac4 : panfrost: Rename VARYING_DISCARD to CONSTANT
    • c6bdd97 : panfrost: Split out v6/v7 format tables
    • 29bb281 : panfrost: Add v7-specific depth formats
    • 8ca1478 : panfrost: Add miscellaneous missing Midgard formats
    • a21d3de : panfrost: Add missing depth/stencil formats
    • f6fb9ef : panfrost: Add v7 special colour formats
    • cae1be9 : panfrost: Remove panfrost_is_z24s8_variant
    • 870283b : panfrost: Remove duplicated format arg for ASTC
    • 44155fa : panfrost: Complete format_to_bifrost_blend
    • e1624b6 : panfrost: Use panfrost_blendable_formats for blending
    • 41f06ac : panfrost: Use panfrost_blendable_formats for SFBD
    • 293198e : panfrost: Use panfrost_blendable_formats for MFBD
    • d9b16ca : panfrost: Add a blendable format table
    • 666677d : panfrost: Use consistent swizzle names in XML
    • 1e18d99 : panfrost: Add MALI_EXTRACT_INDEX helper
    • 1b48b9e : panfrost: Don't double-compose swizzles
    • a575bb1 : draw: Remove draw_install_aaline_stage dead code.
    • 0a4c1db : gallium/u_threaded: merge consecutive draw calls within batches
    • d0916cc : gallium/u_threaded: move a structure up to be used later
    • a44868b : radeonsi: implement multi_draw for compute-based primitive culling
    • cc24ec8 : radeonsi: set NOT_EOP for back-to-back draws on gfx10+
    • ca40dc0 : radeonsi: add support for multi draws
    • 0ce6885 : radeonsi: implement multi_draw but supporting only 1 draw
    • ae8d892 : radeonsi: don't check info-count == 0
    • d9c4ca2 : radeonsi don't get count from pipe_draw_info in si_num_prims_for_vertices
    • 7cc939f : radeonsi: add num_draws parameter into si_need_gfx_cs_space
    • 095ee8f : winsys/amdgpu: remove incorrect assertion check against max_check_space_size
    • 73f2d3c : gallium: add pipe_context::multi_draw
    • d9c12c6 : gallium: move pipe_draw_info::start/count to the beginning and pad empty space
    • 8943c80 : radv: Fix variable name collision.
    • 41be85a : ci/x86: speed up piglit testing
    • e54c7f4 : android: aco: add aco_form_hard_clauses.cpp to Makefile.sources
    • a1d6c03 : etnaviv: don't import allocated scanout resources via from_handle
    • 3862cec : etnaviv: pass correct layout to etna_resource_alloc for scanout resources
    • ec21148 : etnaviv: simplify etna_screen_bo_from_handle
    • 94ec412 : etnaviv: do proper cpu prep/fini when clearing allocated buffer
    • 1817901 : etnaviv: cosmetic etna_resource_alloc fixes
    • 98db7c4 : etnaviv: free tgsi tokens when shader state is deleted
    • 3fd5124 : etnaviv: tex_desc: fix TS compression enable
    • cf871e9 : etnaviv: update headers from rnndb
    • 3ba753d : etnaviv: blt: properly program surface TS offset for clears
    • 866bb22 : etnaviv: drm: fix BO refcount race
    • 8b0218b : etnaviv: drop etna_pipe_wait(..)
    • b750118 : radeonsi: implement inlinable uniforms
    • 6d058ac : aco: Fix accidental copies, attempt two
    • b03c86a : intel/dev: Bump Max EU per subslice/dualsubslice
    • 5957b0c : glthread: pin driver threads to the same L3 as the main thread regularly
    • d8ea509 : util: completely rewrite and do AMD Zen L3 cache pinning correctly
    • 4f2c230 : util: add util_get_current_cpu using sched_getcpu and Windows equivalent
    • 9758b1d : util: add util_set_thread_affinity helpers including Windows support
    • 3433d19 : st/mesa: remove random L3 pinning heuristic for glthread
    • 96d9f77 : util: consolidate thread_get_time functions
    • 53a1592 : util: remove unused util_get_L3_for_pinned_thread
    • cc16113 : android: fix libsync dependencies (v2)
    • a09717c : glsl: add extra pp tokens workaround and enable for CoR
    • ce0b72a : intel/fs: Don't emit_uniformize when getting a constant SSBO index
    • 0fe5490 : v3d/format: use XYZ1 swizzle for three-component formats
    • f7d1460 : gallivm: zero init the temporary register storage.
    • 1761379 : aco: handle SDWA in the optimizer
    • ecc5b59 : aco: don't allow destination opsel for v_cvt_pknorm
    • bb890f2 : aco: fix combine_inverse_comparison()
    • 7e4aa8c : aco: fix printing of some sdwa sels
    • 70320f4 : aco: assert a label only uses one of the members in ssa_info's union
    • 3dfbed2 : aco: create s_clause on GFX10+
    • f4c090a : aco: refactor split_store_data() to always split into evenly sized elements
    • daec83c : intel/genxml: don't generate identical code for different branches
    • e96f33c : intel/tools: fix invalid type in argument to printf
    • 13859c7 : gallium/dri: fix dri2_from_planar for multiplanar images
    • 6c118ae : gallium/dri: fix dri2_query_image for multiplanar images
    • 0f3594c : gallium/dri: allow create image for formats that only support SV or RT binding
    • 25f9848 : nv50/ir/nir: don't use designated initializers
    • e07c546 : v3dv/format: use XYZ1 swizzle for three-component formats
    • d8562b7 : virgl: Correctly align size of blobs
    • 7dc17ae : glsl: Update loop_terminator constructor to accept parameters.
    • f74ef15 : aco/ngg: Incorporate GS invocations into workgroup size calculation.
    • 09b9e52 : aco/ngg: Export a zero-area triangle when primitive count is 0.
    • 73449f9 : aco: Add a few assertions about LDS usage.
    • b6654ad : aco: Make emitting reduction instructions a bit more convenient.
    • 8d62462 : aco: Add some validation for PSEUDO_REDUCTION instructions.
    • 260f9c5 : aco/ngg: Put shader query reduction operand into a VGPR.
    • 9757c3c : aco: Assert that workgroup barriers are not used inappropriately.
    • 915f291 : ci/bare-metal: suppress 'No such file or directory'
    • ecdcf22 : aco: switch aco_print_asm to a FILE *
    • a293fad : aco: refactor repeated instruction disassembly
    • ed2449d : aco: move individual instruction disassembly to its own helper
    • 7ded846 : docs/features: Update extensions for swr
    • 2088f35 : docs/features: Update extensions for softpipe
    • 483657d : aco: use mubuf helper in select_gs_copy_shader
    • ec7ecfe : aco: use control flow creation helpers in select_gs_copy_shader
    • 57d977a : aco: round bytes_written to dwords if larger than 4 bytes
    • 41839d3 : aco: default to a definition size of 32
    • 768186e : docs: s3tc - S3TC
    • d30470e : docs: clang - Clang
    • f3095f9 : docs: gcc - GCC
    • 4aded09 : docs: fbo - FBO
    • e8d0313 : docs: hud - HUD
    • f408343 : docs: api - API
    • a22cedf : docs: anistropy - anisotropy
    • 689145e : docs: eg. - e.g.
    • 580b9d1 : docs: ie. - i.e.
    • fef8a4b : radv: remove call to nir_lower_pack()
    • 212be2a : radv: lower pack_[64/32]_* via nir_lower_alu_to_scalar()
    • bd0468e : nir: add options to lower nir_op_pack_[64/32]_* via nir_lower_alu_to_scalar()
    • 121fa01 : ac/nir: implement nir_op_[un]pack_64_4x16
    • 543f507 : aco: implement nir_op_unpack_[64/32]_*
    • 79cd22b : v3dv: enable alphaToOne feature
    • eb104e9 : radv: Do not access set layout during vkCmdBindDescriptorSets.
    • 29999e6 : radv: Fix 1D compressed mipmaps on GFX9.
    • 06cf838 : intel/mi_builder: Support gen11 command-streamer based register offsets
    • fdb1997 : Fix VMware capitalization.
    • 0e7d45c : util: use dllexport for mingw too
    • e030ab5 : lavapipe: configure suffix in icd json
    • 3017d88 : gallium: use libpipe_loader_links
    • 0355d10 : wsi: move drm code to wsi_common_drm.c
    • 72ce22f : lavapipe: fix usleep usage in lvp_device
    • d399c3e : intel/dev: Add device info for ADL-S
    • 26e53e3 : aco: ignore the ACO-inserted continue in create_continue_phis()
    • fa5acbb : CI: remove llvmpipe cl flake test
    • 8d03cfa : anv: Drop warning about gen12 not being supported
    • a188533 : panfrost: AFBC compress Z16 depth buffers
    • 388c99b : panfrost: Z16 depth buffer support
    • 4a20ed6 : panfrost: Move zs format handling code out of the !afbc case
    • 06d1f7c : docs: Specify when branch points happen
    • 328a350 : vulkan/util,vulkan/wsi,radv: Add typed outarray API
    • 437995b : aco: remove all-undef phi opt
    • 70ff262 : aco: use v_mov_b32_sdwa for some 16-bit constants
    • b882598 : aco: remove some unused optimizations
    • d20a752 : aco: use Builder::copy more
    • e54c111 : aco: always use p_parallelcopy for pre-RA copies
    • 6db5fbf : aco: allow literals on sub-dword p_parallelcopy
    • 74e2e9b : aco: don't use bld.copy() in handle_operands()
    • a834d9e : aco: expand vectors passed as copy operands
    • e092f34 : aco: copy-propgate through p_create_vector during value numbering
    • 0f31fa1 : aco: skip value numbering of copies
    • 72b307a : aco: don't do divergent break+discard
    • d4503a9 : aco: update phi_map in add_subdword_operand()
    • 23fb54b : aco: Clean up some C++ usages
    • 79347f5 : radv: enable VK_AMD_mixed_attachment_samples on GFX6-GFX7
    • ee1d30a : radv: flush CB before and after FMASK_DECOMPRESS or DCC_DECOMPRESS
    • 3fee91d : docs: add link to extension spec
    • 76f74bd : CI: Only run OpenCL tests when we need to
    • 48e83f7 : radv: do not perform a FMASK expand for non-writeable MSAA images
    • cb12879 : aco: fix GFX8 16-bit packing
    • 666817c : v3dv: grow meta descriptor pool dynamically
    • 6b874eb : ci: Run git_archive job if all_paths matches
    • e5cd5e9 : pan/mdg: fix LOCAL_STORAGE wls_instances packing
    • e4b170a : v3dv: enable the logicOp feature
    • af0435c : Revert "radeonsi: use staging buffer uploads for most VRAM buffers"
    • 006ce73 : freedreno/gmem: Respect max-height limits too
    • 9f5c8ff : freedreno: Rework GMEM limit init
    • 6c6693e : pan/bi: Fix ms_idx type to catch missing ms_index source
    • eaed477 : pan/bi: Add ult support
    • f77bbc9 : pan/bi: Lower {i,u}{min,max} instructions
    • 1c8b8e3 : pan/bi: Add support for load_point_coord
    • 55c0dd9 : pan/bi: Add support for load_front_face
    • 0a582b5 : panfrost: Preload primitive flags when gl_FrontFacing is accessed
    • b92eadb : ci: Add "check mr" job to needs: of build jobs
    • 27ce5d9 : aco: remove isel_context::allocated
    • 6f68cac : virgl: Always enable emulated BGRA and swizzling unless specifically told not to
    • 3d9ffdc : nir/lower_memcpy: Don't mask the store
    • 67238f9 : freedreno: Disallow tiled if SHARED and not QCOM_COMPRESSED
    • 63bdbfd : freedreno: Update import/export traces
    • 5f686c3 : st/va: fix build with old libva
    • 638ebde : glx: get rid of memory leak
    • 26f58e8 : mapi: do not return thread-specific data for wrong thread
    • 61d40ae : mapi: do not call thread-unsafe dispatch getter
    • 65d6f25 : mapi: remove unused function
    • cf083f1 : aco: use do_pack() for self-intersecting operations.
    • d96f387 : aco: improve code sequences for 16bit packing
    • 40bfb08 : aco: refactor GFX6_7 subdword copy lowering
    • 06b41ca : iris: add support for fence signal capability
    • aa1d298 : iris: handle PIPE_FD_TYPE_SYNCOBJ type
    • bddaa93 : ac/nir: remove bindless image atomic format check
    • 2ff9784 : docs: document zink's gl 3.0 requirements
    • be693be : docs: do not document required minimum
    • dd0f941 : zink: verify geometry shader feature
    • 82512a1 : gallivm: lower flrp for all sizes.
    • b8a9bd9 : gallivm: get correct min/max behaviour for kernels.
    • 9845c16 : gallivm: add support for 8/16-bit mul_hi
    • 4b6d332 : gallivm: handle sub-32 bit masked stores.
    • 81d8ca7 : gallivm: add b2i8/b216 support
    • 0f78ca9 : gallivm: add 16-bit split/merge support.
    • 4ecdc5e : gallivm: fix 64-16 f2f16
    • 842a539 : llvmpipe: fix 8/16 bit global stores
    • d1ea49d : anv: report latest extension spec versions
    • 303a74c : android: panfrost: Move nir_undef_to_zero to util
    • 6a189c8 : util/xmlconfig: Disable for Windows like Android
    • fb56fb0 : gallivm: add load/store scratch support.
    • 9c1f6ed : frontends/va: Initialize drm modifier on import.
    • bef6007 : gallium: update abs_delta segementation parameter
    • 9c67f3d : frontends/omx/enc: fix omx h264 encoding force-keyframe-period issue.
    • 390887f : mesa: Remove the key parameter from the _mesa_HashDeleteAll callback
    • cc6dcc6 : mesa: Remove the key parameter from the _mesa_HashWalk callback
    • ca752b0 : i965: Get the gl_perf_query_object Id from the object
    • f71488d : mesa: Store the atlas Id in the gl_bitmap_atlas structure
    • 0e17bb5 : mesa: Open-code hash walk in _mesa_HashPrint
    • 9d3b802 : scons/windows: Support build with LLVM 11.
    • f21dd3b : scons: gallium/auxiliary: Unconditionally compile NIR regardless of LLVM
    • 3ba786f : spirv: Fix OpCopyMemorySized
    • a071590 : frontends/omx/h265: Check the pps set before the scaling data
    • 08762d5 : frontends/omx/dec: Use the known codec profile when allocating buffers
    • 56f90a6 : pan/bi: Account for bool32 ld_ubo reads
    • 3cdca15 : panfrost: Don't advertise MSAA on Bifrost
    • f0e6580 : panfrost: Drop PIPE_CAP_GLSL_FEATURE_LEVEL for Bifrost
    • b9a136c : panfrost: Implement BGRA textures
    • 048e431 : panfrost: Fix component order XML
    • 25b66e6 : panfrost: Calculate thread count on Bifrost
    • fb32406 : panfrost: Don't export queries
    • f4ecc43 : panfrost: Record architecture major version
    • 81b28eb : pan/bi: Use nir_undef_to_zero
    • 86b2b4e : panfrost: Move nir_undef_to_zero to common util/
    • f042109 : pan/bi: Pipe through tls_size
    • b43b153 : pan/bi: Implement spilling
    • e33b297 : pan/bi: Pack LOAD/STORE
    • 74be83d : pan/bi: Add bi_foreach_clause_in_block_safe helper
    • 2ff54ca : pan/bi: Factor out singleton construction from scheduler
    • 2d0f46a : pan/bi: Implement bi_spill_register
    • 82dbc4e : pan/bi: Add helpers for working with singletons
    • 1db83fc : pan/bi: Add bi_rewrite_index_src_single helper
    • 9e915fd : pan/bi: Add bi_fill
    • 486a820 : pan/bi: Add bi_spill helper
    • 171bf19 : pan/bi: Add spills/fills parameters
    • 7246dd8 : pan/bi: Implement bi_choose_spill_node
    • ab9abc9 : pan/bi: Add no_spill flag to IR
    • ad8a849 : pan/bi: Stub spilling
    • 8477678 : pan/bi: Fix handling of small constants in bi_lookup_constant
    • 82a62a6 : pan/bi: Drop 64-bit constant support
    • d2bfcba : pan/mdg: Cleanup mir_rewrite_index_src_single
    • b19d9c8 : panfrost: Drop panfrost_vt_emit_shared_memory
    • e615209 : panfrost: Use canonical characterization of tls_size
    • a33827d : panfrost: Get rid of the non-native wallpering bits
    • c89f659 : panfrost: Use native wallpapering on Bifrost
    • edd98aa : panfrost: Add support for native wallpapering on Bifrost
    • a677e34 : panfrost: Split panfrost_load_midg()
    • 8892c9c : panfrost: Pass the texture payload through a panfrost_ptr
    • 3f91d81 : panfrost: Rename gtransfer to transfer
    • 1b3b289 : panfrost: Rename panfrost_transfer to panfrost_ptr
    • bf3cd28 : panfrost: Use real name for attribute's unknown field
    • 6b68c82 : panfrost: Build blit shaders on Bifrost too
    • 69c864b : panfrost: Make {midgard,bifrost}_compile_shader_nir() return a program object
    • b83c293 : pan/bi: Add basic support for txf_ms
    • d33c8af : pan/bi: Support the case where TEXC needs 0 or 1 staging reg
    • 4aff27a : pan/bi: Add support for load_sample_id
    • e2d6156 : pan/bi: Print blend descriptor source properly
    • 74c1580 : pan/bi: Make sure we don't print special index as a register
    • a194dcc : panfrost: Replace unkown renderer state fields by their real names
    • 7486b5d : panfrost: Add specialized preload descriptors
    • d769697 : panfrost: Add the bifrost tiler internal state field
    • 2f1947b : panfrost: Fix tiler job injection
    • 519643b : panfrost: Adjust the renderer state definition
    • f2ae8d1 : freedreno/a6xx: Implement user clip/cull distances
    • b4224c3 : tu: Implement clip/cull distances
    • 47f825a : ir3: Handle clip+cull distances
    • 9e063b0 : ir3: Switch tess lowering to use location
    • 4ca38a1 : nir/lower_clip_cull: Store array size for FS inputs
    • cb6ce4a : iris: fix the order of src and dst for fence memcpy
    • d972a6a : nir: get rid of OOB dereferences in nir_lower_io_arrays_to_elements
    • 33c9d4b : radv: replace RADV_TRACE_FILE by RADV_DEBUG=hang
    • 9516f93 : radv: re-order GPU hang report dumps by usefulness
    • 719b9b3 : radv: dump GPU hang report logs into $HOME/radv_dumps_pid
    • ef6d374 : radv: dump UMR ring and waves into the hang report
    • 1847724 : radv: add radv_dump_cmd() helper
    • bf66bbd : ac: add an option to dump GPU info to a file
    • 5f16c5b : v3dv: properly describe swap_color_rb
    • d8cabe0 : v3dv: compute swap_rb flag after applying all swizzles
    • 40788be : v3d/compiler: fix BGRA vertex attributes for vec2/float size.
    • c15468d : broadcom/cle: fix vec size dump when set to 0
    • 53fc3eb : glsl: Initialize lower_shared_reference_visitor members.
    • 0eccd15 : frontends/va: Return P010/P016 as possible surface formats when encoding
    • b50f40f : src/util/disk_cache_os.c: Add missing headers for open/fcntl
    • ddca93d : anv: Enable stencil buffer compression on Gen12+
    • dc22d6b : anv: Pass correct stencil aux usage during MSAA resolve
    • 993a2a7 : anv: Return optimal aux state for stencil buffer compression
    • 815e6c8 : anv: Don't track clear bo for stencil buffer compression
    • d34ab50 : anv: Get aux usage from plane while clearing stencil buffer
    • c76ebc0 : anv: Set stencil_aux_usage flag
    • be2ca24 : anv: Handle compressed stencil buffer transition on Gen12+
    • c2a4102 : anv: Return number of layers/levels attached to anv_image
    • 6795668 : nir: Rename replicated-result dot-product instructions
    • 8cee9ce : spirv: switch to util_bswap32 to improve portability
    • 7e28fbd : ci: Add the new timeout-prone softpipe-gl test to the skips list.
    • 18cef9c : docs: Fix "Hosted by" link and drop duplicate.
    • 1dd6495 : docs: Add a link to the linux kernel DRM docs under "Developer Topics"
    • 2f39dac : docs: Drop extra link to old DRI wiki in the "Help" section.
    • 31e9de9 : loader/dri3: Allocate up to 4 back buffers for page flips
    • 16a7cc4 : loader/dri3: Keep current number of back buffers if frame was skipped
    • 60585fc : loader/dri3: Only allocate additional buffers if needed
    • ea32691 : anv: fix source/destination layers for 3D blits
    • 87934f0 : blorp: allow blits with floating point source layers
    • e067078 : blorp: identify copy kernels in NIR
    • 6694b42 : ac/nir: abort when an unknown intrinsic is reached
    • fa490cb : ac/nir: ignore set_vertex_and_primitive_count intrinsic
    • b739bb7 : compile/nir: Correct printing dest_type
    • 4e2fe34 : aco: fix determining if LOD is zero for nir_texop_txf/nir_texop_txs
    • 4735c8a : nir/loop_analyze: adjust force unrolling to only include interesting modes
    • 542feb9 : ci: disable check commits job for now
    • 8f9c596 : v3dv: handle buffer to linear depth/stencil image copies in blit path
    • 61b8a49 : v3dv: handle compressed image to buffer copies on the blit path
    • 462429b : v3dv: fix Z coordinate for 3D blits
    • a18411d : v3dv: do not attempt to blit from a linear image source
    • 2118c9b : v3dv: fix multi-layered buffer to image copies on the blit path
    • 0a4fc19 : v3dv: skip unnecessary tile loads when blitting
    • c83d6ff : gallivm/nir: handle nir_op_flt in lp_build_nir_llvm
    • 91b64da : driconf: allow higher compat version for Brink
    • edd0b00 : driconf: disable GLX_OML_swap_method by default on Brink
    • d2cb3f8 : driconf: add a way to override indirect-GL extensions
    • 7eedc79 : driconf: add a way to override GLX extensions
    • 97f6d5e : glx: let users force-enable/disable indirect GL extensions
    • e196449 : glx: initial plumbing to let users force-enable/disable extensions
    • 74722c3 : glx: rework __glXCalculateUsableExtensions to be more readable
    • 1331b86 : glx: stop using hardcoded array sizes for bitfields
    • af25f47 : glx/extensions: split set_glx_extension into find_ and set_
    • 4ba255d : dri/DRI2ConfigQueryExtension: add support for string options
    • f9763b2 : driconf: initialize the option value before using it
    • 0e3a424 : driconf: bump the maximum string size from 25 to 1024
    • 81797fc : android: util,ac,aco,radv: Cross-platform memstream API
    • 025050b : glsl: Initialize ir_if_to_cond_assign_visitor members in constructor.
    • 8cf0024 : nir: Use a switch in nir_lower_explicit_io_instr
    • 8033200 : docs/v3d: Add a little stub of v3d documentation.
    • 97dd039 : docs/vc4: Add information on the hw documentation available.
    • c4290a5 : docs/vc4: Move my old vc4 wiki's documentation into docs.mesa3d.org.
    • 4a4b854 : docs/vmware: Move the vmware driver docs into the drivers section.
    • 313f951 : docs: Move the gallium driver documentation to the top level.
    • 9a644d7 : docs: specify redirects in conf.py instead
    • 47f52e8 : docs: specify redirects relative to docs-root
    • 7daf953 : docs: verify that targets for relative redirects exist
    • a85c318 : docs: create leading directories for redirects
    • b2c737c : virgl: Fixes portal2 binary name in tweak config
    • 3d51c27 : st/mesa: initialize lower alpha func to ALWAYS
    • 33ccf0e : nir: drop unused alpha_ref_float
    • 42ee423 : nir: drop support for using load_alpha_ref_float
    • 1404b8b : vc4: do not report alpha-test as supported
    • 8ad9318 : v3d: do not report alpha-test as supported
    • 4c41d19 : ci: Add jobs running ci-fairy checks
    • 7b09d50 : ci: Add empty needs: to pages job
    • 9d1d3a8 : ci: Move test-docs job to deploy stage
    • 549ae5f : st/mesa: make sure prog-info is up to date for NIR (v2)
    • 1126b5c : Revert "st/mesa: don't pass NIR to draw module if IO is lowered"
    • 2335200 : nir: consider load_color intrinsics as both inputs and sysval in gathering
    • e6e3d89 : glx: Move glXGet{ScreenDriver,DriverConfig} to common code
    • ce8530d : glx: Delegate the core of glXGetScreenDriver to the GLX screen vtable
    • 3bb7ebf : glx: move __glXGetUST into the DRI1 code
    • eb6877d : radv,aco: fix use of texop_samples_identical in the resolve meta path
    • eacdad7 : ac/nir: do not sign-extend the result of texop_samples_identical
    • 11d70e7 : docs/features: add v3dv driver
    • 354e375 : frontends/va/postproc: Un-break field flag
    • fd03813 : aco/isel: Miscellaneous cleanups using the new Stage API
    • 34bc947 : aco: Clean up symbol names and comments related to NGG
    • 86c227c : aco: Use strong typing to model SW-HW stage mappings
    • fdbc45d : nir: Only validate in passes that might have changed things.
    • c54c423 : glsl: relax rule on varying matching for shaders older than 4.00
    • c97e82b : panfrost: Only enable occlusion queries when active
    • 04e955f : panfrost: Precise occlusion query support
    • 584b107 : st/mesa: Drop the TGSI paths for drawpixels and use nir-to-tgsi if needed.
    • 57effa3 : st/mesa: Drop the TGSI paths for PBOs and use nir-to-tgsi if needed.
    • da4c4c0 : gallium/ntt: Add default compiler options for non-native-NIR drivers.
    • f5841c3 : v3dv: Fix assert using assign instead of compare.
    • e7ce744 : radv: Advertise VK_KHR_shader_terminate_invocation.
    • 7642166 : aco: Add VK_KHR_shader_terminate_invocation support.
    • 22a6396 : ci: Enable Werror on meson-arm64-build-test.
    • b03fdca : turnip: Add error path handling for descriptor pool init.
    • d384f3b : turnip: Handle the error path for tu/drm's vkResetFences().
    • 296468e : turnip: Handle some error paths in allocating CS space from a command buffer.
    • 9b156ef : freedreno/fdperf: Silence a compiler warning about current counter.
    • a512e9e : freedreno/tools: Fix compiler warnings about using sz in the error paths.
    • 91c5bbc : freedreno/cffdec: Fix format overflow warning.
    • e4dba52 : llvmpipe: enable CL images
    • cda192d : llvmpipe: fix sampler/image binding for clover.
    • 8743718 : nvc0/CL: enable images
    • ee035c7 : nouveau: hide SVM support behing a variable for now as kernel space is broken
    • 1a775b7 : clover/nir: set kernel_image cap
    • 4f24dee : clover/nir: Add an image lowering pass
    • 5e31fad : clover/nir: Calculate sizes of images and samplers properly
    • f6c46e8 : clover/device: use PIPE_MAX_SHADER_SAMPLER_VIEWS for max_images_read
    • 43a42b6 : clover: clCreateImage: calculate image row_pitch and slice_pitch when not provided
    • f2bdb69 : clover: support custom driver strides
    • 9583ce0 : clover: validate image_row_pitch and image_slice_pitch in clEnqueueMapImage
    • 1ff4db1 : clover: Fix incorrect error check in clGetSupportedImageFormats
    • 3aead71 : clover: use pipe_image_view for images instead of set_compute_resources
    • eb96571 : compiler/types: Allow images and samplers in get_explicit_type_for_size_align
    • 0021d3a : compiler/types: Assert non-zero alignments in get_explicit_type_for_size_align
    • ef68f74 : nir/lower_io: Assert non-zero power-of-two alignments
    • 589d918 : spirv: Add 0.5 to integer coordinates for OpImageSampleExplicitLod
    • 2015a10 : anv,iris: Use the data cache for UBO pulls on Gen12+
    • cdc546a : iris: Flush caches based on brw_compiler::indirect_ubos_use_sampler
    • fa483d8 : android: gallium/auxiliary: Deduplicate nir_to_tgsi.c inclusion
    • 0f82c99 : docs: Document how to build and install Android drivers.
    • c4f8d42 : meson: Don't enable libunwind by in 'auto' mode on Android.
    • f2a27d7 : meson: Don't try to build GLX by default on Android.
    • e3c330c : freedreno: Use Android's libsync instead of libdrm's.
    • 81a0f1e : meson: Only require libexpat when a part of the build needs it.
    • 442a769 : meson.build: xxf86vm is not needed for -Dglx-direct=false
    • d0f8fe5 : softpipe: Switch to using NIR as the shader format from mesa/st.
    • 6444f57 : softpipe: Fix buffer overflows in SSBO atomics.
    • 34cc6a8 : gallium: Add a nir-to-TGSI pass.
    • d867e7c : nir: Add an option to not lower source mods for f64/u64/i64.
    • c730fea : nir: Add a call to get a struct describing SSA liveness per instruction.
    • a206b58 : nir: Add a block start/end ip to live instr index metadata.
    • 2f5d184 : nir: Replace nir_ssa_def-live_index with nir_instr-index.
    • b6cb184 : nir: Introduce nir_metadata_instr_index for nir_index_instr() being current.
    • b05c107 : ci: Enable NIR_VALIDATE everywhere.
    • 4ca1030 : radv: move all NIR pass outside of ACO
    • 9aa89b3 : ac/nir: handle non-const offset with txf/txf_ms
    • e690a1b : ac/llvm: don't lower bool to int32, switch to native i1 bool
    • ed0eb51 : util: Fix rwlock Windows include for MinGW
    • afeb0c3 : genxml: drop gen10
    • 6d32fca : Revert "radv/aco: disable NGG GS support because it randomly hangs the GPU"
    • d8435c1 : aco/ngg: Add assertion to make sure we always know the vertex count.
    • d9cb9ff : nir: Emit set_vertex_and_primitive_count for inactive streams.
    • 07fae31 : radv: add missing 'discardtodemote' option in the debug list
    • 4408131 : vulkan/wsi: fix possible random stalls in wsi_display_wait_for_event
    • 85053c0 : panfrost: Fix stride for AFBC_FORMAT_MOD_BLOCK_SIZE_32x8.
    • 8dd03a7 : anv: Advertise VK_KHR_shader_terminate_invocation
    • 8251eed : vulkan: Update XML and headers to 1.2.158
    • 7e55f0e : clover/nir: add a constant folding pass before lowering mem const
    • aca31ba : isl: Enable Tigerlake HDC:L1 caches via MOCS in various cases.
    • 02fe825 : isl, anv, iris: Add a centralized helper to select MOCS based on usage
    • 103ad42 : anv: Set only one ISL usage bit (RT/texture) for CopyBuffer sources
    • 4df98c3 : turnip: Only link libdrm in the DRM case, not KGSL.
    • f63ce9b : turnip: Don't link the WSI code if we don't have a WSI extension.
    • 8f3313f : turnip: Use Mesa's libsync.h instead of libdrm's libsync.h.
    • 8a72666 : turnip: Drop a dead error checking path in device init.
    • 27b8887 : android: Add pre-4.7 Android kernel compatibility to our libsync header.
    • e8c89a9 : util: Import a copy of drm's libsync.h
    • ab57df2 : ci/android: Switch build to using platform SDK version 26.
    • e92f4ac : android_stub: Update platform headers to include gralloc1.h.
    • 7279988 : radv: report latest extension spec versions
    • cf17d62 : meson: Add xcb-fixes to loader when using x11 and dri3. Fixes undefined symbol for xcb_xfixes_create_region in loader_dri3_helper.c
    • 25627ff : mesa: Pass the correct caller string to _mesa_lookup_or_create_texture
    • 9b847d4 : mesa/st: Silence unused parameter warnings in st_context.c
    • 4cd9305 : i965: Silence unused parameter warnings
    • 3c87ac1 : isl: Fix the aux-map encoding for D24_UNORM_X8
    • 9a06238 : anv: Implement VariableDescriptorCount
    • 215218f : anv: Add a descriptor_count to descriptor sets
    • dd1971c : anv: Bump the number of update-after-bind descriptors to 1M
    • 3c0ba7b : zink: reject resource-imports with modifiers
    • 068b4bf : doc/features: remove zink entries for GL 3.3 items
    • af8d488 : util,ac,aco,radv: Cross-platform memstream API
    • 4c2e7b0 : util,radv,radv/winsys: Cross-platform rwlock API
    • 9d4999e : anv: Ignore continue flag in primary cmd buffers
    • b3529e5 : ac/nir: implement missing nir_op_pack_half_2x16_split
    • 57d3178 : radeonsi: enable NGG culling by default on Navi1x PRO cards
    • 67a67ea : radeonsi: enable NGG on Navi14 PRO cards
    • bc3c74e : radeonsi: tweak LATE_ALLOC_GS numbers for faster NGG culling
    • b38b5aa : radeonsi: pack LDS better for NGG culling
    • 8094367 : radeonsi: write VS/TES system values into LDS after culling
    • 1de0bf0 : radeonsi: remove indirection when loading position at the end for NGG culling
    • 823ee12 : radeonsi: optimize out LDS bank conflicts in the NGG culling shader
    • 7648060 : radeonsi: enable NGG culling by default on gfx10.3 dGPUs
    • e7d3f22 : radeonsi: don't disable NGG culling on gfx10.3
    • 6810e6e : Revert "radeonsi/gfx10: disable vertex grouping"
    • 7d1fb5f : v3dv: Initialize time before usage by free_stale_bos.
    • 8eb246d : v3dv: Remove unsigned comparison to zero.
    • 5e27e04 : intel/isl: Drop redundant unpack of unorm channels
    • cf11ebf : st/mesa: Add missing sentinels in format_map[]
    • 0c70268 : zink: mark ARB_sample_shading as supported
    • bf9a1e0 : zink: add a pipe_context::get_sample_position hook
    • b009bd9 : zink: add ntv handling for ARB_sample_shading
    • 441b32b : zink: add helper for vec-type input variables in ntv
    • 7c51299 : zink: implement ARB_draw_indirect
    • 3e3b6e9 : zink: set 3.3 complete in features.txt
    • 80c6664 : zink: bump GLSL to 3.30
    • 44baddf : zink: set 3.2 complete in features.txt
    • 26b076f : zink: GLSL 1.50
    • b27ec49 : zink: mark off GL 3.1 as done in features.txt
    • 815f0c0 : zink: bump to glsl 1.40
    • 8028991 : zink: enable gs pipe caps
    • 4c734da : zink: add gallium handling for geometry shaders
    • d59e4b8 : zink: round out handling for streamout buffer stride setting during draw
    • 9915c54 : zink: remove ADJACENCY prim types from primconvert path
    • d656c60 : zink: add handling for gs in ntv
    • 9671cd9 : zink: re-transform gl_Position for gs input
    • ea2279d : zink: add ntv handling for geometry shader variables
    • 423a342 : zink: handle shader io vars more generically for use with gs
    • 182f7f9 : zink: add some spirv_builder functions we'll be using for geometry shaders
    • 5934fc1 : r600/sfn: Update state docu
    • 9f5e5fb : r600/sfn: Fix the parameter component type
    • 5f66197 : r600/sfn: Use register keep-alive also when scanning the shader
    • 9eda717 : r600/sfn: fix remapping of deleted attributes
    • f79b7fc : r600/sfn: use 32 bit bools
    • 92a4d45 : r600/sfn: use cacheless op for coherent image write
    • 02e39b7 : r600/sfn: Handle mem barrier and image barrier by using ACK
    • f664fc1 : r600/sfn: use shared pointer to GPR for FS sysvalues
    • 876b975 : r600/sfn: Rework get_temp_register to return a smart pointer to GPRValue
    • 92bc066 : r600/sfn: fix mega fetch count for SSBO/Image atomics result fetch
    • d79d20d : r600/sfn: Fix keepalive patch
    • 8625413 : r600/sfn: Fix ssbo resource offset for buffer loads
    • ed66eaf : r600/sfn: Don't reuse registers for workgroup ID and local invocation ID
    • bafb2bb : glsl: Initialize add_uniform_to_shader member var in constructor.
    • aea74ea : zink: fix stencil wrapping
    • b17e264 : glsl: Initialize lower_ubo_reference_visitor members in constructor.
    • e32eadc : v3dv: Fix assert using assign instead of compare.
    • e7e24d5 : intel/fs: Handle nir_intrinsic_terminate
    • 886d2d1 : spirv: Handle SpvOpTerminateInvocation
    • 4dfd292 : spirv: Update headers and metadata from latest Khronos commit
    • f6d5dd8 : nir: Add nir_intrinsic_terminate and nir_intrinsic_terminate_if
    • fdb65b8 : aco: add missing SCC clobber in get_buffer_size
    • 6be2bbe : Revert "iris: Use the data cache for indirect UBO pulls"
    • a5529eb : anv: Go back to using the sampler for UBO pulls
    • d91cb31 : vc4: Enable nir_lower_io for uniforms
    • 3e877ca : vc4: Add missing range_base/range at nir_load_ubos in yuv_blit fs.
    • 4cfdd42 : vc4: Add missing load_ubo set_align in yuv_blit fs.
    • f91b2fe : nir/opt_load_store_vectorize: add some tests for discard/demote behaviour
    • f8e971f : nir/opt_load_store_vectorize: don't vectorize stores across demote
    • 19aaafd : intel: Remove Gen10-specific device entries
    • b0e3af4 : intel: Remove Gen10-speicific perf support
    • fc04733 : anv: Don't generate Gen10-specific functions
    • 5490f5c : iris: Don't generate Gen10-specific functions
    • 495dc51 : intel/isl: Don't generate Gen10-specific functions
    • 1f00b0f : intel: Remove Gen10-specific cache config code
    • 9c19a38 : i965: Don't build Gen10-specific files and libraries
    • 3220ff7 : i965: Remove Gen10-specific state setup and workarounds
    • 262ca98 : intel/compiler: Remove Gen10-specific code
    • e22f106 : intel: Disable all support for Gen10
    • 112f23f : i965: Rename gen10_emit_isp_disable to gen7_emit_isp_disable
    • fe26109 : i965: Make MOCS index tables static const
    • 02a0819 : i965: Silence many unused parameter warnings in genX_state_upload.c
    • 5a5f30b : i965: Silence many unused parameter warnings in genX_blorp_exec.c
    • b9c326b : intel: Silence many unused parameter warnings in blorp_genX_exec.h
    • de77fab : i965: Allow viewport array extensions with allow_higher_compat_version
    • 9223842 : ci: Unskip fragment_ops tests on Bifrost
    • cdfb1d9 : zink: add last few format maps for ARB_vertex_type_2_10_10_10_rev
    • 8e96b4f : zink: add VK_BUFFER_USAGE_INDEX_BUFFER_BIT to vertex buffer creation
    • c364124 : zink: handle null attachment for ARB_texture_buffer_object samplers
    • 69dc0f2 : zink: assert valid format in zink_create_sampler_view()
    • 00db994 : zink: ensure resource tracking for sampler buffers in render batches
    • 63b299d : zink: implement ARB_texture_buffer_object
    • 4c2b02f : zink: increase descriptor pool sizes for other descriptor types we'll be using
    • 4231cc2 : glsl: more accurately handle swizzle in 64bit varying split with no left value
    • d75d12f : aco: don't use v_pack_b32_f16 if 16-bit input denormals are flushed
    • d4b3e86 : aco: propagate literals into sub-dword pseudo instructions on GFX9+
    • 1a65224 : aco: implement 16-bit literals
    • 91d9c55 : panfrost: Add blend shader support to bifrost
    • f04e5ef : panfrost: Add missing tile-buffer formats to the format enum
    • 685d45f : pan/bi: Special-case load_input for blend shaders
    • 0d40460 : pan/bi: Reserve r0-r3 in blend shaders
    • 3432d0a : pan/bi: Special-case BLEND instruction emission for blend shaders
    • 2f3f5da : pan/bi: Collect return addresses of blend calls
    • 8da0a1d : pan/bi: Add load_output support
    • c774896 : panfrost: Flag blend shader function as an entry point
    • 6d3fce5 : panfrost: Scalarize nir_load_blend_const_color_rgba
    • 8d707cd : panfrost: Add a "Bifrost Internal Blend" descriptor
    • e6186c2 : pan/bi: Support indirect jumps
    • 1a1d9cc : pan/bi: Add support for load_blend_const_color_{r,g,b,a}_float
    • 16179c8 : pan/bi: Rework blend descriptor access handling
    • 6dd2a76 : pan/bi: Get rid of the regs argument in bi_assign_fau_idx()
    • f25850b : pan/bi: Use canonical name for FAU RAM sources
    • 111cf7f : pan/bi: Copy blend shader info from compile_inputs
    • 6c61f0b : panfrost: Extend compile_inputs to pass a blend descriptor
    • d8326ce : panfrost: Fix fixed-function blend on bifrost
    • 442f48f : v3d/compiler: implement load interpolated input intrinsics
    • 3ec165b : broadcom/compiler: track partially interpolated fragment inputs
    • 71ed8c5 : iris: Fix doubling of shared local memory (SLM) sizes.
    • 341f5bf : intel/compiler, anv: Delete cs_prog_data-slm_size
    • e881290 : broadcom/compiler: use nir io semantics
    • 9b01598 : nir/lower_io_to_scalar: update io semantics on per-component inst
    • ac5f0ee : broadcom/compiler: support varyings with struct types
    • ebd5b55 : docs/release-calendar: plan 20.3 release
    • f8117f7 : intel/fs: Allow constant-propagation into SAMPLEINFO and IMAGE_SIZE
    • 438a409 : docs: update calendar and link releases notes for 20.1.10
    • 713b666 : docs: add release notes for 20.1.10
    • 0d9216a : isl: Allow CCS for 8bpp surfaces with 3+ miplevels
    • f94ba6b : iris: Add fast-clear restriction for 8bpp surfaces
    • 1affcea : docs: update calendar and link releases notes for 20.2.1
    • dee2fdb : docs: add SHA256 sums for 20.2.1
    • 3c89e7b : docs: add release notes for 20.2.1
    • bb00a68 : radv: fix optimizing needed states if some are marked as dynamic
    • 68daac2 : docs: Document how to replicate a CI build locally.
