aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
| * Force reset displayId for secondary displayHuan Song2020-04-282-4/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | Rebooting the framework will allocate two new secondary displays. But the previous allocated displays still exits. So, delete the previous allocatd displays when allocating new ones. Bug: 149930332 Test: lunch gcar_hl_md_emu_x86-userdebug; m -j; run emualtor, adb reboot Change-Id: I8d506b15bb3d28d53fafa2657325ad1ef9c2d0dd
* | Snap for 6429305 from 4eedfee63685a066868138c834ab0687ad49a29a to rvc-releaseandroid-build-team Robot2020-04-254-26/+21
|\| | | | | | | Change-Id: If755e4825ff15dc4b5622bb4beb2f46a98a6fa33
| * Merge "Retire cbmananger from EmuHWC2" into rvc-devTreeHugger Robot2020-04-243-11/+21
| |\
| | * Retire cbmananger from EmuHWC2Roman Kiryanov2020-04-233-11/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Replace it with GraphicBufferAllocator from fw. Bug: 154346346 Test: boot Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: Ibd1b559e9721ad229198ddda83ca45ffc6905642 Change-Id: Ia4668c2d047ca5918d690e7b43b002619f1a3ef9
| * | remove some unused variablesMaciej Żenczykowski2020-04-241-15/+0
| |/ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes: device/generic/goldfish-opengl/android-emu/android/base/ring_buffer.c: 426:23: warning: unused variable 'yield_backoff_us' [-Wunused-const-variable] 427:23: warning: unused variable 'sleep_backoff_us' [-Wunused-const-variable] 435:14: warning: unused variable 'start_us' [-Wunused-variable] 436:14: warning: unused variable 'curr_wait_us' [-Wunused-variable] 478:14: warning: unused variable 'step_shift' [-Wunused-variable] which in turn makes ring_buffer_curr_us() an unused function and thus can be deleted. Test: builds, and: git grep 'yield_backoff_us|sleep_backoff_us|start_us|curr_wait_us|step_shift' finds nothing Bug: 153035880 Signed-off-by: Maciej Żenczykowski <maze@google.com> Change-Id: Ia45892ff8bcf215c87a531acc18b343ccb732ad5 Merged-In: Ia45892ff8bcf215c87a531acc18b343ccb732ad5
* | Snap for 6416172 from 8e14cbcaccea7650064d357704e3688d025a30a0 to rvc-releaseandroid-build-team Robot2020-04-224-108/+10
|\| | | | | | | Change-Id: I55700b19c97b32cafa2d18130a9f8aae7ca1f74d
| * Retire hwc2/MiniFenceRoman Kiryanov2020-04-174-108/+10
| | | | | | | | | | | | | | | | | | | | sp<MiniFence> is replaced by shared_ptr<unique_fd>. Bug: 154346346 Test: boot Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: Id02e959b897e6e667e092705815d1082da15a2b4 Change-Id: Ic2ba6df78bdc3138e0c617feece75fccb250200f
* | Snap for 6398428 from 670129a1a6d9d57b9be9f6e836dfa326cccebd3a to rvc-releaseandroid-build-team Robot2020-04-1628-5/+314
|\| | | | | | | Change-Id: I6a2c83a0c9d93500421c631ff12f57c323472cf4
| * vulkan: always tear down host visible memory blocks properly (guest)Lingfeng Yang2020-04-1521-3/+189
| | | | | | | | | | | | | | Test: Chrome Fuchsia cc_unittests don't flake bug: 154138988 Change-Id: I2e736dfb3d940b604970a4e62b5447a34734dea3
| * android vulkan: disallow noflags memory types again for AHB (for now)Lingfeng Yang2020-04-153-0/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | bug: 134773396 This CL makes BasicVulkanGpuTests pass again, via disallowing NVIDIA's sysmem memory types (as they dont work well with dedicated/external). The better (upcoming) solution is to query in the beginning what memory types correspond to dedicated + external on the host in the first place, instead of guessing based on heuristics. Change-Id: Ic259eecd4cb8c2e38a5737c98ac22837febf7896
| * Allow ASTC format in glTexStorageYahan Zhou2020-04-151-0/+29
| | | | | | | | | | | | BUG: 151393557 Test: run hearthstone Change-Id: Ic24beccb9653d397b78a8c270e8b4ec5f7019469