    • 0767af3 : ci/android: Switch to using the Android NDK.
    • ad61899 : symbols-check: Add __cxa_guard_* to the list of approved symbols.
    • 4722491 : glsl/tests: Make the tests skip on Android binary execution failures.
    • f51ce21 : meson: Drop adding -Wl,--gc-sections to project c/cpp arguments.
    • d5a7231 : aco/isel: Remove now unused VS-related code from create_null_export
    • c22c702 : aco/isel: Remove some dead code
    • bf51b11 : aco/isel: Always export position data from VS/NGG
    • f29c81f : aco: use VOP2 for v_cvt_pkrtz_f16_f32 if possible
    • 7240ede : aco: use VOP2 version of v_cvt_pkrtz_f16_f32 on GFX_6_7_10
    • 2f12590 : radv,aco: lower_pack_half_2x16
    • dae1e6f : aco: use v_cvt_pkrtz_f16_f32 for pack_half_2x16
    • 9185b7c : aco: add validation rules for p_split_vector
    • aec872c : aco: use p_split_vector for nir_op_unpack_half_*
    • f503699 : nir/opt_algebraic: optimize unpack_half_2x16_split_x(ushr, a, 16)
    • a38a497 : aco: use p_create_vector for nir_op_pack_half_2x16
    • 3c2abd7 : aco: expand create_vector more carefully w.r.t. subdword operands
    • d887eb1 : aco: propagate SGPRs into VOP1 instructions early.
    • 3424e17 : zink: unify code for emitting named uint-based variable instructions
    • 20d73a9 : aco: adjust an assertion about the wavesize in emit_gfx10_wave64_bpermute()
    • 112e66f : aco: compute the CS workgroup size from the shader NIR info
    • e3e8d13 : radv: move compiler statistics to ACO
    • 97afb2a : aco: remove unused radv_shader.h includes
    • 408195e : aco: remove useless occurences of radv_nir_compiler_options
    • 8a6f60f : aco: remove stub lower_wqm() prototype
    • 23e731f : zink: export PIPE_CAP_MAX*_VARYINGS values
    • d50e855 : zink: add feature-documentation
    • f85488a : zink: redo slot mapping again for the last time really I mean it
    • 4f144dc : zink: don't leak sampler view textures
    • 270969b : zink: explicitly flag fb attachments as being written to in render passes
    • 8dfb941 : zink: add more explicit fencing for transfer maps
    • e3ed624 : zink: optimize transfer_map for resources with pending reads/writes
    • c6687ee : zink: add a mechanism to track current resource usage in batches
    • 48b988e : radv: fix ignoring the vertex attribute stride if set as dynamic
    • 28d02b9 : ac,amd/llvm,radv: Initialize structs with {0}
    • b84d1a0 : radv/aco: disable NGG GS support because it randomly hangs the GPU
    • 21422b1 : nir/opt_uniform_atomics: remove useless returns
    • 7758664 : radv: Only close local_fd when valid
    • 4ca6faa : util: Hide timespec_passed on Windows
    • 1026e2a : radv: Increased const usage
    • 5abac85 : intel/fs: Rework scratch handling on Gen9+
    • e557af9 : intel/fs/ra: Use a set to track added spill/fill instructions
    • f650c4c : intel/fs/ra: Sanity-check our IP counts
    • d80d0a6 : intel/fs/ra: Store the last non-spill VGRF node
    • 2af6528 : intel/fs/ra: Refactor handling of Gen7 scratch reads
    • 74a1843 : intel/fs/ra: Increment spill_offset as part of the emit_spill loop
    • 06ebf23 : intel/fs: Add a SCRATCH_HEADER opcode
    • 24b64c8 : intel/fs: Copy the PTSS from g0 for scratch reads/writes
    • f103012 : intel/batch_decoder: Don't clame vec4 vs/gs/tcs shaders on Gen11+
    • e1efc53 : v3dv/device: Support loader interface version 3.
    • 866f949 : v3dv: fix buffer copies to compressed images on the blit path
    • 82b9789 : v3dv: drop a couple of obsolete comments
    • ec78ee0 : v3dv: limit blit framebuffer dimensions to max coordinates
    • 43063d8 : v3dv: generate proper UUIDs for device and driver
    • c7814bc : v3dv: fix blit path for copies from 3D compressed images
    • 24e46f7 : v3dv: fixes for barriers in secondary command buffers
    • 376b009 : v3dv: implement workaround for GFXH-1918
    • 1e81bb0 : v3dv: implement workaround for GFXH-1461
    • 6fb45a0 : v3dv: flag tmu_dirty_rcl in primaries when linking secondaries that have it set
    • 30534fc : v3dv: only advertise one memory type
    • 49549f4 : v3dv: always program a reasonable internal depth type for copies/clears
    • 1ed8252 : v3dv/pipeline_cache: extend pipeline cache envvar
    • 93e3b00 : v3dv/pipeline_cache: set a max size for the pipeline cache
    • 7eb8eb1 : v3d/compiler: allow to batch spills
    • f7af9eb : v3dv: free noop job if needed when finishing the queue
    • cec3d5e : v3dv: clean-up after obtaining an XCB connection
    • deb0dce : v3dv: don't leak dumb BO handles allocated for swapchain images
    • e7ea918 : v3dv/meta_copy: fix TFU blitting when using 3D images
    • db737a0 : v3dv: honor VkPipelineDepthStencilStateCreateInfo::depthWriteEnable
    • 966b14a : v3dv: fix sampling from stencil aspect of a combined depth/stencil image
    • abca6c9 : v3dv/formats: properly return unsupported for 1D compressed textures
    • 0df0449 : v3dv: signal semaphore/fence if needed after acquiring a swapchain image
    • e14cdd6 : v3dv: do not expose VK_IMAGE_USAGE_SAMPLED_BIT for swapchains
    • e6fb805 : v3dv: fix size computed by vkGetImageSubresourceLayout for 3D images
    • 83336ad : v3dv: fix offset computed by vkGetImageSubresourceLayout for array images
    • 4cc7d6a : v3dv: expose DRM modifiers based on supported features
    • fdcc36a : v3dv: handle VK_STRUCTURE_TYPE_PHYSICAL_DEVICE_IMAGE_DRM_FORMAT_MODIFIER_INFO
    • 287af80 : v3dv/meta_copy: handle mirroring z component bliting 3D images
    • 9518b67 : v3dv: fix color clear pipeline destruction for 32-bit architectures
    • 4acf598 : v3dv: hook up robust buffer access
    • 23c727d : v3d/compiler: add a lowering pass for robust buffer access
    • 4401dde : broadcom/compiler: rename QUNIFORM_GET_BUFFER_SIZE to QUNIFORM_GET_SSBO_SIZE
    • 6f55abc : v3dv: handle QUNIFORM_GET_UBO_SIZE
    • d93d903 : v3d/compiler: implement nir_intrinsic_get_ubo_size
    • f4c754b : nir: add a nir_get_ubo_size intrinsic
    • 4823313 : v3dV: move meta init/finish to meta implementation files
    • db0bb36 : v3dv: don't cache subpass color clear pipelines
    • 07828c0 : v3dv: fix 3D image blits
    • 0d04651 : v3dv: only require texel-size alignment for linear images
    • 57a337f : v3dv: lower interpolateAt functions in NIR and enable sample rate shading
    • 6004ad9 : nir/lower_io: add an option to lower interpolateAt functions
    • 4394efd : v3dv/device: enable largePoints
    • a0153c9 : v3dv/device: fix point-related VkPhysicalDeviceLimits
    • ff02458 : v3d/limits: add line width and point size limits
    • 7620a1b : v3dv/cmd_buffer: set instance id to 0 at start of tile
    • fa7fbdb : v3dv/pipeline: set 16bit return_size for shadows always
    • 229cce4 : v3dv/pipeline: track if texture is shadow
    • c1bd8e0 : v3dv: Call nir_lower_io for push constants
    • 8f06967 : v3dv/pipeline: use derefs for ubo/ssbo
    • 17c5a09 : v3dv/device: fix compute_heap_size for the simulator
    • 73b4720 : v3d/simulator: add v3d_simulator_get_mem_size
    • 02b9670 : broadcom/compiler: allow GLSL_SAMPLER_DIM_BUF on txs emission
    • e8ceb8f : v3dv/meta: fix hash table insertion
    • 5a59a33 : v3dv/pipeline: fix combined_index_map insertions
    • edec743 : v3dv/debug: add v3dv_print_v3d_key
    • ee605bd : v3dv/device: warn when the pipeline cache is disabled
    • 54efbe0 : v3dv/device: add assert for texture-related limits
    • 29ef4dd : v3dv: handle multisample rasterization with empty framebuffers
    • 644a15e : v3dv: implement nir_texop_texture_samples
    • 6ee88f0 : v3dv: enable sample rate shading if fragment shader reads gl_SampleID
    • 1c4c7d9 : broadcom/compiler: track if the fragment shader forces per-sample MSAA
    • 5de282b : v3dv/descriptor: remove v3dv_descriptor_map_get_image_view
    • 888cefc : v3dv/uniforms: handle texture size for texel buffers
    • 531ea35 : broadcom/compiler: implement nir_intrinsic_load_sample_pos
    • b7ca467 : v3dv/formats: fix exposing FEATURE_UNIFORM/STORAGE_TEXEL_BUFFER_BIT
    • c68ed39 : v3dv: handle multisample image clears
    • 4a63b1a : v3dv: handle multisample resolves for formats that don't support TLB resolves
    • 049ea81 : v3dv: handle multisample resolve of integer formats
    • a4b6845 : v3dv: fix blitting of signed integer formats
    • 50351df : nir/glsl: add a glsl_ivec4_type() helper
    • 7bbce22 : v3dv: amend tile size tables with smallest tile sizes available
    • f16da2c : v3dv/device: fix minTexelBufferOffsetAlingment
    • 947fa70 : v3dv: add v3dv_limits file
    • 81b713e : v3dv/descriptor: support for UNIFORM/STORAGE_TEXEL_BUFFER
    • 14d74c0 : broadcom/compiler: handle gl_SampleMask writes in fragment shaders
    • 3310c15 : v3dv: handle multisampled image copies with the blit path
    • f219795 : v3dv: add a blit fallback path for vkCmdResolveImage
    • 6c27dbd : v3dv: setup texture shader state correctly for multisampled images
    • 0061781 : v3dv: handle multisampled image copies in the TLB path
    • 8a1f671 : v3dv: implement vkCmdResolveImage for whole images
    • b48e01e : v3dv: implement subpass multisample rendering and resolve
    • d953eab : v3dv: process VkPipelineMultisampleStateCreateInfo properly
    • d87941c : v3dv: consider MSAA when computing frame tiling
    • 7c9b40e : v3dv/pipeline: handle properly OUT_OF_HOST_MEMORY error when allocating p_stage
    • d64ff26 : v3dv/pipeline: pre-generate more that one shader variant
    • a87d2c7 : v3dv/pipeline: when looking for a variant, check first current variant
    • 62c32d6 : v3dv/pipeline: remove custom variant cache
    • 35381a4 : v3dv/pipeline_cache: add default pipeline cache
    • 48a64f2 : v3dv/pipeline: provide a shader_sha1 to private ShaderModules
    • cead2a6 : v3dv/pipeline_cache: MergePipelineCaches implementation
    • 8e461b8 : v3dv/pipeline_cache: support to serialize/deserialize cached NIRs
    • a00fe4c : v3dv/pipeline_cache: cache v3dv_shader_variants
    • 63b6b63 : v3dv/pipeline: add basic ref counting support for variants
    • 2326d5b : v3dv/pipeline_cache: cache nir shaders
    • 1d2ae87 : v3dv/pipeline_cache: bare basic support for pipeline cache
    • fd70e8e : v3dv/device: proper pipeline cache uuid
    • 8f46750 : v3dv/device: get proper device ID under simulator
    • ffaab55 : v3dv/device: add vendorID/deviceID get helpers
    • a272315 : v3dv: fix dynamic state after meta operation
    • 33a6fdb : v3dv: enable fillModeNonSolid
    • 558987a : v3dv: enable shaderClipDistance
    • 47e02a2 : v3dv: add a fast path for vkCmdClearAttachments
    • c5d0780 : v3dv: only clear depth/stencil attachments if any aspect needs clearing
    • c4564f6 : v3dv: emit new shader state if viewport is dirty
    • e7678f7 : v3dv: only care about barriers between GPU jobs
    • 08193cc : v3dv: avoid prime blit path when presenting WSI images
    • 9789189 : v3dv: only use per-buffer clear bit for cases were we are already storing
    • add8778 : v3dv: ignore stencil load operation if attachment format doesn't have stencil
    • 1e1903f : v3dv: use a binning sync for CL jobs waiting on a semaphore
    • 0db95de : v3dv: improve pipeline barrier handling
    • adbce77 : v3dv: don't support sRGB buffer formats
    • d08da4a : v3dv: handle VK_FORMAT_B8G8R8A8_UNORM vertex attributes
    • 5a2ef59 : v3d/compiler: support swapping R/B channels in vertex attributes.
    • a2538b2 : v3dv: make sure we emit vertex attributes in location order
    • aebfdfa : v3dv/device: handle primary nodes for newer kernels
    • 3bd02a7 : v3dv: fix color border clamping with specific formats
    • f156c5d : v3dv: fix regressions for cubemap array load/store
    • 9eb1b57 : v3dv: use swizzle X001 with D/S formats
    • 336d9e1 : v3dv/image: fix TEXTURE_SHADER_STATE depth for cube arrays
    • cb1e492 : v3dv: handle unnormalized coordinates in samplers
    • 0385da9 : v3dv: fix blit_shader() to honor the region's aspect mask
    • bf20a5e : v3dv: fix srcSubresource description for image to buffer blits
    • 6053e85 : v3dv: fix textureSize() for cube arrays
    • b6c2fde : v3dv: fix vkCmdCopyBuffer unaligned TLB access
    • 56fcc08 : v3dv: handle draw after barrier
    • 8116d65 : v3dv: improve handling of too large image sizes
    • b546155 : v3dv: add workaround for GFXH-1602
    • f6e7290 : v3dv: fix GFXH-930 workaround
    • e01d6af : v3dv/device: expose support for image cube array
    • 10b79bf : nir: include texture query lod as one of the ops that requires a sampler
    • b6ccdd5 : v3dv/uniforms: support for some QUNIFORM_IMAGE_XXX
    • 42b1cac : v3dv/pipeline: lower_image_deref
    • 4074110 : v3dv/descriptors: support for DESCRIPTOR_TYPE_STORAGE_IMAGE
    • 9190756 : v3dv: implement support for shader spilling
    • 182e7c2 : v3dv: lower frexp
    • 4b86903 : v3dv: lower unpack_{u,s}norm_2x16
    • e59e706 : v3dv: don't reset descriptor state after a meta operation
    • 4b9e3bb : v3dv: handle empty set layouts
    • 9833a5a : v3dv: initial support for input attachments
    • 19d3639 : v3dv/pipeline: set load_layer_id to zero
    • 6e0cec5 : v3dv/pipeline: iterate used textures using the combined index map
    • b9dd7e3 : v3d/tex: avoid to ask back for a sampler state if not needed
    • 0dbf040 : v3dv: pipeline initialization fixes for disabled rasterization
    • d809d9f : v3dv: don't free BOs from imported memory objects
    • ddb352b : v3dv: assert that our framebuffers are single sampled
    • 95caf46 : v3dv: return a proper error for too large buffer allocations
    • b356d3d : v3dv: implement indirect compute dispatch
    • 1d6edcc : v3dv: always emit index buffer state for new jobs
    • 2f1c151 : v3dv: handle unsized arrays in SSBOs
    • 7e99068 : v3dv: implement compute dispatch
    • d0b1bb3 : v3dv: handle separate binding points for compute and graphics
    • 67d5b0c : v3dv: support compute pipelines
    • 682cfd8 : v3dv/bo: add dump stats info
    • f78c99f : v3dv/bo: add a maximum size for the bo_cache and a envvar to configure it
    • 2adea94 : v3dv/bo: adding a BO cache
    • e1a11b8 : v3dv: always map full BOs
    • 7fb9afc : v3dv: try to use TFU path when creating tiled images from linear buffers
    • 1f8343b : v3dv: add a CPU path for buffer to image copies
    • 1e57995 : v3dv: add a TFU path for buffer to image copies
    • e1c8041 : v3dv: try harder to skip emission of redundant state
    • 420d13b : v3dv: actually enable early Z
    • ca1e29a : v3dv: fix release build warnings
    • d727881 : v3dv: fix a few cases where we were ignoring suballocated buffers
    • 9bc4613 : v3dv: regen BO lists for CLs inside cloned jobs
    • 2361702 : v3dv/meta-copy: add uintptr_t casting to avoid warning
    • efb1880 : v3dv: fix BCL start offset in presence of chained BOs
    • 204c46a : v3dv: warn users that this is not a conformant driver
    • b2e5631 : v3dv: add stubs for missing API implementations
    • 9a89484 : v3dv/descriptor: use descriptor pool bo for image/samplers
    • 18553b5 : v3dv/descriptor: add general bo on descriptor pool
    • 509c8a6 : v3dv: don't leak attachment state
    • 3092ddf : v3dv: support vkCmdClearAttachments in secondary command buffers
    • ca4b42a : v3dv: implement vkCmdWaitEvents for secondary command buffers
    • 6a34ef6 : v3dv: add basic support for secondary command buffers
    • fa03471 : v3dv: fix bogus command buffer allocation scopes
    • ce2a799 : v3dv: handle OOM properly during command buffer recording in more places
    • 814dd98 : v3dv: ensure BCL space is available before emitting packets
    • 45af201 : v3dv: check that GPU device matches requirements
    • ca86c7c : v3dv: assert command buffers are executable when submitting to a queue
    • 8e4cf16 : v3dv: remove some unnecessary / unused functions
    • e048eba : v3dv: drop the extra BO handling from the command buffer
    • bbdfc52 : v3dv: make TLB clearing paths return true/false
    • a5cede6 : v3dv: fix vkResetCommandPool
    • 27d360c : v3dv: don't leak BOs from CLs when using BRANCH
    • e723225 : v3dv/pipeline: support for specialization constants
    • 03f5fae : v3dv/cmd_buffer: move variant checking to CmdDraw
    • 57a254c : v3dv: implement events
    • 10dbb1e : v3dv: make the driver more robust against OOM
    • 57d0ff8 : v3dv: fix depth/stencil clears on hardware
    • ce681aa : v3dv: fix the command buffer private object framework for 32-bit
    • 1957689 : v3dv: fix dynamic blend constants
    • 98ade03 : v3dv: implement wide lines
    • 21936e8 : v3dv: ignore dynamic updates of depth bounds state
    • 9aaf07e : v3dv: implement depth bias
    • 88a5943 : v3dv: drop blit path for depth/stencil formats
    • 5ba9f81 : v3dv: do not rewrite blit spec for combined depth/stencil in get_blit_pipeline
    • b0af413 : v3dv: implement partial image to buffer copies
    • f1b1d5b : v3dv: only require 4-byte alignment for linear images
    • fe0ccdb : v3dv: fix image addressing calculations to account for suballocation
    • 83eb1f2 : v3dv/meta-copy: ensure valid height/width with compressed formats
    • d3c6bd6 : v3dv: always return true from a fallback path if it can handle the case
    • 21633aa : v3dv: implement partial buffer copies to depth/stencil images
    • df364f9 : v3dv: support blitting both depth and stencil aspects at the same time
    • 0a3c7ac : v3dv: implement partial buffer copies to color images
    • ee4be1b : v3dv: handle copies from/to compressed formats
    • abea0ec : v3dv: limit software integer RT clamp to rgb10a2
    • 3432e51 : v3d: fix Tile Rendering Mode Cfg (Color) packet description
    • e270d2f : v3dv: implement stencil aspect blits for combined depth/stencil format
    • eac43d0 : v3dv: implement fallback for partial image copies
    • 1b4a9c7 : v3dv: properly return OOM error during pipeline creation
    • 2894d6a : v3dv: handle texture/sampler shader state bo failure with OOM error
    • 05adac2 : v3dv: use the private object framework in the meta clear path
    • f49d9d7 : v3dv: fix leaks during recording of meta blits
    • bfadd70 : v3dv: add framework for private driver objects
    • e18865b : v3dv: support blits with 1D and 3D images
    • 7fecf7a : v3dv: remove incorrect assert
    • 8b791a5 : v3dv: don't support 1D depth/stencil for transfer sources or sampling
    • 94de6a2 : v3dv: don't support blitting of combined depth/stencil formats
    • c441993 : v3dv: support depth blits
    • 9e30475 : v3dv: handle miplevel correctly for blits
    • 05cf7b0 : v3dv/blit: fix integer blits from larger to lower bit size
    • 6308628 : v3dv: don't leak state BO from samplers
    • 71a75a5 : v3dv: don't leak the texture shader state BO from image views
    • a9ce6ad : v3dv: don't leak the compiler from the physical device
    • 55729ea : v3dv: don't leak prog_data from shader variants
    • 8e0a341 : v3dv: don't leak default pipeline attributes BO
    • 89e426a : v3dv: don't leak host memory allocated for shader variants
    • 77bccfd : v3dv: don't leak NIR code in pipelines
    • c414a24 : v3dv: move early-Z update to pre-draw
    • ffacf92 : v3dv: require optimal tiling for features that reqiure sampling
    • 831aa5d : v3dv: implement shader draw fallback for vkCmdBlitImage
    • 5fc55e3 : v3dv: save and restore push constant state during meta operations
    • 441d391 : v3dv: save and restore descriptor state during meta operations if needed
    • 689bac3 : v3dv: meta operations can happen outside a render pass
    • bd53ee6 : v3dv: reset subpass index at render pass end
    • 30f1fc2 : v3dv: implement TFU blits
    • d1b60d2 : v3dv: add a bunch of API stubs
    • 71ba6a1 : v3dv: simplify handling of no-op jobs
    • 7ea4601 : v3dv: submit a no-op job if a command buffer doesn't have any jobs.
    • 53657b0 : v3dv: implement occlusion queries
    • 8c09324 : v3dv: reset all state to dirty when we start a new job for a command buffer
    • 5c82656 : v3dv/format: expose correctly if a texture format is filterable
    • 781dd7f : v3dv: fix subpass merge tests
    • b94946c : v3dv/uniforms: fill up texture size-related uniforms
    • 9b98d36 : v3dv/descriptor: handle not having a sampler when combining texture and sampler id
    • 56b611a : v3dv: emit instanced draw calls when requested
    • f41857e : v3d/compiler: implement nir_intrinsic_load_base_instance
    • ee917d2 : v3dv/descriptor_set: combine texture and sampler indices
    • 9d8b1b0 : v3dv/descriptor: move descriptor_map_get_sampler, add and use get_image_view
    • 1b80bac : v3dv: handle partial clears of just one aspect of combined DS targets
    • b31ea83 : v3dv: simplify partial clearing code
    • 360adff : v3dv: fix incorrect attachment reference
    • 03a150c : v3dv: fix incorrect attachment reference
    • 93fd897 : v3dv: compute tile granularity for each subpass
    • f7059e4 : v3dv: set render area for partial clears to match clear rect
    • ad64add : v3dv: create a v3dv_cmd_buffer_subpass_resume helper
    • debdea6 : v3dv: push/pop more state during meta operations
    • dedff74 : v3dv/descriptor_set: support for immutable samplers
    • 53de889 : v3dv: assert on subpasses that use input or resolve attachments
    • 79b07b8 : v3dv: handle stencil load/store operations
    • 7e96f9b : v3dv: restrict render pass clears to the render area
    • ad0a68d : v3dv: use the TLB to clear attachments even if we have an active scissor
    • 4db5573 : v3dv: don't always assert that we have an active job
    • c9b87e9 : v3dv: fix v3dv_GetRenderAreaGranularity to account for attachment bpp
    • 22d6326 : v3dv: check the render area against the clip window
    • e5642a3 : v3dv: store the clip window in the command buffer state
    • d6d8bfb : v3dv: implement proper caching for partial clear pipelines
    • c09c8c0 : v3dv: implement partial depth/stencil attachment clears
    • 7a39e5e : v3dv: implement partial color attachment clears
    • 152a641 : v3dv: improve asserts for VkPipelineColorBlendStateCreateInfo handling
    • c6dca4c : v3dv: allow to create shader modules from NIR
    • f47d082 : v3dv/pipeline: fix adding texture/samplers array elements to texture/sampler map
    • 6e39565 : v3dv/cmd_buffer: allow return in the middle of variant update if needed
    • 07addb4 : v3dv/cmd_buffer: update shader variants at CmdBindDescriptorSets/CmdBindPipeline
    • f76dad8 : v3dv/pipeline: add support for shader variants
    • b6d4148 : v3dv/uniforms: filling up QUNIFORM_TMU_CONFIG_P0/P1
    • 272f0d9 : v3dv/descriptor_set: added support for samplers
    • 3491053 : v3dv/pipeline: sampler lowering
    • 522f711 : v3dv/pipeline: unify local allocator name
    • 477b142 : v3dv/uniforms: cleaning up, moving udpate ubo/ssbo uniforms to a function
    • f2d1652 : v3dv/cmd_buffer: push constants not using descriptor anymore
    • ca186f7 : v3dv: fix a1r5g5b5 format
    • 45e86d8 : v3dv: don't emit the subpass RCL for jobs that have emitted their own
    • 32d261f : v3dv: fix job subpass index for vkCmdClearAttachments jobs
    • 010589a : v3dv: fix clearing of 3D images
    • 3faa7db : v3dv: fix incorrect image slice selection
    • 7de80fa : v3dv: use compatible TLB formats if possible during copies and clears
    • cfc0e15 : v3dv: make sure we only expose transfer features for formats we can use
    • 19841fd : v3dv: check support for transfer usage flags
    • 9e7403d : v3dv: improve assert handling for fallback paths on meta copy/clear operations
    • d471c47 : v3dv: assert on vkCreateComputePipelines
    • df2bb5b : v3dv: disable depth/stencil testing if we don't have a depth/stencil attachment
    • e86f381 : v3dv: drop incorrect assertion on number of clear values at render pass begin
    • a81e63a : v3dv: use perp end caps rasterization mode for lines
    • 69ce1f4 : v3dv: only emit config bits and varyings packets if needed
    • e903075 : v3dv/cmd_bufffer: rename and split emit_graphics_pipeline
    • 6bd4e69 : v3dv: stencil state fixes
    • c5784fb : v3dv: drop redundant emission of stencil state
    • 7c1910f : v3dv: rewrite dirty state handling
    • c525849 : v3dv: only emit blend state if the pipeline is dirty
    • 1e54b62 : v3dv: implement dynamic state for blend constants
    • 4c15131 : v3dv: always flush draw calls if we are doing sRGB blending
    • 761b0c0 : v3dv: add an 'always flush' mode
    • ebbd326 : v3dv: only expose blending on formats that support it
    • d73bb59 : v3dv: implement color blending
    • 4c8531a : v3dv: fix depth/stencil clear color
    • 03048b2 : v3dv: fix copies and clears of 3D images
    • 2721c55 : v3dv: use vk_error() for all queue/submit errors
    • 2fc31a4 : v3dv: return OOM error if we fail to import or export sync objects
    • 8ed2e53 : v3dv: support submits without a command buffer
    • d9152cb : v3dv: only export the last job sync object once
    • 5cfc3b5 : v3dv: handle the case where we fail to allocate a new job gracefully
    • e5adc33 : v3dv: don't leak job allocations
    • a1337c0 : v3dv: work around viewport Z scale hardware bug
    • db629c0 : v3dv: fix viewport Z
    • 57f5fa0 : v3dv: fix scissor outside viewport
    • 5ba6fd3 : v3dv: fix image tiling configuration
    • f986a60 : v3dv: drop assert for map of a mapped buffer
    • 75b1dea : v3dv: drop incorrect assertion
    • 3e3b8db : v3dv: don't assume that VkPipelineColorBlendStateCreateInfo is provided
    • 2bec99c : v3dv: lower {i,u}mulExtended
    • 040a3dd : v3dv: lower usubborrow and uaddcarry
    • 236e5a7 : v3dv: split fragment shader array outputs
    • 51c5396 : v3dv/device: tweak ssbo/ubo device limits
    • 521f862 : v3dv: fix incorrect sizing of the vertex attribute state array
    • 0e5571c : v3dv: call nir_lower_io_arrays_to_elements_no_indirects on vertex shaders
    • 1f41a12 : v3d/compiler: implement nir_op_fquantize2f16
    • cb7f148 : v3dv/descriptor: take into account pPushConstantRanges
    • d734dba : v3dv: implement interpolation qualifiers
    • 02de1d9 : v3dv: add a no-op fragment shader if we don't have one
    • 1811c03 : v3dv/pipeline: clean up io lowering
    • 8da4de7 : v3dv/pipeline: revamp nir lowering/optimizations passes
    • 518714a : v3dv: drop incorrect assertion
    • 26bfbf2 : v3dv: don't reset loader data on command buffers
    • 9b4a1fb : v3dv: assign driver locations on fragment shader output variables
    • fadccc1 : v3dv: trivial refactors in a few meta copy helpers
    • 3138094 : v3dv: more frame tiling refactors
    • 2a3fa91 : v3dv: rewrite frame tiling setup
    • 44cda4d : v3dv: fix subpass tracking in the command buffer state
    • 4a8fdb8 : v3dv: fix framebuffer format when computing fragment shader key
    • 6a90938 : v3dv: take memory format from appropriate miplevel for image load/store
    • 5d1089b : v3dv/descriptor: support for dynamic ubo/ssbo
    • 404220c : v3dv: fix supertile coverage when render are size is 0.
    • 44370b4 : v3dv: implement vkGetRenderAreaGranularity
    • 1af2606 : v3dv: fix fill buffer with VK_WHOLE_SIZE
    • d71a5af : v3dv: don't support image formats that we can rendet to or texture from
    • 8e163ab : v3dv: fix image clearing with VK_REMAINING_*
    • 1ecd188 : v3dv: don't use TLB path for formats that are not supported for rendering
    • 7f3e79e : v3dv: don't swap R/B channels for VK_FORMAT_R5B6G5_UNORM_PACK16
    • aa9ad02 : v3dv: implement vkResetCommandPool
    • c4f7455 : v3dv: meet requirements for supported format features and properties
    • e338503 : v3dv: declare that we support robust buffer access
    • 5dca6dc : v3dv: adjust a few limits to comply with CTS minimum requirements
    • d6afbcb : v3dv: implement host-side event handling functions
    • aba2a66 : v3dv: implement vk{Create,Destroy}BufferView
    • 43c1fa4 : v3dv: destroy wsi device during physical device termination
    • eac594b : v3dv: implement vkQueueWaitIdle
    • 7c4827c : v3dv: honor swizzle for non-copy operations of color formats
    • 204d06c : v3dv: add assertions for unimplemented fallback paths
    • 79569fb : v3dv: implement vkResetCommandBuffer
    • ea09600 : v3dv: initialize in_sync_bcl in our submits
    • 8355484 : v3dv: add support for primitive restarts on indexed draw calls
    • 914a23a : v3dv: implement indirect draws
    • 06fa292 : v3dv: add an assert to catch applications trying to clear invalid aspects
    • 9a2d6cc : v3dv: vkCmdCopyBufferToImage for depth/stencil formats
    • f13dd2e : v3dv: implement vkCmdCopyBufferToImage for color formats
    • e9085e5 : v3dv/cmd_buffer: support for push constants
    • c821273 : v3d/compiler: handle GL/Vulkan differences in uniform handling
    • 4e2817f : v3dv: implement vkCmdClearDepthStencilImage
    • 11310ee : v3dv: fix buffer automatic stride for image to buffer copies
    • 8da9da8 : v3dv: implement vkCmdClearColorImage
    • 1aa5fc0 : v3dv: implement vkCmdCopyImage
    • efbf91c : v3dv: fix copy size for image to buffer copies
    • 1ee270f : v3dv: refactor common code in meta copy operations
    • d39901f : v3dv: add a concept of a fake framebuffer for meta-copy operations
    • 7a2544e : v3dv: move the framebuffer setup code for buffer copy/fill to a helper
    • 5ed78d9 : v3dv: implement vkCmdFillBuffer
    • b3f9a3b : v3dv: implement vkCmdUpdateBuffer
    • 2f0fbc7 : v3dv: implement vkCmdCopyBuffer
    • 18824bf : v3dv: don't advertise texel buffer support yet.
    • c7404f4 : v3dv: Add more supported formats to our format table
    • 83e8325 : v3dv: no need to manually add assembly bo to the job
    • 3dff3f3 : v3dv: take the number of layers from the framebuffer
    • 0dc4416 : v3dv: ignore image view aspects for depth/stencil attachments
    • 50652cb : v3dv: fix clockwise primitive setting
    • 1c69384 : v3dv: implement indexed draws
    • f487888 : v3dv: implement vkCmdClearAttachments
    • 255d424 : v3dv: add a helper to get the Z/S buffer from an aspect mask
    • 6869229 : v3dv: add a helper to compute the hardware clear color
    • aa8a61f : v3dv/pipeline: null check for pCreateInfo-pDepthStencilState
    • 34cf475 : v3dv/descriptor_set: support for array of ubo/ssbo
    • 9afd24f : v3dv: initial descriptor set support
    • 7d6fbea : v3dv: debug nir shader also after spirv_to_nir
    • 29a7d99 : v3dv: fix the mess with dynamic state handling
    • 134afdf : v3dv: implement dynamic stencil states
    • d84458b : v3dv: clamp stencil masks and reference value to supported limits
    • 590f293 : v3dv: implement early Z optimization
    • 72040f9 : v3dv: fix viewport state from pipeline
    • 8d143a0 : v3dv: support copying depth/stencil aspects to buffer
    • 79b1b68 : v3dv: fix indentation
    • b44d434 : v3dv/bo: add a bo name
    • 55acd9f : v3dv: implement stencil testing
    • bb6e9d2 : v3dv: support depth testing on combined depth/stencil formats
    • c608e13 : v3dv: select correct internal type for depth/stencil formats
    • 0e59771 : v3dv: select the depth/stencil buffer from the attachment aspect mask
    • df34b8f : v3dv: compute subpass ranges for attachments at render pass creation time
    • a7a402c : v3dv: don't always skip tile buffer stores
    • 4d0e497 : v3dv: implement support for depth testing
    • c005a18 : v3dv: don't swap RB channels when copying images to buffers
    • 9a72a79 : v3dv: rename drm device fields so they are more explict
    • 356cf82 : v3dv: allocate winsys BOs properly
    • 1d92c31 : v3dv: implement device detection on actual hardware
    • 1ae844c : v3dv/pipeline: adding some nir-based linking
    • dc45d57 : v3d/cmd_buffer: emit flat_shade/noperspective/centroid flags
    • 6cac991 : v3dv: stubs for Create/DestroyPipelineCache
    • 6ab4c51 : v3dv: hook up WSI support
    • c75846e : v3dv: implement DRM modifier setup for WSI
    • 4825a76 : v3dv: implement vkGetImageSubresourceLayout
    • 54374f3 : v3dv: fix copy image to buffer
    • beee181 : v3dv: implement VK_KHR_external_memory{_fd,_dma_buf}
    • 0fd434d : v3dv: implement VK_KHR_external_memory_capabilities
    • 931b247 : v3dv/format: add R32G32B32A32_SFLOAT format
    • 0fd72b6 : v3dv: implement VK_KHR_get_physical_device_properties2
    • ecd0b1f : v3dv: include Vulkan version 1.1 as unsupported.
    • 5935ef0 : v3dv: implement vkGetPhysicalDeviceSparseImageFormatProperties
    • 85f1f06 : v3dv: provide default values for input attributes
    • aad44cc : v3dv: vertex input support
    • 8c4fcc2 : v3dv/pipeline: lower fs/vs inputs/outputs
    • 4587506 : v3dv: support queue submissions with multiple command buffers
    • 047ee91 : v3dv: implement fences
    • 56b4eac : v3dv: implement semaphore waits and signals on queue submissions
    • d24dbd8 : v3dv: implement vk{Create,Destroy}Semaphore
    • 0b08f83 : v3dv: implement vkDeviceWaitIdle
    • 4279145 : v3dv: prepack VCM_CACHE_SIZE
    • 5226928 : v3dv: partial prepack of the gl_shader_state_record
    • 0e49c62 : v3dv/cmd_buffer: emit CFG_BITS
    • b60e5ef : v3dv: handle VK_ATTACHMENT_UNUSED properly in more places
    • 04e1a6f : v3dv/cmd_buffer: rename render pass RCL emission helpers to be more explicit
    • a904b6f : v3dv: create a helper to start a new frame
    • d902e68 : v3dv: rewrite the attachment needs clearing condition
    • 8a5135e : v3dv: fix tile buffer loading
    • 22e431e : v3dv: merge subpasses into the same job when possible
    • 0b5df86 : v3dv: only clear attachments on the first subpass that uses them
    • 2271489 : v3dv: rewrite attachment state tracking
    • 6379a55 : v3dv: add a note on interactions between clearing and scissor
    • 9f3fd1a : v3dv: fix clipping against render area
    • 2baeb91 : v3dv: do not automatically emit a binner flush when finishing jobs
    • 7019700 : v3dv: implement vkCmdPipelineBarrier
    • 856d102 : v3dv: use the correct miplevel slice for the tile load operation
    • 3da6e6a : v3dv: implement vkCmdNextSubpass
    • 1c8226c : v3d/compiler: update uses_vid/uses_iid check
    • be1531a : v3dv/cmd_buffer: start jobs with CmdBeginRenderPass
    • 5259175 : v3dv: add the concept of a job
    • aef5a5c : v3dv: implement vkCmdCopyImageToBuffer
    • 94b4e18 : v3dv: rename and make compute_tile_size_for_framebuffer() public
    • 00cefce : v3dv: split framebuffer internal bpp calculations from tiling calculations
    • 2b02117 : v3dv: add v3dv_write_uniforms
    • a3b440c : v3dv/cmd_buffer: cache viewport translate/scale
    • e252e53 : v3dv: clif format dumping support
    • 15fb14c : v3dv/cmd_buffer: add shader source bos to cmd_buffer
    • 8431196 : v3dv/cmd_buffer: start to emit draw packets
    • 67aa802 : v3dv/cmd_buffer: emit shader_state packets
    • 746cfff : v3dv: flush at the end of each subpass
    • e163ea1 : v3dv: emit tile loads
    • 29b6e29 : v3dv/cmd_buffer: emit Viewport packets
    • 7bfe1db : v3dv/cmd_buffer: emit Scissor packets
    • d479f86 : v3dv/cmd_buffer: init command buffer dynamic state during pipeline bind
    • f752cdb : v3dv/pipeline: start to track dynamic state
    • 5285d83 : v3dv: CmdSetViewport and CmdSetScissor implementation
    • fca4dce : v3dv: initial stub for CmdBindPipeline
    • 6014562 : v3dv: initial CreateGraphicsPipeline/DestroyPipeline implementation
    • 6cb24a3 : v3dv/format: add v3dv_get_format_swizzle
    • 62ca997 : v3d/compiler: num_tex_used on v3d_key
    • 3f00638 : v3dv: setup color clear values at subpass setup time
    • 9e6bf4e : v3dv: be more conservative resetting command buffer state
    • cdc9660 : v3dv: implement vkQueueSubmit
    • 570ff24 : v3dv: clear set of BOs in the command buffer on reset
    • bd3c28b : v3dv: keep track of the numbre of BOs in a command buffer
    • bb72768 : v3dv: create the command buffer BO set before we init CLs
    • 6e4471c : v3dv: implement vkEndCommandBuffer
    • e6e80d3 : v3dv: emit the render command list
    • 61399b2 : v3dv: emit scissor to render area and precompute hw color clear values
    • 98697e1 : v3dv: precompute more tiling info at framebuffer creation time
    • 3b8aeb7 : v3dv: plug leak when destroying device
    • e6a7d88 : v3dv: make v3dv_layer_offset public
    • 14f72ee : v3dv: implement vkDestroyImage and vkDestroyImageView
    • 4075f8e : v3dv: revert the decision that the command buffer takes ownership of BOs
    • ec1d7e4 : v3dv: add the tile state and alloc BOs to the command buffer BO list
    • ae235e0 : v3dv: store base mip level in the image view
    • 52d9b8c : v3dv: add a few more API stubs
    • bab28ab : v3dv: make the command buffer own the command list BOs
    • 0020eb4 : v3dv: implement vkCmdBeginRenderPass
    • 51705fc : v3dv: compute tile size for framebuffer
    • cd24292 : v3dv: make v3dv_bo_alloc allocate memory for the bo struct
    • 291cd6e : v3dv: implement vkGetPhysicalDeviceImageFormatProperties
    • 3cf527a : v3dv: start handling command buffer status
    • 9bb6d6f : v3dv: implement vkBeginCommandBuffer
    • 9ac3261 : v3dv: add a concept of a command list
    • dc005f2 : v3dv: create a v3dv_bo struct and reference it from v3dv_device_memory
    • d9a7ba3 : v3dv: implement vk{Allocate,Free}CommandBuffers
    • 1d8d8c0 : v3dv: implement vkCreateCommandPool
    • e07b657 : v3dv: implement vk{Create,Destroy}Framebuffer
    • c95904f : v3dv: implement vk{Create,Destroy}RenderPass
    • cdba536 : v3dv: implement vkCreateRenderPass
    • 582ebb0 : v3dv: implement vkBindBufferMemory
    • 5ee1550 : v3dv: implement vkGetBufferMemoryRequirements
    • b478e8c : v3dv: implement vk{Create,Destroy}Buffer
    • dee9976 : v3dv: implement vkCreateImageView
    • 16ec12b : v3dv: implement vkBindImageMemory
    • e068314 : v3dv: implement vkUnmapMemory
    • cff9f08 : v3dv: implement vkMapMemory
    • d1d97ea : v3dv: implement vkFreeMemory
    • 85632e4 : v3dv: implement vkAllocateMemory
    • 33bb395 : v3dv: initialize mememory heaps in the physical device
    • 4c80097 : v3dv: implement vkGetImageMemoryRequirements
    • dc1fbeb : v3dv: Create/DestroyShaderModule implementation
    • a201c45 : v3dv: stubs for graphics pipeline methods
    • 46fa01f : v3dv/debug: add v3dv_debug
    • ccdc5a7 : v3dv/debug: plug v3d_debug
    • 0fcce6c : v3dv: implement vkCreateImage
    • d5dd157 : v3dv: add stub for vkDeviceWaitIdle
    • 7a045cf : v3dv: implement vkEnumerate{Instance,Device}LayerProperties
    • 871ac1e : v3dv: add some basic support for format properties
    • 63e23a9 : v3dv: add stubs for the format table and vkGetPhysicalDeviceFormatProperties
    • 9cc736f : v3dv: add dummy implementations for the packet definition generator
    • 8d9772f : v3dv: implement vkGetDeviceQueue
    • 41a5719 : v3dv: implement vkCreateDevice
    • 1c9f9d8 : v3dv: implement vkGetPhysicalDeviceQueueFamilyProperties
    • 37472a6 : v3dv: amend vkEnumerateInstanceExtensionProperties to handle layers
    • c2dd7bf : v3dv: implement vkEnumerateDeviceExtensionProperties
    • 10e8967 : v3dv: implement vkGetPhysicalDeviceFeatures
    • e75d462 : v3dv: implement vkGetPhysicalDeviceMemoryProperties
    • 0f0a014 : v3dv: add a comment to clarify how we should implement uuid / deviceID retrieval
    • 9de4ab1 : v3dv: retrieve device name from device info
    • 4e88e2d : v3dv: add support to use v3d simulator
    • 0e0f18a : v3dv: Implement vkGetPhysicalDeviceProperties
    • fc52dc8 : v3dv: pretend to initialize a physical device
    • 36e07a0 : v3dv: memory management stubs
    • fd81dc6 : v3dv: implement vkEnumeratePhysicalDevices
    • dbce41f : v3dv: implement vkDestroyInstance
    • be1575c : v3dv: implement vkCreateInstance
    • 20363b0 : v3dv: add support for valgrind macros
    • 2f383f9 : v3dv: add support for VK_EXT_debug_report
    • e4af07b : gitlab-ci: add broadcom vulkan driver
    • e5034f0 : v3dv: add v3d vulkan driver skeleton
    • bdb1e5a : glsl: Initialize ir_to_mesa_visitor members in constructor.
    • 6529cbd : svga: Remove unused printf argument.
    • 12c20cb : util/xmlconfig: Initialize xmlconfig member options in constructor.
    • 77c1b01 : zink: do not report SSBOs as halfway supported
    • 044d213 : scons: fix SPIR-V - NIR build
    • c122315 : aco: fix get_ssbo_size with a vgpr resource
    • a7114f3 : nir/opt_uniform_atomics: don't optimize atomics twice
    • bc43650 : nir/opt_uniform_atomics: optimize image atomics
    • bcf7a70 : aco: use nir_opt_uniform_atomics
    • f83bc5b : nir: add pass to optimize uniform atomics
    • 37b6b09 : nir: allow divergence information to be updated when inserting instruction
    • e1120f2 : nir: move divergence analysis options to nir_shader_compiler_options
    • bb5c0ba : aco: implement last_invocation
    • 1a912a5 : nir: add last_invocation intrinsic
    • 8850a63 : radv/aco,nir/lower_subgroups: don't lower elect
    • 36da9c4 : aco: implement elect
    • bf77f53 : aco: optimize more uniform reductions/scans
    • c31abab : nir: update ubo locations in nir_lower_uniforms_to_ubo
    • 47c3582 : glsl: fix up location setting for variables pointing to a UBO's base
    • ef0baba : zink: ARB_uniform_buffer_object is now implemented, so add cap and feature doc
    • 1b6a80a : zink: support loading any UBO
    • 67e4bcb : zink: support non-const offsets
    • 47bdda0 : zink: use nir_lower_ubo_vec4 to simplify things a bit
    • b9ca492 : aco: implement missing nir_op_unpack_half_2x16_split_{x,y}_flush_to_zero
    • 1070bba : android: fix SPIR-V - NIR build
    • 86ededf : docs: fix 20.2.0 relnotes
    • e57cd6d : docs: fix release calendar
    • b04f57c : docs: fix relnotes index
    • c02e933 : radv: add missing u_atomic.h include
    • 1fb3e1f : radv: Fix mipmap extent adjustment on GFX9+.
    • 85dfe61 : zink: enable VK_EXT_shader_stencil_export
    • 25d13b0 : zink: add VK_EXT_pipeline_creation_cache_control
    • bab0d86 : zink: support VK_EXT_extended_dynamic_state
    • 66fc9f5 : zink: support VK_EXT_blend_operation_advanced
    • fa2fd72 : zink: add VK_EXT_custom_border_color
    • 9c43ccd : zink: enable VK_KHR_vulkan_memory_model extension
    • 66d7bb0 : radv: fix adjusting vertex alpha
    • 207138b : docs: Add sh256 sums for 20.2.0
    • ddad8d9 : docs: add release notes for 20.2.0
    • 7d72110 : docs/release-calender: Add 20.2 stable releases
    • 63f61b0 : docs: update calendar and link releases notes for 20.2
    • b32a8f8 : radv: move lower_io_arrays_to_elements before lower_io_to_scalar_early
    • b0829c6 : radv: replace RADV_ALPHA_ADJUST by AC_FETCH_FORMAT
    • 5000c34 : ac/llvm: move AC_FETCH_FORMAT to non-LLVM code
    • ab067ad : radeonsi/gfx10: fix stream index for multi-stream overflow query
    • cf3b638 : radv: remove RDR2 discard workaround
    • 037d9fb : spirv: replace discard with demote for incorrect HLSL-SPIR-V translations
    • 18f9fc9 : spirv: add and use a generator id enum
    • 875ff84 : radv/winsys: Expand scope of allbos lock.
    • ea77869 : radv: Fix event write cmdbuffer allocation when tracing.
    • 98f538d : radv: remove one leftover TODO in the shader info pass
    • cec12d4 : radv/llvm: reduce LDS size for tess by using NIR IO assigned locations
    • 47e26bf : radv/llvm: reduce the ESGS itemsize by using NIR IO assigned locations
    • 569b894 : radv/llvm: switch to NIR IO assigned locations
    • 6387341 : ac/nir: pass the variable location to store_tcs_outputs
    • 8f8ee5b : ac,radv,radeonsi: stop multiplying driver_location by 4
    • 0a90dab : radv/llvm: stop assigning driver_location in NIR-LLVM
    • 181d5f5 : nir: Allow more deref modes in phis
    • 002a23e : android: util: Move xxd.py to util
    • 7346933 : vc4: enable lower_isign for VC4
    • b7d16ad : pan/bi: Range check newc/oldc when rewriting
    • c9b2dfd : pan/bi: Disable mediump output lowering
    • b11cb41 : panfrost: Temporarily disable FP16 on Bifrost
    • 0bfc7cb : panfrost: Set "shader modifies coverage?" flag
    • 3f919a7 : panfrost: Update XML for Bifrost early-z/FPK
    • 8cf798d : pan/bi: Implement FETCH
    • 7f54284 : pan/bi: Add bi_emit_lod_cube helper
    • 125d18d : pan/bi: Map NIR tex ops to Bifrost ops
    • f42eb33 : panfrost: Fix faults on block-based formats on Bifrost
    • 5363719 : pan/bi: Use new block dimension helper
    • 8bb1d61 : panfrost: Add panfrost_block_dim helper
    • aa6a5d2 : pan/bi: Don't terminate helper threads
    • 5fdd0c6 : pan/bi: Fix message type printing
    • 3ecffd9 : panfrost: Set helper_invocation_enable for Bifrost
    • 7fdb090 : pan/bi: Implement txb
    • 5fa5c01 : panfrost: Add some missing Bifrost texture XML
    • 239e4e8 : panfrost: Fix Bifrost high LOD clamp
    • 3943bce : panfrost: Fix Bifrost filter selection
    • 401d594 : panfrost: Add missing XML for Bifrost samplers
    • ea85e3a : pan/bi: Prefer 'texture_index' to 'image_index'
    • 3089501 : pan/bi: Rewrite to fit dest = src constraint
    • 9384023 : pan/bi: Pack TEXC
    • 4836366 : pan/bi: Handle nir_tex_src_lod
    • f815131 : pan/bi: Add data register passing infrastructure
    • 3bf4e60 : pan/bi: Stub out TEXC handling
    • dcce3fe : pan/bi: Add texture operator descriptor
    • d99df25 : pan/bi: Pack skip bit for texture operations
    • 39ec3eb : pan/bi: Encode skip bit into IR
    • 5cf53d1 : pan/bi: Streamline TEXC/TEXS naming/selection
    • 6ed1bdf : pan/bi: Use canonical texture op names in IR
    • 93f9052 : pan/bi: Fix simple txl test
    • c3796c9 : pan/bi: Expose GL 2.1 on Bifrost
    • a204eac : pan/bi: Handle vector moves
    • a227798 : pan/bi: Pass flow_control through directly
    • 0f181f4 : pan/bi: Use canonical flow control enum
    • d232864 : pan/bi: Use canonical term dependency
    • 2b9484c : pan/bi: Use canonical term "message type"
    • 800ee3d : pan/bi: Print message types as strings
    • 77a4e39 : pan/bi: Add missing message types
    • d2fac19 : pan/bi: Expand clause type to 5-bit
    • 785344e : pan/bi: Use canonical name for staging registers
    • 4131bc3 : pan/bi: Use canonical next_clause_prefetch
    • 6c1cabc : pan/bi: Canonicalize terminate_discarded_threads
    • a597588 : pan/bi: Use canonical floating-point modes
    • c8b9a05 : pan/bi: Cull unnecessary edges on the CF graph
    • 026a295 : pan/bi: Drop if 0'd combine lowering
    • 405544e : pan/bi: Fix memory corruption in scheduler
    • 7c351a6 : pan/decode: Ensure mappings are zeroed
    • 3ed7472 : pan/bi: Add copy for register COMBINEs
    • 556bb17 : ci: piglit: conditionally build OpenCL tests
    • d5e5f72 : vc4: Enable lower_umax and lower_umin
    • e7127b3 : nir/algebraic: optimize iand/ior of (n)eq zero when umax/umin not available
    • 210db65 : panfrost: Add a debug flag to disable AFBC
    • 205f1d7 : radeonsi: disable SDMA on gfx6-7 and gfx10.3 to decrease CPU overhead
    • 7a01629 : radeonsi: update the DMA perf test
    • 4182fbc : winsys/amdgpu: apply the VM alignment optimization to the physical alignment too
    • 99446b8 : winsys/amdgpu: rework the VM alignment optimizations
    • a4e4644 : ac/surface: fix valgrind warnings in DCC retile tile lookups
    • b0df97b : CI: build our own spirv tools
    • d166188 : ci: fix deqp clone + fetch
    • 5de5693 : disk_cache: build option for disabled-by-default
    • 8e98145 : radv: use radv_optimize_nir() less in radv_link_shaders()
    • 55254f2 : radv: move optimizations in shader_compile_to_nir() to after io_to_scalar
    • 5f2671b : nir: return progress from nir_lower_io_to_scalar_early
    • fd4d0b4 : panfrost: Move the blend shader cache at the context level
    • a5005c3 : panfrost: Get rid of the constant patching done on blend shader binaries
    • c973994 : panfrost: Let compile_blend_shader() allocate the blend shader object
    • dbc33e8 : panfrost: Don't leak NIR blend shaders
    • 8a5b885 : panfrost: Allocate blit_blend with ralloc()
    • 0a74a04 : panfrost: Pass compile arguments through a struct
    • 78ec522 : panfrost: Move the blend constant mask extraction out of make_fixed_blend_mode()
    • 4441e80 : panfrost: Constify the rt_fmts arg passed to pan_lower_framebuffer()
    • da132d8 : radv: Set fce metadata correctly on DCC initialization.
    • 5ae3656 : aco/ngg: Calculate workgroup size of NGG shaders.
    • 61280bb : aco/ngg: Allocate NGG GS space early for const vertex/primitive counts.
    • e8a0409 : aco/ngg: Use more efficient LDS layout to help reduce bank conflicts.
    • 9bf92d4 : radv/aco: Enable NGG GS by default.
    • dd73719 : aco/ngg: Add shader query support to NGG GS.
    • df62c8f : aco/ngg: Place workgroup barrier outside control flow for NGG GS.
    • 1129575 : aco/ngg: Implement NGG GS output.
    • 62b5012 : aco/ngg: Implement workgroup reduce / exclusive scan for NGG GS.
    • c29e288 : aco/ngg: Create LDS layout for NGG GS.
    • 2680329 : aco/ngg: Setup NGG GS.
    • 9c3d840 : aco/ngg: Allow NGG GS to create VS exports.
    • b67878f : aco/ngg: Allow NGG GS to load per-vertex GS inputs.
    • 8f25d9f : aco/ngg: Allow NGG GS to store ES outputs.
    • b57b1a0 : aco/ngg: Clean up and reorganize NGG VS/TES code.
    • 3645a31 : aco/ngg: Make primitive export packing less prone to error.
    • 0bfe049 : aco/ngg: Refactor ngg_emit_prim_export in preparation for NGG GS.
    • b08ced0 : aco/ngg: Refactor gs_alloc_req in preparation for NGG GS.
    • ecfabfd : aco: Add wave-specific opcode for s_lshl and s_flbit.
    • 57d8799 : aco: Optimize thread_id_in_threadgroup when there is just one wave.
    • 5e31fb4 : aco: Use thread_id_in_threadgroup helper for ES outputs.
    • 924f816 : aco: Extract thread_id_in_threadgroup to a separate function.
    • b1964ad : aco: Extract lanecount_to_mask to a separate function.
    • 0b8e7be : aco: Clarify missing export error message in assembler.
    • c1d11bb : aco: Add loop creation helpers.
    • 17ad2ad : radv/aco: Use new GS lowering options for ACO with NGG GS.
    • f11f4a2 : nir: Add ability to count primitives per stream.
    • aac5adc : nir: Count vertices per stream.
    • 70b94ad : nir: Add ability to overwrite incomplete GS primitives.
    • c977c36 : nir: Add ability to count emitted GS vertices per primitive.
    • 2be9901 : nir: Add ability to count emitted GS primitives.
    • 73dd86c : radv,anv: use CLOCK_MONOTONIC_FAST when CLOCK_MONOTONIC_RAW is undefined
    • fe894dc : util/os_misc: os_get_available_system_memory() for FreeBSD
    • 9aa5c7c : radv: use the same NIR compiler options for both compiler backends
    • 63049b0 : radv/llvm: do not lower sub
    • abce271 : lima/parser: Fix varyings decoding in RSW
    • 6db3df5 : aco: dump the program if the disassembler failed
    • a7d3be7 : clover: don't call validate spirv at all if not needed
    • 91ace07 : util/disk_cache: remove unused function param
    • 5931d78 : swr/rasterizer: Remove BuilderGfxMem member mpTrackMemAccessFuncTy.
    • 4b51313 : meson: Use more portable compiler option -std.
    • 3a1f22c : turnip: Add support for GetSwapchainGrallocUsage2ANDROID().
    • 5a595cd : turnip: Detect Qualcomm gralloc and its UBWC flag on gralloc surfaces.
    • 9a14e74 : turnip/kgsl: Add support for importing dma-bufs.
    • b732e4f : turnip/kgsl: Fix last minute breakage of the build.
    • 290f3fe : Revert "anv: Add driconf option to disable compression for 16bpp format"
    • cce6fc3 : anv: Enable multi-layer aux-map init for HIZ+CCS
    • 624a2aa : freedreno/ir3: Don't leave holes the UBO upload plan.
    • ddf468f : freedreno/ir3: Clean up the UBO upload plan setup.
    • 330c68d : ci/softpipe: Add another flaky GS test to the skips list.
    • 0b96b7b : util/hash_table: add function for reserving size in a hash table
    • 06a5edf : nir/opt_deref: Fix the vector bitcast optimization
    • 54eae33 : clover: Stop leaking NIR shaders
    • cc00d57 : util/xmlconfig: eliminate memory leak
    • 1138984 : radv/llvm: remove dead code for 64-bit GS inputs
    • a41bed2 : radv/llvm: do not lower nir_op_fsat
    • 31a0574 : ac/nir: implement nir_op_fsat
    • 0747f21 : ci: Disable pm_runtime and max clocks in LAVA jobs
    • 7cae9f0 : ci: Test Panfrost on Khadas VIM3 boards
    • b74cfa7 : ci: Update kernel for LAVA
    • 9814643 : freedreno/a6xx: Support PIPE_FORMAT_R8_G8B8_420_UNORM for texturing
    • 11563da : freedreno/a6xx: Generalize pointers in struct fd6_pipe_sampler_view
    • 826a102 : st/mesa: Add NV12 lowering to PIPE_FORMAT_R8_G8B8_420_UNORM
    • e23bcb6 : util/formats: Add PIPE_FORMAT_R8_G8B8_420_UNORM
    • 3b78437 : iris: Use the data cache for indirect UBO pulls
    • b54d37a : anv: Use the data cache for indirect UBO pulls on Gen8+
    • 89f3d11 : anv: Plumb the device into *bits_for_access_flags
    • 3a33560 : anv: Use format_for_descriptor_type for descriptor buffers
    • d2185f0 : anv: Add a device parameter to format_for_descriptor_type
    • 3d22de0 : intel/fs: Add an option to use dataport messages for UBOs
    • 0d462db : intel/fs: Add an alignment to VARYING_PULL_CONSTANT_LOAD_LOGICAL
    • 459f68a : i965: Take an isl_format in emit_buffer_surface_state
    • caea5a6 : intel/dev: fix 32bit build issue
    • dd9c34a : intel/nir: Lower load_global_constant in lower_mem_access_bit_sizes
    • fd04f85 : intel/nir: Don't try to emit vector load_scratch instructions
    • 0a172dc : ci: enable piglit testing of clover/llvmpipe.
    • 0cb1775 : CI: Don't run pixmark-piano twice on radeonsi
    • 2fa7c79 : spirv: Move nir_lower_libclc to src/compiler/spirv
    • 43390a5 : clover: Use core libclc loader
    • ef453f5 : spirv: Add a shared libclc loader
    • 5ffdb10 : meson: generalize libclc usage
    • ed7e962 : clover/meson: use dep.get_variable instead of deprecated get_pkgconfig_variable
    • d3dcb11 : util/xxd.py: Add an option for binary files
    • 82bbf4c : util: Make xxd.py output char array instead of string
    • 22ffc05 : util: Move xxd.py to util
    • 3ff513e : glsl/xxd.py: fix imports
    • 449f1fe : gallium/pipe: Add a GALLIUM_PIPE_SEARCH_DIR override env var
    • 1d3600c : iris: Fill out compute caps and enable clover support
    • 9df9f94 : iris: Add support for load_work_dim as a system value
    • 67ee9c5 : iris: Handle runtime-specified local memory size
    • d9adb55 : iris: Add pipe-loader support
    • 8de380d : broadcom/compiler: add V3D_DEBUG_RA option
    • bcb8dd7 : broadcom/common: increase V3D_MAX_TEXTURE_SAMPLERS, add specific OpenGL limit
    • 5f7810d : aco/isel: Fix out-of-bounds write in visit_load_input
    • d38094b : radv: Respect alignment requirements in descriptor set layouts
    • 3b55913 : radv: Clean up CreateDescriptorSetLayout
    • 76add35 : radv: Fix unaligned memory access when writing specialization map entries
    • 984dcfc : radv: Avoid calling memcpy with null pointers
    • 0ef2f1d : nir: Fix unaligned pointer access
    • 6a9dc75 : nir: Fix undefined behavior due to signed integer multiplication overflows
    • 137d7d2 : ci: move to using clang 10 for meson + clover
    • ae7189d : st/mesa: pass inlinable uniforms to drivers if they requested it
    • 593517a : gallium: add pipe_context::set_inlinable_constants
    • 3f1b35a : nir: add new helper passes that lower uniforms to literals
    • 10a7682 : util: add _mesa_set_create_u32_keys where keys are not pointers
    • 7f2ea2a : panfrost: Fix fixed-function blend on Mali v6
    • 8389976 : panfrost: XML-ify the blend descriptors
    • ca392e8 : panfrost: Rework the render target layout to use overlapping structs
    • 713419e : panfrost: Rework fixed-function blending
    • f204404 : panfrost: gen_pack: Add a no-direct-packing attribute
    • 4205c95 : panfrost: gen_pack: Support overlapping structs
    • 3d09e26 : panfrost: gen_pack: Drop support for opaque structs
    • 01121c7 : panfrost: Drop the with_opaque specifier on midgard blend desc
    • 7bb85ea : panfrost: Get rid of the with_opaque qualifier on the renderer state desc
    • f734e67 : panfrost: Adjust the renderer state definition
    • 51331d6 : panfrost: Adjust the primitive desc definition
    • d343f23 : panfrost: Adjust the draw descriptor definition
    • 1e7d82c : nir/algebraic: always lower idiv to shifts if bitops are allowed
    • 315df8d : glthread: handle glInterleavedArrays
    • bd70b61 : mesa: remove api_loopback to remove call indirections
    • 4bd94bd : mesa: don't use GET_DISPATCH because it doesn't work with glthread
    • fd3209a : clover: move tokenize function to algorithm
    • ee5b46f : clover/spirv: support CL_KERNEL_COMPILE_WORK_GROUP_SIZE
    • 2bb2ba7 : clover/spirv: parse arg_info
    • 4f99826 : clover: avoid adding an extra space to compiler options
    • 811b17a : clover: implements notification callback on program builds
    • c04d5e7 : clover: implements clGetKernelWorkGroupInfo CL_KERNEL_COMPILE_WORK_GROUP_SIZE
    • aadd134 : clover: add CL_KERNEL_ATTRIBUTES for clGetKernelInfo
    • ef0f8ec : clover: bind sampler_t type to module::argument::sampler
    • 9aea6e3 : clover: implements clGetKernelArgInfo
    • d7d66e3 : clover: implements clEnqueueFillImage
    • c0f03f6 : clover: implements clEnqueueMigrateMemObjects
    • 4b96eb0 : etnaviv: Make sure to track different pipe_screens for different DRM device descriptions
    • ac16110 : panfrost: Fix a warning
    • ccb3d7d : panfrost: bifrost: disassemble: Fix decoding of next_regs
    • 19561f3 : radv: remove trailing whitespace
    • 8096e0d : aco: remove trailing whitespace
    • 3c5eb1f : aco: more uses of nir_get_io_offset_src()
    • 1211d05 : aco: bail out if the NIR IO base offset isn't zero
    • 9d5ec7f : etnaviv: stop leaking the dummy texure descriptor BO
    • 6a44bda : intel/uuid: use git-sha1/package for the driver UUID
    • 456fa9b : iris: plumb device/driver UUID generators
    • e9a766a : intel: split driver/device UUID generators
    • 713386a : turnip: Enable support for KHR_incremental_present
    • 8b20540 : omx/tizonia: fix build
    • f5f0c01 : gallium/util: remove empty file u_half.h
    • 7169795 : gallium/util: remove redundant util_float_to_half_rtz
    • b42c6ff : util: remove util_float_to_half and util_half_to_float wrappers
    • 24a62dd : util: move util_half_to_float code into _mesa_half_to_float_slow
    • ffcdf76 : util: implement F16C using inline assembly on x86_64
    • 4c54f05 : nir/constant_folding: init nir_const_value to zero
    • 335132c : lavapipe: drop dri,dricommon deps.
    • 9c25689 : intel: drop likely/unlikely around INTEL_DEBUG
    • e06da55 : anv: drop likely/unlikely around INTEL_DEBUG
    • 4b9e38c : i965: drop likely/unlikely around INTEL_DEBUG
    • e6d26fb : iris: drop likely/unlikely around INTEL_DEBUG
    • 4015e18 : intel: add INTEL_DEBUG expected value in declaration
    • bcfec61 : anv: Add driconf option to disable compression for 16bpp format
    • 4959320 : anv: Factor out dri option initialization code in separate function
    • 7c6f1d3 : clover/nir: extract constant buffer into its own section
    • adbfff6 : clover: bind constant buffer if one is provided
    • af85985 : clover/util: add id_type_equals to support symbols with multiple sections
    • 7fb4ab9 : driconf: Restore the ability to override driconf with the environment.
    • 455bfec : driconf: Make sure that the range check on the defaults actually works.
    • 24b4a63 : CI: Re-enable Panfrost T7xx CI
    • 3dc00c3 : gallium/u_threaded_context: fix use-after-free in transfer_unmap
    • 706fef7 : gallium/u_threaded_context: always flush asynchronously if requested
    • 28256d3 : gallium/u_threaded_context: don't call memcpy in tc_set_constant_buffer
    • ed3c5fe : radeonsi: implement GL_INTEL_blackhole_render
    • bad7b38 : radeonsi: remove dead variable postponed_kill
    • d5575ca : gallium/util: add set_frontend_noop into driver_noop and u_threaded_context
    • 72fa337 : util/idalloc: resize if ID is too large for reservation
    • 155b1b1 : st/mesa: enable GL name reuse for queries based on the driconf option
    • 6f7446b : driconf: enable force_glsl_extensions_warn for viewperf
    • a8b4e69 : driconf: force the vendor string to NVIDIA to fix viewperf energy tests
    • 5827fe8 : gl_marshal.py: inline print_sync_dispatch
    • bcd6b0c : mesa: factor out layout parsing for glInterleavedArrays
    • 60825a5 : nir/constant_folding: Fold load_deref of nir_var_mem_constant
    • 481b753 : nir: Validate constant initializers
    • 1ada835 : nir/constant_folding: Use nir_shader_instruction_pass
    • 9df1ff3 : nir/constant_folding: Use the builder
    • 3b11b37 : radv: remove VkPipelineCacheCreateInfo::flags assert
    • 3f3a5f3 : iris: Fix a fast-clear skipping optimization
    • c416adf : anv: remove VkPipelineCacheCreateInfo::flags assert
    • 9ad4b8b : intel/dev: add a small non installable tool to print device info
    • 61c095f : zink: enable pipe caps for ARB_timer_query
    • bb60335 : zink: add pipe_context::get_timestamp hook
    • b551ee2 : zink: handle TIME_ELAPSED queries
    • 7a40b73 : zink: handle timestamp queries
    • 79f3544 : intel/perf: fix crash when no perf queries are supported
    • 82563b6 : util: drop non-posix header fnmatch
    • f0707e9 : CI: Disable Panfrost T7xx CI
    • 2b977ad : xmlconfig: fix scandir_filter
    • a0c13c9 : spirv: Make the clc_shader const
    • e33f9db : turnip/kgsl: Add strerror decode in BO init failure.
    • 5d3aeaf : turnip: Report device loss through _mesa_loge() instead of fprintf.
    • 50f25da : turnip: Always enable TU_DEBUG=startup on debug drivers.
    • a4d9a9d : turnip: Extend the coverage of TU_DEBUG=startup.
    • 80869f0 : turnip: Mark the vk_errorf helper as bring printflike.
    • 01de452 : turnip: Use mesa's normal PRINTFLIKE macro instead of our own.
    • a7bc2f8 : turnip: Don't expose VK_ANDROID_native_buffer on non-Android.
    • d82826a : anv: Implement VK_EXT_transform_feedback on Gen7
    • 9831888 : android: panfrost: use python3 for generated sources rules
    • 2792117 : android: pan/bi: Use new disassembler (v2)
    • 8163c81 : turnip: implement timestamp fences/semaphores for kgsl backend
    • 462d623 : broadcom/simulator: update to a newer simulator
    • e3b814d : zink: hook up driconf
    • f984654 : ac/nir,radv: fix invalid IR when loading inline uniform blocks
    • 9badadf : radv: remove dead deref code in the shader info pass
    • bc099b8 : radv: remove unused gs.writes_memory in the shader info pass
    • 9896337 : ac,radv,radeonsi: remove unused parameters in the shader ABI IO
    • b7f9940 : radeonsi: remove dead code in TCS/TES/GS since const_index is always 0
    • e3296e0 : ac/nir: remove dead shader IO code
    • 456f63e : ac/nir: remove dead global load/store/atomic derefs code
    • 6b0695c : radv/llvm: enable lower_unpack_half_2x16
    • f11d4f3 : ac/llvm: implement nir_op_unpack_half_2x16_split_{x,y}
    • cdf6d93 : radv/llvm: lower VS IO
    • b340299 : radv/llvm: gather VS input usage mask from load_input
    • 1c4a213 : radv/llvm: lower TCS IO
    • 80cefee : radv/llvm: gather TCS outputs from the output variables
    • 9615273 : radv/llvm: lower TES IO
    • 6e33941 : radv/llvm: lower GS IO
    • ed55bc4 : radv/llvm: assign driver locations for VS, TCS, TES and GS correctly
    • 1ea48c1 : ac/nir: remove dead load/store deref code for temporary variables
    • 7a8f5ea : ac/llvm: adjust dmask when image stores are shrinked using the format
    • 961a8d7 : vc4: Avoid negative scissor caused by no intersection
    • 9a899e5 : v3d: Update the TEXTURE_SHADER_STATE when there’s a new buf for a tex
    • e2f037e : v3d: Split the creating of TEXTURE_SHADER_STATE into a helper function
    • d25504d : v3d: Make the function to set tex dirty state for a stage global
    • ef918f0 : turnip: remove pre-emption marker
    • 3b3a3af : glsl: Initialize ast_node member field location.path in constructor.
    • 81cd4c8 : intel/vec4: Remove leftover code from Gen8+ removal.
    • 0a7bd14 : freedreno: Move rsc NULL check to before rsc dereferences.
    • 1b0fec4 : iris: Fix the constant data address calculation
    • 23ff2a0 : nv50/ir: Initialize Source members.
    • 7e86267 : radv: Fix asserts using assign instead of compare.
    • 9660427 : lima: Print usage if --help is any of the arguments.
    • 2fb2749 : android: Disable the user XML config parsing.
    • 7edb2ae : util/xmlconfig: Drop use of XML_Char in parsing.
    • 6f3352b : driconf: Stop quoting true/false in boolean option definitions.
    • 8a05d6f : driconf: Make the driver's declarations be structs instead of XML.
    • 9fbf206 : driconf: Use DRI_CONF_OPT_I for remaining int options
    • 23c3eb1 : driconf: Delete disjoint range support.
    • 4f37161 : util/xmlconfig: Indent to Mesa style.
    • 91ccbb3 : util/xmlconfig: Drop silly open-coded strdup.
    • df3d3ea : driconf: Make a DRI_CONF_OPT_S() for string options.
    • 4a57384 : r200: Reuse DRI_CONF_OPT_F for texture_blend_quality.
    • c47c08f : driconf: Fix extra quoting on "Jimenez'".
    • b1e730b : driconf: Eliminate the DRI_CONF_OPT_BEGIN_B macro.
    • 8427e56 : intel/fs: Don't use NoDDClk/NoDDClr for split SHUFFLEs
    • b23541c : android: util: add log.c to Makefile.sources
    • f9977f8 : panfrost: decode: Flag pandecode_log_typed() as PRINTFLIKE
    • 12b00b7 : panfrost: decode: Fix decode_bifrost_constant() prototype
    • 7568c97 : radv: Use atomics to read query results.
    • 7cc55df : pan/bi: Fix assert when writing vertex outputs
    • 447cef4 : zink: use sizeof(vec4) multiplier for nir_lower_uniforms_to_ubo
    • 1c8e80a : zink: correctly set up ubo bindings and buffer indices
    • 59027a2 : zink: don't emit ubos or bindings for ubo variables
    • d369c00 : zink: emit ubo variables sized based on the overall ubo block size
    • 76ac341 : zink: always emit descriptor set 0 in ntv
    • 9a85392 : zink: fix shader buffer size caps to use 65536
    • 0e3e323 : zink: run nir_lower_uniforms_to_ubo conditionally
    • 7767c3d : zink: correctly handle ARB_arrays_of_arrays in ntv for samplers
    • 18fd627 : aco/tests: add disassembler tests to reproduce the add3+clamp crash
    • 01704dd : aco: apply the clamped integer addition disassembly workaround for v_add3
    • c9192d1 : pan/mdg: map uabs_i/usub to i/uabsdiff
    • cea032a : pan/mdg: remove unused arg from ALU_CHECK_CMP and ALU_CASE_CMP
    • 37c1b9c : aco: fix get_buffer_resource_flags()
    • b2e1fc8 : nir: Add a pass to lower vec3s to vec4s
    • 9190f82 : nir/opt_deref: Add an optimization for bitcasts
    • 80e6ac3 : nir/opt_deref: Add an instruction type switch
    • 769ede2 : nir: Add component mask re-interpret helpers
    • 57e7c5f : nir/opt_load_store_vectorize: Use bit sizes when checking mask compatibility
    • f6667cb : nir: Add a memcpy optimization pass
    • e363da3 : nir: Handle memcpy in copy_prop_vars and combine_stores
    • 100a5ac : nir/find_array_copies: Properly discard copies for casts
    • 1c49299 : gallium/util: allow scissored blits for stencil-fallback
    • e7e0468 : gallium/util: fix memory-leak
    • 03bea54 : meson: fix power8 option
    • 57b4f60 : turnip: don't initialize GRAS_LRZ_CNTL/RB_LRZ_CNTL tu6_init_hw()
    • 3c07a14 : turnip: enable LRZ
    • 1d83f5a : turnip: disable LRZ on vkCmdClearattachments() 3D fallback path
    • 2f79e00 : turnip: disable LRZ on vkCmdClearAttachments()
    • 27743b0 : turnip: emit correct LRZ fast clear setup
    • 0ca87ed : turnip: add support to clear LRZ
    • 0b2cfd0 : turnip: add LRZ valid tracking for secondary command buffers
    • 517b26b : turnip: add LRZ tracking to command buffer state
    • fdad1ca : turnip: disable LRZ depending on fragment changes
    • d1fa40b : turnip: disable LRZ writes when blend is enabled
    • 38f008e : turnip: disable LRZ on specific cases
    • 6089b00 : turnip: create LRZ buffer
    • 138d292 : turnip: add environment variable to disable LRZ
    • e3c7748 : ci/bare-metal: Move the "POWER_GOOD not seen in time" check to the right time.
    • 98bb74b : nir: Fix a misspelling
    • 038fcbc : glsl: don't duplicate state vars as uniforms in the NIR linker
    • 535fd6d : freedreno/cffdec: fix decoding of bindless descriptors
    • a8ac61b : intel/fs: NoMask initialize the address register for shuffles
    • 5c2129d : virgl: fix stride + layer_stride inconsistency
    • 87383e3 : virgl: query blob mem
    • 3b54e58 : virgl: support PIPE_CAP_BUFFER_MAP_PERSISTENT_COHERENT
    • cd31f46 : virgl/drm: add resource create blob function
    • e01ec6e : virgl/drm: query for resource blob and host visible memory region
    • 7b7f210 : drm-uapi: virtgpu_drm.h: resource create blob + host visible memory region
    • c73c0cc : virgl: add flags to (*resource_create) callback
    • 1aac47d : Revert F16C series (MR 6774)
    • 4a0164e : android: gallium/virgl: cleanup virgl_driinfo.h gen rules
    • d7fbf94 : android: gallium/radeonsi: cleanup si_driinfo.h gen rules
    • a648aea : android: gallium/iris: cleanup iris_driinfo.h gen rules
    • cb95065 : nir: Add lowering from regular ALU conversions to the intrinsic
    • bc7ed03 : clover/nir: Call nir_lower_convert_alu_types
    • 7d97f3d : spirv: Implement vload[a]_half[n] and vstore[a]_half[n][_r]
    • a85afb7 : spirv/opencl: Drop dest_type from handle_v_load_store
    • 8610af1 : spirv: Handle all OpenCL conversion ops with full rounding
    • 8e84582 : spirv: Add some conversion handling helpers
    • 383ecfb : nir: Add a passes for nir_intrinsic_convert_alu_types
    • d5cb51e : nir: Add builder helpers for OpenCL type conversions
    • 588bb66 : nir: Add a conversion and rounding intrinsic
    • 0aa08ae : nir: Split NIR_INTRINSIC_TYPE into separate src/dest indices
    • 4c70f1b : gallivm/nir: fix non-32 bit find lsb/msb
    • e8f1cc4 : llvmpipe/cs: add in shader shared size.
    • 35b162e : gallivm/nir: make sure to mask global reads.
    • 545d852 : intel/gen9: Enable MSC RAW Hazard Avoidance
    • 237f4d9 : radeonsi: restructure si_pipe_set_constant_buffer
    • d5cb7bd : radeonsi: call nir_lower_bool_to_int32 last because it breaks nir_opt_if
    • fd6bbdc : radeonsi: use staging buffer uploads for most VRAM buffers
    • 701f7ae : radeonsi: move si_set_active_descriptors_for_shader into si_update_common_shader_state
    • f5912c6 : radeonsi: kill disabled clip distances and planes at per-channel granularity
    • 30c3b2c : radeonsi: simplify NGG culling enablement and add radeonsi_shader_culling option
    • b02bef0 : intel/blorp: Conditionally clear full surface depth and stencil
    • d5849bc : anv: Skip HiZ and CCS ambiguates which preceed fast-clears
    • e9d5ec3 : anv: Use more temp vars in cmd_buffer_begin_subpass
    • 7c92e41 : anv: Allow HiZ clears for multi-view
    • 03af98a : radeonsi: support for external buffers (ext_external_objects)
    • df63491 : radv/aco: lower IO for all stages outside of ACO
    • 2c32251 : radv: gather output usage mask from store_output for VS, TES and GS
    • 79368ab : ttn: Fix number of components for IF/UIF
    • b00a023 : ac/nir: fix nir_intrinsic_shared_atomic_fadd
    • 8227b08 : ac/llvm: fix invalid use of unreachable in ac_build_atomic_rmw()
    • 892e74d : radv: fix gathering writes_memory for global store/atomic operations
    • e94fd4c : lavapipe: rename vallium to lavapipe
    • 5e8791a : radeonsi: Fix dead lock with aux_context_lock in si_screen_clear_buffer.
    • 90e42f8 : add one last 20.1 release to coincide with expected 20.2.1
    • fe16e40 : docs: update calendar and link releases notes for 20.1.9
    • 00d87db : docs: add release notes for 20.1.9
    • 49ec863 : freedreno/ir3: Enable the i/o vectorizer on UBOs.
    • e3f4655 : nir: Make nir_lower_ubo_vec4() handle non-vec4-aligned loads.
    • 618556a : nir: Drop the high_offset argument to the load_store_vectorizer filter.
    • 5f757bb : nir: Make the load_store_vectorizer provide align_mul + align_offset.
    • 9c5a793 : nir/gl_nir_lower_buffers: Set up align_mul/offset on UBOs.
    • ffbfc1e : nir/nir_lower_uniforms_to_ubo: Set better alignments on our new instructions.
    • c88c89f : nir: Print the alignment information on casts.
    • 6c1c571 : nir: Document a bit about how align_mul/offset work.
    • bd60e31 : freedreno/ir3: Make sure we run the opt loop after lowering UBOs to vec4.
    • 8dc8922 : turnip: implement legacy API functions separately
    • 25ebd7f : Revert "nir/lower_goto_if: Add a route::outside set"
    • 57c9fc3 : nir/lower_goto_ifs: Always include level dom_frontiers in prev_frontier
    • 7749983 : nir/lower_goto_ifs: Add asserts for SSA forks
    • dc010cb : nir/lower_goto_ifs: Use rzalloc
    • fa3c38c : spirv: Only run repair_ssa if structured
    • 719c680 : nir/dominance: Use _mesa_set_clear instead ofhand-rolling it
    • b6a4172 : nir/lower_goto_ifs: Don't destroy SSA form in the process
    • 6f134a6 : nir/validate: Improve the validation of blocks
    • 4fb2edd : gallium/util: remove empty file u_half.h
    • 7a1deb1 : gallium/util: remove redundant util_float_to_half_rtz
    • 2b6a172 : util: remove util_float_to_half and util_half_to_float wrappers
    • 5af8139 : util: move util_half_to_float code into _mesa_half_to_float_slow
    • 87900af : util: implement f16c - fast half-float conversions
    • 7dbb1f7 : nir/cf: Better handle intra-block splits
    • 5e2e882 : nir: Disallow goto and goto_if in clone and [de]serialize
    • 9a48ed8 : nir/copy_propagate: Copy-prop into jump conditions
    • 7f0cd6f : nir/opt_if: Use early returns in opt_if_merge()
    • 656e428 : nir/opt_if: Remove open-coded nir_ssa_def_rewrite_uses()
    • c6f871b : nir/lower_returns: Use nir control flow insertion helpers
    • f103bde : ttn: Use nir control flow insertion helpers
    • f504eb6 : radv: Use nir control flow insertion helpers
    • b2ede62 : intel/nir: Use nir control flow helpers
    • 55e2b34 : radeonsi: Fix imports with displayable DCC.
    • 636f770 : bin/gen_release_notes.py: escape special rST characters
    • ae7975e : docs: cpu - CPU
    • 5fe6124 : docs: Sandybridge - Sandy Bridge
    • 302fc31 : docs: vmware - VMWare
    • 37bb6dd : docs: ubuntu - Ubuntu
    • 0c90662 : docs: scons - SCons
    • 7f4f441 : docs: quote "git log"
    • 50e26e5 : docs: git - Git
    • 24e8e53 : docs: android - Android
    • ca613a5 : docs: drm - DRM
    • 9f59b19 : docs: wayland - Wayland
    • 8d2090e : docs: x11 - X11
    • 0db0152 : docs: cmake - CMake
    • b8b0e5e : docs: gpu - GPU
    • 12d925d : docs: visual studio - Visual Studio
    • 0e9d87e : docs: windows - Windows
    • 0df19e3 : docs: linux - Linux
    • 557ee09 : docs: microsoft - Microsoft
    • d84420f : docs: mingw - MinGW
    • a14d647 : docs: ninja - Ninja
    • eb1e7e6 : docs: chocolatey - Chocolatey
    • 0732efb : docs: mako - Mako
    • fafd1f4 : docs: bison - Bison
    • 1e75294 : docs: flex - Flex
    • d95e2e4 : docs: python3 - Python 3
    • e198fb6 : docs: meson - Meson
    • 0835ee9 : docs: correct reference to meson.build
    • 5ed6432 : docs: llvm - LLVM
    • a1904c9 : docs: bpp - BPP
    • 647a828 : docs: webservice - web service
    • 648cf63 : docs: existant - existent
    • 766705d : docs: softwara - software
    • 92a594b : spirv: Delete the legacy offset/index UBO/SSBO lowering
    • 1cadbe5 : vallium: Stop using lower_ubo_ssbo_access_to_offsets
    • 657d49a : spirv: Use derefs for push constants
    • d3fa745 : anv,radv,tu,val: Call nir_lower_io for push constants
    • ac7537f : nir/lower_io: Add support for push constants
    • 7a2b4ce : nir: Allow creating variables with nir_var_mem_push_const.