| * Fix SkQP tests (android 10 cts r3)Lingfeng Yang2020-04-155-2/+79
| | | | | | | | | | | | | | | | | | | | | | | | bug: 139956393 This addresses a bug where in Q CTS r3 (more recent than that we've used to validate before), SkQP test cases fail due to glTexSubImage2D not accounting for different settings for offset and row length. Test: ./cts-tradefed run cts -m CtsSkQPTestCases on android 10 r3: 0 failures Change-Id: I771d81bc831b4ea86bc4a985f311c687cb60694a
* | Snap for 6364069 from c83652e242b7989a61805e43f90c96e2270a370a to rvc-releaseandroid-build-team Robot2020-04-047-36/+480
|\| | | | | | | Change-Id: Ic9b525aa484d1befea0cee290e274aef5f6e598d
| * vulkan: 1.2.1 CTS fixesLingfeng Yang2020-04-024-24/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | - Set the default API version to 1.1 (assume 1.1 is supported on the host), so that CTS tests don't exercise the rare case where we have a 1.0 instance and a 1.1 device - Don't suppress memory type indices with properties == 0. This works around a driver bug where some ycbcr copy tests only work with a memory type index that has properties == 0 Fixes: 152769369 Merged-In: I2dc9df05ea93f8562c1c34a0bddaf28a5a83c414 Change-Id: I98db7e30120d4486eb466243b8d9e47fd8bc461b
| * vulkan: suppress sampler field of VkDescriptorImageInfo (guest)Lingfeng Yang2020-04-024-5/+171
| | | | | | | | | | | | | | | | | | | | | | | | when in the immutable sampler case. Test: vulkanhal_unittests (needs host side CL) dEQP-VK.ycbcr* in 64 bit mode dEQP-VK.binding_model* bug: 145153816 Merged-In: Id501fc5a383d5519dd99886b5dd43b6057034b15 Change-Id: I84fd8b4f2630b86e80da987b3e010824392599c3
| * vulkan: Properly free guest-side handles for descriptor sets (guest) am: ↵Lingfeng Yang2020-04-024-9/+306
| | | | | | | | | | | | | | | | | | | | e9e77d5f9b bug: 145153816 Merged-In: Ic9d8f42b075ef26322252ecb067a5fcc58f3eab0 Change-Id: I4da6b39a3237d0abb0bc09672e51852c439d6759 (cherry picked from commit c0a57f1a032ab5382f362d6b3bd71a5dcb6d9039)
* | Snap for 6360479 from 1b5bf91372c6f2cbc09cdaed39974839cf813b17 to rvc-releaseandroid-build-team Robot2020-04-031-1/+6
|\| | | | | | | Change-Id: I8de486cc58b44c7c2ba45cb2fe96b9580369d577
| * vulkan: hack: use 65536 as min alignment for virtualized sub-mappings am: ↵Lingfeng Yang2020-04-021-1/+6
| | | | | | | | | | | | | | | | | | d782b7e019 Fixes: 152769369 Change-Id: I1e1e05847dbe062b066c3d5d8d1cb477dda1a5d1 (cherry picked from commit 569010328698638be79ff3626919bcfc609217af)
* | Snap for 6353066 from cb2d0b02498b85b78f420198d58081ce132279d9 to rvc-releaseandroid-build-team Robot2020-04-012-2/+8
|\| | | | | | | Change-Id: Iaeab9000365cff5c4c252fac2eee0c5ee4485c3f
| * Add OMX_COLOR_FormatYUV420Planar to GoldfishMapper::isSupportedImplRoman Kiryanov2020-03-302-2/+8
| | | | | | | | | | | | | | | | | | | | Allocator and Mapper must agree on supported formats. Bug: 152554068 Test: presubmit Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: I157b79835b4bf0b74e8f08451a2a39f1eb595e83 Change-Id: Ic515d07d8ad1411a33b560b59636b65b0519a5df
* | Snap for 6314400 from 40db38be5b13dc2ed7d1b444b8864983968914cb to rvc-releaseandroid-build-team Robot2020-03-2011-46/+118
|\| | | | | | | Change-Id: I4c517887e896c31f76f9867702ee7de61b48c31a
| * (reland) Use the shared slots host memory allocatorRoman Kiryanov2020-03-1911-46/+118
| | | | | | | | | | | | | | | | | | | | This time we check if the host side advertises this feature. Bug: 149254427 Test: boot, host side tests Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: I2915e00035d5b392a605a6bb7321d9730cf13b45 Change-Id: Id364a9e3c151df288053dc3c9c8c29477d73a210
* | Snap for 6309957 from 18f51fad475b21f42dac8c2f1688c4a44791f91c to rvc-releaseandroid-build-team Robot2020-03-192-3/+3