    • b0504ed : llvmpipe: use an alternate env var to enable clover.
    • 760ba65 : gallivm/nir: handle non-32-bit mul high
    • 5d4502c : gallivm/nir: fix up non 32-bit load stores
    • 2891136 : gallivm: fix 64-bit CL intrinsics.
    • 728061b : turnip: signal fence and semaphore in AcquireNextImage2KHR
    • e192f8f : turnip: share code between semaphores/fences + fence import/export
    • 0497c9c : turnip: remove remaining uses of drmSyncobj helpers
    • c4d5010 : turnip: rework ImportSemaphoreFdKHR
    • 8343c32 : turnip: rework GetSemaphoreFdKHR
    • fb76af2 : turnip: semaphores simplification (only syncobj semaphores supported)
    • 1dfb5a9 : turnip: set MSM_SUBMIT_SYNCOBJ_RESET for submit pWaitSemaphores
    • 2a3f6e0 : turnip: always create permanent syncobj for semaphore
    • dcc278c : turnip: clean up tu_device_memory
    • d12cdc9 : gallivm: fix pow(0, y) to be 0
    • 89a1a3e : r600/sfn: extend life range of all variables by one
    • 02245e3 : r600/sfn: Add support for more barrier instructions
    • 48edb98 : r600/sfn: Fix IDX register ID
    • 7683e27 : r600/sfn: Fix enabling the right interpolator for inerpolate_at_sample
    • 16a3a56 : freedreno: Don't leak LRZ bo's
    • 7454ae4 : freedreno/registers: Add a couple things used on kernel side
    • 27c8d97 : freedreno/drm: Also clean ring_cache
    • 98093e2 : freedreno/a6xx: Small cleanup
    • 5be3cfd : freedreno: Don't leak border_color_buf reference
    • 69a3ef6 : freedreno/drm: drop bo's dev reference
    • 6a20c63 : zink: unify all occurrences of waiting on a fence
    • e31381a : zink: combine all surface layout-setting for src/dst into util function
    • 24f19f4 : radv: Write correct dispatch size for RGP.
    • 1b39d74 : ci: Enable remaining (non-rotate) mustpass CTS tests
    • 8abe3bb : ci: cherry-pick deqp fix for config choosing
    • aee1c08 : ci/deqp-runner: Allow overriding width/height/config
    • efaea65 : intel: Add support for i945g to intel_stub_gpu.
    • f590148 : Add driver override to envvars.rst
    • 2584d48 : gallium/vl: add chroma_format arg to vl_video_buffer functions
    • b121b1b : gallium/vl: do not call transfer_unmap if transfer is NULL
    • 8d27577 : tu: Enable multi-position output
    • 64ad5a1 : ir3, tu: Link per-view position correctly
    • 6982e85 : ir3, tu: Run optimization loop twice
    • 41a5a21 : tu: Refactor shader compilation flow
    • 67ac166 : tu: Write multiview control registers in binning pass
    • a3363c3 : gallium: Make pipe_viewport_state swizzle_x/y/z/w bit-fields 8 bits wide
    • 291cfb1 : radv: move lowering of FS outputs outside of ACO
    • 4dae9e5 : radv: call nir_io_add_const_offset_to_base for FS outputs
    • 85074ec : radv/aco: Set I/O variable locations outside ACO.
    • cd1fab4 : radeonsi: enable support for GL_NV_shader_atomic_int64
    • 40c1f98 : mesa,glsl: add support for GL_NV_shader_atomic_int64
    • 8930b3b : gallium: prepare framework for supporting GL_NV_shader_atomic_int64
    • 6d4f33e : turnip: initial implementation of VK_KHR_push_descriptor
    • 992d247 : turnip: delete unused/broken pipeline layout hashing code
    • 560cff8 : turnip: remove unused cmd_buffer/device arguments in descriptor sets
    • d4d5382 : panfrost: Add support for rbg16 formats
    • dfc1d54 : CI: Re-enable Panfrost T7xx
    • 51e2b31 : nir: Handle per-view io in nir_io_add_const_offset_to_base()
    • bc8a5c0 : nir: Add per_view to IO semantics
    • 5a88db6 : nir/lower_io_arrays: Fix xfb_offset bug
    • df955ce : nir: Count i/o slots correctly for per-view variables
    • a9b2ac8 : anv: Use nir_lower_multiview pass
    • ffe946d : nir: Add nir_lower_multiview pass
    • b4c07a8 : gallium/util: allow scaling blits for stencil-fallback
    • 69654f0 : gallium/util: fix texture-coordinates for stencil-fallback
    • a21e1e6 : gallium/util: set right dst-dimensions
    • 778fe02 : radv/llvm: call nir_lower_io_to_vector with FS to fix array tests
    • 2c08404 : docs: add INTEL_SHADER_ASM_READ_PATH description
    • 1a19fae : radeonsi: call nir_io_add_const_offset_to_base only once per shader
    • a0e35c7 : nir/lower_io: change nir_io_add_const_offset_to_base to use bitfield modes
    • 7685c37 : st/mesa: use roundf instead of floorf for lod-bias rounding
    • 8f24a14 : docs: gallium - Gallium
    • 9d34c99 : docs: docker - Docker
    • 8a5eebd : docs: debian - Debian
    • 40cb542 : docs: flavours - flavors
    • 915d4e2 : docs: stabilisation - stabilization
    • 6ec9a7e : docs: consistantly - consistently
    • b1874a2 : docs: clippping - clipping
    • af9b073 : docs: drop outdated gallium-docs comment
    • cff7de4 : kmsro: Add mediatek entry point
    • fefb3e9 : panfrost: Add preliminary support for Mali G72
    • c420db1 : panfrost: Adjust quirks for bifrost v6
    • 89e3c3a : mesa: bptc fixes for decompress rgba_unorm and rgb_float
    • 1b4d968 : ac/llvm: fix invalid IR if image stores are shrinked using the format
    • 1588644 : radv: lower deref operations for global memory for both backends
    • 10b73e2 : ac/nir: implement nir_intrinsic_global_atomic_*
    • 82cf607 : ac/nir: implement nir_intrinsic_{load,store}_global
    • f01656c : CI: Temporarily disable Panfrost T7xx
    • 62ba074 : pan/mdg: Fix memory leak on error path.
    • 287ef94 : gallium/swr: Remove unreachable code.
    • 70945dd : aco: Initialize mad_info member literal_idx.
    • 93b8cdf : anv: Check file descriptor before closing.
    • 1d71b1a : intel/vec4: Remove everything related to VS_OPCODE_SET_SIMD4X2_HEADER_GEN9
    • 2a49007 : intel/vec4: Remove all support for Gen8+ [v2]
    • 60e1d0f : intel/compiler: Remove INTEL_SCALAR_... env variables
    • d0ce24c : intel/vec4: Remove inline lowering of LRP
    • 86bab92 : intel/compiler: Don't fallback to vec4 when scalar GS compile fails [v2]
    • 92f0886 : intel/compiler: Silence unused parameter warning in brw_surface_payload_size
    • 9bcdca2 : intel/vec4: Silence unused paramter warnings in brw_vec4_generator.cpp
    • bc32d41 : r600/sfn: Add support for helper invocations
    • 6aa196b : r600/sfn: remove old code to track uniforms as it is no longer needed
    • 31e42fb : r600/sfn: replace hand-backed literal check by NIR function
    • 6784cea : r600/sfn; go back to not lowering uniforms to UBOs
    • 9a6b11a : r600/sfn: Fix indirect const buffer access
    • 73c5f45 : r600/sfn: Fix interpolate at sample
    • a8435f7 : r600/sfn: use fine gradient evaluation for interpolate_at_offset
    • 8540523 : r600/sfn: save some instructions when doing multisample on sample 0
    • 0c5dfd2 : r600/sfn: Support group memory barrier
    • 972dd73 : r600/sfn: lower to scalar for some optimizations and vectortize later
    • d55404b : r600/sfn: Handle nir_op_b2b32
    • b2df9c2 : r600/sfn: Fix emitting shared atomics with constant sources
    • e8c5f8b : nir/lower_clip: Add i/o semantics for load/store intrinsics.
    • 4ed9049 : ci/freedreno: Add trace tests for glxgears, 0 A.D., and xonotic.
    • 8538b7d : ci/freedreno: Sort the traces in the .yml of expectations
    • a55dc27 : turnip: Replace tu_log*() with mesa_log*()
    • 13ea7db : mesa: Promote Intel's simple logging façade for Android to util/
    • 78165ea : radv: Record cache flushes for RGP.
    • cc73182 : radv: Include flushes in the barrier.
    • eff6710 : radv: Fix RGP Asic CU info for GFX10+.
    • 8349b82 : radv: Use correct alignment for SQTT buffer sizes.
    • 929aac5 : radv: Fix emitting SQTT userdata.
    • b2d042b : zink: remove old extension infrastructure
    • 08d51e9 : zink: use the new, generated extension infrastructure
    • 79c0023 : zink: use the new extension infrastructure in device creation
    • 0ed17a2 : zink: hook zink_device_info.py to build system
    • f1432fd : zink: generate extension infrastructure using a python script
    • 39098a2 : nir/lower_memory_model: do not break with global atomic operations
    • de14090 : nir/lower_memory_model: return progress when visiting instructions
    • 756441b : pan/mdg: Infer whether to disassemble shaders from info.internal
    • 90eaaad : panfrost: Mark blit shaders as internal
    • ee073cb : nir/load_store_vectorize: improve vectorization with identical operations
    • d9c4ec9 : panfrost: Remove extra printf arguments.
    • f2b94ae : glsl/linker: Add support for XFB varying lowering in geometry shader
    • 5bea0e8 : docs: recognisable - recognizable
    • b8f1075 : docs: url - URL
    • 7ee8a3a : docs: gitlab - GitLab
    • 0894b59 : docs: initialisation - initialization
    • 2dccdb1 : docs: timeplan - time plan
    • 9890927 : docs: behaviour - behavior
    • 3318043 : docs: unecessarily - unnecessarily
    • fa834fd : docs: make two acronyms upper-case
    • 58512ce : docs: frambuffer - framebuffer
    • e8c1b78 : docs: scons - SCons
    • 3fffa27 : docs: apis - APIs
    • 2550193 : docs: everytime - every time
    • e48b77b : docs: remove webmaster article
    • d2dfbf0 : docs: store prefixes in redirects
    • 12760d0 : docs: show 'Edit on GitLab'-link
    • 96ac711 : docs: escape backquote character
    • 9283141 : lima: set clear depth value to 0x00ffffff as default
    • 9a22109 : android: pan/bi: fix typo in bifrost_gen_disasm.c gen rules
    • 2e9599b : panfrost: Fix bifrost tiler job emission
    • 6abb4cd : panfrost: Fix bifrost tiler descriptor definition
    • 4017116 : ci: add dEQP-VK.info.device_extensions to the list of skipped tests
    • a3543ad : clover: set LLVM min version to 8.0.1
    • bf80fb7 : clover/nir: Call the memcpy lowering pass
    • 8323c03 : spirv: Add support for OpCopyMemorySized
    • a3177cc : nir: Add a lowering pass to lower memcpy
    • b2899f7 : nir: Add a new memcpy intrinsic
    • 8cbb664 : spirv: Use the new types helpers
    • de295b6 : compiler/types: Add glsl_baseN_t_type(bit_size) helpers
    • 4cb6a8d : gallivm: add support for lowered IO in vertex shaders
    • 4654609 : meson/anv: Use variable that checks for --build-id
    • 5f1855d : vtn/opencl: Switch fma to conditionally use libclc for 32bit floats
    • 8fb25c4 : vtn/opencl: Only use libclc ldexp when lower_ldexp is set
    • 09bca4c : vtn/opencl: Switch some nir-sequence ops to use libclc
    • 03fd217 : vtn/opencl: Switch division-related ops to use libclc
    • 2624faf : vtn/opencl: Switch exp/pow/log to use libclc
    • 4cc9ad4 : vtn/opencl: Switch non-native trig to use libclc
    • 6aee495 : vtn/opencl: Hook up OpenCL async copy and group wait opcodes via libclc
    • 00261d8 : vtn/opencl: Rework handle_instr to be able to handle core SPIR-V opcodes via libclc
    • b08fd45 : vtn/opencl: Implement a lot of opcodes via libclc
    • f33b417 : clover: handle libclc shader (v3)
    • 14bc2dc : util: add missing extern C
    • 6436e3a : vtn/opencl: Add infrastructure for calling out to libclc
    • 45d43ad : clover/nir: add libclc lowering pass
    • 2013903 : vtn/opencl: Rework type handling for CL extension opcodes
    • 938be4d : nir/vtn: Add handling for SPIR-V event variables
    • c9a6b94 : vtn: add an option to create a nir library from spirv
    • 2826a5a : nir: Add glsl_base_type unsigned - signed version helper
    • 93db59e : nir: Add an internal flag to shader_info
    • b4a087c : driconf: Use nesting macros for defining options.
    • e5d2481 : virgl: Clean up the driconf definition of GLES_SAMPLES_PASSED_VALUE.
    • 1eb79df : util/xmlconfig: Add a unit test of the code.
    • 7f3e881 : blorp: Ensure aligned HIZ_CCS_WT partial clears
    • 7898993 : meson: add support for loongson's mips/mips64 arch.
    • b54a0bb : freedreno/layout: add tile_all flag to the layout
    • 0206fb3 : nir/liveness: Consider if uses in nir_ssa_defs_interfere
    • dcba32b : turnip: implement VK_EXT_extended_dynamic_state
    • b2fa2d9 : turnip: move A6XX_RB_ALPHA_CONTROL write to init_hw
    • d1588c7 : turnip: fix wrong indentation in tu6_draw_common
    • a18c84e : nir/instr_set: hash intrinsic sources
    • b8c31ac : lima: fix glCopyTexSubImage2D
    • 7b9cf77 : kmsro: Extend to include imx-dcss
    • 77486db : intel/fs: Disable sample mask predication for scratch stores
    • cbef2dc : zink: only set stencil-ref for back if two-sided
    • 4b2525b : zink: store base-object of DSA-state
    • 9e94dcc : zink: correct typo in stencil-setup
    • d1d27e9 : radeonsi: remove redundant info.uses_fbfetch
    • 7b1e01f : radeonsi: remove redundant variables from struct si_compute
    • 16afaf0 : radeonsi: assume that constant load_local_group_size has been optimized out
    • 0492ada : radeonsi: run NIR optimizations that glsl_to_nir runs but other places might not
    • 8a1d556 : radeonsi: get input/output usage flags from shader_info directly
    • 0280cc7 : radeonsi: get information about FS color outputs from shader_info directly
    • 343e5ca : radeonsi: use info.system_values_read
    • cce3556 : radeonsi: call nir_shader_gather_info after lowering and optimizing NIR
    • 8be46d6 : radeonsi: fix indirect dispatches with variable block sizes
    • a7e3cc7 : etnaviv: simplify linear stride implementation
    • f7583bd : ci: Extend meson-clang coverage by compiling all gallium drivers
    • 1dd4505 : panfrost: gen_pack: Fix gnu-empty-initializer errors
    • ea77958 : nir: gather information about fbfetch and dual source color
    • a6abf17 : nir: fix input/output info gathering for lowered IO
    • ef98c17 : nir: gather fs.uses_sample_qualifier from lowered IO
    • 7b108e6 : nir: set system_values_read for all intrinsics
    • abe9588 : nir: gather tess.tcs_cross_invocation info from lowered IO intrinsics
    • 10be706 : nir: gather indirect info from lowered IO intrinsics
    • 3173367 : spirv: vtn_fail with a nice message on unsupported rounding modes
    • 03e7b75 : gallium/dri2: Move image-texture assignment after image NULL check.
    • 4bd0df0 : r600/sfn: Initialize GPRValue member m_pin_to_channel.
    • 974981c : gallium/drm: Make the pipe loader handle the driconf merging.
    • 8739238 : gallium/drm: Define the DRM entrypoints in drm_helper.h
    • bc10b49 : gallium/drm: Refactor the stub screen create functions.
    • 9ec28b8 : gallium/drm: Deduplicate screen creation for the dynamic (clover) pipe loader.
    • 472a20c : radeonsi: Only call nir_lower_var_copies at the end of the opt loop
    • 7971918 : zink: reorder create_stream_output_target to fix failure case leak
    • bc5c043 : zink: apply viewport count when creating pipelines
    • cc939fe : zink: correctly set up fb-sized scissors for each viewport
    • 8a2ddcb : zink: set multiviewport cap in ntv when gl_ViewportIndex is a written output
    • 09be30b : zink: move viewport count to zink_gfx_pipeline_state
    • 818bd61 : zink: implement ARB_instanced_arrays
    • 0051551 : zink: verify that src and dst aspects are the same in resource_copy_region hook
    • 6feec2d : zink: clamp min created fb size to 1x1
    • 9ce1e99 : zink: use correct layer count when creating framebuffer
    • e30367a : zink: use correct number of samples on framebuffer in set_framebuffer_state
    • f6c8721 : zink: use correct value for color buffer sample count when creating renderpass
    • 140f53e : Revert "nir: replace lower_ffma and fuse_ffma with has_ffma"
    • d8cdcd4 : radv: use syncobj for wsi fence
    • 1aa1b9f : radv: fix incorrect ResetFences path for WSI fence
    • 2479689 : radeonsi/tmz: add workaround for mpv/vaapi subtitles
    • 2c66435 : radeonsi/tmz: add a tmz variant for sctx::eop_bug_scratch
    • 1338955 : radeonsi: disable primitive discard if tmz is in use
    • 8e2768b : radeonsi/tmz: add tmz variant for sctx::tess_rings
    • 2589888 : radeonsi/tmz: add tmz variant of sctx::wait_mem_scratch
    • de088da : radeonsi/tmz: fail si_texture_transfer_map if tex is encrypted
    • 1b0d660 : radeonsi/tmz: allow secure job if the app made a tmz allocation
    • f5ec617 : ac/gpu_info: add detection of TMZ support
    • 5e4aece : radeonsi: introduce SI_RESOURCE_FLAG_INTERNAL / RADEON_FLAG_DRIVER_INTERNAL
    • 131412c : radeonsi/tmz: allocate depth/stencil buffers as encrypted
    • 0a59d2a : radeonsi/tmz: add safety assert when tmz is enabled
    • 59b267d : radeonsi/tmz: use secure job if using an encrypted z/s buffer
    • 5bc24fe : radeonsi/tmz: use secure job if framebuffer has dcc
    • cec0bc7 : turnip: rework fences to use syncobjs
    • c232067 : turnip: require syncobj support
    • 89ffe85 : turnip: add a fd field to tu_device
    • ec4fe92 : turnip: delete unused tu_fence_signal function
    • 9e13a16 : mesa: handle GL_FRONT after translating to it
    • 005d9d2 : ci: adjust RadeonSI rules
    • 4c71cda : vulkan/wsi/display: add option for display fence to signal syncobj
    • 32d7548 : radeonsi: always inline draw-related functions that have only one use
    • f24b589 : radeonsi: lift the conditional for skipping si_upload_vertex_buffer_descriptors
    • 0b2f75f : radeonsi: add unlikely statements into si_draw_vbo
    • 8ab15c9 : radeonsi: move si_upload_vertex_buffer_descriptors into si_state_draw.c
    • 12b1e8a : radeonsi: reorganize the code around the gfx9 scissor bug
    • 532259b : radeonsi: don't call emit_cache_flush after uploading bindless descriptors
    • d647065 : radeonsi: move a displaced comment in si_draw_vbo
    • 816a867 : radeonsi: call si_upload_graphics_shader_descriptors before the big conditional
    • 0051f2c : radeonsi: use optimal order of operations when setting up a compute dispatch
    • e560028 : panfrost: Remove redundant casts of viewport position
    • 3d0ae7a : panfrost: Handle non-positive viewport positions
    • 2aa48bb : panfrost: Clamp uniform buffer size
    • 939ddf3 : nir: replace lower_ffma and fuse_ffma with has_ffma
    • f128450 : radeonsi: fuse or lower ffma optimally on all chips
    • 771aad3 : nir: split lower_ffma into lower_ffma16/32/64
    • 21174de : nir: split fuse_ffma into fuse_ffma16/32/64
    • c58b46e : radeonsi: force linear for textures with height=1 (gfx6-8)
    • 5f27777 : radeonsi: add a tweak for PS wave CU utilization for gfx10.3
    • b5debe1 : radeonsi: break a binning batch on a new PS if bins can use multiple state sets
    • 972fb03 : radeonsi: move binning parameters into si_screen
    • 8b134fb : radeonsi: set KEEP_TOGETHER_ENABLE if needed
    • 28a7616 : pan/bi: Remove old register mode definitions
    • 7a0f3b6 : pan/bi: Use canonical register packing
    • 514da97 : pan/bi: Rename port - slot
    • c749803 : pan/bi: Decode all 32-bit register modes
    • bdb33f7 : pan/bi: Pass 'first' through disassembler
    • 4d3d5a6 : pan/bi: Add bifrost_reg_mode enum
    • b392c07 : pan/bi: Disassemble PC-relative addresses
    • 08a9e5e : pan/bi: Decode M values in disasm
    • 9de34e1 : pan/bi: Track M values of disassembled constants
    • 43c6623 : pan/bi: Inline dump_instr
    • 5ff3fea : pan/bi: Annotate disassemble with format names
    • ab5cc3e : pan/bi: Annotate stop bit (canonically "Z-bit")
    • 24a74b3 : radv: add a tweak for PS wave CU utilization for gfx10.3
    • 6865a6a : radv: set KEEP_TOGETHER_ENABLE if necessary on GFX10+
    • 68bb91e : radeonsi: don't use nir_opt_access
    • 87f91e2 : st/nir: call nir_opt_access before gl_nir_lower_buffers
    • e893102 : radv: Add VK_FORMAT_E5B9G9R9_UFLOAT_PACK32 rendering support.
    • 43735c0 : radv: Simplify radv_is_hw_resolve_pipeline.
    • bf0c82b : radv: Disable NGG on APUs.
    • 12128fb : lima: add natively supported vertex buffer formats
    • c669db0 : radv: advertise VK_KHR_copy_commands2
    • 0401507 : radv: add support for CmdResolveImage2KHR()
    • 9a700af : radv: cleanup selecting the hardware resolve path
    • fe81971 : radv: add support for CmdCopyImageToBuffer2KHR()
    • f9915ac : radv: add support for CmdCopyImage2KHR()
    • 69dfcfb : radv: add support for CmdCopyBufferToImage2KHR()
    • 22a08da : radv: add support for CmdCopyBuffer2KHR()
    • eccc73f : radv: add support for CmdBlitImage2KHR()
    • cc709a0 : radv: add VK_KHR_copy_commands2 but leave it disabled
    • 93f158f : radv: bump the advertised patch version to 145
    • 43c7de6 : radv: remove useless assignment of MAX_API_VERSION
    • e8a4071 : gallium/util: add blitter-support for stencil-fallback
    • 2970d38 : gallium/util: add shader for stencil-blits
    • 924e276 : nir_lower_system_values: Fix load_global_invocation_id to use base_work_group_id even with no base_global id
    • ee4cee6 : android: Disable trying to read/write to the disk cache.
    • 25b36d9 : mesa: Make the android_stub be a set of non-installed shared libraries.
    • 40a50e9 : radeonsi: remove KILL_PS_INF_INTERP/CLAMP_DIV_BY_ZERO, use screen::options
    • 9818840 : radeonsi: move debug options from si_disk_cache_create to si_get_ir_cache_key
    • cd12fcf : radeonsi: set TRUNC_COORD=0 for Total War: WARHAMMER to fix it
    • a1f46d7 : ci: Use ignore_scheduled_pipelines anchor in .radeonsi-rules
    • e781cc7 : tu: Expose shaderStorageImage*WithoutFormat
    • 37054a3 : ir3: Don't use the format to get the image type
    • 6ebc20f : tu: Expose shaderImageGatherExtended
    • 205f4e9 : tu: Expose shaderStorageImageExtendedFormats
    • 5d3a72c : r600/sfn: Be a bit more verbose when logging skipped FS outputs
    • 2abb9e5 : r600/sfn: Acquire the number of FS outputs and the write_all info early
    • c2c5f71 : r600/sfn: Take dual source blending output indices into account
    • 44b45a3 : r600/sfn: Handle number of color outputs taking dual source blending into account
    • 234891c : r600/sfn: Keep info about dual-source blend in FS
    • 0a59530 : r600/sfn: Sort the outputs of the FS according to data index
    • a57f2d0 : r600: Add flag for dual-source blending to shader key
    • 1625a9b : r600/sfn: remove a useless if-condition
    • 4f4e90b : r600/sfn: run late algebraic optimizations
    • 8a9876a : r600/sfn: use cnde instead of cnde_int
    • 952cf95 : etnaviv: fix comment for source of etna_mesa_debug
    • 9750164 : nir: Rename get_buffer_size to get_ssbo_size
    • f100cf0 : aco: stop multiplying driver_location by 4
    • fd872c3 : aco: remove dead indirect fs input loading
    • 7f51a0c : aco: use nir's constant source helpers more
    • 430cc90 : aco: use nir_get_io_offset_src() in visit_load_input()
    • 9bba790 : aco: use io semantics to get an intrinsic's slot
    • d58a1a8 : aco: Use NIR IO semantics for tess factor IO locations.
    • 10540c2 : anv: implement VK_KHR_copy_commands2
    • 4c1c546 : vulkan: bump headers/registry to 1.2.154
    • f2b17de : nir/lower_samplers: Clamp out-of-bounds access to array of samplers
    • 0ba82f7 : nir/large_constants: Eliminate out-of-bounds writes to large constants
    • 66669eb : nir/lower_io: Eliminate oob writes and return zero for oob reads
    • 8e1193b : radeon: rename RADEON_TRANSFER_* - RADEON_MAP_*
    • 3176534 : gallium: rename transfer flags - map flags in comments
    • 3586068 : gallium: rename pipe_transfer_usage - pipe_map_flags
    • 22253e6 : gallium: rename PIPE_TRANSFER_* - PIPE_MAP_*
    • 44f1b1b : radeonsi: add 16-bit ALU vectorization
    • 6702e54 : radeonsi: don't scalarize 16-bit vec2 ALU opcodes
    • e823742 : radeonsi: vectorize IO for better ALU vectorization
    • 98a52fe : radeonsi: implement 16-bit FS color outputs
    • d3c63d6 : radeonsi: set flags for FP16 in shaders
    • c5277f0 : radeonsi: don't lower pack for better 16-bit vectorization
    • 581a351 : ac/nir: implement nir_op_[un]pack_[64/32]_*
    • 3022cf3 : glsl: Initialize ir_constant member const_elements in all constructors.
    • a5967f3 : pan/bi: Fix typo.
    • ade72e6 : st/mesa: Deal with empty textures/buffers in semaphore wait/signal.
    • 08add9f : turnip/kgsl: Associate fences with submits.
    • e807584 : turnip: Add kgsl backend
    • cc3bf00 : intel/compiler: fixup Gen12 workaround for array sizes
    • 54b9013 : wsi/x11: Create a present queue for VK_PRESENT_MODE_FIFO_RELAXED_KHR
    • 26a83be : gallium/auxiliary/vl: Include src region in scale_y calculation
    • 9a4daac : frontends/va: Enabled packed headers for H.264 encoder
    • 7b9414f : radeon: Pass HEVC encode crop parameters to the encoder
    • 245f619 : gallium: Parse packed HEVC SPS encode header for crop parameters
    • 0a3f92b : r600/sfn: Fix typo in comment
    • 07cb38f : r600/sfn: Fix component count for fdph
    • eb3199d : r600/sfn: make number of source components a local variable
    • d78e7b7 : gallium/auxiliary/vl: Odd Dimensions are failing
    • 05b6612 : radv: do not lower UBO/SSBO access to offsets
    • 3808fa4 : radv: add a helper for loading meta descriptors
    • 0836e90 : ci: Don't create test-docs job if the pages one exists in the pipeline
    • efec576 : ci: Prevent pages job from running in pre-merge pipelines
    • 8cb450f : ci: Always use CI_PROJECT_NAMESPACE instead of CI_PROJECT_PATH
    • 208876b : ci: Add "is for Marge Bot" YAML anchor
    • 0fcb8b3 : ci: Add "is pre-merge pipeline" YAML anchor
    • b47f1bb : ci: Add "is post-merge pipeline" YAML anchor
    • 22a7c25 : ci: Add "is forked branch" YAML anchor
    • 24f5329 : ci: Add "is forked branch or pre-merge pipeline" YAML anchor
    • a33026e : ci: Add "is post-merge pipeline, not for Marge Bot" YAML anchor
    • f79ae94 : ci: Add "is pre-merge pipeline for Marge Bot" YAML anchor
    • ac56d66 : ci: Add "is master branch of main project" YAML anchor
    • c3ae82b : ci: Add "is scheduled pipeline" YAML anchor
    • 2b99e15 : radv: fix transform feedback crashes if pCounterBufferOffsets is NULL
    • d2c18b7 : aco: use bit vectors for liveness sets
    • ec2185c : aco: keep track of temporaries' regclasses in the Program
    • f007115 : radeonsi: change vendor name to AMD
    • b73e165 : radeonsi: reduce PIPE_SHADER_CAP_MAX_CONST_BUFFER_SIZE value
    • 1826367 : radeonsi: move GL vendor workaround to drirc
    • fc6df02 : gallium: add PIPE_CAP_MAX_TEXTURE_MB
    • 2228835 : radv,aco: fix reading primitive ID in FS after TES
    • b1c3f63 : panfrost: decode: Misc formatting improvements
    • 6b92303 : panfrost: Avoid copying job descriptors around when we can
    • d289209 : panfrost: XML-ify the compute job descriptor
    • 54d716a : panfrost: Rename the Blend dither disable flag
    • 89fafe9 : panfrost: XML-ify the fragment job descriptor
    • 9121e7d : panfrost: XML-ify the write value job descriptor
    • eb92354 : panfrost: XML-ify the job header descriptor
    • 5d5f755 : panfrost: XML-ify the multi-target framebuffer descriptors
    • efce73d : panfrost: XML-ify the bifrost tiler descriptors
    • 95eb7d9 : panfrost: XML-ify the single target framebuffer descriptor
    • e855698 : panfrost: XML-ify the midgard tiler descriptor
    • fb95729 : panfrost: Add Tiled linear mode to the Block Format enum
    • 76096c7 : panfrost: Clarify what TILED mode is
    • 3a06fc3 : panfrost: XML-ify the local storage descriptor
    • 706974c : panfrost: decode: Use pan_{unpack,print}() when applicable
    • aa2670c : panfrost: decode: Add a macro to dump unpacked descriptors
    • 670e818 : panfrost: decode: Rework the DUMP_{CL,ADDR}() macros
    • 6249ae7 : panfrost: decode: Make the indentation consistent with auto-generated print helpers
    • fd0941c : panfrost: gen_pack: Allow enum/define values expressed in hexadecimal
    • 51d5d44 : panfrost: gen_pack: Add a log2 modifier
    • 718b521 : panfrost: gen_pack: Add an align() modifier
    • 1b27817 : panfrost: gen_pack: Allow empty structs
    • c7a1019 : panfrost: gen_pack: Add the aggregate concept
    • fa7d097 : panfrost: gen_pack: Move the group get_length() logic to its own method
    • 62c0ef0 : panfrost: gen_pack: Add pan_{unpack,print}() helpers
    • ddd82c3 : panfrost: gen_pack: Fix __gen_unpack_uint()
    • 66f25dd : panfrost: gen_pack: Minor formatting improvement
    • 2aa5838 : Revert "panfrost: Drop implicit blend pooling"
    • 8ae4cec : Revert "radv: emit {CB,DB}_RMI_L2_CACHE_CONTROL at framebuffer time"
    • 0a84c59 : Revert "radv: set BIG_PAGE to improve performance on GFX10.3"
    • d78df70 : radv,radeonsi: Disable compression on interop depth images
    • ecc19e9 : ac/surface: Fix depth import on GFX6-GFX8.
    • f63ffc1 : intel/fs/swsb: SCHEDULING_FENCE only emits SYNC_NOP
    • e1fc232 : nir: Add a pass for lowering CL-style image ops to texture ops
    • 3fc425b : spirv: Plumb access qualifiers through from image types
    • 1e90210 : spirv: Access qualifiers are not a bitfield
    • afa5c2e : clover: Pull the stride from pipe_transfer for image maps
    • 27516ba : pan/mdg: Use nir_shader_instructions_pass for fdot2 lowering
    • 7a348b4 : pan/mdg: Use nir_shader_instructions_pass for nir_lod_errata
    • 6064b8e : pan/mdg: Return a bool from midgard_nir_lod_errata
    • 7acf364 : panfrost: Set modifier_constant to true for exported resources
    • 95ee0ba : lima: fix vertex shader uniform buffer size
    • 6a19260 : lima: allocate new bo for stream draw
    • 017ca86 : radeonsi: Move display dcc dirty tracking to framebuffer emission.
    • c6c1fa9 : radeonsi: Put retile map in separate buffers.
    • 4656e68 : radeonsi: fix user fence GPU address
    • 9aa86eb : glsl_type: Add packed to structure type comparison for hash map
    • 7b4eaac : radv: fix vertex buffer null descriptors
    • ec13622 : radv: fix null descriptor for dynamic buffers
    • 090fc59 : mesa: fix glUniform* when a struct contains a bindless sampler
    • 77af1ca : etnaviv: add disk cache
    • 6a0d7f6 : etnaviv: shuffle some variant fields
    • 12b8ad8 : swr: fix _BitScanForward64 on unix
    • 82c49a6 : swr: missing _BitScanForward64 on 32 bits win
    • 5805f5a : glsl: take EXT_gpu_shader4 in to account when adding round
    • cde5b86 : turnip: Release bo_mutex lock before potential error path.
    • ffbdbd6 : panfrost: Delete debug allocated syncobj.
    • fcc506e : radesonsi: Remove unsigned comparison to zero.
    • 1b86271 : disk_cache: Fix filename leak on error path.
    • 183ca88 : android: pan/bi: Use new packing
    • 792b517 : android: pan/bi: Use new disassembler
    • d74fe47 : glx: Allow depth-30 pbuffers to work without a depth-30 pixmap format
    • a163cba : anv: Fix the target_bo assertion in anv_reloc_list_add
    • 85cc295 : radv: initialize with expanded cmask if the destination layout needs it
    • 6f2b695 : nir: remove ubo_r600 instrinsic since ubo_vec4 is used now
    • 18e9781 : r600/sfn: Use load_ubo_vec4 lowering pass
    • 98eb00f : r600: enable lowering uniforms to UBO
    • 4ac4cdb : aco: fix incorrect assertion in emit_vop3a_instruction()
    • 207219d : turnip: Add support for a615.
    • 45bff4e : util: Makefile.sources: Add disk_cache_os.{c,h}
    • 2aaa156 : nir/lower_io: don't reduce range if parent length is zero
    • a51aa6f : pan/bi: Drop *FMIN reference
    • 865e252 : pan/bi: Remove unused packing data structures
    • ec43629 : pan/bi: Remove unused prints
    • 2ff5387 : pan/bi: Use new packing
    • 3fadd82 : pan/bi: Move packing helpers to dedicated file
    • d2158a5 : pan/bi: Use src1/dest_invert instead of src_invert[]
    • 08b105d : pan/bi: Use 8-bit shifts
    • 3d63a47 : pan/bi: Pass blend descriptor explicitly in IR
    • 67d8956 : pan/bi: Track compute_lod in IR
    • 8dd3a81 : pan/bi: Add format field to IR
    • 6f5b788 : pan/bi: Introduce segments into the IR
    • 33710ff : pan/bi: Add dummy carry/borrow argument for iadd/isub
    • 42ec4aa : pan/bi: Use canonical syntax for special constants
    • 5a569d0 : pan/bi: Use canonical syntax for registers/uniforms/imms
    • f8fc210 : pan/bi: Use new disassembler
    • deab752 : pan/bi: Export dump_src
    • 0504181 : pan/bi: Add bi_disasm_dest_* helpers
    • 944cb8b : pan/bi: Add disassembly prototypes
    • 82f3315 : pan/bi: Add disassembler generator
    • cf8f79a : pan/bi: Add packing generator
    • 58da316 : pan/bi: Add ISA parser
    • 07a5ec8 : pan/bi: Add XML describing the instruction set
    • 2044b1f : pan/bi: Lower flrp16
    • b75a92e : anv: assert that the target bo is valid when adding a reloc list
    • 355f241 : anv: improve error message when failing to open device path
    • 203b934 : docs: update calendar and link releases notes for 20.1.8
    • fdff24d : docs: add release notes for 20.1.8
    • 123bdb6 : gallium+mesa/st: lower uniforms based on compiler flag instead of packed uniforms cap
    • feb463d : llvmpipe: set lower_uniform_to_ubo compiler flag
    • 80cde3a : intel/compiler: Set lower_uniform_to_ubo compiler flag
    • 7ab804d : freedreno/ir3: set lower_uniforms_to_ubo compiler flag
    • b155b68 : radeonsi: set compiler flag lower_uniforms_to_ubo
    • 2c9fee9 : nir: Add option lower_uniforms_to_ubo
    • 8939c64 : gallium/aux: remove unused u_blit.[ch]
    • 9ba2365 : vc4: remove unused header
    • 9992797 : v3d: remove unused header
    • 39e7bc2 : r600/sfn: Fix comparison with different signedness
    • 2d6316c : r600/sfn: more fixing of vec4 fetching
    • f46e04a : r600/sfn: Fix source swizzle for gradient queries
    • 02581ac : r600/sfn: Fix bitfield ops and 2x16 split_y
    • dfa45e2 : r600/sfn: Fix split_alu_modifiers
    • e85c0bd : r600/sfn: Lower *sign opcodes in nir
    • c22e0fb : r600/sfn: avoid some copies
    • f2280e0 : r600/sfn: Fix ordering of tex param moves
    • daaa719 : r600/sfn: clone shader before lowering to registers and src/dest modifiers
    • 09558ad : r600/sfn: Fix loading vertex attributes
    • e46b2ad : r600/sfn: correct ring op patching
    • 688680d : r600/nir: fetch sources and split uniforms before emittting alu instructions
    • 85f39ca : r600: revert disabling llvm draw
    • f1730be : radeonsi: fix guardband handling for large values
    • c493bb9 : radeonsi: fix quant_mode selection for large negative values
    • 7d85396 : radeon/vcn: set dec-bs_ptr = NULL on unmap
    • eb60849 : r600/uvd: set dec-bs_ptr = NULL on unmap
    • 15038fe : docs/release-calendar: update 20.2
    • babe86f : docs: add another 20.1.x release to allow for more overlap with 20.2
    • fd990aa : docs/download: mention tarball GPG signatures and link to the keys
    • 4790811 : Revert "radeonsi: move L2_CACHE_CONTROL registers into si_emit_framebuffer_state"
    • b23013d : Revert "radeonsi: set BIG_PAGE fields on gfx10.3"
    • 758ab39 : radeonsi: clean up ffma handling
    • 57bf4c2 : nir,radeonsi: move ffma fusing to late optimizations for better codegen
    • a3512dd : st/mesa: don't enable NV_copy_depth_to_color if NIR doesn't support FP64
    • 782016e : libgl-gdi: Fix unused-variable warnings.
    • bf849b0 : spirv: Handle OpTypeOpaque
    • efff734 : turnip: multiViewport and VK_EXT_shader_viewport_index_layer
    • 52534c3 : freedreno/ir3: add view_zero to shader key
    • e732750 : freedreno/ir3: allow layer/viewport output for VS/GS/DS
    • 338745c : frontends/va: Derive image from interlaced buffers in some cases
    • fcb5583 : frontends/va: Derive image from interlaced buffers
    • 8c7ca97 : frontends/va: Add support for NV12/P010/P016 to vaDeriveImage
    • 7bfaeaa : aco: pass -fno-exceptions and -fno-rtti
    • 51a6261 : iris: remove additional pipe control done before hiz for older gens
    • 320d80f : intel/genxml: make sure test assert are compiled in
    • c7719b8 : nv50/ir: Initialize Converter members.
    • e607477 : freedreno: Check file descriptor before write.
    • 50f1cd4 : ac/llvm: Fix nonportable sizeof.
    • 16e7884 : android: aco/isel: Move context initialization code to a dedicated file
    • f3150ab : etnaviv: call nir_opt_shrink_vectors(..) in opt loop
    • b10f34a : pan/bit: Fix unit tests
    • c4f518c : pan/bit: Set d3d=true for CMP tests
    • 322ddbd : panfrost: Fix nonzero stencil mask on vertex/compute
    • f3109c4 : turnip: avoid heap allocations in QueueSubmit when semaphores are used
    • 00914e2 : nir/algebraic: fold some nested comparisons with ball and bany
    • 8a2a9e9 : glsl: fix per_vertex_accumulator::fields size
    • 26299c8 : aco: Add base argument to emit_mbcnt.
    • f3780e7 : aco: Clean up emit_mbcnt.
    • efa1c76 : aco: Fix emit_boolean_exclusive_scan in wave32 mode.
    • 12567de : glsl: mark some builtins with correct glsl(es) version check
    • 4868ce1 : lima: dont split vec3 unaligned load inputs
    • bea6290 : turnip: device global bo list
    • 20a4235 : anv, iris: Set MediaSamplerDOPClockGateEnable for gen12+
    • e3101c9 : anv/image: Disable multi-layer CCS_E on TGL+
    • 656d8ed : nir/opt_vectorize: don't lose exact and no_*_wrap flags
    • eceb4fb : freedreno: fix fence-fd leak
    • ef980ac : radeonsi: fix max syncobj wait timeout
    • 3182209 : v3d/compiler: fix V3D double-rounding of .8 fixed-point XY coordinates
    • 367ac07 : disk_cache: move cache item loading code into disk_cache_load_item() helper
    • 8598dc1 : disk_cache: add new OS specific helper disk_cache_evict_item()
    • 7893dc4 : disk_cache: move get_cache_file() to an OS specific helper
    • 593ef16 : disk_cache: create new helper for writing cache items to disk
    • b4a3a80 : disk_cache: move evict_lru_item() to an OS specific helper
    • e123647 : disk_cache: move munmap into an OS specific helper
    • 889356e : disk_cache: move index mmap into OS specific helper
    • 62c4339 : disk_cache: add disk_cache_enabled() helper
    • 4339ecd : disk_cache: move cache dir generation into OS specific helper file
    • 65d0fa0 : nv50/ir: Remove duplicate mask assignment.
    • a2a330a : nv50/ir: Add fallthrough statement.
    • 50d3358 : nir/algebraic: add late optimizations that optimize out mediump conversions (v3)
    • b86305b : nir/algebraic: collapse conversion opcodes (many patterns)
    • cdd498b : nir: add new mediump opcodes f2[ui]mp, i2fmp, u2fmp
    • 385b4db : nir: enforce 32-bit src type requirement for f2fmp and i2imp
    • 3d3df8d : nir: remove redundant opcode u2ump
    • 26fc5e1 : nir/algebraic: expand existing 32-bit patterns to all bit sizes using loops
    • 3c8934a : nir/algebraic: add flrp patterns for 16 and 64 bits
    • 3064feb : pan/mdg: Obey f2fmp size restriction in fuse_io_16
    • 5cd6235 : blorp: Drop trailing whitespace in blorp_clear.c
    • 1eff389 : blorp: Fix alignment test for HIZ_CCS_WT fast-clears
    • 40f7afc : nir: fix lower_mediump_outputs to not require variables
    • c2ae39e : nir: add mediump flag to IO semantics
    • 5e9e457 : glx/dri3: Implement GLX_EXT_swap_control_tear
    • 60ebeb4 : glx: Implement GLX_EXT_swap_control for DRI2 and DRI3
    • b8239ab : glx: Reject glXSwapIntervalMESA greater than INT_MAX
    • e648442 : glx: Collect all the non-applegl extensions in the GetProcAddress table
    • 72a0f74 : glx: Use GLX_FUNCTION2 only for actually aliased function names
    • 77d2562 : wsi/x11: Hook up VK_PRESENT_MODE_FIFO_RELAXED_KHR
    • 2f62a44 : ci: do not build libdrm for vc4, reedreno and etnaviv
    • 52becd3 : turnip: rework vertex buffers draw state handling
    • 69e65b9 : intel/tools: fix possible randomly increased verbosity of error2aub
    • 18eb853 : intel/compiler: quiet Coverity warnings
    • 5b6fd2a : intel/tools: handle ftell errors
    • 46a82aa : intel/tools: fix possible memory leak in the error path
    • 5ea0b6a : intel/compiler: initialize remaining fields of various classes
    • 40b964d : intel/compiler: remove unused fs_validator::param_size
    • d3cd249 : anv: fix minor gen_ioctl(I915_PERF_IOCTL_CONFIG) error handling issue
    • 265a3b9 : driconf: add option to reuse GL names
    • a56849d : mesa: use _mesa_HashFindFreeKeys for GL functions
    • 60ffadc : mesa: add _mesa_HashFindFreeKeys
    • fefc6d2 : mesa: add GL name reuse support
    • 3485212 : mesa: add a isGenName parameter to _mesa_HashInsert
    • 553d371 : util/idalloc: add lowest_free_idx to avoid iterating from 0
    • e808d38 : util/idalloc: add util_idalloc_reserve
    • 87ef970 : mesa: move u_idalloc from gallium/aux/util to util
    • 911ce37 : freedreno/a6xx: Fix MSAA clear
    • 6e4d0a4 : freedreno: Clear gs/tcs/tes state for clear blits
    • 18f5d36 : freedreno/a6xx: Fix fd6_draw_vbo() return
    • 46edc08 : freedreno/a6xx: Skip empty tile_setup
    • 74f1c50 : freedreno: Don't bypass fd_draw_vbo() in clear fallback
    • 97e49b2 : freedreno: Fix rast state for multisample clear
    • b0722cb : gallivm: disable brilinear for lod bias and explicit lod.
    • cd4fb5a : freedreno/fdl: Add layout test for the Android CTS's MSAA mustpass surface.
    • 14131ed : freedreno/cffdec: Add support for texturator's 2DMS layout setup.
    • 2f39727 : freedreno/cffdec: Fix up texturator parsing scripts for XML changes.
    • bd647f8 : freedreno: Add another new sysmem flake.
    • 0bc36ef : spirv: Initialize spirv_test member shader.
    • 2f61d7c : svga: Fix unused printf argument.
    • 5879691 : freedreno: Fix file descriptor leak.
    • a4885d2 : pan/mdg: Fix spilling of non-32-bit types
    • 89401e5 : nir: More NIR_MAX_VEC_COMPONENTS fixes
    • c5dd54e : nir/idiv_const: Use the modern nir_src_as_* constant helpers
    • d86e38a : nir: More NIR_MAX_VEC_COMPONENTS fixes
    • 14b60ea : iris: Support I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS
    • dd215c8 : iris: Support MC modifier in plane count queries
    • 80a8bb5 : intel/common: Drop unused gen_aux_map_add_image
    • 8272085 : iris: Support planar resource imports for MC
    • 1ddb1a6 : intel/common: Add get_aux_map_format_bits()
    • c3c99f4 : st/mesa: Don't map all P01X DRM formats to P016
    • 9efd714 : intel/isl: Add YUV format info for the aux-map
    • 1ea3a54 : intel/isl: Support ISL_AUX_USAGE_MC in surface states
    • e15543b : intel/isl: Describe I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS
    • d686835 : gallium/dri2: Support I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS import
    • a42bf9a : gallium/dri2: Report I915_FORMAT_MOD_Y_TILED_GEN12_MC_CCS num_planes
    • 7ee5da9 : nir_dominance: Use uint32_t instead of int16_t for dominance counters
    • 03eec29 : st/nir: Call st_glsl_to_nir_post_opts before interface unification
    • cdc025c : mesa: Generate more errors from GetSamplerParameter
    • 06c6920 : mesa: Fix GL_CLAMP handling in glSamplerParameter
    • 0f61f01 : ci/bare-metal: Allow wget of the kernel/dtb for kernel development.
    • b4317fc : ci/bare-metal: Update the kernel to msm-next-pgtables
    • 802d361 : turnip: Fix truncation of iovas to 32 bits in queries.
    • 329c317 : turnip: Fix truncation of CS shader iovas to 32 bits.
    • 3b3772d : freedreno: Make the pack struct have a .qword for wide addresses.
    • 021523d : turnip: Fix a compiler warning in release builds of the query code.
    • 5a95cc0 : turnip: remove some unnecessary regs init
    • 3d0ab65 : turnip: delete unused "tu_cmd_buffer_upload"
    • 3b144d5 : turnip: fix the type of tu_shader_module code field, delete unused sha1
    • 6f51192 : turnip: delete unused tu_image fields
    • bd53a25 : turnip: delete tu_physical_device path field
    • e394ff9 : aco: fix validation of sub-dword parallel-copies
    • 834b449 : aco: fix value numbering of reductions
    • 5a8447c : aco: don't apply constant to SDWA on GFX8
    • 29cb3cb : aco: workaround disassembler bug of v_writelane_b32 with literal
    • 36e58a1 : aco: fix v_writelane_b32 with two sgprs
    • 641d45b : nir/opt_loop_unroll: fix is_access_out_of_bounds with vectors
    • c0d04cd : panfrost: AFBC to linear layout conversion
    • a3431ff : panfrost: Move tiled-linear conversion checking to a new function
    • c356453 : panfrost: Seperate resource setup and bo creation
    • 23ad952 : panfrost: Correctly set modifier_constant
    • 361396c : panfrost: Remove old comment on broken depth reload
    • 5857a01 : panfrost: Cleanup panfrost_get_param
    • 811b9cb : docs/features: Add missing Panfrost extensions
    • cf2eebd : radv,gallium: Add driconf option to reduce advertised VRAM size.
    • cc431f4 : iris: Add missing range_base/range to our nir_load_ubos.
    • 4df0eef : llvmpipe: include gallivm perf flags in shader cache.
    • fefeaee : aco/isel: Compile all helper functions with static linkage
    • 793dc66 : aco/isel: Move add_startpgm to aco_instruction_selection.cpp
    • 47de553 : aco/isel: Move context initialization code to a dedicated file
    • 1eac0b5 : aco/isel: Remove unused definitions
    • 150de63 : aco/isel: Consistently use references for input parameters in emit_load
    • dab0af0 : aco/isel: Simplify nested branching code
    • 757de68 : aco/isel: Turn the function template emit_load into a proper function
    • 3bd7c3c : intel/nir: Call validate_ssa_dominance at both ends of the NIR compile
    • 7cedc41 : spirv: Run repair_ssa if there are discard instructions
    • 45bcb10 : nir: Add a dominance validation pass
    • 6cef804 : nir/opt_if: fix opt_if_merge when destination branch has a jump
    • 1ed78bd : nir: Use explicit deref information to provide real UBO ranges.
    • 41b5aaf : freedreno/ir3: Apply the max upload limit to initial range setup
    • f74c3b0 : freedreno/ir3: Use the new NIR UBO ranges in UBO analysis.
    • f3b33a5 : nir: Add a range_base+range to nir_intrinsic_load_ubo().
    • 3a93568 : nir: Update the comment about nir_lower_uniforms_to_ubo()'s multiplier.
    • 1388c25 : radeonsi: inline trivial PS functions
    • 9fc9615 : ac,radeonsi: lower 64-bit IO to 32 bits and remove all dead code
    • e7d6062 : radeonsi: remove swizzle == ~0 dead code in si_llvm_load_input_gs
    • fc44861 : radeonsi: eliminate unused shader outputs for separate NGG geometry shaders
    • 5c445be : radeonsi: set outputs_written_before_ps for geometry shaders too
    • 0b6448b : aco/isel: refactor emit_vop3a_instruction() to handle 2 operand instructions
    • 5b31056 : aco/isel: refactor code and remove unnecessary v_mov
    • 28df8ff : swr: Use ElemenCount constructor for LLVM 11
    • 6049dc1 : aco: improve fsign selection
    • 0e9425a : aco: fix one-off error in Operand(uint16_t)
    • f4210df : gitlab-ci: Enable unit test report for arm64_a630_traces
    • 88fc3e2 : gitlab-ci: Enable unit test reports for lava traces jobs
    • 7e82e25 : gitlab-ci: Enable unit test reports for normal runner traces jobs
    • d24ffd8 : tracie: Produce JUnit XML results
    • dd47192 : tracie: Make tests independent of environment
    • 171e94f : mesa/st: enable EXT_color_buffer_half_float when formats supported
    • 8447b0f : mesa: add EXT_color_buffer_half_float plumbing
    • 331e5f1 : mesa: refactor floating point texture fbo completeness check on gles
    • bb42dee : egl/android: HAVE_DRM_GRALLOC path fixes (v2)
    • 9d02d65 : android: freedreno/common: add libmesa_git_sha1 static dependency
    • cc86a13 : android: freedreno: Implement pipe screen's get_device/driver_uuid()
    • 9b34918 : radeonsi: Work around Wasteland 2 bug.
    • e4d75c2 : nir/opt_shrink_vectors: shrink image stores using the format
    • bd42816 : nir/lower_io: Fix the unknown-array-index case in get_deref_align
    • 11bd64b : ci: Remove any existing results directory before running piglit
    • 013a2b1 : spirv2nir: Rework argument handling
    • a5e427f : spirv: Improve the "Entry point not found" error message
    • 2c94a97 : swr: fix build with mingw
    • c56fbed : radeonsi: kill point size VS output if it's not used by the rasterizer
    • 2345051 : radeonsi: deduplicate setting key.mono.u.vs_export_prim_id
    • 1dd243d : radeonsi: use shader_info::cs::local_size_variable to clean up some code
    • 757f790 : radeonsi: remove redundant si_shader_info::uses_derivatives
    • f3f08bc : radeonsi: remove redundant si_shader_selector::max_gs_stream
    • 2b4fa68 : radeonsi: remove redundant GS variables in si_shader_selector
    • 7960668 : radeonsi: remove redundant si_shader_info::writes_memory
    • 83cdffd : radeonsi: rename num_memory_instructions - num_memory_stores
    • c8ab589 : radeonsi: reduce type sizes in si_shader_selector
    • 99c4e61 : radeonsi: remove redundant si_shader_info::uses_kill
    • 8df349a : radeonsi: merge uses_persp_opcode_interp_sample/uses_linear_opcode_interp_sample
    • 7b3e24c : radeonsi: remove unused si_shader_info::uses_(vertexid|basevertex)
    • f02cd0e : radeonsi: remove redundant si_shader_info:*(clip|cull)* fields
    • d15a7d1 : radeonsi: remove redundant si_shader_info::const_buffers_declared
    • 0dabcb9 : radeonsi: remove redundant si_shader_info::images_declared
    • c1af2f4 : radeonsi: remove redundant si_shader_info::shader_buffers_declared
    • cb63e4a : radeonsi: remove info::samplers_declared, image_buffers, msaa_images_declared
    • cb7bc98 : radeonsi: stop using TGSI_PROPERTY_FS_COLOR0_WRITES_ALL_CBUFS
    • 89cf878 : radeonsi: stop using TGSI_PROPERTY_CS_LOCAL_SIZE
    • ebe8238 : radeonsi: stop using TGSI_PROPERTY_FS_DEPTH_LAYOUT
    • ce0291d : radeonsi: stop using TGSI_PROPERTY_FS_COORD_PIXEL_CENTER
    • c597e39 : radeonsi: stop using TGSI_PROPERTY_FS_POST_DEPTH_COVERAGE
    • 5aeaf3d : radeonsi: stop using TGSI_PROPERTY_FS_EARLY_DEPTH_STENCIL
    • ba7b87a : radeonsi: stop using TGSI_PROPERTY_CS_*
    • e627528 : radeonsi: stop using TGSI_PROPERTY_GS_*
    • a8e7024 : radeonsi: stop using TGSI_PROPERTY_TES_VERTEX_ORDER_CW
    • eeea2f5 : radeonsi: stop using TGSI_PROPERTY_TES_SPACING
    • 34b8e60 : radeonsi: stop using TGSI_PROPERTY_TES_POINT_MODE / TES_PRIM_MODE
    • af59f06 : radeonsi: stop using TGSI_PROPERTY_TCS_VERTICES_OUT
    • a37d897 : radeonsi: stop using TGSI_PROPERTY_VS_WINDOW_SPACE_POSITION / VS_BLIT_SGPRS_AMD
    • 46bb051 : radeonsi: stop using TGSI_PROPERTY_NEXT_SHADER
    • 4c5f5d2 : lima/ppir: Skip instruction merge when having more than one successor
    • 927a395 : swr: Fix crashes on non-AVX hardware
    • afa1fba : vulkan/device_select: don't pick a cpu driver as the default
    • be31398 : android: nv50/ir: Add nv50_ir_prog_info_out serialize and deserialize
    • d853929 : ci: Run deqp-gles3 and deqp-gles31 on RadeonSI
    • 3d5bed0 : radeonsi: fix user fence space when MCBP is enabled
    • c638301 : radeonsi: fix syncobj wait timeout
    • be28183 : android: util: fix missing include path
    • fb42d73 : android: amd/registers: switch to new generated register definitions
    • e8d55e6 : ac/llvm: fix b2f for v2f16
    • d9a77f9 : ac/llvm: add better code for fsign
    • ca74603 : ac/llvm: add better code for isign
    • cecb6d1 : ac/llvm: remove dead code handling for fmod
    • 44528f5 : ac/llvm: fix bcsel for v2*16
    • 735c390 : ac/llvm: fix amdgcn.rsq for v2f16
    • dd8561f : ac/llvm: fix amdgcn.fract for v2f16
    • ab614ab : ac/llvm: fix amdgcn.rcp for v2f16
    • 84500ee : ac/llvm: remove stub prototype for fmed3
    • f852942 : Revert "ac: generate FMA for inexact instructions for radeonsi"
    • a407123 : radeonsi: move nir_shader_compiler_options into si_screen
    • 1476bea : radeonsi: remove redundant no-signed-zero-fp-math LLVM attribute
    • 4242073 : egl/x11: Free memory allocated for reply structures on error
    • 4b7f93b : glsl_to_nir: fix crashes with int16 shifts
    • 57fba85 : spirv: fix emitting switch cases that directly jump to the merge block
    • 11fbd98 : compiler/types: Fix deserializing structs with = 15 members
    • 50ff8a7 : freedreno/regs: add 7nm DSI PHY/PLL regs
    • 1d018bc : gallivm: add InstSimplify pass
    • 64b0b7c : intel/compiler: fix typo in a comment
    • 95ce619 : intel/compiler: print dispatch width when shader fails to compile
    • e5f735a : intel/compiler: move extern C functions out of namespace brw
    • d4c6e3f : intel/compiler: use the same name for nir shaders in brw_compile_* functions
    • 0dda209 : intel/compiler: match brw_compile_* declarations with their definitions
    • a7ece63 : nir/algebraic: add 16-bit versions of a few 32-bit patterns
    • 00b28a5 : nir/algebraic: trivially enable existing 32-bit patterns for all bit sizes
    • ac55b1a : nir: get ffma support from NIR options for nir_lower_flrp
    • efd439e : iris: Better determine map_would_stall for Z/S
    • 2c30079 : iris: Drop a use of the need_resolve boolean
    • 5b82d8c : iris: Avoid resolving Z/S reads in transfer_map
    • 2982d7c : iris: Make iris_has_color_unresolved more generic
    • e7908a9 : iris: Fold a condition into no_gpu for consistency
    • 73eb24a : aco: handle unaligned loads on GFX10.3
    • a083560 : nir/load_store_vectorizer: Add unit tests for alignment handling.
    • c6f7219 : nir/load_store_vectorizer: Use more imm helpers in the tests.
    • 91a8a07 : nir/load_store_vectorize: rework alignment calculation
    • 8faf85f : aco: fix byte_align_scalar for 3 dword vectors
    • 663c4d5 : intel/fs: add hint how to get more info when shader validation fails
    • ab07926 : intel: add INTEL_DEBUG=shaders
    • 7acc7ec : ac/llvm: fix unaligned VS input loads on gfx10.3
    • aaa492e : nir: fix const-cast warning on MSVC
    • 0f4d094 : meson: drop leftover PTHREAD_SETAFFINITY_IN_NP_HEADER
    • ebf2576 : radv,aco: disable opts if VK_PIPELINE_CREATE_DISABLE_OPTIMIZATION_BIT
    • bb80ed8 : panfrost: Fix gnu-empty-initializer errors.
    • 3cf6325 : iris: Fix aux assertion in resource_get_handle
    • fd2ee49 : ci/bare-metal: Use python for handling fastboot booting and parsing
    • 0453a46 : ci/bare-metal: Fix capturing of serial output as job artifacts.
    • 24f5f11 : ci/bare-metal: Log why our run restarts when it does.
    • 785d3ca : ci/bare-metal: Include a timestamp in our serial reads.
    • ff42b7e : ci/bare-metal: Fix detection of "POWER_GOOD not seen in time" fails
    • b7787ce : ci/bare-metal: Use re.search() instead re.match() for our line matching.
    • 2182bbf : aco: Fix integer overflows when emitting parallel copies during RA
    • f18fc34 : radv: Fix various non-critical integer overflows
    • a99ae19 : aco: remove omod_success/clamp_success
    • dc79152 : aco: fix mad splitting after applying output modifiers
    • fdadbdb : radv: remove descriptor_indexing fails from expected fails
    • 21fbffc : anv: Set alignments on UBO/SSBO root derefs
    • c7dec05 : spirv: Drop the OpenCL type layout code
    • 796d3fe : clover/nir: Use lower_vars_to_explicit for uniform and global
    • e1ed5a1 : spirv: Stop counting inputs in entry_point_wrapper
    • 8bea5aa : clover: Use args.size() to compute new var locations
    • 9641f48 : nir: Allow uniform in nir_lower_vars_to_explicit_types
    • 3719b69 : nir: Allow var_mem_global in nir_lower_vars_to_explicit_types
    • 3135984 : spirv: Propagate alignments to deref chains via casts
    • 207b462 : spirv: Add pointer helper vars to OpCopyMemory
    • e7fbec0 : nir/opt_deref: Remove restrictive alignment information from casts
    • 99bb2a4 : nir/opt_deref: Don't remove casts with alignment information
    • beefd37 : nir/lower_io: Apply alignments from derefs when available
    • 480329c : nir: Add a helper for getting the alignment of a deref
    • 0654a9e : nir: Handle all array stride cases in nir_deref_instr_array_stride
    • 878a8da : nir: Add alignment information to cast derefs
    • a0b82c2 : nir/glsl: Add an explicit_alignment field to glsl_type
    • ca11b17 : compiler/types: Make booleans 32-bit for cl_size/align
    • 3be890a : nir: Expose the packed attribute attached to glsl_type objects
    • 42f97f8 : nir/glsl: Consider block interfaces as structs when it comes to size/align calculation
    • be68de8 : glsl: Propagate packed info in get_explicit_type_for_size_align()
    • d0c2958 : spirv: Propagate packed information to glsl_type
    • 938d6ce : spirv: Don't accept CPacked decoration on struct members
    • bcfeead : clover: Call nir_lower_mem_constant_vars
    • 9414cbc : nir: Don't bail too early in lower_mem_constant_vars
    • a7a0315 : intel/nir: Stop using nir_lower_vars_to_scratch
    • 38a83a3 : nir/lower_indirect_derefs: Add a threshold
    • c897cd0 : intel/compiler: Handle all indirect lowering choices in brw_nir.c
    • 9005c9c : zink: generically handle matrix types
    • 2c5c55e : gallium/util: use uint sampler for stencil-reads
    • 98e866c : radeonsi: optimize out the loop in si_get_ps_input_cntl
    • 6ecb8b6 : radeonsi: replace TGSI_SEMANTIC with VARYING_SLOT and FRAG_RESULT
    • a803008 : radeonsi: replace TGSI_INTERPOLATE with INTERP_MODE
    • 99fe3ef : compiler: add INTERP_MODE_COLOR for radeonsi
    • 6925401 : radeonsi: remove si_shader_selector::type
    • 23fdb79 : radeonsi: change PIPE_SHADER to MESA_SHADER (si_dump_descriptors)
    • 9663079 : radeonsi: precompute si_*_descriptors_idx in si_shader_selector
    • cbfc13b : radeonsi: change PIPE_SHADER to MESA_SHADER (si_shader_dump_disassembly)
    • 62aaa0d : radeonsi: remove unused si_shader_context::type
    • 5151421 : radeonsi: change PIPE_SHADER to MESA_SHADER (si_get_shader_part)
    • a90d1df : radeonsi: change PIPE_SHADER to MESA_SHADER (si_compile_llvm)
    • 3c54d73 : radeonsi: change PIPE_SHADER to MESA_SHADER (debug flags)
    • b4b323c : radeonsi: change PIPE_SHADER to MESA_SHADER (si_shader_context::type)
    • b1cb72c : radeonsi: change PIPE_SHADER to MESA_SHADER (si_shader_selector::type)
    • 1439153 : radeonsi: simplify handling color interp modes in si_emit_spi_map
    • 562b8c1 : radeonsi: don't execute LDS stores for TCS outputs that are never read
    • 08ee721 : radeonsi: don't lower indirect IO in GLSL
    • f4d0565 : radeonsi: remove in/out/uniform variables from NIR after lowering IO
    • 81d106d : radeonsi: lower IO intrinsics - complete rewrite of input/output scanning
    • 408fc4e : ac/nir: handle all lowered IO intrinsics
    • 44eaee6 : radeonsi: clean up code for loading VS inputs
    • ed9391d : radeonsi: get color interpolation info from shader_info
    • 0464ee7 : radeonsi: don't crash if input_usage_mask is 0 for a VS input
    • bf23ff8 : freedreno: fence_server_sync() fixes
    • aae1e68 : freedreno: Fix missing rsc-seqno updates
    • e953f6c : docs: shift 20.2 rc dates by two weeks to match reality
    • cdb666b : docs: update calendar and link releases notes for 20.1.7
    • aca2bda : docs: add release notes for 20.1.7
    • bbaa62e : iris: Re-emit push constants if we have a varying workgroup size
    • fe18a0f : intel/nir: Lower load_num_work_groups to 32-bit if needed
    • 5799da4 : intel/fs: Use a single untyped surface read for load_num_work_groups
    • 8e8701b : intel/fs: Don't copy-propagate stride=0 sources into ddx/ddy
    • 95415a5 : st/mesa: fix lowered IO - don't call st_nir_assign_vs_in_locations twice
    • 8c43edf : nir: fix a bug in is_dual_slot in nir_io_add_const_offset_to_base
    • 536727c : iris: Patch constant data pointers into shaders
    • 5ee3242 : anv: Patch constant data pointers into shaders with using softpin
    • a06955f : nir/builder: Add load/store_global helpers
    • a5acf3d : anv: Properly cache brw_stage_prog_data::relocs
    • 91becd8 : intel/fs: Add support for a new load_reloc_const intrinsic
    • 8d8a381 : intel/eu: Add a mechanism for emitting relocatable constant MOVs
    • 272ab28 : intel/eu: Include brw_compiler.h in brw_eu.h
    • aff078e : anv: Stop storing the shader constant data side-band
    • 90b6745 : intel/fs,vec4: Stuff the constant data from NIR in the end of the program
    • 91348d1 : intel/eu: Add some new helpers
    • bc2c5f9 : iris: Use gen_disassemble
    • 3727743 : intel/compiler: Get rid of struct gen_disasm
    • 54ba0da : intel/compiler: Get rid of the global compaction table pointers
    • 191f8a4 : spirv: Deal with glslang not setting NonUniform on constructors.
    • 965b844 : spirv: Deal with glslang bug not setting the decoration for stores.
    • 61b714a : radv: Avoid deadlock on bo_list.
    • 6b75262 : radv: Fix threading issue with submission refcounts.
    • e4dadb5 : intel/tools: Disassemble WAIT's argument as a destination
    • 329dee1 : gallium/tgsi_exec: Fix up NumOutputs counting
    • ed745fe : gallium/tgsi_exec: Add missing DFLR opcode support.
    • 9121afe : nir/clone: Add a helper for cloning most instruction types
    • f25e169 : nir/opt_vectorize: Add a callback for filtering of vectorizing.
    • 479d9c9 : nir: Add simplistic lowering for bany_equal/ball_inequal.
    • 9ed3338 : gallium/ureg: Set the next shader stage from the shader info.
    • 500b073 : gallium/tgsi: Add a helper for initializing ureg from a shader_info.
    • 34b3e1f : gallium/tgsi: Add some missing opcodes to tgsi_ureg.
    • 1ca9f22 : gallium/tgsi: Add support for PRIMITIVEID as a system value.
    • 6e7c006 : gallivm: Report the unsupported intrinsic instead of just assert(0);
    • 615f1c8 : ci: Do not create manual test-docs job in post-merge pipelines
    • ed9ac3d : intel/fs,vec4: remove unused assignments
    • 8e8356e : intel/compiler: mark debug constant as const
    • c7a9dc7 : intel/compiler/test: use TEST_DEBUG env var consistently
    • 289cb6b : intel/perf: don't generate logically dead code
    • 0789f4b : i965: verify format width in blorp_get_client_bo
    • 5e676ac : iris: verify color component width in convert_fast_clear_color
    • 0685815 : iris: handle os_dupfd_cloexec failure
    • 612ef74 : freedreno/computerator: Use a render node
    • 87fa645 : intel/compiler: Fix pointer arithmetic when reading shader assembly
    • c601e73 : mesa/st: Don't set alpha if ALPHA_TEST is lowered
    • 075e40e : turnip: Implement VK_EXT_host_query_reset
    • b92be73 : turnip: Support pipeline statistics query
    • 170da45 : turnip: Refactor structs of tu_query
    • 03fbff1 : intel/assembler: Add labels support
    • bc4a127 : intel/disasm: Label support in shader disassembly for UIP/JIP
    • 6cbd476 : intel/disasm: brw_label and support functions
    • afa39d0 : intel/disasm: Change visibility of has_uip and has_jip
    • e94c224 : anv: refresh cached current batch bo after emitting some commands
    • a1c2bd6 : radeonsi: use radeonsi_clamp_div_by_zero for SPECviewperf13, Road Redemption
    • b844552 : radeonsi,driconf: add clamp_div_by_zero option
    • 32f46a5 : ac/llvm: add option to clamp division by zero
    • f8c0e20 : radeonsi: enable PIPE_CAP_NO_CLIP_ON_COPY_TEX
    • d94bec5 : mesa/st: introduce PIPE_CAP_NO_CLIP_ON_COPY_TEX
    • 54fed1c : gallium/swr: Fix compilation TCS/TES compilation issues
    • 639605e : gallium/swr: Fix compilation with LLVM 12
    • abbfc5b : spirv: fix retrieving dest type for OpFragmentMaskFetchAMD
    • 4500e6e : vulkan: make VK_TIME_DOMAIN_CLOCK_MONOTONIC_RAW_EXT conditional
    • 7fbeb26 : radv: remove seccomp includes
    • b30bd6f : util/os_misc: os_get_available_system_memory() for OpenBSD
    • 5b1ed09 : anv: use os_get_available_system_memory()
    • 033dcb2 : util/os_misc: add os_get_available_system_memory()
    • 81b163f : anv: use os_get_total_physical_memory()
    • 0369dd9 : freedreno/a6xx: Add ARB_depth_clamp and separate clamp support.
    • 5c0d34c : ci: Enable KHR-GL30 CTS testing on freedreno a630.
    • ef807a5 : ci: Use the same VK-GL-CTS tree for GL/GLES as VK.
    • eb02b31 : ci: Bump vulkan CTS version to 1.2.3.2, and keep the GL CTS around.
    • b9df3fe : panfrost: Drop PIPE_CAP_MAX_COMBINED_HW_ATOMIC_COUNTER/BUFFERS
    • d7b6d2e : panfrost: add support for atomics
    • 8e221f5 : panfrost: add atomic ops infrastructure
    • 703b03b : panfrost: add support for src[3] in LOAD_STORE ops
    • f611900 : panfrost: introduce LDST_ATOMIC property
    • 16664fc : panfrost: add LDST_ADDRESS property to atomic ops
    • 8e8805b : panfrost: add atomic_cmpxchg opcode
    • 1101261 : panfrost: fix undefined value access on mir_set_intr_mask()
    • 526f356 : clover: Use 64-bit offsets for shader_in on 64-bit GPUs
    • c93ade9 : nir/lower_explicit_io: Assert that compute address sizes match derefs
    • 6678f1e : spirv: Drop the constant_as_global as option
    • 26a4c8f : clover/nir: Use nir_var_mem_constant for __constant memory
    • dfa63f2 : llvmpipe: Add support for load_global_constant
    • 4087b53 : nouveau/nir: Implement load_global_constant
    • ff2f44d : intel/fs: Implement nir_intrinsic_load_global_constant
    • 1bdf850 : spirv: Use nir_var_mem_constant for UniformConstant data in CL
    • bb8d8ba : nir: Allow opt_large_constants to be run with constant_data_size 0
    • 4360a8a : nir/lower_io: Add support for nir_var_mem_constant
    • ef142c6 : nir/lower_io: Add a build_addr_for_var helper
    • 965c268 : nir/lower_io: Use the variable mode for load_scratch_base_ptr checks
    • ff124e3 : nir: Add a load_global_constant intrinsic
    • e4f07f8 : nir: Add a new nir_var_mem_constant variable mode
    • b9927c8 : nir/serialize: fix serialization of system values
    • 5c45eaf : clover/spirv: fix vec3 alignment
    • 6e52c6d : nvc0/cl: hande 64 bit pointers in nvc0_set_global_handle
    • 5eacaa9 : spirv: fix 64 bit atomic inc and dec
    • 1c9efcd : nvc0/ir: fix load propagation for sub 4 byte addressing
    • 036f1c2 : clover/llvm: undefine __IMAGE_SUPPORT__ for devices without image support
    • 7dc3983 : clover/nir: use offset for temp memory
    • d7b65cf : nv50/ir: fix cas lowering for 64 bit
    • d421af3 : clover/nir: Lower function_temp to scratch.
    • 8e259da : egl/x11: simplify dri2_initialize_x11()
    • bd385c4 : egl/wayland: simplify dri2_initialize_wayland()
    • f7e0cdc : egl/surfaceless: simplify dri2_initialize_surfaceless()
    • 83b5c1a : egl/android: simplify dri2_initialize_android()
    • cccb497 : intel/fs: Fix MOV_INDIRECT and BROADCAST of Q types on Gen11+
    • 70cbddc : nir: use enum operator helper for nir_variable_mode and nir_metadata
    • 76a1fb3 : util: add helpers to define bitwise operators on enums for C++
    • 24dfd79 : nir: use nir_var_all to get rid of casting
    • 15da983 : nir: Improve the comment on num_inputs and friends
    • 4d18e71 : nir: Rename num_shared to shared_size
    • 471f260 : spirv: Delete some dead workgroup variable handling code
    • 1459cce : spirv: add some tests for volatile/available/visible
    • fb6b243 : spirv: Support big-endian strings
    • a8d8fbb : spirv: implement SpvMemoryAccessVolatileMask
    • f81e1d2 : spirv: implement Volatile image operand
    • 2ba3ffa : spirv: implement Volatile memory semantic
    • 23bfba8 : spirv: implement MakePointerAvailable/MakePointerVisible for OpCopyMemory
    • e01d1a9 : spirv: add vtn_emit_make_{visible,available}_barrier helpers
    • b85c38a : spirv: make OpLoad/OpStore visibility/availablity barriers acquire/release
    • 5b92392 : spirv: fix Uniform and Output MemoryAccessMakePointer{Visible,Available}
    • 857b9c5 : spirv: Add a vtn_get_mem_operands() helper
    • 959539f : intel/isl: Drop unnecessary check on 16bpp depth format
    • 829699b : anv: implement shareable timeline semaphores
    • a965ffa : anv: add new gem/drm helpers
    • a0c07e4 : include/drm-uapi: bump headers
    • a6291b1 : freedreno/ir3: rework setup_{input,output} to make struct varyings work
    • c694af4 : freedreno/ir3: improve handling of aliased inputs
    • acb6163 : freedreno/ir3: remove indirect input load
    • 0097354 : radv: Allow triggering thread traces by file.
    • 0d862da : radv: Centralize enabling thread trace.
    • ec5e918 : vulkan/wsi/x11: wait for acquirable images in FIFO mode
    • d0bc1ad : vulkan/wsi/x11: add sent image counter
    • 68957a8 : aco: Add README which explains about what ACO is and how it works.
    • b4d4c54 : aco: Fixup markdown formatting of the README-ISA.
    • 086fafc : aco: Move README to README-ISA
    • bc123c3 : nir/algebraic: mark some optimizations with fsat(NaN) as inexact
    • a6c4db9 : ci: Restrict "success" job to pipelines for MRs
    • d383133 : docs: Stop claiming to implement OpenVG
    • 0308e63 : ci: Don't exclude "success" job from mesa/mesa pipelines for MRs
    • f13f32f : ci: Create test-docs job in mesa/mesa pipelines for MRs
    • 52cac06 : gallivm: fix build on LLVM 12 due to LLVMAddConstantPropagationPass removal
    • 8b1ab78 : radv: dump GPU info into the hang report
    • aa675cd : radv: improve reporting faulty pipelines when a GPU hang is detected
    • dd1f035 : radv: force RADV_DEBUG=syncshaders when RADV_TRACE_FILE is used
    • 004119d : vulkan: Fix memory leaks.
    • 7036b38 : ANDROID: Bring back gbm target, renamed to gbm_mesa
    • 50e14c3 : radv/winsys: Fix memory leak.
    • b778e7b : nir/vtn: Convert constant samplers to variables with data
    • 397730e : nir/vtn: Add intrinsics for CL image format/order queries
    • 5ce6821 : nir/vtn: ImageSizeLod op can be applied to images
    • 701cb9d : nir/vtn: Handle integer sampling coordinates
    • a196f05 : nir/vtn: Use return type rather than image type for tex ops
    • de36b5b : nir/vtn: Add support for kernel images to SPIRV-to-NIR.
    • ce6f662 : nir/vtn: Add type constant to image intrinsics
    • 865a2ad : clover/nir/spirv: Use uniform rather than shader_in for kernel inputs
    • 41d0a81 : pan/mdg: Fix perspective combination
    • c8ac01a : pan/mdg: Fix discard encoding
    • 9efdbb2 : anv: fix robust buffer access
    • 73d2c6c : nir: Switch the indexing of block-live_in/out arrays.
    • 5d2b2b5 : nir: Invalidate live SSA def information when making new SSA defs.
    • d3b652f : ci: Switch to using gold as the linker.
    • 8c4fb19 : pipe-loader: Use real galliumvl if radeonsi is being linked.
    • 221aa00 : turnip: Make sure we include the build id.
    • 2d1c607 : ci: Make a missing device name correctly bail out of deqp-runner.sh.
    • ee288f2 : nir: add shared/global atomics to nir_get_io_offset_src()
    • e281ee0 : nir: Look up the shader when printing a single instruction.
    • a3a8322 : nir: Fix printing of individual instructions with io semantics.
    • 4c050f2 : panfrost: enable DrawTransformFeedback*
    • e8b3bc1 : intel/nir: Lower things with 4 components in lower_mem_access_bit_sizes
    • 55ae704 : intel/fs: Add support for vec8 and vec16 ops
    • b6a013c : anv: VK_INTEL_performance_query interaction with VK_EXT_private_data
    • 5d52c26 : vulkan: Don't pointlessly depend on libxcb-dri2
    • 8239fe5 : zink: add note about buffer-image copy functions not handling multisample
    • d84a754 : zink: use u_transfer_helper to split/merge interleaved depth/stencil formats
    • 7059708 : broadcom/qpu_instr: wait is not a read or write vpm instruction
    • efd29d4 : docs/features: Add missing Panfrost extensions
    • 8ff6e7c : panfrost: Set PIPE_CAP_MAX_TEXTURE_GATHER_COMPONENTS
    • 7dab574 : pan/mdg: Implement texture gathers
    • 661123b : pan/mdg: Add disassembly for shadow gathers
    • f6e19dd : pan/mdg: Refactor texture op/mode handling
    • 80ebf11 : pan/mdg: Implement nir_intrinsic_load_sample_mask_in
    • 41184f8 : pan/mdg: Bounds check swizzle writing globals
    • 2486fe6 : pan/mdg: Scalarize 64-bit
    • 3e2cb21 : pan/mdg: Implement i/umul_high
    • fd41dec : panfrost: Emit texture/sampler points for compute
    • 55765f8 : util/u_thread: include pthread_np.h if found
    • c66c5b3 : util: futex fixes for OpenBSD
    • 0398caa : meson: conditionally include -ldl in gbm pkg-config file
    • c97af23 : meson: don't build with USE_ELF_TLS on OpenBSD
    • 9ac1686 : meson: don't advertise TLS support if glx wasn't build with it
    • f9a7e6e : meson: build with _ISOC11_SOURCE on OpenBSD
    • 6e9c066 : util/anon_file: add OpenBSD shm_mkstemp() path
    • 7eab684 : util: unbreak endian detection on OpenBSD
    • 8301a43 : radv: dump shader stats with VK_KHR_pipeline_executable_properties
    • 35a42ff : vallium: disable VK_KHR_shader_float16_int8.
    • 8f6eb35 : gallivm/nir: add some f16 support
    • b31e846 : gallivm/nir: allow 64-bit arit ops
    • dc6313b : llvmpipe/fs: multisample depth/stencil bad ir generated
    • 6d60cd4 : llvmpipe: add array/3d clearing support
    • 2d63857 : llvmpipe: blend has effects even if no colorbuffers.
    • 727bb21 : gallivm: use common code to select texel type
    • 6596957 : gallivm/nir: add subpass sampler type support
    • 21324c4 : gallivm/nir: lower frexp/ldexp
    • d7aed02 : gallivm/nir: fix const compact
    • 96a5ced : gallivm/nir: add quantize to f16 support
    • 1d4a560 : gallivm/nir: add indirect swizzle output loading support
    • 5177ffe : gallivm/nir: lower tg4 offsets.
    • 87a638d : gallium/nir/tgsi: fix nir-tgsi info conversion for samplers/image
    • 668e4c2 : vallium: limit buffer allocations to gallium max.
    • 8570088 : vallium: handle 3D image views properly.
    • 99330e5 : llvmpipe: add reference counting to fragment shaders.
    • ac6b8e4 : nir: Take a variable remap parameter in nir_inline_function_impl
    • b6f3189 : nir: Use a switch in nir_inline_function_impl
    • b7db9ee : intel/nir: Clean up lower_alpha_to_coverage a bit
    • b6fdb14 : intel/nir: Rewrite the guts of lower_alpha_to_coverage
    • 72dc06e : intel/nir: Pass the nir_builder by reference in lower_alpha_to_coverage
    • 373e25e : ci: Add a release build with -Werror enabled.
    • 60fb460 : virgl: Fix unused var warnings in release build from assertions.
    • c24f792 : lima: Fix uninitialized var warning from using assert() as unreachable().
    • f6456d7 : lima: Fix unused var/function warnings in release build from assertions.
    • 65f4841 : etnaviv: Fix unused var warning in release build from assertions.
    • ab6a09f : zink: Fix unused var warnings in release build from assertions.
    • ca73c3b : nv50: Fix uninitialized var warnings from using assert() as unreachable().