|\| | | | | | | Change-Id: I4e0f97ac825445a3643747195fbdb16c59e02c4c
| * goldfish-opengl: use RGBA for YCBCR_420_888bohu2020-03-171-2/+2
| | | | | | | | | | | | | | | | | | | | To avoid reformat on the host side. Host side is using RGBA anyway. Bug: 145244672 Test: presubmit Change-Id: I1f6ef415816c631b3d1087ba306ad83ba398ed11 Signed-off-by: Roman Kiryanov <rkir@google.com>
| * Rename mapper to be emulator specificLingfeng Yang2020-03-171-1/+1
| | | | | | | | | | | | | | | | | | This might cause conflicts with other mapper @ 3.0 implementations Bug: 145244672 Test: presubmit Change-Id: Ie3228fb61c926d7d9dc5378206f3904730d276e7 Signed-off-by: Roman Kiryanov <rkir@google.com>
* | Snap for 6305839 from 64c1e39e83a2a0614bf92e20a94bd59408b6dea6 to rvc-releaseandroid-build-team Robot2020-03-186-100/+29
|\| | | | | | | Change-Id: Iff6734eba6d24f9a1d92c1bdab545cabc8f813b9
| * Revert "Use the shared slots host memory allocator"Roman Kiryanov2020-03-176-100/+29
| | | | | | | | | | | | | | | | | | | | | | This reverts commit fa6d39c4cc1df808f46f83d20408e7f57ea0ca82. Bug: 151752077 Bug: 149254427 Test: none Merged-In: I21124384d4e462305a186685a82c541e01a4cef9 Signed-off-by: Roman Kiryanov <rkir@google.com> Change-Id: Id0eeae8d1e13afba67ca36bb64e18f1444fdbf61
| * Revert "Fix issues around GoldfishAddressSpaceBlockProvider::open"Roman Kiryanov2020-03-173-18/+1
| | | | | | | | | | | | | | | | | | | | | | This reverts commit be31206b746f7632690686302dc6379d9ed10c4e. Bug: 151752077 Bug: 149254427 Test: none Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: I11f8492aca9c2b0a29074307ae1fec90c665f189 Change-Id: I69e85a425fbeee370b1308b4f7da373a650e5894
| * Fix issues around GoldfishAddressSpaceBlockProvider::openRoman Kiryanov2020-03-163-1/+18
| | | | | | | | | | | | | | | | Bug: 149254427 Test: host side tests Merged-In: I16812a5a398fa72189c7d4fa18944ec1e31b72f2 Signed-off-by: Roman Kiryanov <rkir@google.com> Change-Id: Ib9de7df1feca19ace7af7f782589afee3d00cb00
* | Snap for 6301911 from 994843fbd9533ec575a2f0411bde791344fc62f0 to rvc-releaseandroid-build-team Robot2020-03-178-40/+109
|\| | | | | | | Change-Id: Iace6f7fcfed3ac0b96807fdebc74b48be9307794
| * Use the shared slots host memory allocatorRoman Kiryanov2020-03-166-29/+100
| | | | | | | | | | | | | | | | Bug: 149254427 Test: boot Signed-off-by: Roman Kiryanov <rkir@google.com> Merged-In: I6bb166e46007e868a67144e404af5b6a5e31cd74 Change-Id: I16f537bddc448b05a3910f15107023352158cdb3
| * Make vkEnumerate*ExtensionProperties conform to the Vulkan specTheodore Dubois2020-03-161-4/+8
| | | | | | | | | | | | | | | | | | | | | | | | Vulkan specifies that "on return the [pPropertyCount] is overwritten with the number of structures actually written to pProperties," but this wasn't happening. Bug: 150953274 Test: presubmit Merged-In: I9c576e05fa66eb2a1210c2b432d3ed20adf2cdd0 Change-Id: Iff26b63d77da0389ecf3b59fde72c1cac8d75472 Signed-off-by: Roman Kiryanov <rkir@google.com>
| * don't yield in ring_buffer_yieldLingfeng Yang2020-03-161-7/+1
| | | | | | | | | | | | | | | | | | | | system call yields are unpredictable Bug: 149254427 Test: presubmit Merged-In: Ibb9a6844c66a4f8e8d010c20a5c43e30a7932337 Change-Id: I6a2cf6c410f75a67e59f1be77d5b96ca4b1a8c24 Signed-off-by: Roman Kiryanov <rkir@google.com>
* | Snap for 6279178 from 8c6155f6388cef947b58db781a62d119ef929c96 to rvc-releaseandroid-build-team Robot2020-03-101-0/+1
|\| | | | | | | Change-Id: Iaa7201e4dd2d2bd6478fe8f730d2225e547c3496
| * Flush host connection on rcFBPostLingfeng Yang2020-03-091-0/+1
| | | | | | | | | | | | | | | | | | | | | | | | Fixes glitches with app launching / switching to/from client mode Thanks to huans@ for the find! Bug: 147760640 149598490 Test: run emulator with hwcomposer disabled Change-Id: I4a0f7a146d7f437e0b76b4b7958be0a12529d28a