    • 689e36e : vc4: Fix unused var warnings in release builds from assertions.
    • 2647126 : nvc0: Fix compiler warning about unused var that gets asserted.
    • b9972fb : gallium/tests: Fix compiler warning about unused vars in trivial tests.
    • 4c24c82 : panfrost: Fix remaining release-build warnings.
    • ecab580 : panfrost: Fix OOB array access compiler warning.
    • ccb5e0b : drm-shim: Fix unused variable warnings from asserts in release build.
    • 8d1d0c1 : freedreno/a6xx: move ubwc clear to blitter
    • 8d9ab0a : freedreno/a5xx+a6xx: use sysmem path for nondraw batches
    • 21b9070 : freedreno/batch: replace lrz_clear with prologue
    • 2807997 : freedreno/batch: split out helper for rb alloc
    • 40eceb5 : ANDROID: Remove conflicting gbm target
    • 9522eb7 : panfrost: Store transient BOs in a dynamic array
    • d87ab72 : panfrost: Avoid accessing pan_pool fields directly
    • 4047c69 : panfrost: Rename panfrost_create_pool() into panfrost_pool_init()
    • b8f2d71 : CI: Disable Windows again
    • 7fbded8 : pan/decode: Drop prefix braces
    • eb261a8 : panfrost: Remove mali_vertex_tiler_postfix
    • 7602891 : pan/decode: Use unpack for vertex_tiler_postfix_pre
    • 2425bac : pan/decode: Use generation for vertex_tiler_postfix
    • 8004906 : pan/decode: Fix awkward syntax
    • c088a3b : pan/decode: Print shader-db even for compute
    • 6855228 : pan/decode: Drop mali_vertex_tiler_postfix arg
    • 38ae088 : pan/decode: Drop scratchpad size dump
    • c92be29 : panfrost: Use nir_builder_init_simple_shader for blits
    • 4467e79 : panfrost: Inline bifrost_tiler_only
    • 0c1f44b : panfrost: Drop bifrost_payload_fused
    • 02e768e : panfrost: XMLify invocations
    • 0225ed4 : panfrost: Add invocation XML
    • b60d567 : panfrost: XMLify primitive information
    • 3bae493 : panfrost: Drop point sprite from shader key
    • b17b6cc : panfrost: Cleanup point sprite linking
    • 7d328e7 : panfrost: Simplify ZSA bind
    • 90cc815 : panfrost: Use pack for draw descriptor
    • 33895ac : panfrost: Use draw pack for compute jobs
    • e5c77cb : panfrost: Detangle postfix from varying emits
    • 3a4d930 : panfrost: Inline panfrost_vt_set_draw_info
    • 3f61da7 : panfrost: Inline panfrost_vt_init
    • 136fd5c : panfrost: Don't call panfrost_vt_init for compute
    • 1513392 : panfrost: Avoid postfix dep for vertex_data
    • 8b5f9fc : panfrost: Remove postfix parameter from UBO upload
    • 1357eec : panfrost: Inline vt_update_{rasterizer, occlusion}
    • b716936 : panfrost: Separate postfix from emits
    • 59b6e3c : panfrost: Use draw pack for blit
    • d6a5df0 : panfrost: Add XML for mali_vertex_tiler_postfix
    • e213a86 : panfrost: Add padded type for instance fields
    • 45c59db : panfrost: Drop blend indirection
    • 87e3510 : panfrost: Emit explicit REPLACE for disabled colour writeout
    • 75a274d : panfrost: Add opaque midgard_blend XML
    • 0e13f80 : CI: Re-enable VS2019 build
    • a280810 : CI: Windows: Use separate config file for Docker
    • f4c1c21 : egl: simplify eglSwapInterval() fallback logic
    • 1d3fb7a : egl: inline eglSwapInterval() fallback
    • 957bbc6 : anv: simplify push constant emissions
    • 7953402 : anv: move push constant allocation tracking into gfx pipeline state
    • 8076c75 : aco: fix wrong source position for constant with nir_op_cube_face_coord
    • f4d8616 : radv: set BIG_PAGE to improve performance on GFX10.3
    • d6bc0f2 : radv: emit {CB,DB}_RMI_L2_CACHE_CONTROL at framebuffer time
    • 31f75aa : nir/lower_discard_to_demote: Use nir_shader_instructions_pass().
    • 265dcb3 : nir/nir_lower_wrmasks: Use the nir_lower_instructions_pass() helper.
    • c5e64c0 : nir/lower_io_to_scalar: Convert to use nir_shader_instructions_pass().
    • dfb8465 : nir/opt_undef: Convert to use nir_shader_instructions_pass().
    • e5d4bbd : nir/lower_vec_to_movs: Convert to use nir_shader_instructions_pass().
    • a62098f : nir: Add a helper for general instruction-modifying passes.
    • a82f664 : nir/opt_copy_prop_vars: Quiet valgrind warning about overlapping memcpy.
    • 98222db : egl: document which driver hooks are only required by extensions
    • cf663d0 : egl: move extension driver functions after core functions
    • 12c9418 : docs/egl: correct/update DRI2 mention with the shiny new DRI3
    • ab2e597 : docs/egl: add some more documentation
    • 7897c31 : docs/egl: add haiku driver
    • c84d304 : docs/egl: complete list of dri2 platforms
    • 3909e9d : docs/egl: move section around
    • 3704b02 : docs/egl: fix typo
    • 608c87a : nir/vtn: SPIR-V bit count opcodes (core and extension) dest size mismatches nir
    • a54695d : nir: Add bit_count to lower_int64 pass
    • d91f85f : nir: Remove 32bit restriction for uadd_carry optimization
    • 9232887 : nir: Implement mul_high lowering for bit sizes other than 32
    • ea71574 : nir_lower_bit_size: Support lowering ops with differing source/dest sizes
    • 0d595f7 : nir: Relax opt_if logic to prevent re-merging 64bit phis for loop headers
    • 43d22c8 : nir: Add a lowering pass to split 64bit phis
    • 6b1a56b : iris: Drop stale syncobj references in fence_server_sync
    • e98c7a6 : iris: Reorder the loops in iris_fence_await() for clarity.
    • 45793c0 : ci: Fix up rules for post-merge / main project branch pipelines
    • 502b9da : aco: add ACO_DEBUG=novn,noopt,nosched for debugging purposes
    • 4d40a71 : radv: Fix 3d blits.
    • 003ea78 : st/mesa: remove useless code for lowered IO in st_nir_assign_vs_in_locations
    • 8684061 : ci: Test the KHR-GL* CTS cases with softpipe.
    • d2cf6a8 : aco: sink get_alu_src() in bfe lowering
    • 14d748e : aco: fix sgpr ubfe/ibfe if the offset is too large
    • 454bc59 : aco: remove 64-bit SGPR ubfe/ibfe
    • eb3c16e : aco/tests: add tests for long jumps
    • 192b9f4 : aco: shorten disassembly for repeated instructions
    • ae6330d : aco/tests: add test for GFX10 0x3f bug
    • fe2dc41 : aco: create long jumps
    • 156fd58 : aco: reserve 2 sgprs for each branch
    • e8ac145 : aco: keep loop live-through variables spilled
    • 75d6c30 : aco: fix spills_entry heuristic for branch blocks in init_live_in_vars()
    • fc9f502 : aco: fix regclass checks when fixing to vcc/exec with Builder
    • a537c9e : aco: don't fix break condition for break+discard to exec
    • 1a5444b : aco: don't consider the first partial spill if it's the wrong type
    • 8f6a900 : aco: consider branch definitions in spiller
    • cacb388 : radv: print a warning when RADV_TRAP_HANDLER is used
    • 60975eb : etnaviv: Add lock around pending_ctx
    • da660c9 : etnaviv: Remove etna_resource_get_status()
    • 785e270 : etnaviv: Fix disabling early-z rejection on GC7000L (HALTI5)
    • 0d8ae4a : radv: fix setting EXCP_EN for different shader stages
    • 05a0349 : v3d: set instance id to 0 at start of tile
    • 2e8b6f6 : v3d/packet: fix typo on Set InstanceID/PrimitiveID packet
    • 15bdbf3 : freedreno/a5xx: Don't set the VARYING flag for fragcoord-only programs.
    • 36bd3e9 : gallium/dri2: Report correct YUYV and UYVY plane count
    • 7323062 : ci: Mark the rest of compswap as flaky on freedreno.
    • 27e6117 : nir: Report progress properly in nir_lower_bool_to_*
    • 39d0072 : freedreno/a6xx: fix hang with large render target
    • 0f3c12c : freedreno: add env var to override tiles-per-pipe
    • 688ca54 : freedreno: add env var to override GMEM size
    • 758fdb9 : freedreno/gmemtool: add tile_alignw/h and a650
    • a618908 : radeonsi: fix tess levels coming as scalar arrays from SPIR-V
    • e512f2c : mesa: add NV_copy_depth_to_color support for nir
    • da9d721 : nvc0: Add shader disk caching
    • 82dd683 : nv50/ir: Add nv50_ir_prog_info serialize
    • 3e99af6 : nv50/ir: Add prog_info_out print
    • f20a210 : nv50/ir: Add nv50_ir_prog_info_out serialize and deserialize
    • 5ecc06e : nv50/ir: Use a bit field in info_out structure
    • f292499 : nv50/ir: add nv50_ir_prog_info_out
    • fa8e628 : nv50/ir: remove symbol table support for compute shaders
    • 83139ac : radv: Update CTS version.
    • 4e30191 : radv: Remove conformance warnings with ACO.
    • 2a9ffc2 : nir: Move new edgeflag assert into the io_lowered case
    • 2fcfcca : nir: Copy semantics to nir_intrinsic_load_fs_input_interp_deltas
    • 3cda333 : panfrost: Drop mali_shader_meta
    • 3d7ce13 : pan/decode: Use unpacks for state descriptor
    • 1b377c2 : panfrost: Use pack for fragment shaders
    • c86b51a : panfrost: Use opaque pack for vertex shaders
    • 100b15c : panfrost: Add optional opaque packs to GenXML
    • a7b2317 : panfrost: Use pack for Bifrost test state
    • 5b3b2a6 : panfrost: Use pack for blit shaders
    • a36ac2b : panfrost: Inherit default values from structs
    • 754d543 : panfrost: XMLify the rest of shader_meta
    • 80f1d61 : panfrost: Use preuploaded shader descriptors
    • 1e4c49e : panfrost: Upload shader descriptors at CSO create
    • 8c14482 : panfrost: Allocate a state uploader
    • da677a4 : panfrost: Ensure shader-db state is zero-initialized
    • e0a6af9 : panfrost: Simplify shaderless packing
    • 55d9c25 : panfrost: Prepack fragment properties/preload
    • a29fb64 : panfrost: Pack vertex properties when compiling
    • 5393d73 : panfrost: Derive UBO count from shader_info
    • 661b469 : panfrost: XMLify beginning of shader descriptor
    • 3bb61e2 : panfrost: Derive texture/sampler_count from shader
    • d353b15 : panfrost: Quiet pandecode error
    • a0857e9 : panfrost: Support SHADERLESS mode everywhere
    • c9858fb : panfrost: Identify additional SFBD flags
    • 7a95ed2 : panfrost: XMLify Bifrost preload
    • 1574866 : panfrost: Group SFBD code tighter
    • 518537d : panfrost: Drop redundant NULL check
    • acf77cb : panfrost: XMLify bifrost1
    • 1b7d4f1 : panfrost: Remove midgard1 bitfield
    • 0268e84 : panfrost: Simplify bind_blend_state
    • 96a9153 : panfrost: Fold work_count packing for blend shaders
    • 41aad9a : panfrost: Use pack for shaderless
    • 19ded1e : panfrost: Use packs for fragment properties
    • d475412 : panfrost: Pack compute Midgard properties
    • 87c5951 : panfrost: XMLify Midgard properties
    • 68503f3 : panfrost: Group SFBD state together
    • 58ae50f : panfrost: Clean up blend shader errata handling
    • 8073e27 : panfrost: Rename shader emit functions
    • 9a2df30 : panfrost: Specialize compute vs frag shader init
    • 9f83217 : panfrost: Bake the initial tag into the shader pointer
    • 606f05b : panfrost: Clamp shader-uniform_count
    • e451421 : panfrost: Size UBO#0 accurately
    • e5689a5 : panfrost: Combine frag_shader_meta_init functions
    • bf6d548 : panfrost: XMLify blend equation
    • 6beac11 : panfrost: Honour load_dest/opaque flags
    • 919818a : panfrost: Simplify make_fixed_blend_mode prototype
    • 94c9f87 : panfrost: XMLify blend flags
    • bbec4ff : panfrost: Separate shader/blend descriptor emits
    • cfef6de : panfrost: Hoist blend finalize calls
    • 6fd6221 : panfrost: Decode nested structs correctly
    • aac5a55 : pan/decode: Drop legacy 32-bit job support
    • bc6d6fd : scons: bump c++ standard to 14 to match meson
    • 0fd3320 : radv: allocate the TMA BO into 32-bit addr space
    • d243d21 : radv: align the TMA BO size to 256
    • 2dbb872 : gitlab-ci: Test the traces from bgfx
    • 88b935b : mesa: fix formatting of messages printed using _mesa_log
    • fd10cb8 : anv: fix transform feedback surface size
    • 5e4d69e : st/mesa: Fix EGLImageTargetTexture2D for GL_TEXTURE_2D
    • 855afe0 : meson: Fix lmsensors warning message.
    • c6861b9 : freedreno: Drop UNIFORM_BUFFER_OFFSET_ALIGNMENT to 32
    • a79dad9 : nir,amd: remove trinary_minmax opcodes
    • 1fa43a4 : freedreno: handle case of shadowing current render target
    • 743ad7b : freedreno: add debug helper to dump buffers
    • 6f9bb6c : freedreno/a6xx: refactor debug logging
    • 440aab9 : gallium/tgsi: add helper tgsi_get_interp_mode
    • 2519472 : st/mesa: handle lowered IO in st_nir_assign_vs_in_locations
    • 2424485 : st/mesa: don't generate NIR for ARB_vp/fp if NIR is not preferred
    • 493fdcf : st/mesa: don't pass NIR to draw module if IO is lowered
    • 484a60d : nir: generate lowered IO in nir_lower_passthrough_edgeflags
    • 7980f3e : nir: add interpolation qualifiers for color sysvals into shader_info
    • 4d36dce : nir: add shader_info::io_lowered
    • 844167d : nir: properly identify texcoords for lowered IO in nir_lower_drawpixels
    • 2b1ef5d : nir: print IO semantics (v2)
    • 01ab308 : nir: update IO semantics in nir_io_add_const_offset_to_base
    • 17af070 : nir: gather all IO info from IO intrinsics
    • 502abfc : nir: save IO semantics in lowered IO intrinsics
    • ec1fa1d : intel/perf: fix raw query kernel metric selection
    • f1da3bb : intel/perf: store query symbol name
    • a9019d1 : nir/load_store_vectorizer: Clean up unit test swizzle assertions.
    • 2b25240 : freedreno/ir3: Replace our custom vec4 UBO intrinsic with the shared lowering.
    • 7361659 : nir: Add a lowering pass for backends wanting load_ubo with vec4 offsets.
    • d84a16e : nir: Add nir_[iu]shr_imm and nir_udiv_imm helpers and use them.
    • e78b887 : nir: Make the nir_builder *_imm helpers consistently handle bit size.
    • f820dde : aco: Fix convert_to_SDWA when instruction has 3 operands.
    • 0d194a7 : aco: Fix unused variable warning by adding ASSERTED.
    • 089bc22 : r300: Use util_blend_factor_uses_dest
    • 7a776f4 : si: Use util_blend_factor_uses_dest
    • e055463 : gallium: Add util_blend_uses_dest helper
    • e059305 : gallium: Add util_blend_factor_uses_dest helper
    • 48a910d : radv: use the trap handler to detect faulty shaders/instructions
    • 8e97a61 : radv: enable the trap handler and configure the shader exceptions
    • af3230e : radv: add initial trap handler support with RADV_TRAP_HANDLER=1
    • 8fd2f5c : radv: add a small interface for creating the trap handler shader
    • a0814a8 : aco: skip unnecessary compiler pass for the trap handler program
    • 9c46e6f : aco: add a helper for building a trap handler shader
    • a6146aa : aco: validate that SMEM operands can use fixed registers
    • baa9268 : aco: add TBA/TMA/TTMP0-11 physical registers definitions
    • 5e841e8 : nir: add iabs-lowering code
    • 9e5e3be : egl: drop invalid shebang
    • 32dc71b : clover: Call clang with -O0 for the SPIR-V path
    • e3e45e2 : clover/spirv: Don't call llvm::regularizeLlvmForSpirv
    • c84e278 : intel/nir: Allow splitting a single load into up to 32 loads
    • febe762 : intel/fs: Fix an assert in load_scratch
    • 63dd1e9 : iris: Always re-upload sysvals when we have kernel inputs
    • b4b39b8 : iris: Normalize all compute shaders to MESA_SHADER_COMPUTE
    • d0a8ad7 : iris: ref/unref the GLSL type singleton in screen_create/destroy
    • b9678aa : iris: Stop advertising clover-only caps
    • 6ec2869 : iris: Stop advertising PIPE_SHADER_IR_NIR_SERIALIZED
    • 4fd2a45 : clover/nir: add support for global invocation id offsets
    • 47e52d8 : nir: fix nir_variable_create for kernels
    • 368e9a0 : spirv: Use new global invocation offset system value
    • d3faac7 : nir: Add options to nir_lower_compute_system_values to control compute ID base lowering
    • 2e1df6a : nir: Move compute system value lowering to a separate pass
    • 5807414 : compiler/nir: make lowering global-id to local-id optional
    • 41e4eb9 : nir: Add new system values and intrinsics for dealing with CL work offsets
    • 6b1515c : nir: Populate some places where existing system values were missing
    • 2da1178 : ci/bare-metal: Try rebooting chezas again if they get stuck during tftp.
    • c27075e : ci/bare-metal: Retry booting chezas instead of failing when !POWER_GOOD
    • c636481 : ci/bare-metal: Convert the main cros-servo boot code to python
    • b437408 : ci/bare-metal: Use a new serial buffer tool.
    • ca7d66e : virgl: update protocol headers
    • 42e29fe : virgl: move protocol headers to a common place
    • a4c708d : util/ralloc: fix ralloc alignment.
    • fd7e1ed : svga: fix draw elements with 8-bits indices
    • f41848a : winsys/svga: fix display corruption after surface_init
    • 090239e : st/mesa: increase size of gl_register_file bitfields
    • e5899c1 : nir: rename nir_op_fne to nir_op_fneu
    • 85b7403 : bifrost: use nir_intrinsic_has_type
    • 2133e64 : aco: use nir_intrinsic_has_access
    • 7530f66 : nir: add and use nir_intrinsic_has_ helpers
    • 9c1e0d8 : aco: fix non-rtz pack_half_2x16
    • e802bff : glsl: Eliminate out-of-bounds triop_vector_insert
    • 5922d57 : glsl: Eliminate assigments to out-of-bounds elements of vector
    • e93979b : ir_constant: Return zero on out-of-bounds vector accesses
    • b243a74 : anv: support fd==-1 in ImportSemaphoreFdKHR
    • f153151 : aco: add ACO_DEBUG=force-waitcnt to emit wait-states
    • 258ef4d : wgl: Switch to Win10 version defines to enable usage of Win10 WGL callbacks
    • f811ddf : radv: ignore BB labels when splitting the disassembly string
    • 89b56f1 : anv: take depth in to account in anv_GetImageSubresourceLayout
    • 3dd1a81 : iris: Add missing newline to debug log message
    • fb525ad : ci/freedreno: List more common flakes reported recently.
    • c19b7fc : ci/freedreno: Move our skips lists over to being known-flakes lists.
    • 30da82c : ci/deqp-runner: Add a post-deqp-run filter list for known flakes.
    • 4bb59fc : ci/deqp-runner: Drop unused "count" variable
    • 32fd2ee : ci/deqp-runner: Drop stale comment from deqp-runner.sh.
    • 42d7bbf : nir: Use 'unsigned' instead of enum types in nir_variable::data
    • 1ccd681 : nir: Add an LOD parameter to image_*_size
    • a104902 : broadcom/compiler: Enable PER_QUAD for UBO and SSBO loads.
    • c3258f9 : broadcom/compiler: Add a constant folding pass after nir_lower_io
    • bd87cda : broadcom/compiler: support nir_intrinsic_load_sample_id
    • adda97c : clover/spirv: pass list of supported extensions to the translator
    • a624fae : clover/nir: Register callback for translation messages (v2)
    • 2402466 : clover/llvm: Use the highest supported SPIR-V version (v4)
    • 6ed8759 : meson: Raise minimum version for SPIR-V OpenCL deps (v4)
    • ec6bad1 : clover/spirv: Print linked SPIR-V module if asked
    • a9ca737 : clover/spirv: Remove unused tuple header
    • e9ddb9b : clover/spirv: rework handling of spirv extensions
    • 1cfbc5c : panfrost: Fix alignment on Bifrost
    • b708a1a : tu: Enable VK_KHR_multiview
    • c0c7dbd : tu: Implement multiview pipeline state
    • c884afc : tu: Add multiview lowering pass
    • 7b53ac1 : tu: Implement multiview query interactions
    • ff5f460 : tu: Improve timestamp queries
    • 6c446fe : tu: Implement multiview clear/resolve interactions
    • 99a87e5 : tu: Parse multiview render pass info
    • f01a0dc : tu: Translate VkRenderPassMultiviewCreateInfo to VkRenderPassCreateInfo2
    • 5ef960e : ir3: Add support for gl_ViewIndex in VS & FS
    • 4b163ff : freedreno/a6xx: Add multiview registers
    • 6a8aaf7 : freedreno/a6xx: disable LRZ when color channels are masked
    • 4de027d : freedreno/cffdump: add arg to filter by process name
    • b440c28 : nir: Shrink store intrinsic num_components to the size used by the writemask.
    • 5f26c21 : nir: Expand opt_undef to handle undef channels in a store intrinsic.
    • a29b7b6 : nir/opt_undef: Handle a couple more normal store intrinsics.
    • b3c822a : radv: Move nir_opt_shrink_vectors() into the opt loop.
    • 94f4ecb : panfrost: Reduce attribute buffer allocations
    • 9a6934d : panfrost: Don't reserve for NPOT w/o instancing
    • 09ea7c0 : panfrost: Don't overallocate attributes
    • 3df90c7 : panfrost: Fix attribute buffer underallocation
    • 373a204 : panfrost: Pass alignments explicitly
    • 1cb47f8 : panfrost: Free batch-dependencies
    • 680fb05 : panfrost: Use memctx for sysvals
    • 8dd38e5 : pan/mdg: Free previous liveness
    • 62637a9 : panfrost: Free hash_to_temp map
    • da6d0e3 : panfrost: Free NIR of blit shaders
    • 9146f59 : panfrost: Free cloned NIR shader
    • ba4fc67 : panfrost: Fix blend leak for render targets 5-8
    • 4ff4b04 : panfrost: Keep finalized blend state constant
    • 93200ce : panfrost: Drop depth-only case in blend finalize
    • 8249e2b : panfrost: Explicitly handle nr_cbufs=0 case
    • dc7fbe1 : panfrost: Drop implicit blend pooling
    • d8deb1e : panfrost: Share tiler_heap across batches/contexts
    • 01d12c3 : pan/decode: Don't try to dereference heap mapping
    • b41692c : panfrost: Avoid minimum stack allocations
    • 17c617c : panfrost: Introduce invisible pool
    • 40c0d7a : panfrost: Pre-allocate memory for pool
    • 3fed1c7 : iris: Fix headerless sampler messages in compute shaders with preemption
    • 58817bd : aco: fix file leak in ra_fail()
    • 74ac8f3 : nir/opt_large_constants: Fix a type/deref_type typo
    • e5bac3c : gitlab: ask for more detailed info about GPU
    • b553c7d : radv: fix null memcpy and zero-sized malloc
    • b50ae77 : radv: align pipeline cache entry and header sizes
    • 4f08af6 : radv: don't pass null to _mesa_sha1_update
    • 30fca3b : nir: fix memory leak in nir_cf_list_clone
    • 27ec38d : nir: fix potential left shift of a negative value
    • 391eeb7 : util: add a alignof() macro
    • 721cb6c : gitlab: ask inxi output to be in code blocks
    • e901b90 : radv,aco: report ACO errors/warnings back via VK_EXT_debug_report
    • c2b1978 : aco: rework the way various compilation/validation errors are reported
    • bc723df : aco: rename DEBUG_VALIDATE to DEBUG_VALIDATE_IR
    • 709dffa : anv: null check for buffer before reading size
    • 8d38b25 : util: Explicitly call the unpack functions from inside bptc pack/unpack.
    • edf0aeb : util: Expose rgba unpack/fetch functions as external functions as well.
    • 9d503b3 : iris: Drop buffer support in resource_from_handle
    • c54bb23 : iris: Add and use iris_resource_configure_main
    • 51e42e2 : iris: Move size/offset calculations out of configure_aux
    • 8b2fc91 : iris: Drop old comment on clear color BO allocation
    • fd3c206 : iris: Drop unused resource allocation optimization
    • 2555321 : iris: Drop iris_resource_alloc_separate_aux
    • 0dc0a79 : egl: drop another indentation level in _eglFindDisplay() by inverting an if
    • 8b2fc1d : egl: drop an indentation level in _eglFindDisplay() by replacing break/if with a goto
    • b5d36e9 : gitlab-ci: fix quoting of variables passed down to bare-metal runners
    • 72fac11 : gitlab-ci: fix testing whether a variable with a given name is set or not
    • 64fb3e6 : docs: update calendar and link releases notes for 20.1.6
    • 8be321f : docs: add release notes for 20.1.6
    • bd38ea7 : v3d/compiler: add v3dv_prog_data_size helper
    • a880f97 : compiler/types: Allow interfaces in get_explicit_type_for_size_align
    • df95963 : nir/large_constants: Handle incomplete derefs
    • 9f3c595 : nir/find_array_copies: Handle cast derefs
    • a0a0bcf : nir/builder: Add a nir_iand_imm helper
    • ac95bb4 : nir: Initialize nir_ssa_def::live_index
    • 8f7784e : clover/nir: Use the correct address mode for shared
    • b2226f7 : clover/nir: Stop computing the global address format twice
    • 884d202 : clover/nir: Stop setting ubo_addr_format
    • 5ef9c11 : nv50/ir/nir: support load_work_dim
    • 080bee3 : nvc0: handle nr being 0 in nvc0_set_global_bindings
    • b283fb3 : nv50/ir/nir: fix global_atomic_comp_swap
    • 4403176 : clover/nir: support int64 atomics if the device supports it
    • 38e904f : nv50/ir/nir: assert on unknown alu ops
    • ada8c76 : nve4: fix uploading unaligned sized input buffers
    • 918e444 : clover/nir: Call vars_to_explicit_types for shared memory
    • 1e7c2d1 : nir/glsl: Add glsl_get_cl_type_size_align helper
    • a27823e : freedreno/ir3: Fix assertion failures dumping CS high full regs.
    • ce335dc : freedreno/cffdec: When .mergedregs is set, don't count half regs.
    • 803ec06 : freedreno/ir3: Fix compiler warning from the setjmp fails path.
    • c1a6e34 : nir/lower_input_attachments: Support loading layer id via gl_ViewIndex
    • 76f711d : tu: Use an input for the layer when lowering input attachments
    • c777162 : radv: Use an input for the layer when lowering input attachments
    • d243bf1 : nir/lower_input_attachments: Support loading layer id as an input
    • e728957 : nir/lower_input_attachments: Refactor to use an options struct
    • 340c73d : nir/spirv: Add the option to keep ViewIndex as an input
    • c1e6957 : Revert "gitlab-ci: reuse container_post_build when building the test images"
    • 7e3a788 : freedreno/a6xx: fix occlusion query with more than one tile
    • 5550bc0 : freedreno/a6xx: only generate streamout for draw pass shader
    • ee7949b : freedreno/registers: SC_WAIT_WC is not a6xx
    • 60dae98 : nir/opt_remove_phis: optimize out phis with undef
    • 326eb56 : egl/x11_dri3: implement EGL_KHR_swap_buffers_with_damage
    • eae181e : egl/x11_dri3: enable & require xfixes 2.0
    • 6c9cf84 : ci: Run deqp-gles2 on RadeonSI
    • ff3ea3b : pan/mdg: Fix auxiliary load/store swizzle packing
    • 529f79d : pan/mdg: Fix printing of r26 ld/st sources post-RA
    • 14e7749 : pan/mdg: Identify barrier out-of-order field
    • af97331 : pan/mdg: Handle 32-bit offsets from store_shared
    • 06c3103 : pan/mdg: Ensure barrier op is set on texture
    • 415eb43 : panfrost: Fix shared memory size computation
    • 39bf1fb : panfrost: Fix WRITES_GLOBAL bit
    • 8626d4c : android: freedreno: Another build fix
    • 41bb645 : radv: restrict exported symbols with static llvm
    • a457256 : freedreno: Destroy syncobj too when destroying fence
    • b08fcab : gitlab-ci: reorder container_post_build call for arm64_test image
    • 40ab6d7 : gitlab-ci: reuse container_post_build when building the test images
    • 1cf7a39 : panfrost: Hoist assert from bind to create
    • 3b129dd : panfrost: Remove useless comment
    • a2e4159 : panfrost: Hoist add_fbo_bo call
    • 433fa39 : panfrost: Drop QUADS primitive convert
    • 4b21c8b : panfrost: Drop panfrost_invalidate_frame
    • 1362371 : panfrost: Drop ZSA null checks in draws
    • ec35159 : panfrost: Drop rasterizer null checks in draw calls
    • e46d4fe : panfrost: Decontextualize rasterizer
    • d26f62c : radv: track and report if a logical device is lost
    • c9c53da : ci: Update kernel used in LAVA to 5.8-based drm-misc
    • 91cd7fe : ci: Fix URL to imagediff page in traces dashboard
    • 9222174 : ci: Test Panfrost with more traces
    • 7d5f4d3 : ci: Split traces.yml file per driver
    • eb78777 : llvmpipe/nir: lower some more packing at nir level.
    • 95d611e : llvmpipe: lower mul 2x32_64
    • 25ba200 : llvmpipe: lower cs local index from id
    • 3cc5b75 : gallivm/nir: split tex/sampler indirect offsets
    • e616223 : gallivm/nir: fixup compact TCS variable storage.
    • 597d0e9 : gallium/nir/tgsi: add support for compact variables
    • fe9667e : llvmpipe: lower uadd_carry/usub_borrow.
    • 7893dfa : gallivm/sample: fix lod query on array textures.
    • da4f221 : gallivm/nir: add imod support
    • 407fa32 : gallium/nir/tgsi: reindent some code in the nir-tgsi info (v2)
    • e137bbb : llvmpipe/cs: update compute counters not fragment shader.
    • 6c7383d : vallium: fix input attachment lowering variable shadowing
    • e0e9712 : freedreno: support GL_EXT_semaphore
    • f6187aa : freedreno: Enable GL_EXT_memory_object and GL_EXT_memory_object_fd
    • e3c39e5 : freedreno: Implement pipe screen's get_device/driver_uuid()
    • 6bef126 : freedreno: implement pipe screen's resource_from_memobj
    • 03fdf41 : freedreno/layout: Move hard-coded minimum width for UBWC to a macro
    • 536ec9d : freedreno: Refactor fd_resource_create_with_modifiers() into a helper
    • 0eecd3d : freedreno: Implement memory object create/destroy for GL_EXT_memory_object
    • e48f27e : st: Pass TextureTiling option from texture to memory obj
    • b479de8 : spirv: Don't emit RMW for vector indexing in shared or global
    • f038b3a : radv: Fix handling of attribs 16-31.
    • 0067f89 : radv: Override the uniform buffer offset alignment for World War Z.
    • a1d4721 : driconf: Support selection by Vulkan applicationName.
    • f88b9eb : egl/android: Add support for CrOS buffer info perform op
    • acb235c : egl/android: Use droid_create_image_from_prime_fds() in get_back_bo()
    • cad9d42 : egl/android: Move droid_create_image_from_prime_fds() function up
    • 5c8b67e : egl/android: Simplify droid_create_image_from_name() path
    • d47b6f3 : egl/android: Drop unused ctx argument
    • 9cc788f : egl/android: Look up prime fds in droid_create_image_from_prime_fds()
    • 972f36d : egl/android: Call createImageFromDmaBufs directly
    • 7c98066 : freedreno: Add afuc regression test
    • d145fcc : freedreno/afuc: Install asm/disasm
    • f0b8718 : freedreno/afuc: Make 0 a valid number
    • 66dd248 : freedreno/afuc: Handle xmov modifiers
    • b2b1923 : freedreno/afuc: Add iret
    • a2c14ac : freedreno/afuc: Handle setsecure opcode
    • 0acc394 : freedreno/afuc: Fix printing preemptleave on a5xx
    • 165009b : spirv: Add support for the CL Round instruction
    • 6d28270 : meson: Make some warnings handled as errors with MSVC
    • 45f2252 : meson: Add MSVC narrowing-int-to-char warnings
    • d8edf8e : Revert "CI: Disable Panfrost T720/T760 CI"
    • e229e79 : mesa: utils: add "force_gl_vendor" option support
    • fcc7a1e : gallium/aux: reorder vertex attributes in triangle fans according to PV
    • f5cce89 : gallium/indices: translate primitive-restart values
    • 21952f1 : gallium/indices: introduce u_primconvert_config
    • c2e9cd3 : gallium/indices: implement prim-restart for triangle fans
    • dfc0a62 : gallium/indices: use prim_restart-helper for polygon
    • 3aa4f64 : gallium/indices: implement prim-restart for line-loops
    • ac96225 : gallium/indices: generalize primitive-restart logic
    • 2122b90 : gallium/indices: don't expand prim-type for 8-bit indices
    • 0c85d6c : gallium/util: factor out primitive-restart rewriting logic
    • b6ebf6e : spirv: Only require bare types to match when copying variables
    • 16f1d8b : glsl: fix crashes on out of bound matrix access using constant index
    • 7dcb1d2 : st/mesa: Replace UsesStreams by ActiveStreamMask for GS
    • ab5a36b : meson: fix trivial s/dir/dri/ typo
    • 9ef192c : CI: Disable Panfrost T720/T760 CI
    • 5a26015 : st/wgl: do not reject PFD_SUPPORT_GDI
    • ac9c94f : CI: Skip another flaky GS test on softpipe
    • e513d8d : CI: Skip flaky CS test on VirGL
    • 3e447c1 : CI: Disable Panfrost T860 and AMD Stoney tests
    • 80163bb : nir/vtn: Support OpOrdered and OpUnordered opcodes
    • 81e6ad8 : nir/vtn: Handle LessOrGreater deprecated opcode
    • c349e35 : nir/vtn: Support SpvOpIsFinite via fisfinite
    • af59e4c : nir: Add fisfinite op
    • e6e6f52 : nir/vtn: Support SpvOpIsNormal via fisnormal
    • 9ebbed6 : nir: Add fisnormal op
    • 199bea0 : nir: Fix i64tof32 lowering
    • f9e773b : llvmpipe: enable GL 4.5
    • 2e40b2c : llvmpipe/blit: for 32-bit unorm depth blits just copy 32-bit
    • 18e464c : compiler/nir: Add new flags to lower pack/unpack split instructions
    • 9e9d280 : mesa: Move ATTRIBUTE_NOINLINE for glsl_to_tgsi_visitor::visit_expression for MSVC
    • c60cea0 : glsl/test: Don't run whitespace tests in parallel
    • aeba69d : drisw: Port the MIT-SHM check to XCB
    • ee905aa : nir/vtn: CL SPIR-V callers should specify address modes
    • a097bf1 : nir/builder: Make nir_get_ptr_bitsize take a nir_shader
    • 627c8e1 : nir: Add nir_address_format_32bit_index_offset_pack64
    • 113458d : nir: Add nir_address_format_32bit_offset_as_64bit
    • a1ed83f : nir: Optimize mask+downcast to just downcast
    • fd9b33d : nir: Fix serialize/deserialize of void samplers/images
    • c17b58a : glsl: Add 'bare' shadow sampler type
    • 418c4c0 : compiler/nir: extend lower_fragcoord_wtrans to support VARYING_SLOT_POS
    • 7aaddf1 : compiler/nir: rewrite lower_fragcoord_wtrans to use nir_lower_instructions
    • c9f4b14 : Revert "CI: temp disable t720/t760 jobs."
    • d79c946 : nv50/ir/nir: rework indirect function_temp handling
    • fd265bb : panfrost: Update CI expectations
    • 7a13351 : panfrost: Drop union mali_attr
    • 4e3fe54 : panfrost: Drop hand-rolled pandecode for attribute buffers
    • ec58cda : panfrost: Use packs for varying buffers
    • 9732b64 : panfrost: Remove mali_attr_meta
    • eb2762a : pan/bit: Use packs for Bifrost unit tests
    • 7ef205d : panfrost: Use MALI_ATTRIBUTE_LENGTH
    • 59fa269 : panfrost: Use pack for general varying
    • 6c85063 : panfrost: Use pack for XFB varying
    • 40b4ee9 : panfrost: Factor out general varying case
    • b805cf9 : panfrost: Pass varying descriptors by reference
    • d5a264f : panfrost: Reword comment
    • 44397c5 : panfrost: Use packs for vertex built-ins
    • e646c86 : panfrost: Use packs for vertex attribute buffers
    • c9bb5dc : panfrost: Inline panfrost_vertex_instanced
    • e9dc225 : panfrost: Hoist instance_shift/instance_odd fetch
    • 27f8b87 : panfrost: Use packs for attributes
    • 6caf789 : panfrost: Move attr_meta emission to the draw routine
    • 8236fa3 : panfrost: Make attribute-buffer map explicit
    • 9cc81ea : panfrost: Simplify offset fixup proof
    • 4fc90f7 : panfrost: Use better packs for blits
    • 3e7fd2a : panfrost: Add XML for attribute buffers
    • 76de3e6 : panfrost: Merge attribute packing routines
    • 2c8a722 : panfrost: XMLify attributes
    • cdc3276 : panfrost: XMLify mali_channel
    • 668ec24 : panfrost: Squash 22-bit format field in attr_meta
    • c5f18ce : panfrost: Drop NXR format
    • e03622e : panfrost: Set STRIDE_4BYTE_ALIGNED_ONLY
    • 350dd2a : pipebuffer: Remove unused buffer event in slab bufmgr
    • 8a012f4 : nir: handle load_input_vertex in nir_get_io_offset_src
    • 83953f9 : nir: remove nir_strip stub declaration
    • f3a9781 : compiler: add glsl_print_type
    • b7c04b1 : ac/llvm: handle static/shared llvm init separately
    • 916110e : st/mesa: Enable clip planes lowering for geometry shaders
    • 584f273 : st/mesa: factor ucp-lowering logic into helper
    • ff4c991 : compiler/glsl: Initialize local variable to zero to fix MSVC RTC error
    • b38879f : vallium: initial import of the vulkan frontend
    • 8004fa9 : vulkan/wsi: add sw support. (v2)
    • 5a9eba4 : util/format: add some ZS helpers for vallium
    • 3abf548 : gallivm/nir: add load push constant support
    • 94e4a17 : llvmpipe: add support for memory allocation APIs
    • ecb617a : gallium: add a resource flag to say no over allocation.
    • e6c7ca2 : gallium: add an interface for memory allocations.
    • 50b0395 : radv: Use getter instead of setter to extract value.
    • 0919635 : gallium: Use unpack_rgba() instead of fetch_rgba in translate_generic
    • 5b8d67c : util: Move fetch_rgba to a separate function table.
    • 9cc8436 : util: Mark the format description getter functions as const.
    • 35b22b5 : util: Make all 3 fetch_rgba functions occupy the same function slot.
    • 80babbb : uitl: Add R1_UNORM to the list of noaccess (no pack/unpack) formats.
    • 6668d7a : egl: drop unused egldriver.h header
    • 1f01fa1 : egl: drop now empty egldriver.c
    • 890d033 : egl: inline _eglInitializeDisplay() into eglInitialize()
    • 0e7ae39 : egl: inline _eglGetDriverProc() into eglGetProcAddress()
    • 735b35e : egl: drop unused ${drv}_driver()
    • c74628f : egl: replace `&_eglDriver`/`NULL` tested against `NULL` with simple `true`/`false`
    • 37be01b : egl: drop unused _EGLDriver from MesaGLInteropEGL{QueryDeviceInfo,ExportObject}()
    • 5eb5847 : egl: replace replace _EGLDriver with _EGLDisplay-Driver in eglapi.c
    • 372c6c4 : egl: replace _EGLDriver with _EGLDisplay-Driver in _eglGetSyncAttrib()
    • b9fb63a : egl: drop unused _EGLDriver from _eglSurfaceAttrib()
    • b981bab : egl: drop unused _EGLDriver & _EGLDisplay from _eglQueryContext()
    • 3ef8aac : egl: drop unused _EGLDisplay from _eglSetDamageRegionKHRClampRects()
    • 0345a61 : egl: drop unused _EGLDriver from _eglGetConfigs()/_eglChooseConfig()/_eglGetConfigAttrib()
    • 67c7456 : egl: drop unused _EGLDriver from SetBlobCacheFuncsANDROID()
    • 1bbb0c7 : egl: drop unused _EGLDriver from QueryDmaBuf{Formats,Modifiers}EXT()
    • a71dcd2 : egl: drop unused _EGLDriver from ExportDMABUFImage{,Query}MESA()
    • 429e936 : egl: drop unused _EGLDriver from QueryBufferAge()
    • 155541b : egl: drop unused _EGLDriver from PostSubBufferNV()
    • f010568 : egl: drop unused _EGLDriver from CreateWaylandBufferFromImageWL()
    • 845a37c : egl: drop unused _EGLDriver from {Bind,Unbind,Query}WaylandDisplayWL()
    • 9d5162e : egl: drop unused _EGLDriver from {Create,Export}DRMImageMESA()
    • 692b9c6 : egl: drop unused _EGLDriver from DupNativeFenceFDANDROID()
    • f49805c : egl: drop unused _EGLDriver from {Create,Destroy,ClientWait,Wait,Signal}SyncKHR()
    • 740ec08 : egl: drop unused _EGLDriver from {Create,Destroy}ImageKHR()
    • a7d15d2 : egl: drop unused _EGLDriver from GetProcAddress()
    • b8d1c46 : egl: drop unused _EGLDriver & _EGLDisplay from WaitNative()
    • 4e65469 : egl: drop unused _EGLDriver from WaitClient()
    • 648f603 : egl: drop unused _EGLDriver from SetDamageRegion()
    • d202fd6 : egl: drop unused _EGLDriver from CopyBuffers()
    • 82035b2 : egl: drop unused _EGLDriver from SwapBuffers{,WithDamageEXT,RegionNOK}()
    • 347840c : egl: drop unused _EGLDriver from SwapInterval()
    • 2f4a6ad : egl: drop unused _EGLDriver from {Bind,Release}TexImage()
    • 982ac69 : egl: drop unused _EGLDriver from QuerySurface()
    • 1a17f1d : egl: drop unused _EGLDriver from MakeCurrent()
    • 6b4c474 : egl: drop unused _EGLDriver from Create{Window,Pixmap,Pbuffer}Surface() & DestroySurface()
    • f27a32a : egl: drop unused _EGLDriver from {Create,Destroy}Context()
    • 435ad51 : egl: drop unused _EGLDriver from Terminate()
    • ad61d4f : egl: drop unused _EGLDriver from Initialize()
    • e87e68d : egl: replace _EGLDriver with _EGLDisplay-Driver in _eglQuerySurface()
    • f3aa7b2 : egl: replace _EGLDriver param with _EGLDisplay-Driver in dri{2_x11,3}_create_window_surface()
    • 485f8f8 : egl: replace _EGLDriver param with _EGLDisplay-Driver in _eglReleaseDisplayResources()
    • a76423e : spirv/vtn_cfg.c: Include util/debug.h for env_var_as_boolean.