* | Snap for 6224475 from f10e5a3645cfafaaad106639669b2dfaf1c4028e to rvc-releaseandroid-build-team Robot2020-02-213-1/+21
|\| | | | | | | Change-Id: Id9dbce52f61528ca5695aa195be96794c7a4b5f0
| * goldfish-codecs: fix crash and adapt to allocator3 am: 2294d506f9 am: ↵Automerger Merge Worker2020-02-203-1/+21
| |\ | | | | | | | | | | | | | | | 1715fdcf2e am: caf84356b3 Change-Id: I8a73c0071a63e9359f39c4241e4087b83e726497
| | * goldfish-codecs: fix crash and adapt to allocator3 am: 2294d506f9 am: 1715fdcf2eAutomerger Merge Worker2020-02-203-1/+21
| | |\ | | | | | | | | | | | | Change-Id: I4d49f9acd6dd66dc22b74d9a7cc10ba9aadf1153
| | | * goldfish-codecs: fix crash and adapt to allocator3 am: 2294d506f9Automerger Merge Worker2020-02-203-1/+21
| | | |\ | | | | | | | | | | | | | | | Change-Id: Ibe40d2e576949db5e395fc64fdfe1d2c80b05f78
| | | | * goldfish-codecs: fix crash and adapt to allocator3bohu2020-02-193-1/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Test: emulator -qemu -append qemu.hwcodec.avcdec=2 stagefright -S buck.mp4 BUG: 124388359 Change-Id: Id4e96cc804f6025e9390421a77125168a5f92376
* | | | | Snap for 6212682 from 798085ef2a80b450ae0104cb3d10a446a73a193c to rvc-releaseandroid-build-team Robot2020-02-174-7/+27
|\| | | | | | | | | | | | | | | | | | | Change-Id: I00af10643140f92e61bf2e8de27389c0009bd561
| * | | | Merge "goldfish-codecs: Handle decoding to buffer mode" am: a1da9d5548 am: ↵Bo Hu2020-02-164-7/+27
| |\| | | | | | | | | | | | | | | | | | | | | | | | | | | | 6703c9ff66 am: 0437643925 Change-Id: Iee33398919eb30e8285c0297d0be2ea5e27fd636
| | * | | Merge "goldfish-codecs: Handle decoding to buffer mode" am: a1da9d5548 am: ↵Bo Hu2020-02-164-7/+27
| | |\| | | | | | | | | | | | | | | | | | | | | | | | | | | 6703c9ff66 Change-Id: I2ed4d4cdd3710b8053f4d74ae2b1e3acad393877
| | | * | Merge "goldfish-codecs: Handle decoding to buffer mode" am: a1da9d5548Bo Hu2020-02-164-7/+27
| | | |\| | | | | | | | | | | | | | | | Change-Id: I597ac469e07a15a2dd1f11594e37f4f19f748cfe
| | | | * Merge "goldfish-codecs: Handle decoding to buffer mode"Bo Hu2020-02-164-7/+27
| | | | |\
| | | | | * goldfish-codecs: Handle decoding to buffer modebohu2020-02-154-7/+27
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When decoding to buffer mode, there will not be any surface to render decoded frames into. Instead, the client just expect the decoded frame to be put in the buffer. In that case, the host side should simply put the decoded frame back to the guest, regardness of rendering mode(by guest or by host). Test: atest android.media.cts.DecoderTest#testH264Decode320x240 atest android.media.cts.DecoderTest#testCodecResetsH264WithoutSurface BUG: 124388359 Change-Id: I83685a897c0044232c9e7b312f5368d647c7211e
* | | | | | Snap for 6211803 from e3600a655be86f0fa50713a6d259a58e8e18eb74 to rvc-releaseandroid-build-team Robot2020-02-1624-14/+398
|\| | | | | | | | | | | | | | | | | | | | | | | Change-Id: I3ce265ac4f1a695181eb943a69553b2dcf0ca906
| * | | | | Merge "virtio-gpu: hostmem" am: 1abd773eb6 am: 9c5857eb75 am: 1e9c2894e9Lingfeng Yang2020-02-1524-14/+398
| |\| | | | | | | | | | | | | | | | | | | | | | Change-Id: I007a3a1ec91409cbf384ff51087cec4930868007
| | * | | | Merge "virtio-gpu: hostmem" am: 1abd773eb6 am: 9c5857eb75Lingfeng Yang2020-02-1524-14/+398
| | |\| | | | | | | | | | | | | | | | | | | | | Change-Id: I9849ca03a293691242c44495475c4fbc09936084
| | | * | | Merge "virtio-gpu: hostmem" am: 1abd773eb6Lingfeng Yang2020-02-1524-14/+398
| | | |\| | | | | | | | | | | | | | | | | | | | Change-Id: Ia6216c52538ea421042971517fcee0c2cbfa18aa
generated by cgit v1.2.3 (git 2.25.1) at 2024-05-19 22:20:57 +0000