    • 430e331 : llvmpipe: only read 0 for channels being read
    • f1001a3 : CI: temp disable t720/t760 jobs.
    • 031f9e1 : anv: don't fail userspace relocation with perf queries
    • dc667b1 : nv50/ir/nir: fix smem size
    • 384c4df : clover/nir: fix mem_shared by using address_format_32bit_offset
    • aa86611 : android: util/format: fix generated sources rules
    • b41e74a : spirv: Add a MESA_SPIRV_FORCE_UNSTRUCTURED environment variable
    • 217def3 : spirv: parse unstructured CFG
    • 467b90f : spirv: extract switch parsing into its own function
    • 96dff31 : spirv: Move the emit a 'return value' store logic into own function
    • 4638de8 : nir/lower_goto_if: Add some debug prints
    • d57573d : nir/lower_goto_if: Add a route::outside set
    • d2bf850 : nir/lower_goto_if: Replace a tripple loop with a double loop
    • b892d47 : nir/lower_goto_if: Add a block_for_singular_set helper
    • 193765e : nir/lower_goto_if: Sort blocks in select_fork
    • 0fad20f : nir/lower_goto_if: Rework some set union logic
    • d161798 : nir/lower_goto_if: Rework handling of skip targets
    • 2455d03 : nir/lower_goto_if: Use util/list instead of exec_list
    • f69d732 : nir/lower_goto_if: Clean up ralloc usage
    • 202b4df : nir/lower_goto_if: Document some data structures
    • b273611 : nir: Add a structurizer
    • 025bdba : nir: Add goto_if jump instruction
    • 92db942 : nir: Add and use nir_foreach_block_unstructured helpers
    • 4768ea1 : nir: Add a structured flag to nir_shader
    • 19f3573 : spirv: rename vtn_emit_cf_list to vtn_emit_cf_list_structured
    • 16f8589 : util/set: add _mesa_set_intersects
    • acf756a : android: panfrost: Redirect cmdstream includes through GenXML (v2)
    • 92f65e2 : turnip: add missing tu_bo_list_add in CmdWriteTimestamp
    • ddac593 : turnip: call packing functions directly for pack_gmem_clear_value
    • 201331e : turnip: rework format_to_ifmt
    • aa5ebf0 : turnip: fix CmdBlitImage with D32_SFLOAT_S8_UINT
    • f83e895 : turnip: delete a blit_image TODO that has already been resolved
    • 2ca1ccc : turnip: remove dead tu_minify/typed_memcpy functions
    • 9937254 : android: panfrost/bifrost: add libpanfrost_lib static dependency
    • bce1a7e : android: panfrost: Redirect cmdstream includes through GenXML
    • 0980d1d : radv: Clean up setting the surface flags.
    • d19bc94 : radv: When importing an image, redo the layout based on the metadata.
    • a0d67b7 : anv: add no reloc flags on empty and simple bo paths.
    • 00b9a34 : anv: Mark anv_dump_{start,finish} as PUBLIC
    • f024bc8 : zink: invalidate pipeline hash on more changes
    • 65d7172 : util/macros: Add ATTRIBUTE_NOINLINE definition for MSVC
    • 58af311 : u_debug_stack_test: Fix MSVC compiling by using ATTRIBUTE_NOINLINE
    • 4932985 : iris: Make iris_bo_import_dmabuf take a modifier
    • c111e90 : iris: Don't call SET_TILING for dmabuf imports
    • 0c97e60 : turnip: implement VK_EXT_custom_border_color
    • 84ed2d0 : util/format: expose generated format packing functions through a header
    • a322924 : panfrost: add missing dependency on midgard_pack.h
    • 8557b1a : panfrost: Fix border colour
    • 8277334 : nir/lower_idiv: Port recent LLVM fixes to emit_udiv
    • d452c04 : aco: do not set valid_mask for POS0 exports on GFX 10.3
    • 0c621dc : panfrost: XMLify enum mali_format
    • 78445ce : panfrost: XMLify exception access
    • 99d17fb : panfrost: XMLify MSAA writeout mode
    • c9bdba2 : panfrost: XMLify Block Format
    • 8c1ac96 : panfrost: Drop unused mali_channel_swizzle
    • ad0b32c : panfrost: XMLify Bifrost textures
    • f008a63 : panfrost: XMLify Midgard textures
    • b10c3c8 : panfrost: XMLify Bifrost samplers
    • f74186b : panfrost: XMLify Midgard samplers
    • c4c8c8e : panfrost: Don't mask coverage mask to 4-bits
    • 6afd4ad : panfrost: Simplify depth/stencil/alpha
    • 1675d92 : panfrost: Simplify zsa == NULL case
    • 931e7b4 : panfrost: XMLify stencil test
    • fa94967 : panfrost: XMLify UBOs
    • 7f487e0 : panfrost: XMLify viewport
    • ef20691 : panfrost: XMLify wrap modes
    • e9ca320 : panfrost: XMLify stencil op
    • d182a60 : panfrost: XMLify mali_func
    • 32dbc80 : panfrost: XMLify draw_mode
    • 4b7056b : panfrost: XMLify job_type
    • d2ddd4d : pan/decode: Add helper to dump GPU structures
    • 88dc4c2 : panfrost: Redirect cmdstream includes through GenXML
    • 1ad9219 : panfrost: Build midgard_pack.h via meson
    • 75cc5b8 : panfrost: Adopt gen_pack_header.py via v3d
    • 9cb3a9e : panfrost: Add stub midgard.xml
    • 33e6920 : util: Fix up indentation in the generated format tables code.
    • 9fd0f45 : util: Change a codegenned switch statement to a nice little table.
    • 4064a6c : util: Split the pack/unpack functions out of the format desc.
    • 6723add : docs: update calendar for 20.2.0-rc2
    • 6de5acd : docs: update calendar for 20.2.0-rc1
    • 4c84600 : radeonsi: fix compute-based culling with VERTEX_COUNTER_GDS_MODE == 1
    • 2b78cb2 : radeonsi: disable NGG culling on gfx10.3 because of hangs
    • 61042b1 : zink: handle more draw modes
    • ced9e6e : zink: change pipeline hashes to index based on vk primitive type
    • 2701f88 : anv/nir: Unify inputs_read/outputs_written between geometry stages
    • 2a12175 : panfrost: Implement panfrost_query_dmabuf_modifiers
    • 24c921b : panfrost: Ensure AFBC slices are aligned
    • 4d0241d : panfrost: Allocate enough space for tiled formats
    • 8c0cb56 : panfrost: Enable YTR where allowed
    • d285ea6 : panfrost: Implement YTR availability check
    • acb8dcf : panfrost: Choose AFBC when available
    • c54d45d : panfrost: Import staging routines from freedreno
    • 348d374 : panfrost: Respect modifiers in resource management
    • 56f9cc9 : panfrost: Account for modifiers when creating BO
    • 965537d : panfrost: Use modifier instead of layout throughout
    • 6fed69c : panfrost: Introduce create_with_modifier helper
    • 32bd588 : panfrost: Remove hint-based AFBC heuristic
    • 340017e : panfrost: Set `initialized` more conservatively
    • 2277702 : gallium/dri2: Support Arm modifiers
    • 37872ed : zink: print error when getprocaddr fails for extension functions
    • 371f6f4 : radv: Update CI expectations for the recent descriptor indexing regressions.
    • 3d7d76c : iris: Add support for MESA_SHADER_KERNEL in the disk cache
    • 65eeb06 : iris: Upload kernel inputs with system values
    • f5e7be3 : iris: Copy dest size from the original intrinsic in setup_uniforms
    • 43429ef : iris/disk_cache: Stop assuming stage == cache_id
    • 6dfe41c : iris: Add a kernel_input_size field for compiled shaders
    • e39326e : iris: Use blob_write_uint32 for num_system_values
    • 003b04e : intel/compiler: Allow MESA_SHADER_KERNEL
    • e2b6ccb : intel/compiler: Use C99 array initializers for prog_data/key sizes
    • 8e1de8e : intel/cs_intrinsics: Handle 64-bit intrinsics
    • 8d62735 : iris: Add support for serialized NIR
    • baa4cf9 : iris: Implement set_global_binding
    • 17280a8 : iris: no-op implement set_compute_resources
    • ac6e207 : gitlab-ci: test Fossilize with GFX1030
    • 431a3cf : radv/winsys: add null winsys entries for Sienna Cichild/Navy Flounder
    • 638a09b : radv: fix emitting the border color pointer on the compute queue
    • 96cfc68 : util: Fix memory leaks in unit test.
    • ef66e02 : src/mesa: add GL_NV_half_float extension support (v2)
    • 8abf59f : dri_util: Update internal_format to GL_RGB8 for MESA_FORMAT_B8G8R8X8_UNORM
    • fdb97d3 : aco: execute branch instructions in WQM if necessary
    • 678cb6d : nir: nir_range_analysis needs to be updated for vec16
    • 7b4c24e : aco: don't move memory accesses to before control barriers
    • e4c6204 : radv: limit LATE_ALLOC_GS to prevent a GPU hang on GFX10
    • 0256250 : radv/gfx10: add missing initialization of registers
    • 044b238 : etnaviv: completely turn off MSAA
    • 5cf7eec : nir/lower_ssbo: Don't set align_* for atomics
    • 2d010d3 : anv: add a check for depthStencilState before using it
    • ffc8f2b : anv: fix up dynamic clip emission
    • 240c074 : anv: centralize vk to gen arrays
    • 3d6e4a2 : freedreno/decode: try harder to not crash in disasm
    • cbfce48 : freedreno/crashdec: handle section name typos
    • 8d437b2 : freedreno/ir3: add more disasm stats
    • 7aa74ab : freedreno/ir3: add tracking for # of instructions per category
    • 2cd0c8d : gallium/u_transfer_helper: add util functions for doing deinterleaving during map
    • 2368813 : gallium: add pipe_transfer_usage for z/s only mappings
    • cf52b40 : intel/fs: work around gen12 lower-precision source modifier limitation
    • ee06e47 : intel/fs: Assert if lower_source_modifiers converts 32x16 to 32x32 multiplication
    • 166630f : android: pan/bi: Separate disasm/compiler targets
    • 41c9a2e : android: pan/mdg: Separate disassembler and compiler targets
    • b9e58be : android: panfrost: Move pandecode into lib/
    • de352f5 : android: panfrost: Rename encoder/ to lib/
    • 6e70508 : aco: set constant_data_offset correctly in the case of merged shaders
    • 0f8ef37 : radeon/vcn: fix jpeg decode for navi10
    • e362ccb : radv: Do not consider layouts fast-clearable on compute queue.
    • 2fa83dc : radv: Add forcecompress debug flag.
    • a777b25 : intel/perf: export performance counters sorted by [group|set] and name
    • 59716e4 : intel/perf: split load_oa_metrics
    • 2fbab5a : intel/perf: fix performance counters availability after glFinish
    • 9fa6480 : intel/perf: streamline error handling in read_oa_samples_until
    • 03e8b35 : intel/perf: fix how pipeline stats are stored
    • e7e6f70 : intel/perf: fix calculation of used counter space
    • 6d988ad : radv: report a better error message when QueueWaitIdle() failed
    • 3691ef5 : radv: report errors back to the application via VK_EXT_debug_report
    • cc4b9c2 : radv: rework the error function helpers a bit
    • 11781c0 : radv: report the spirv-nir logs back to the application
    • ff4f620 : radv: Fix assert that is too strict.
    • f7e7cf6 : radeon/radeon_vce: fix out of target bitrate in CBR mode (H.264)
    • 4975b3e : r600: Enable compute shaders for NIR code path
    • 834a61d : r600/sfn: Force a minimum of 4 GPRs, it seems to fix atomics
    • 3e3068a : r600/sfn: handle querying SSBO size
    • ac87cc2 : r600/sfn: Correct ssbo instruction handling
    • 11a861c : r600/sfn: correct allocating and emitting of atomics
    • 702619c : r600/sfn: Add a mapping table for atomics
    • 56dedf0 : r600/sfn: add r600 specific lowering pass for atomics and use it
    • df2acf8 : r600/sfn: Sort uniforms by binding and offset
    • a828f6c : r600: Set PIPE_CAP_NIR_ATOMICS_AS_DEREF to true
    • a03e24a : gallium + mesa/st: Add PIPE_CAP_NIR_ATOMICS_AS_DEREF and use it
    • dd003ab : meson: bump required glvnd version
    • 7fbadfc : driconf: fix force_gl_vendor description
    • 800816d : egl/entrypoint-check: add check that GLVND and plain EGL have the same entrypoints
    • 351d513 : egl/entrypoint-check: split sort-check into a function
    • 04bd58f : Correct a typo in threads_win32.h
    • e3069c4 : pick-ui: specify git commands in "resolve cherry pick" message
    • b238d17 : zink: Fix 32-bit compilation
    • 08f3dcf : i965: add support for force_gl_vendor
    • ab92c11 : freedreno/ir3: don't install ir3_compiler cmdline tool
    • 3b6e867 : freedreno/registers: add some missing regs to build
    • 1e28745 : meson/freedreno: Fix lua requirement
    • 0cdd411 : radeonsi: various fixes for gfx10.3
    • e2e700f : radeonsi: remove the NGG hack decreasing LDS usage to deal with overflows
    • 97456e8 : radeonsi: add a common function for getting the size of gs_ngg_scratch
    • 68b3e92 : radeonsi: don't count unusable vertices to the NGG LDS size
    • 64c741f : radeonsi: fix applying the NGG minimum vertex count requirement
    • 7a468fc : radeonsi: increase minimum NGG vertex count requirement per workgroup on gfx 10.3
    • 633d2aa : radeonsi: use the same units for esgs_ring_size and ngg_emit_size
    • b6fb09f : radeonsi: use correct wave size in gfx10_ngg_calculate_subgroup_info
    • 61c671c : Revert "radeonsi: honor a user-specified pitch on gfx10.3"
    • 15bd3f3 : ac/gpu_info: set num_tiles_pipes on gfx10+ too
    • 9333a85 : radeonsi: enable ETC2 hw acceleration on Raven2
    • 1185b3f : zink: pre-hash gfx-pipeline-state
    • 7be12df : zink: rename zink_gfx_program::stages to 'modules'
    • 6196f2a : zink: always compile shaders in pipeline order
    • c312299 : zink: start using per-stage flags for new shaders, refcount shader modules
    • 76d3645 : zink: use ZINK_SHADER_COUNT instead of PIPE_SHADER_TYPES - 1 everywhere
    • 0f059d5 : zink: split up creating zink_shader objects and VkShaderModule objects
    • a03d17e : zink: refcount zink_gfx_program objects
    • 8772c69 : zink: adjust zink_shader struct to contain full streamout info
    • e481d3e : zink: move shader state methods for pipe_context into zink_program.c
    • d83be29 : intel/dump_gpu: add an option to capture a single frame
    • a6a5b0e : intel/dump_gpu: fix --platform option
    • 9f0db06 : anv: track the current frame and write it into the driver identifier BO
    • 42cb068 : intel/dump_gpu: only map in GTT buffers not previously mapped
    • 22b512e : intel/dump_gpu: add an only-capture option
    • f0e1382 : intel/dump_gpu: set default device_override
    • c4bfa6d : intel/dump_gpu: further track mapping of BOs
    • 63c90ac : intel/dump_gpu: only write BOs mapped by the driver
    • 54c6849 : zink: set primitive restart cap
    • 5959178 : zink: use util_draw_vbo_without_prim_restart for unsupported prim modes
    • f4583b4 : zink: move 8bit index handling out of u_primconvert path
    • 5edaf08 : zink: basic primitive restart support for strip/fan topologies
    • 8531830 : radv: Add ETC2 support on RAVEN2.
    • bea8930 : radv: allow to force-enable LLVM internally for a specific shader stage
    • e63a788 : etnaviv: call nir_lower_bool_to_bitsize
    • c5b2319 : panfrost: Drop compiler cmdstream deps
    • 1d0b3ef : panfrost: Treat texture dimension as first-class
    • 5a3088e : panfrost: Inline max rt into compilers
    • 54f9654 : pan/bi: Drop use of MALI_POSITIVE
    • c94013c : panfrost: Add panloader/ to .gitignore
    • 52e3a67 : panfrost: Reduce bit dependency to disassembly only
    • 14bb72c : pan/bi: Separate disasm/compiler targets
    • b792d61 : pan/mdg: Separate disassembler and compiler targets
    • d62a6e7 : panfrost: Move pandecode into lib/
    • 1c62b55 : panfrost: Rename encoder/ to lib/
    • 956b096 : panfrost: Don't export exception_status
    • 4ddc6e5 : panfrost: Remove panfrost-misc.h
    • 64f3c9d : panfrost: Inline panfrost-misc.h into panfrost-job.h
    • ebf7e87 : pan/decode: Remove shader replacement artefact
    • e4df95b : panfrost: Move format stringify to decode.c
    • b80e753 : panfrost: Remove blend prettyprinters
    • b43bcb0 : anv: fix incorrect realloc failure handling
    • 21b47cb : aco: fix C++11/C++14 compilation
    • 782ba8d : st/mesa: Treat vertex outputs absent in outputMapping as zero in mesa_to_tgsi
    • a92cfa6 : Bump development version and clear new_features
    • 7f06d19 : anv: Advertise shaderIntegerFunctions2
    • 9e39795 : panfrost: signal syncobj if nothing is going to be flushed
    • c6aadba : radv: Don't use both DCC and CMASK for single sample images.
    • e2c614a : appveyor: Use Python3.
    • 0f9fb7f : appveyor: Upgrade pip.
    • 2e66545 : util: Fix SCons build.
    • 56ab105 : freedreno: Add more asserts for DST_OFF/NUM_UNIT in indirect const uploads.
    • 3e970de : freedreno: Increase the NUM_UNIT on compute's consts in indirect dispatch.
    • f07e25b : freedreno/ir3: Clean up instrlen setup.
    • 8f637d6 : freedreno: Split ir3_const's user buffer and indirect upload APIs.
    • 1542999 : freedreno: Rename emit_const_bo() to emit_const_ptrs().
    • 51acfe2 : freedreno/ir3: Simpify the immediates from an array of vec4 to array of dwords.
    • e873c4d : freedreno/ir3: Merge the redundant immediate_idx/immediates_count fields
    • 5e922fb : glsl_to_nir: fix bitfield_extract with 16-bit operands
    • 92f5e94 : glsl: improve precision determination for calls
    • 282a1e6 : glsl: don't lower to mediump for desktop OpenGL
    • 01e0085 : glsl: don't create conversion opcodes for array types
    • 5020403 : glsl: don't lower atomic functions to mediump
    • 93076f6 : glsl: don't inline intrinsics for mediump
    • 48a6255 : glsl: fix constant expression evaluation for 16-bit types
    • f2d5f48 : glsl: lower_precision - fix assertion failure with dereferences of constants
    • a88fd7b : docs: update calendar and link releases notes for 20.1.5
    • 2d3f81f : docs: add release notes for 20.1.5
    • a4c4e01 : glsl: remove LowerPrecisionTemporaries
    • c4e0cae : gallium: replace 16BIT_TEMPS cap with 16BIT_CONSTS
    • 0a763c0 : glsl/lower_precision: split out const lowering
    • 4f06054 : freedreno/ir3: lower local_index using local_id
    • e094790 : freedreno/ir3: ir3_cmdline updates
    • 283ad85 : radeonsi: call nir_split_array_vars/shrink_vec_array_vars/opt_find_array_copies
    • 47beee2 : radeonsi: reorder NIR optimizations
    • 0294eae : radeonsi: extend workaround for KHR-GL45.texture_view.view_classes on gfx9
    • caa9824 : st/mesa: optimize DEPTH_STENCIL copies using fragment shader
    • 141b295 : freedreno: allow fence_fd fences to be recycled
    • 07a49bf : radeonsi: disable SDMA on gfx9
    • 879444a : ci: Add a build test for the Android platform
    • db504c4 : radv/android: Remove unused variable
    • a7fe711 : vulkan: Allow global symbol HMI for Android
    • 6b3f56f : anv: Add stub for anv_gem_get_tiling() for Android
    • ff0dbf2 : anv: Pass device to setup_gralloc0_usage for error reporting
    • 5096ebf : meson: Define ANDROID and ANDROID_API_LEVEL when compiling for Android
    • e05e60b : turnip: Make tu_android.c compile again
    • a725899 : mapi: Mark TLS symbols as optional in glapi-symbols.txt
    • 932f51d : ci: Include enough Android headers to let us compile test EGL
    • 5ae7098 : gallium/android: Rewrite backtrace helper for android
    • d0d14f3 : util: Add unit test for stack backtrace caputure
    • 848e7b9 : util: Move stack debug functions to src/util
    • e487043 : gallium: Switch u_debug_stack/symbol.c to util/hash_table.h
    • e1a58ae : mapi/test: Change type to unsigned for offset
    • c874930 : egl/android: Remove unused variable
    • 1541ef6 : ci: Use a rootfs tarball for NFS root, instead of a ramdisk (for LAVA)
    • 8fab7d7 : radv: set BYPASS_VTX_RATE_COMBINER_GFX103 on GFX 10.3
    • 05b09d6 : radv: fix sample shading on GFX 10.3
    • 8b36823 : radv: increase minimum NGG vertex count requirement per workgroup on GFX 10.3
    • 22c8079 : radv: do not honor a user-specified pitch on GFX 10.3
    • 7f25f1f : nir/lower_tex: handle query lod with nir_lower_tex_packing_16 at lower_tex_packing
    • 6e2e775 : radv/llvm: enable VK_KHR_memory_model
    • 4f3630b : ac/nir: fix coherent global loads/stores
    • 4640e7d : ac/nir: consider an image load/store intrinsic's access
    • da38e99 : radv/aco: enable VK_KHR_memory_model
    • 389c95a : spirv: set ACCESS_COHERENT for ssbo/global/image atomic load/store
    • 71572eb : nir/lower_tex: skip lower_tex_packing for the texture samples query
    • 1938e25 : freedreno/computerator: Set SP_MODE_CONTROL to the same value as vulkan/GL
    • c92ae9d : freedreno/a6xx: Document the bit for the magic 32bit-uniforms-as-16b mode.
    • db25c18 : freedreno/ir3: Fix the type of half-float indirect uniform loads.
    • 13b3c40 : nir: Print the constant data size associated with a shader.
    • 041bae2 : nir: Add a little more docs about NIR's constant_data.
    • 2e833b1 : nir/lower_amul: Use num_ubos/ssbos instead of recomputing it.
    • add2b44 : turnip: Only include msm_drm in tu_drm.c
    • 4b9c0cd : turnip: Move remaining drm code to tu_drm.c
    • 7dfeae1 : turnip: Collapse some tu_drm wrappers
    • 59e5f95 : turnip: Move tu_bo functions to tu_drm.c
    • 270599c : turnip: Move device enumeration and feature discovery to tu_drm.c
    • d33f46e : anv: fix allocation of custom border color pool
    • fea3e49 : aco: replace MADs in isel with FMA on GFX10.3
    • 41c901b : aco: disable SMEM stores on GFX10.3
    • b811b1d : aco: update aco_opcodes.py for GFX10.3
    • 07250a9 : aco: implement subgroup shader_clock on GFX10.3
    • a5303a3 : aco: update vgpr_alloc_granule for GFX10.3
    • 37988b5 : aco: fix max_waves_per_simd on Polaris, VegaM and GFX10.3
    • c68fba9 : aco: update bug workarounds for GFX10_3
    • 4f1242a : aco: don't create v_mad_f32 on GFX10.3
    • 5718f7c : aco: fix waitcnt insertion on GFX10.3
    • b75427c : panfrost: Implement EXT_multisampled_render_to_texture
    • 2c47993 : panfrost: Add MSAA mode selection field
    • 7973b6c : docs/features: Add GL_EXT_multisampled_render_to_texture
    • 0c569e2 : radv: print warnings for famous RADV_PERFTEST options that no longer exist
    • 65d23e7 : radeon/vcn: Corrected vp9 ref associated data incase of target-codec is NULL
    • b98dd70 : spirv: Propagate explicit layout only in types that need it
    • 9a9b35a : lima: Fix lima_screen_query_dmabuf_modifiers()
    • ee2c58d : tu: Implement VK_EXT_conditional_rendering
    • f226b19 : tu: Reset has_tess after renderpass
    • 06332ef : freedreno: Document draw predication packets
    • 9e48e31 : tu: Fix DST_INCOHERENT_FLUSH copy/paste error
    • 95db96d : turnip: implement VK_EXT_4444_formats
    • 036ff53 : util/format: translate A4R4G4B4_UNORM and A4B4G4R4_UNORM vulkan formats
    • 35f8f35 : turnip: rework extended formats to allow more extended formats
    • 0b24783 : ci: Actually upload trace artifacts to MinIO for baremetal
    • 67b1163 : turnip: add support for D32_SFLOAT_S8_UINT
    • a133f7d : ci: Remove kernel module build that slipped in
    • 7633037 : ci: Download traces from MinIO in baremetal runs
    • d4ca45e : ci: Upload traces' reference and actual images to MinIO
    • 4816f6f : spirv: Do more complex unwrapping in get_nir_type
    • 140a549 : compiler/types: Add a struct_type_is_packed wrapper
    • 66d8bbd : freedreno: Fix "Offset of packed bitfield changed" warnings:
    • ba6cdb2 : turnip: delete tu_clear_sysmem_attachments_2d
    • 9b6486b : turnip: fix sysmem CmdClearAttachments 3D fallback breaking GMEM path flush
    • 611f654 : nir/deref: Don't try to compare derefs containing casts
    • ee2f21b : nir: Remove the old nir_opt_shrink_load.
    • 023e666 : i965: Enable vector shrinking in the vec4 backend.
    • 1c9906d : nir: Add a pass to cut the trailing ends of vectors.
    • 5afaec3 : turnip: workaround for a630 d24_unorm_s8_uint fails
    • 6fc5273 : etnaviv: fix nir validation problem
    • caa107c : freedreno/decode: move dependencies up a level
    • 9c33c53 : freedreno/registers: install gzip'd register database
    • 62ebd34 : freedreno/registers: split header build into subdirs
    • e59b241 : freedreno/registers: add .gitignore
    • 24f55eb : freedreno/rnn: rework RNN_DEF_PATH construction
    • bf425b7 : freedreno/rnn: also look for .xml.gz
    • 8efdd38 : radv: Implement VK_EXT_4444_formats
    • 3c2a1af : anv: Implement VK_EXT_4444_formats
    • b44139e : vulkan: Update Vulkan XML and headers to 1.2.149
    • 070fd2b : u_prim_restart: add inline function for getting restart index based on index size
    • 1f24c54 : iris: Zero the add-on clear color BO on import
    • a956093 : ci: Build-test Panfrost tools
    • cd1c21f : zink: implement handling for VK_EXT_calibrated_timestamps
    • 8930b19 : zink: store valid timestamp bits onto zink_screen
    • 3e6366b : zink: handle VK_EXT_vertex_attribute_divisor setup
    • 99b4453 : zink: clamp PIPE_SHADER_CAP_MAX_SHADER_BUFFERS to PIPE_MAX_SHADER_BUFFERS
    • 01d6220 : zink: implement VK_EXT_robustness2
    • b2f4752 : pan/mdg: Test for SSA before chasing addresses
    • b4de9e0 : pan/mdg: Mask spills from texture write
    • 26dd8f8 : swr: Bump maximum 2D texture size to 16kx16k
    • 99cf910 : mesa/st: Actually free the driver part of memory objects on destruction.
    • 63bff6a : gallivm/nir: Lower uniforms to UBOs in llvm draw if the driver didn't request this already
    • b54afde : android: freedreno: move a2xx disasm out of gallium
    • ca114e6 : android: freedreno/common: add support for libfreedreno_common static
    • f065ec4 : android: freedreno/ir3: fix include paths
    • 711c30b : android: freedreno/registers: fix generated headers rules
    • 32e0f7e : anv: toggle on VK_EXT_extended_dynamic_state
    • b9a0544 : anv: dynamic vertex input binding stride and size support
    • e4590c0 : anv: depth/stencil dynamic state support
    • f6fa4a8 : anv: add support for dynamic primitive topology change
    • f426663 : anv: add support for dynamic viewport and scissor with count
    • 9220598 : anv: add support for dynamic cull mode and winding order
    • c34d8ac : anv: handle dynamic viewport count
    • 11f3fb9 : anv: consider dynamic state when creating pipeline
    • 65de778 : anv: add new dynamic states
    • 2260ce6 : anv: add VK_EXT_extended_dynamic_state but leave it disabled
    • a4181fc : meson: fix `-D xlib-lease=auto` detection
    • c147604 : egl: consistently use dri2_egl_display() helper macro
    • 27eea62 : freedreno/afuc: Fix PM4 enum parsing
    • a5daaed : freedreno/afuc: Add missing rnn_prepdb()
    • 8d0e5e0 : freedreno/cffdec: Stop open-coding enum parsing
    • 73241ca : freedreno/rnn: Make rnn_decode_enum() respect variants
    • b67695d : egl/haiku: drop overwritten preset of EGL version
    • 9c6fa94 : egl: const _eglDriver
    • 81c0e2a : egl: Allow software rendering for vgem/virtio_gpu in platform_device
    • eaf27eb : intel/tools: Test notification subregisters
    • ac7ecd2 : intel/tools: Simplify notification register handling
    • 63181df : intel/tools: Don't hardcode notification register
    • 90c18ec : intel/tools: Manually set ARF register file/nr/subnr
    • af6d6f5 : intel/tools: Pass integers, not enums, to stride()
    • c883c48 : intel/compiler: Relax SENDS regioning assertions
    • 363e5ef : intel/tools: Simplify dstregion
    • 3d9c673 : intel/tools: Simplify immediate handling
    • 59801f0 : intel/tools: Make writemask an integer
    • e115c49 : intel/tools: Make swizzle an integer
    • 3e1602c : intel/tools: Simplify register type handling
    • 2851c21 : intel/tools: Don't allow empty type specifier
    • 6809b93 : intel/tools: Remove stray newline
    • fdfbb1e : intel/tools: Fix typos
    • 2dec909 : pan/bit: Remove BI_SHIFT stub
    • aa989ae : pan/bit: Update f32-f16 convert test
    • cf8a8b7 : ci: Set date in LAVA DUTs from NTP servers
    • 7f40db4 : docs: Explain how to set up a personal gitlab runner.
    • 8e62687 : freedreno/a6xx: Fix CP_BIN_SIZE_ADDRESS name
    • 6c11a79 : i965/i915: Add colorspace support to YUV sampling
    • d8fdb8d : nir: Add colorspace support to YUV lowering pass
    • f7f31f0 : mesa: utils: fix check of option is not found
    • f3509c0 : zink: add extension loading framework for spirv builder
    • a91011c : pan/mdg: emit REGISTER_UNUSED on unused ALU src2
    • a206591 : mapi: Return NULL function pointers for GL_EXT_debug_marker
    • 258165b : egl: drop left-over function prototype
    • ed3f1e0 : egl: rename _eglMatchDriver() to _eglInitializeDisplay()
    • 6d6b82a : egl: inline _eglMatchAndInitialize() and refactor _eglMatchDriver()
    • a77050c : egl: fix _eglMatchDriver() return type
    • f91851e : egl: drop unnecessary _eglGetDriver()
    • d24e3ea : egl: replace _eglInitDriver() with a simple variable
    • 3d4deb6 : pan/mdg: remove ins-br_compact and ins-branch_extended
    • 8150c1d : pan/mdg: defer branch packing
    • 140185e : pan/mdg: refactor emit_alu_bundle
    • 0f0f9ee : pan/mdg: remove ins-alu
    • 5299239 : pan/mdg: externalize mir_pack_mod
    • 1a4d165 : pan/mdg: defer register packing
    • bea6a65 : pan/mdg: eliminate references to ins-load_store.op
    • 92c808c : pan/mdg: eliminate references to ins-texture.op
    • 83592de : pan/mdg: apply float outmods to textures
    • 5011373 : pan/mdg: eliminate references to ins-alu.outmod
    • f34815c : pan/mdg: fix comment
    • 5f7e018 : pan/mdg: eliminate references to ins-alu.reg_mode
    • f4c89bf : pan/mdg: eliminate references to ins-alu.op
    • 598527f : pan/mdg: prepare effective_writemask()
    • b1b0ce0 : pan/mdg: fix src_type in instructions that need a implicit zero
    • 75b1f3d : drm-shim: Return -EINVAL instead of abort()ing on unknown ioctls.
    • c77a414 : u_prim_restart: handle indirect draws
    • a58081f : panfrost: Android build fixes 2020 week 31
    • 75a68ee : aco: optimize swizzled SALU 8/16-bit conversions
    • 80c135e : radv: fix build on Android 7 (v2)
    • e1b1148 : nir: Get rid of __[u]int64_to_fp32() and __fp32_to_[u]int64()
    • 025988f : intel: Set int64_options to ~0 when lowering 64b ops
    • 936c58c : nir: Extend nir_lower_int64() to support i2f/f2i lowering
    • bfee35b : nir: Stop passing an options arg to nir_lower_int64()
    • 9e23925 : freedreno: Initialize lower_int64_options to a proper value
    • f1f9fdb : aco: add GFX6/7 subdword lowering tests
    • 000530e : aco/tests: add tests for sub-dword swaps
    • 54394a4 : ci: enable ACO tests
    • d488d0f : aco: add framework for testing isel and integration tests
    • bb7d775 : aco: add a few tests for the assembler and optimizer
    • e6366f9 : aco: add framework for unit testing
    • dd155ae : nir: Allow casts in nir_deref_instr_get[_const]_offset()
    • 6a13823 : nir: Use a switch in build_deref_offset()/deref_instr_get_const_offset()
    • e933ac2 : ci: Generate MinIO credentials within LAVA jobs
    • cb82274 : ci/bare-metal: Capture the first devcoredump a job produces.
    • 3120b4d : ci: Print URL to image diff when a trace replay fails
    • bea34a0 : ci: Upload reference images for traces
    • cb19fe2 : intel/vec4: fix out of bounds read
    • 56228b0 : iris: quiet down static analyzers
    • c3a251f : mesa: quiet down static analyzers
    • 0906d5d : mesa: fix out of bounds access in glGetFramebufferParameterivEXT
    • 28f2585 : util/format: initialize non-important components to 0
    • f13042e : util: fix possible buffer overflow in util_get_process_exec_path
    • 59bb0ff : glsl: catch out of bounds access in the debug version
    • eac0ba7 : util: fix possible fd leaks in os_socket_listen_abstract
    • 62bfc70 : vulkan/util: add struct vk_pipeline_cache_header
    • b5558f2 : freedreno/a6xx: fixup draw state earlier
    • 82b419f : freedreno/a6xx: don't emit a bogus size for empty cb slots
    • ba9d502 : freedreno/ir3: add missing track_ubo_use()
    • d0e32e5 : egl/dri2: only take a dri2_dpy reference when binding a new context/surfaces
    • b9391c1 : freedreno/decode: cffdec warnings cleanup
    • 20e703b : freedreno/rnn: headergen2 warnings cleanup
    • 128cbcd : iris: Delete shader variants when deleting the API-facing shader
    • 9a49d4c : aco: remove isel for GLSL-style barriers
    • cc3bc94 : radv: use scoped barriers
    • a8f8c02 : ac/nir: implement scoped_barrier
    • 6b99cf6 : nir/load_store_vectorize: fix indentation
    • 1a42e7d : nir: Filter modes of scoped memory barrier in nir_opt_load_store_vectorize
    • 5c5555a : nir: Add a find_variable_with_[driver_]location helper
    • fc1363c : nir/gl_nir_linker: Call add_vars_with_modes once for GL_PROGRAM_INPUT
    • 789ec95 : nir/split_per_member_structs: Inline split_variables_in_list
    • d70fff9 : nir: Use a single list for all shader variables
    • 473b0fc : gallium/ttn: Use variable create/add helpers
    • a41a84d : mesa/ptn: Use nir_variable_create
    • e5536e4 : aco: Use nir_foreach_variable_with_modes to walk SSBOs
    • a61be31 : panfrost: Use nir_foreach_variable_with_modes in pan_compile
    • 94f0bae : panfrost/midgard: Make search_var take a nir_shader and mode
    • 568022b : r600/sfn: Use nir_foreach_variable_with_modes in IO vectorization
    • cc51cec : r600/sfn/lower_tex: Get rid of the lower_sampler vector
    • 9296730 : r600/sfn/lower_tess_io: Rework get_tcs_varying_offset
    • e4d812d : freedreno/ir3_cmdline: Rework i/o variable fixup
    • ce6e59b : lima/standalone: Rework i/o variable fixup
    • 52dd84a : freedreno/ir3_lower_tess: Rework var list helpers
    • 77c41ce : nir/gl_nir_linker: Use nir_foreach_variable_with_modes
    • c256cd9 : nir/lower_variable_initializers: Restrict the modes we lower
    • 4d7e064 : nir/split_per_member_structs: Use nir_variable_with_modes_safe
    • 90cf470 : nir/lower_uniforms_to_ubo: Use nir_foreach_variable_with_modes
    • d0d5ef6 : nir/lower_two_sided_color: Use nir_variable_create
    • bb3994c : nir/io_to_vector: Use nir_foreach_variable_with_modes
    • 0a77c67 : nir/lower_io_to_temporaries: Use a separate list for new inputs
    • cd66005 : nir: Use a nir_shader and mode in lower_clip_cull_distance_arrays
    • d5d15c3 : nir/lower_amul: Add a variable mode check
    • 3be0be7 : nir: Split nir_index_vars into two functions
    • 86c9303 : nir/split_vars: Add mode checks to list walks
    • 2f6c263 : st/nir: Rework fixup_varying_slots
    • 4c3a303 : nir/linking: Rework some internal helpers
    • caab46c : nir: Take a shader and variable mode in nir_assign_io_var_locations
    • 964c1c4 : nir: Take a nir_shader and variable mode in assign_var_locations
    • 6f6f7a3 : nir: Add and use a nir_variable_list_for_mode helper
    • e3e1c50 : nir: Add a nir_foreach_gl_uniform_variable helper for GL linking
    • feb32f8 : nir: Add a nir_foreach_uniform_variable helper
    • 92dcda5 : nir: Add a nir_foreach_function_temp_variable helper
    • 2956d53 : nir: Add nir_foreach_shader_in/out_variable helpers
    • 9bf8572 : nir/dead_variables: Respect the modes passed to remove_dead_vars
    • 5746af4 : nir: Take a mode in remove_unused_io_vars
    • fd91744 : docs: change "Fixes:" tag example to match git fixes output
    • aa89693 : freedreno/rnn: add schema validation
    • 3169b8c : freedreno/rnn: schema updates for dynamic/irregular offsets
    • c537947 : freedreno/registers/mdp5: fix validation error
    • 0586238 : freedreno/rnn: fix use-group
    • 14a7ca5 : freedreno/rnn: allow name to be optional in arrays
    • e801069 : freedreno/rnn: add "addvariant" to schema
    • b0e3ef5 : freedreno/rnn: describe copyright element in schema
    • 09a3a2c : freedreno/registers/adreno_pm4: fix validation errors
    • c61ad1f : freedreno/registers/a4xx: fix validation error
    • 17997d5 : freedreno/registers/a2xx: fix validation error
    • 64f1122 : freedreno/rnn: add variants/varset to domain
    • 7b76e5f : freedreno/rnn: relax Hexadecimal to HexOrNumber
    • cb3ed46 : freedreno/rnn: add radix/align
    • 83bc96b : freedreno/rnn: add high/low/pos to registers
    • 61ae65a : freedreno/rnn: add relaxed boolean type
    • e3958ef : freedreno/rnn: update schema for 'pos'
    • dfe9746 : freedreno/rnn: rename schema file
    • 5a8d19c : freedreno/rnn: add error helper
    • 3a91083 : freedreno/rnn: split out helper to find files
    • 751af16 : freedreno/tools: check rnn parse status
    • 6c8b921 : ci: Build kernels and rootfs for x86 devices
    • 5d0ba8b : ci: Split building of libdrm to its own script
    • 8a8afcb : ci: Don't ship vk-build-programs after building dEQP
    • 5262cd8 : ci: Fix URL for glslang
    • 7455403 : ci: Print load stats after running dEQP
    • 735ad2d : ci: Always print status code of HTTP uploads in tracie
    • 903a7d0 : freedreno: Add trace for CP_DRAW_INDIRECT_MULTI
    • 85e4305 : freedreno/cffdec: Handle CP_DRAW_INDIRECT_MULTI like other draws
    • 4b94053 : freedreno/rnn: Support stripes in rnndec_decodereg
    • 9a1924d : tu: Dump CP_DRAW_INDIRECT_MULTI draw BO's
    • c7626ac : bifrost: Fix packing of ADD_FEXP2_FAST
    • a618631 : bifrost: Honor src swizzle in special math ops
    • a400c2f : prevent multiply defined symbols
    • ccfe981 : aco: create acq+rel barriers instead of acq/rel
    • 3d9eb17 : aco: improve workgroup-scope and lower vmem/smem barriers
    • 3af2b9e : aco: improve sync_info for TCS output stores
    • 8a16498 : aco: use storage_scratch
    • 46c4b25 : aco: enable value numbering of s_buffer_load_*
    • 2adb337 : nir,radv/aco: add and use pass to lower make available/visible barriers
    • 7a61480 : aco: consider intrinsic access in visit_{load,store}_image
    • cd392a1 : radv/aco,aco: use scoped barriers
    • d1f992f : aco: rework barriers and replace can_reorder
    • 1bbb64f : aco: add missing add_to_hazard_query
    • 1882b1e : bifrost: Add support for nir_op_iabs
    • 1cdd161 : anv: fix descriptor set free
    • 845a50e : nine: fix incorrect calculation of layer count for 3D textures
    • d136fa1 : panfrost: Allow PIPE_TEXTURE_1D_ARRAY textures
    • 8847183 : i965: propagate error from gen_perf_begin_query to glBeginPerfQueryINTEL
    • 627c019 : iris: propagate error from gen_perf_begin_query to glBeginPerfQueryINTEL
    • 0f9131d : freedreno/rnn: Return success when parsing addvariant
    • aa6fbdd : freedreno/ci: add a2xx trace to CI job
    • 18bc5a8 : freedreno: deduplicate a2xx disasm
    • f39afda : freedreno: move a2xx disasm out of gallium
    • f7bd345 : freedreno: deduplicate a3xx+ disasm
    • 6b379a4 : freedreno: drop shader_t
    • bb98b71 : freedreno/ir3: split out regmask
    • ddcee24 : freedreno: add CI for envytools tools
    • b62e4a8 : freedreno/afuc: warnings cleanup
    • 5125b4b : freedreno/decode: warnings cleanup
    • cbbaafd : freedreno/rnn: warnings cleanup
    • 536f43c : freedreno: slurp in afuc
    • 1ea4ef0 : freedreno: slurp in decode tools
    • 7c0bd84 : freedreno: slurp in rnn
    • b721d33 : freedreno: slurp in rnndb
    • 7de0842 : freedreno: make gen_header.py check parent directory
    • 05b2783 : radv: Fix host-host signalling with legacy timeline semaphores.
    • 9746e84 : radv: cleanup locking around timeline waiting.
    • a0a7087 : bifrost: Add support for nir_op_imul
    • 718d444 : bifrost: Add support for nir_op_uge
    • 946ff9b : bifrost: Add support for nir_op_ishl
    • 539ea08 : bifrost: Add support for nir_op_inot
    • 12dd545 : spirv: Handle most execution modes earlier
    • ef78188 : bifrost: Add lowering for b2i32
    • 1a168c9 : bifrost: Document d3d/gl comparison control bit
    • ec37c71 : bifrost: Emit "d3d" variant of comparison instructions
    • 0ffefad : bifrost: Lower x-bool conversions to != 0
    • 8e8baec : tu: Enable resource dynamic indexing
    • 8bc060a : ir3: Fix incorrect src flags for samp_tex
    • e73a8a2 : ir3: Remove redundant samp_tex validation
    • 3adc23f : ir3: Validate bindless samp_tex correctly
    • d542bfc : tu: Fix descriptor update templates with input attachments
    • 9ece612 : turnip: fix SP_HS_UNKNOWN_A831 value for A650
    • e646e77 : turnip: use patchControlPoints for HS_INPUT_SIZE value
    • da49a45 : turnip: move WFI out of draw state to fix a650 hangs
    • e5f4527 : freedreno/ir3: fix wrong local_primitive_id_start type
    • 7cb4d4f : vulkan/wsi: Convert usage of -1 to UINT32_MAX.
    • e1147ca : vulkan/wsi/x11: report device-group present rectangles with prime.
    • af0bc71 : radv: call radv_nir_lower_ycbcr_textures after first optimizations
    • fffc287 : radv: link with ld_args_build_id
    • d1677c8 : v3d/compiler: request fragment shader clip lowering to be vulkan compatible.
    • 71d5c19 : v3d/compiler: handle compact varyings
    • 17fd191 : nir/lower_clip: make the pass compatible with Vulkan semantics
    • 18fe130 : radv: Fix uninitialized variable in renderpass.
    • e2e89fb : nir/lower_io: assert that offsets are used for shader_in
    • 60915f8 : etnaviv: do register setup only once
    • 7ee146a : etnaviv: move shader_count to etna_compiler
    • 5839a7d : etnaviv: introduce struct etna_compiler
    • 9e596cc : tu: Enable vertex & fragment stores & atomics
    • f7f29a0 : tu: Detect invalid-for-binning renderpass dependencies
    • d6d75fc : tu: Fix hangs for DS with no output
    • 7ad962b : tu: Fix empty blit scissor case
    • 63cf8ad : spirv: Also copy over binding information for atomic counters
    • 6cbdffd : tu: Implement VK_KHR_draw_indirect_count
    • 52ec35f : tu: Add missing wfi to tu6_emit_hw()
    • a0ca688 : tu: Integrate WFI/WAIT_FOR_ME/WAIT_MEM_WRITES with cache tracking
    • cd78a7a : freedreno: Add INDIRECT_COUNT CP_DRAW_INDIRECT_MULTI variants
    • 8da31ee : freedreno: Clean up CP_DRAW_MULTI_INDIRECT definition
    • 1747f9f : turnip: remove extra gmem alignment
    • 8534f49 : etnaviv: explicitly set nir_variable_mode
    • 5b300be : radv: clean up remaining pipeline init functions
    • 5575ce0 : radv: remove useless return value to radv_pipeline_scratch_init()
    • 0721e2d : radv: add radv_pipeline_init_shader_stages_state()
    • f3774ec : radv: constify all radv_pipeline_generate_*() helpers
    • f32cb82 : radv: assign pipeline gfx fields before PM4 emission
    • 2a5fb87 : radv: clean up binning state initialization
    • 08dd704 : radv: clean up adjusting MSAA state if conservative rast is enabled
    • 067b01c : radv: add radv_pipeline_generate_vgt_gs_out()
    • 749d513 : radv: add radv_pipeline_init_input_assembly_state()
    • e31f8b9 : radv: clean up tessellation state emission
    • 9b691bf : radv: remove unnecessary radv_tessellation_state::lds_size
    • a1b237b : radv: set LDS TCS size at shaders creation for GFX9+
    • cf89bdb : radv: align the LDS size in calculate_tess_lds_size()
    • 9c36ebf : radv: remove one unnecessary param to radv_generate_graphics_pipeline_key()
    • 55f6156 : radv: remove no-op si_multiwave_lds_size_workaround()
    • 7f79413 : radv: remove unnecessary radv_tessellation_state::num_patches
    • 83f63ab : radv: clean up radv_compute_generate_pm4()
    • 8bfd827 : radv: reduce the number of allocated dwords for compute CS
    • 62ec897 : radv: clean up PA_SC_CLIPRECT_RULE emission
    • 09c4f76 : radv: clean up VGT_SHADER_STAGES_EN emission
    • 62ffa83 : radv: emit PA_SC_LINE_CNTL as part of the rasterization state
    • 934d6ac : radv: emit more invariant registers as part of the initial gfx state
    • 847e0b8 : radv: remove outdated TODO related to PA_SU_VTX_CNTL.PIX_CENTER
    • 62af9df : radv: remove set but unused radv_pipeline::vertex_elements
    • 76de141 : radv: remove declared but unused radv_pipeline::is_dual_src
    • 8596873 : bin/khronos-update: add workaround for python bug 9625
    • aa5c391 : bin/khronos-update: add support for the SPIRV files
    • ccb91bc : bin/khronos-update: having a folder in include/ is not a requirement
    • de5130f : v3d: Retry with the fallback scheduler when RA fails
    • 1c8167d : v3d: Changed v3d_compile:failed to an enum
    • 56846a2 : nir/schedule: Add an option for a fallback scheduling algorithm
    • 08f1746 : v3d: Mark scheduling dependency for prim id and first output
    • bafd259 : nir/schedule: Add a callback for backend-specific dependencies
    • 260a8f7 : nir/schedule: Store a pointer to the options struct in scoreboard
    • 7665398 : nir/scheduler: Move nir_scheduler to its own header
    • 14a12b7 : spirv: Rework our handling of images and samplers
    • 196db51 : anv,turnip,radv,clover,glspirv: Run nir_copy_prop before nir_opt_deref
    • 66c8628 : spirv: More heavily use vtn_ssa_value in function parameter handling
    • 32ba23f : spirv: Remove a dead case in function parameter handling
    • 9e3213a : spirv: Add a helper for getting the NIR type of a vtn_type
    • 051f8d3 : spirv: Give atomic counters their own variable mode
    • 8a7932b : spirv: Drop the sampled boolean from vtn_type
    • d0be2fe : spirv: Add better checks for SSA value types
    • 47ebb50 : spirv: Hand-roll fewer vtn_ssa_value creations
    • 085ade4 : spirv: Simplify vtn_ssa_value creation
    • 00af112 : spirv/subgroups: Refactor to use vtn_push_ssa
    • ea246c3 : spirv/subgroups: Stop incrementing w
    • c5fcd12 : spirv/glsl450: Use vtn_push_ssa_value
    • 7560ed2 : spirv/alu: Use vtn_push_ssa_value
    • dbb4a24 : spirv: Refactor vtn_push_ssa
    • 8be9f2a : spirv: Use the new helpers in OpConvertUToPtr/PtrToU
    • 37ab323 : spirv: Add a vtn_get_nir_ssa helper
    • e5b29b9 : spirv/amd: Use vtn_push_nir_ssa
    • d8eb6f2 : spirv: Add a vtn_push_nir_ssa helper
    • aaf1f34 : spirv: Rename push_value_pointer to push_pointer
    • ac1e6d5 : spirv: Add a helpers for getting types of values
    • 953b7a3 : spirv: Use nir_bany/ball for OpAny/All
    • 8dfee57 : spirv: Clean up OpSignBitSet
    • 62c53ad : spirv: Fix indentation in vtn_handle_ptr
    • 516fd78 : spirv: Drop the void *ptr from vtn_value
    • af81486 : spirv: Simplify our handling of NonUniform
    • 0d5cd1a : nir/vtn: Add support for 8 and 16 vector ball/bany
    • 456edf0 : nir: Support 8 and 16 component vectors for reduceable intrinsics
    • d572f4d : nir: Support algebraic opts on vectors larger than 4
    • aa581fc : nir: Support vec8/vec16 in nir_lower_bit_size
    • d35b54c : freedreno: sync registers from envytools
    • 1610c69 : tu: Enable VK_EXT_depth_clip_enable
    • 1b3be07 : aco: ensure readfirstlane subdword operands are always dword aligned
    • 4c89bfc : aco: prevent infinite recursion in RA for subdword variables
    • 626081f : aco: don't split store data if it was already split into more elements
    • bd75e99 : aco: ensure to not extract more components than have been fetched
    • 6bc5ce7 : radv: Add timeline syncobj for timeline semaphores.
    • 55d8022 : radv: Add winsys functions for timeline syncobj.
    • 66e598d : radv: Add winsys support for submitting timeline syncobj.
    • fb6b38d : radv: Add thread for timeline syncobj submission.
    • fa97061 : radv/winsys: Add binary syncobj ABI changes for timeline semaphores.
    • ce4064f : nir/ lower_int_to_float: Handle umax and umin
    • 589d866 : ci: Use half as many parallel softpipe / virgl test jobs
    • d9693c6 : ci: Do not mark container / pages jobs as interruptible
    • e74933e : ci: Use FDO_CI_CONCURRENT in run-shader-db.sh as well
    • 429ec82 : ci: Namespace trace artifacts to the job number
    • 772ed65 : nir_ allow nir_lower_clip_halfz to run in tess eval shader
    • 09ecfd9 : nir: allow lower_psiz_mov to run in tessellation stages
    • fced3c4 : Revert "llvmpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS."
    • be6c53b : llvmpipe/ms: fix sign extension bug in rasterizer.
    • 4d73988 : intel/ir/gen12+: Work around FS performance regressions due to SIMD32 discard divergence.
    • 45d159c : glx: Fix build and warnings with -Dglx=dri -Dglx-direct=false
    • ba22f01 : softpipe: Enable PIPE_CAP_TGSI_ANY_REG_AS_ADDRESS;
    • 8e61fd9 : softpipe: Enable PIPE_CAP_TGSI_TEXCOORD.
    • 259a03b : softpipe: Add support for reporting shader-db output.
    • 991def0 : softpipe: Convert to comma-separated SOFTPIPE_DEBUG for debug options.
    • 86cfb62 : softpipe: Refactor pipe_shader_state setup.
    • e67da8d : llvmpipe: enable robust buffer access + GL 4.3, GLES 3.2 and robust buffer access behaviour
    • 6d3cefe : llvmpipe: add device reset query context hook.
    • 3cb3d17 : glx/drisw: add robustness support
    • 80f7b58 : drisw: add robustness extension support.
    • 12c06a0 : llvmpipe/draw: handle constant buffer limits and robustness (v1.1)
    • f0c3a25 : llvmpipe: enable EXT_texture_shadow_lod
    • c30824a : nir/lower_io: Add support for global scratch addressing
    • 4815ae5 : nir/lower_io: Use b2b for shader and function temporaries
    • 3a2975d : nir/lower_io: Choose to set access based on intrinsic metadata
    • c475e29 : nir: Allow for system values with variable numbers of destination components
    • 0338db5 : docs/releasing: improve wording
    • ae2d045 : bin/gen_release_notes: automatically commit release notes
    • 5f649be : post_version.py: fix relnotes links
    • 6c4ad62 : post_version.py: update the files in the current worktree, not the one with the script that we run
    • a28a089 : post_version.py: stop using non-existent functions and fix commit message
    • f5353e0 : post_version.py: drop incorrect conf.py changes
    • 24e118f : post_version.py: don't generate relnotes twice
    • 04e38eb : docs: update calendar and link releases notes for 20.1.4
    • 8a44983 : docs: add release notes for 20.1.4
    • b6b100c : gitlab-ci: Test AMD's Raven with traces
    • eb7642c : i915: Remove a bunch of default handling of pipe caps.
    • e35d3d2 : svga: Remove a bunch of default handling of pipe caps.
    • c32f723 : swr: Use the default behavior of ALLOW_MAPPED_BUFFERS.
    • 39598a1 : swr: Remove a bunch of default handling of pipe caps.
    • e2ffd21 : virgl: Remove a bunch of default handling of pipe caps.
    • e5554e3 : softpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS.
    • 855f3ff : softpipe: Remove a bunch of default handling of pipe caps.
    • 6ec4906 : llvmpipe: Use the default behavior of ALLOW_MAPPED_BUFFERS.
    • ae919b2 : llvmpipe: Remove a bunch of default handling of pipe caps.
    • 292882f : ci: Fix the overwriting of traces.yml for baremetal
    • 262731b : ci: Update checksums for freedreno traces.
    • 4664612 : radeon/vcn: increase render_pic_list size
    • 89d2dac : radeonsi: enable preemption if the kernel enabled it
    • 9e2113c : radeonsi: set up IBs for preemption
    • b8892bc : radeonsi: don't restore states at the beginning of IBs if they're shadowed
    • 95c9048 : radeonsi: add debug code for register shadowing
    • 8af0f91 : radeonsi: add reg shadowing codepaths to GS and tess ring setup
    • 69014d8 : radeonsi: implement CP register shadowing
    • b84dbd2 : radeonsi: reorder code in update_gs_ring_buffers and init_tess_factor_ring
    • babd87f : radeonsi: make cs_preamble_state optional
    • 7a6af4c : winsys/amdgpu: make amdgpu_bo_unmap non-static
    • 5a5467c : ac: add tables for CP register shadowing
    • dc3dade : ac: add helper ac_get_register_name
    • 976edae : radeonsi: sort registers in si_init_cs_preamble_state according to GPU gen
    • 88fe9de : radeonsi: sort registers in si_emit_initial_compute_regs according to GPU gen
    • 1c6eca2 : radeonsi/gfx10: set the correct value for OFFCHIP_BUFFERING
    • d244a25 : radeonsi: add missing initialization of registers
    • fd20e98 : docs: add some very basic documentation about zink
    • 202592a : radv/winsys: be more robust when a CS failed during recording
    • f82eb7a : radv/winsys: return a Vulkan error code when binding virtual buffers/images
    • 4b58e35 : radv/winsys: remove useless check when binding virtual buffers/images
    • f0112fa : radv/winsys: check more allocation failures
    • 7a61e31 : radv: add missing return values check for some winsys calls
    • 348e8b5 : nir/tests: Add tests for opt_if_simplification
    • 6f94b3d : nir/opt_if: Fix opt_if_simplification when else branch has jump
    • 3f783a3 : zink: omit Lod image operand in ntv when not using an image texture dim
    • 0ef5e19 : zink: add some asserts for building access chains in ntv
    • 2af2205 : zink: handle texelFetchOffset with offsets
    • 6587f41 : zink: use helper function to handle uvec/bvec types
    • 7015d2c : aco: fix scratch loads which cross element_size boundaries
    • c33e8d7 : mesa/program: fix shadow property for samplers
    • 4b04800 : ci: Disable trace testing on Mali T760
    • 6c1108d : radv: advertise VK_EXT_shader_atomic_float
    • b8517e5 : ac/nir: add support for nir_intrinsic_shared_atomic_fadd
    • 7615f2d : aco: add support for nir_intrinsic_shared_atomic_fadd
    • 97482c6 : ci: Test with more traces
    • 87e0b2c : ci: Prefix tracie artifacts with the device name
    • ed5374f : ci: Use smaller glxgears trace
    • dfe394b : ci: Upload images of failed replays to MinIO
    • 7dc4fe6 : nvc0: add documentation for nve4+ (Kepler) COPY class
    • 203d565 : nvc0: fix macro define for NVE4_COPY()
    • 3a4024e : anv: properly handle fence import of sync_fd = -1
    • 323d5bb : meson: Add mising git_sha1.h dependency.
    • 12b99d2 : aco: fix includes in aco_ir.cpp
    • 8fff8af : turnip: disable tiling for NV12/IYUV formats
    • b559d26 : freedreno/ir3: Fix SSBO size for bindless SSBO's
    • c11c4d8 : aco: fix copy of uninitialized boolean
    • 40b65a8 : radv: fix invalid conversion warnings in vk_format.h
    • fcd8f69 : aco: print ACO IR before scheduling instead of after
    • bf4b377 : aco: make validate() usable in tests
    • e75946c : aco: move some setup code into helpers
    • 2907fae : egl/dri2: try to bind old context if bindContext failed
    • 8b0b6f9 : dri2: do not conflate unbind and bindContext() failure
    • 6b12999 : dri2: dri2_make_current() fold multiple if blocks
    • 29c39ae : aco: use nir_addition_might_overflow to combine additions into SMEM
    • 72ac3f6 : nir: add nir_unsigned_upper_bound and nir_addition_might_overflow
    • 2694a34 : aco: add NUW flag
    • 3a48471 : aco: allow overflow for some SMEM instructions
    • d169f09 : aco: be more careful combining additions that could wrap into loads/stores
    • 4d0e062 : gitlab-ci/traces: updated paths and checksums for POLARIS10 traces
    • f1e67aa : gitlab-ci: get the last frame from a gfxr trace using gfxrecon-info
    • d15e122 : gitlab-ci: build gfxreconstruct from the "dev" branch
    • bb5b136 : freedreno: Use common guardband helper
    • c9c848d : tu: Use common guardband helper
    • 19895dd : freedreno: Add a helper for computing guardband sizes
    • 86a6597 : panfrost: Remove unused batch_fence-ctx
    • f18e537 : panfrost: Remove unused batch_fence-signaled
    • 64d6f56 : panfrost: Allocate syncobjs in panfrost_flush
    • 85a2216 : panfrost: Skip specifying in_syncs
    • e5ef5a3 : panfrost: Remove wait parameter to flush_all_batches
    • 0c4db88 : panfrost: Avoid wait=true flushing all batches
    • 04ea4f1 : aco: implement b2i8/b2i16
    • 3a7cd7b : nv50/ir: initialize persampleInvocation to false
    • 618b355 : nv50/ir/tgsi: silence warning about unhandled GS_INPUT_PRIM property
    • 6ced98c : radv: disable CPU caching for the upload BO to reduce fetch latency
    • b3eae4e : radv: do not perform read-modify-write with the upload BO
    • d9072a1 : radv: replace discard with demote for Quantic Dream games
    • 51bc11a : aco: always set FI on GFX10
    • 8b3452a : ci: Set XDG_CACHE_HOME to tmpfs for bare-metal runners to avoid NFS.
    • b2cd6a0 : gitlab-ci: Fix needs: of the arm64 LAVA test jobs
    • a1947f0 : gitlab-ci: Upload tracie artifacts to MinIO
    • 20507f8 : gitlab-ci: Download traces from MinIO
    • 087be7e : gitlab-ci: Replay traces on lava devices
    • 576c53d : iris: Fix CCS check in iris_texture_subdata().
    • 0868638 : nir/lower_int64: lower 64-bit amul
    • 4d44848 : anv: Advertise support for VK_EXT_shader_atomic_float
    • 675d7b1 : intel/fs: Use the correct logical op for global float atomics
    • 84086b6 : spirv: Add support for SPV_EXT_shader_atomic_float
    • 2a568c5 : spirv: Update headers and grammar json
    • cc03448 : egl: inline _EGLAPI into _EGLDriver
    • 7b7917a : radeonsi: Inhibit clock-gating for perf counters.
    • 36e6ac6 : anv: Advertise VK_EXT_image_robustness
    • d973e50 : freedreno/ir3: Add missing ld_args_build_id to the ir3_delay unit test.
    • 3688da2 : radv: advertise VK_EXT_image_robustness
    • 096adbe : ci: bare-metal: use nginx to get results from DUT
    • 101400d : mesa: change error code of *TextureSubImage* for incorreect target
    • af92348 : freedreno/ir3: Fix disasm of register offsets in ldp/stp.
    • d6d8dc1 : freedreno/ir3: Refactor cat6 general dst printing.
    • 62dcf75 : freedreno/ir3: Add a bunch more tests for cat6 opcodes.
    • ed3338f : freedreno/ir3: Add a note about the instructions in the disasm test.
    • 4ab3a21 : vulkan: Update Vulkan XML and headers to 1.2.148
    • fd24a95 : ci: Use FDO_CI_CONCURRENT as our -j flags when present in the runner env.
    • 5f79e4e : nir/algebraic: fold some nested bcsel
    • 2724466 : nir/algebraic: propagate b2i out of ior/iand
    • baee5a9 : nir/algebraic: add distributive rules for ior/iand
    • 70d3efe : nir/algebraic: optimize (a 0.0) ? -a : a - fabs(a)
    • 9d22c5e : nir/algebraic: optimize fmul(x, bcsel(c, -1.0, 1.0)) - bcsel(c, -x, x)
    • 56ec814 : nir/algebraic: add some more unop + bcsel optimizations
    • 2fca183 : nir/algebraic: add optimizations for fsign/isign
    • 8e1b75b : nir/algebraic: optimize iand/ior of (n)eq zero
    • e4281db : nir: also move b2i in case of nir_move_copies
    • de0ebaf : nir/algebraic: optimize bcsel(a, 0, 1) to b2i
    • e764192 : pan/mdg: Use the blend RT for blend shader framebuffer fetches
    • 3ec252a : panfrost: 8x MRT support
    • 978f963 : panfrost: Use more tilebuffer sizes
    • c1d3d39 : panfrost: Fake RGTC support
    • fac813d : spirv: don't split memory barriers
    • 28c227c : radv/winsys: always allow GTT placements on APUs
    • d1bba2e : radv: disable CPU caching for IBS to reduce fetch latency
    • d2a3ca2 : radeonsi: adjust epitch for PIPE_FORMAT_R8G8_R8B8_UNORM
    • 87ecfdf : ac/surface: adapt surf_size when modifying surf_pitch
    • 1fa36c1 : d600/sfn: write stream outputs to correct mem ring
    • 21d296a : r600/sfn: Make the pin_to_channel generic
    • 3ea847e : r600/sfn: Only use sample mask if the according shader key is set
    • c91979c : r600: Add shader key item to identify when the sample mask should be used
    • 05df4bf : r600/sfn: Fix default z swizzle for GDS instructions
    • 2779aa3 : r600/sfn: Fix Ring output swizzle masks
    • c18b1c6 : r600/sfn: Add a forced output swizzle for depth write
    • d31ef0b : r600/sfn: correct handling of loading vec4 with fetching constants
    • aca99e6 : r600/sfn: Add option to get a temp value for a specific channel
    • 2586188 : r600/sfn: emit texture instructions in one block
    • deccf02 : r600/sfn: Pipe through requesting a register at a given channel
    • 55cc712 : r600/sfn: lower rotate ALU ops
    • 4708ccb : ci/llvmpipe: reenable gpu shader5 tests
    • 41c7bb6 : llvmpipe: add framebuffer fetching support (v1.1)
    • 0e0b6d4 : llvmpipe/cs: respect render condition
    • 912ad09 : freedreno/ir3/ra: fix array conflicts for split/merged
    • b1465c3 : freedreno/ir3/ra: assign vreg names to all array elements
    • 6317f7d : freedreno/ir3/ra: debug msgs tweak
    • c2d94aa : freedreno/ir3: fix half-reg array stores
    • 5be171b : freedreno/ir3: set array precision on creation
    • 0472ca2 : freedreno/ir3/parser: half-precision relative regs
    • 79b0651 : freedreno: whitespace fix
    • 835201d : freedreno: small comment re-word
    • 2b34323 : zink: free all ntv allocations after creating shader module
    • adc4f38 : zink: free pipeline cache during program destroy
    • 1ff2d19 : zink: destroy descriptor pools on context destroy
    • 7116dec : zink: destroy gfx program when a shader is freed
    • 9e9c8e2 : android: panfrost/encoder: add libmesa_nir static dependency
    • 0ef168d : panfrost: Fix calls to panfrost_flush_batches_accessing_bo
    • 858cc13 : panfrost: Make panfrost_bo_wait take a wait_readers bool
    • 5b38048 : freedreno/ir3: Add unit tests for derivatives disasm.
    • 3d7d5d2 : freedreno/ir3: Fix duplicated fine derivatives instructions.
    • f302ef3 : aco: use s_waitcnt_depctr to mitigate VMEMtoScalarWriteHazard
    • bcf94bb : aco: properly recognize that s_waitcnt mitigates VMEMtoScalarWriteHazard
    • 33f33bb : meson: Enable GCing of functions and data from compilation units by default.
    • 610af7e : panfrost: Enable FP16 by default
    • e5169b1 : gitlab-ci: re-enable all a630 jobs
    • 4080f8b : freedreno/a2xx: Fix compiler warning in disasm.
    • cf55abe : gitlab-ci: Re-add kernels for bare-metal
    • 2aa507f : nir: Set the alignment for SSBO lowering
    • d735f07 : intel/perf: Move perf query register programming to static tables.
    • d4e5693 : intel/perf: Fix unused var warning in release builds.
    • afe07c7 : intel: Fix release-build warnings about sf_entry_size.
    • 086f760 : zink: use ralloc for spirv_builder as well
    • 810bf7d : zink: pass mem_ctx to ralloc_size-call
    • 35beb93 : zink: use ralloc for plain malloc-calls
    • 12b324b : zink: use ralloc in nir-to-spirv
    • 56d9bcd : radv: enable more float_controls features
    • 23631dd : aco: set tcs_in_out_eq=false if float controls of VS and TCS stages differ
    • b36950a : aco: fix nir_op_f2f16_rtne with non-default rounding modes
    • d14f4fa : aco: flush denormals before fp16 fabs/fneg if needed
    • 305cffa : aco: use s_round_mode/s_denorm_mode
    • ca44c00 : panfrost: Set depth_enabled when stencil is enabled
    • 2b796c7 : radv: return better Vulkan error codes when VkQueueSubmit() fails
    • 1829bdd : radv: improve the error messages when a CS submission failed
    • 93047f6 : radv: remove one useless goto in radv_queue_submit_deferred()
    • 0114748 : panfrost: Report TEXTURE_BUFFER_OBJECTS cap when gl3 flag set
    • 4e98656 : nir: Fix lower_two_sided_color when the face is an input
    • 314ba5e : nir: Add a face_sysval argument to nir_lower_two_sided_color
    • 2a6db94 : panfrost: Do per-sample shading when outputs are read
    • c20d166 : pan/mdg: Do per-sample framebuffer loads
    • 25747ce : panfrost: Rename lower_store to is_blend in pan_lower_framebuffer
    • 4a8ad1e : pan/mdg: Don't disassemble blit shaders
    • 3f86564 : gitlab-ci: Only trigger test-docs job automatically for MRs
    • 575ab30 : virgl: set PIPE_CAP_BLEND_EQUATION_ADVANCED
    • a0f42b8 : virgl: Encode barrier for blend_equation_advanced
    • a026364 : virgl: Use alpha_src_factor to store blend_equation_advenced value
    • 0ee1a67 : glsl_to_ir: do lower_blend_equation if PIPE_CAP_FBFETCH
    • d0df56c : st: expose KHR_blend_equation_advanced if PIPE_CAP_BLEND_EQUATION_ADVANCED
    • 377731e : gallium: Add PIPE_CAP_BLEND_EQUATION_ADVANCED
    • 57174c9 : virgl: Reserved last caps of capability_bits
    • 4f8fc0f : glsl_to_tgsi: Set TGSI_PROPERTY_FS_BLEND_EQUATION_ADVANCED
    • b3a3f7c : gallium: add TGSI_PROPERTY_FS_BLEND_EQUATION_ADVANCED
    • 30f4ccf : mesa: treat Color._AdvancedBlendMode as enum
    • 6ffa0e9 : mesa: do not use bitfields for advanced-blend state
    • c36aac5 : st/mesa: fix st_CopyPixels without support for stencil exports
    • e7fd1ce : nvc0_2d: Document SET_PIXELS_FROM_MEMORY_CORRAL_SIZE from rnndb
    • bc69e73 : nv50_2d,nvc0_2d: Document SET_PIXELS_FROM_MEMORY_SAFE_OVERLAP from rnndb
    • 9123458 : docs/features: Mark trivial missed feature
    • e0ef5a5 : egl: mark the rest of the callbacks as mandatory or optional
    • 4dc322c : egl: drop now empty egl_dri2_fallbacks.h
    • bc38fe8 : egl: inline fallback for get_sync_values
    • 2e3eb0c : egl: inline fallback for create_wayland_buffer_from_image
    • c5fb1fb : egl: inline fallback for query_buffer_age
    • 90000b0 : egl: inline fallback for copy_buffers
    • 2d5f12a : egl: inline fallback for post_sub_buffer
    • 1ba5075 : egl: inline fallback for swap_buffers_region
    • 7d5a13e : egl: inline fallback for swap_buffers_with_damage
    • b501ece : egl: drop unused fallback function
    • 43e2d80 : egl: inline fallback for create_pbuffer_surface
    • 179e442 : egl: inline fallback for create_pixmap_surface
    • 81124d8 : driconf: allowlist/denylist
    • 01d60d3 : pan/decode: Open the dump file later
    • 4dcbad4 : gitlab-ci: Run all of GLES3 tests for Panfrost
    • 4e8d7ad : radv: split fence into two parts as enum+union.
    • 56395a8 : radv: optimize creating signaled syncobj with amdgpu_cs_create_syncobj2()
    • 3b7cd73 : radv: fix the error code when allocating a fresh imported syncobj fails
    • dd795ee : radv: fix the error code when exporting a semaphore/fence fails
    • 8aa9d0a : radv: fix destroying the syncobj when exporting a fence FD
    • b5a48a9 : tu: Enable VK_EXT_shader_stencil_export
    • aeca92e : ir3: Handle gl_FragStencilRefARB
    • 981608a : freedreno/a6xx: Add stencilref register info
    • 40b99bb : panfrost: Revert "Disable frame throttling"
    • 19da812 : panfrost: Enable Chromium
    • 96fa8d7 : panfrost: Report CAPs more honestly
    • afa4b32 : panfrost: Fix faults with RASTERIZER_DISCARD
    • 6c6a8b2 : panfrost: Honour cso-compare_mode
    • 9addb82 : panfrost: Avoid integer underflow in rt_count_1
    • 77bb19e : panfrost: Abort on unsupported blit
    • cce3d92 : panfrost: Implement Z32F_S8 blits
    • 6ffebfb : panfrost: Fix sRGB clear colour packing
    • 721b5c6 : panfrost: Set PIPE_CAP_MIXED_COLORBUFFER_FORMATS
    • a0003c3 : panfrost: Overhaul tilebuffer allocations
    • 3d13870 : panfrost: Call util_blitter_save_fragment_constant_buffer_slot
    • e16f59c : llvmpipe: fix position offset interpolation
    • 87e2754 : llvmpipe: fix stencil only formats.
    • 045711d : radeonsi: use PIPE_FORMAT_P010 for 10-bit VP9 decoding
    • b85ef04 : aco: add add_interference() helper
    • 2c7554f : aco: use unordered_set for spill id interferences
    • 47d7e1e : aco: rewrite graph coloring in spiller
    • 5a941f4 : aco: fix underestimated pressure in spiller when a phi has a killed def
    • 293f251 : panfrost: Use Midgard-specific reloads
    • f611af3 : st/mesa: use fragment shader to copy stencil buffer
    • 3af0711 : mesa/main: use p_atomic_inc_return instead of locking
    • a0f8439 : st/mesa: fix corrupted texture levels, when adding more levels than expected
    • b8c0196 : gallium/util: do not use _MTX_INITIALIZER_NP on Windows
    • 485ea7d : radv/winsys: pass the buffer list via the CS ioctl for less CPU overhead
    • 40bea60 : radv/winsys: replace alloca() by malloc() everywhere
    • 97f8ec3 : v3d/compiler: Lower geometry output store base into offset src
    • 081691b : radeonsi: prevent a gfx10_ngg_calculate_subgroup_info failure for TES+NGG GS
    • 5714cd3 : ci: bump piglit checkout for dsa tests
    • 854dbea : mesa: change dsa texture error codes for GL 4.6
    • 34a0310 : panfrost: Extract panfrost_batch_reserve_framebuffer
    • 5d0d8fa : panfrost: Track surfaces drawn per-batch
    • 64734c0 : panfrost: Set zs_samples as necessary
    • 8225604 : panfrost: Handle per-sample shading
    • 080b751 : panfrost: Add rectangle subtraction algorithm
    • e061bf0 : panfrost: Identify zs_samples field
    • 6088891 : panfrost: Include sample count in payload estimates
    • 5c65a27 : panfrost: Add MALI_PER_SAMPLE bit
    • adacf1f : panfrost: Expose panfrost_get_blend_shader
    • 528e132 : panfrost: Force Z/S writeback
    • da2eed3 : pan/mdg: Implement gl_SampleID
    • b2749c1 : pan/mdg: Identify per-sample interpolation mode
    • 59308a3 : pan/mdg: Bump compiler RT maximum
    • 29849ac : Android: Fixes for Q and R
    • 2adb13f : llvmpipe/format: fix snorm conversion
    • 75e01d0 : gallivm/sample: always square rho before fast log2
    • a9c9486 : nv50_2d: regenerate envytools-based rnndb headers
    • 0859dcb : radv: destroy the base object if VkCreateInstance() failed
    • 50fdefc : radv: destroy the base object if VkAllocateCommandBuffers() failed
    • e5f2bf3 : radv: destroy the base object if VkCreateFence() failed
    • d25764d : radv: destroy the base object if VkCreateSemaphore() failed
    • ce7a7ae : radv: destroy the base object if VkCreateEvent() failed
    • 8ef5297 : radv: destroy the base object if VkCreateBuffer() failed
    • 8523164 : radv: destroy the base object if VkCreateImage() failed
    • 2e59680 : radv: destroy the base object if VkCreateRenderPass*() failed
    • 0eec81d : radv: destroy the base object if VkCreateQueryPool() failed
    • 25bacea : mesa/st: release debug_output after destroying the context
    • 7f0b6a5 : mesa: add bool param to _mesa_free_context_data
    • e6f7b43 : mesa: rename _mesa_free_errors_data
    • cbdb87c : gitlab-ci: Fix "triggered by Marge for a merge request" rule
    • 559b26b : intel/ehl: Add new PCI-IDs
    • 7cb2ace : intel/ehl: Rename gen_device_info struct
    • 13c7093 : intel/ehl: Use macro GEN11_LP_FEATURES in device info
    • e08ec89 : intel/ehl: Use GEN11_URB_MIN_MAX_ENTRIES in device info
    • 787c1ed : panfrost: Dual source blend support
    • 334dab0 : pan/mdg: Skip z/s combining for dual-source writes
    • 85954ec : pan/mdg: Dual source blend input/writeout support
    • 0ff6263 : pan/mdg: Add a nir pass to reorder store_output intrinsics
    • 58c0e1d : gallium: Dual source support in blend_factor_to_shader
    • bedd4b4 : compiler: Add dual-source factors to blend_factor
    • 7f9039f : freedreno/ir3: DCE unused arrays
    • 2e4bab8 : freedreno/a6xx: don't enable early-z/lrz if no z-test
    • 37e0e07 : freedreno/ir3/ra: be better at failing
    • afadaae : freedreno/a6xx: bail instead of crash for compile fails
    • b3ca55f : freedreno/ir3: make compile fails more visible
    • 788792f : freedreno/ir3: add missing VS driver params
    • 5e9cd64 : zink: enable tgsi texcoord pipe cap
    • 05362b0 : nouveau: expose HMM
    • 212f1ab : nvc0: support PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
    • c0f7f83 : gallium: add PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
    • 08ceba9 : nouveau: enable HMM
    • 1f37662 : ci: bump libdrm to 2.4.102
    • 9660655 : ci: need to install wget in order to download libdrm
    • 0e90b3d : nir: Support load/store of temps as scratch in nir_lower_explicit_io
    • 99aaf0e : nir: When nir_lower_vars_to_explicit_types is run on temps, update scratch_size
    • bf138c1 : nir_lower_io: Add addr_format_is_offset helper
    • d00487d : freedreno/regs: update a6xx PC regs
    • 2e32a20 : freedreno/regs: update a6xx VPC regs
    • e883aa2 : freedreno/regs: update a6xx RB regs
    • a5c6685 : freedreno/regs: update a6xx GRAS registers
    • be33d58 : gitlab-ci: re-enable arm64_a630_vk
    • f37f1a1 : turnip: remove use of tu_cs_entry for draw states
    • 7f24a69 : turnip: fix inconsistencies with tu6_load_state_size
    • bf997ca : turnip: emit compute pipeline directly in CmdBindPipeline
    • dce6cb1 : turnip: use DIRTY SDS bit to avoid making copies of pipeline load state ib
    • 2e48830 : zink: try to handle multisampled null buffers
    • 1ee598c : zink: handle empty attachments
    • 6493d29 : pan/mdg: Fix non-debug compiliation
    • 89caa48 : Revert https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4580
    • 460300e : nouveau: expose HMM
    • efa847d : nvc0: support PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
    • dd2cc5b : gallium: add PIPE_CAP_RESOURCE_FROM_USER_MEMORY_COMPUTE_ONLY
    • f67cd0b : nouveau: enable HMM
    • 6f4a6ed : ci: bump libdrm to 2.4.102
    • e1e7584 : zink: block resolve blits for depth/stencil buffers
    • 582669f : zink: block vkCmdBlitImage usage for multi sampled blits
    • ab78831 : zink: try copy_region hook for blits where we can't do a regular blit or resolve
    • 34fe561 : mesa/main: use call_once instead of open-coding
    • 0a4aa61 : mesa/main: factor out one-time-init into a helper
    • 38e3cbb : gv100/ir: set ftz bit on floating point operations
    • f95637f : meson: fix android vulkan build
    • 3ed104b : gitlab-ci: Drop dependencies:
    • aa2457f : gitlab-ci: Remove indirect dependencies from needs:
    • bf1376a : tu: Don't invalidate irrelevant state when changing pipeline
    • a161367 : freedreno/a6xx: Add some documentation for shared consts
    • e1fa740 : freedreno/a6xx: Rename and document HLSQ_UPDATE_CNTL
    • dad042b : clover: implements clEnqueueFillBuffer
    • fea109d : clover: add more cl_mem_object_type to pipe_texture_target mapping
    • ac642d8 : radv: add the custom border color BO to the list of buffers
    • 684cfca : freedreno/registers: Rename SP_2D_SRC_FORMAT
    • d5c82c3 : freedreno/a6xx: Split clear and blit texture into different functions
    • f383b2c : freedreno/a6xx: Don't take pipe_blit_info in emit_blit_dst
    • 3158b8b : freedreno/a6xx: Program RB_UNKNOWN_8C01 in setup helper
    • 3ee89e6 : freedreno/a6xx: Move CP_SET_MARKER to setup helper
    • 3961fb7 : freedreno/a6xx: Move REG_A6XX_SP_2D_SRC_FORMAT programming to helper
    • 03ad130 : freedreno/a6xx: Program A6XX_SP_2D_SRC_FORMAT_COLOR_FORMAT based on dst format
    • 5c3c0c5 : freedreno/a6xx: Make blit_control helper a little more helpful
    • 77d4aa7 : freedreno/a6xx: Enable FMT6_10_10_10_2_UNORM blitting
    • 1a58596 : freedreno/a6xx: Separate stencil sysmem clear fix
    • ab61393 : freedreno/a6xx: Don't emit src state when clearing
    • 416513d : freedreno/a6xx: Consolidate