Age | Commit message (Collapse) | Author |
|
simpleperf-release
Change-Id: Ia082f6e268a3b81df2bb261ca6cb0ff9606120c5
|
|
|
|
Add support for customizing the cvd host package with target specific
binaries.
To add a target specific binary to the cvd host package add the
following in one of the target .mk files:
$(call soong_config_append,cvd,binary, test-binary)
Bug: 318734012
Test: m dist and check the cvd host tarball
Change-Id: I6c10ead8b1caefe903521cb712c678e12b95f358
|
|
|
|
|
|
Shared pointers can be copied and stored and the usual semantics of it
imply the underlying data is valid until the shared pointer is
destroyed. This is not true for audio buffers as those buffers are
released as soon as the function returns.
This changes those pointers into const references which explicity
indicate borrow semantics.
Bug: 327434508
Test: m webRTC
Change-Id: I9296390463002cbc0389b17777560ddfa9b98c9e
|
|
device/google/cuttlefish/host/commands/process_sandboxer/main.cpp:65:17: error: no member named 'SimpleAtoi' in namespace 'absl'
65 | CHECK(absl::SimpleAtoi(inherited_fd, &fd));
Bug: 327430106
Change-Id: Ib2c7b77ceead20c2fc54c6a2dd7736c17b930a8d
|
|
simpleperf-release
Change-Id: I359dae1453ced54e9bef9e3c2e24dc0905ffb4ec
|
|
into main
|
|
|
|
|
|
Currently cuttlefish tracks if a restore is happening by checking the
config. Once the config is created, it cannot be modified. This caused
issues when attempting to powerwash or restart the device after it was
restored, since the config would remain the same. This lead to a bug
where powerwash/restart would actually perform a restore instead.
Change the behavior so that if FLAGS_snapshot_path is set, create a file
in assembly called "restore".
When restoring, we check if the file "restore" exists.
If trying to powerwash or restart, delete the file "restore" to ensure a
restore does not happen.
Bug: 323596000
Test: boot CF, snapshot, restore, powerwash.
Change-Id: I0a7b79e5fd9f13504cf0f596e4c02c8069fedfbf
|
|
To pick fetched version if presented
Test: crosvm/qemu run using copied efi loader locally
Bug: 322983262
Change-Id: I3d4c3fd07aae92f69b0f7dee489c9427e80d21b9
Signed-off-by: Dmitrii Merkurev <dimorinny@google.com>
|
|
... as no longer needed after the ANGLE updates from b/315387961
and the SwiftShader updates from b/324625557.
Bug: b/324625557
Test: cvd start --gpu_mode=guest_swiftshader
cts -m CtsMediaDecoderTestCases
Change-Id: Id13943451b2d8940e899e4c104ae9fb425062ef6
|
|
|
|
|
|
|
|
Input config apex sets input_device.config_file.apex instead of using a
hard-coded apex name.
Bug: 315080500
Test: adb shell dumpsys input
# set "touch.orientationAware = 0" in Touchscreen_0.idc
# build/install the input config apex
# Observe the Input configuration
# "Touch Input Mapper" shows "OrientationAware: false"
Change-Id: I873b3590f7cfa59ba56c57789b29f91f0b370f31
|
|
main
|
|
Start adbd once the device resumes as opposed to eagerly when restoring.
Attempts of connecting were causing delays and the device to appear offline
due to failure. This would cause a delay in the device coming online.
Bug: 320760033
Test: boot CF, snapshot, restore
Change-Id: Iec19c7e0f83c547ff8d4819f3f4b30cefb3f5de6
|
|
Without this, `mm` in device/google/cuttlefish fails with
a link error about an undefined formatter symbol.
Bug: 321293146
Test: mm
Change-Id: I87a53431794dc462ed68dd02c2ef860d94dccddc
|
|
--gpu_mode=drm_virgl"" into main
|
|
The test `AudioEffectTest#ManageSourceDefaultEffects` fails on
GMS CF automotive because software hotword detector occupies
the primary input. This test used to be skipped on HIDL HAL
because it did not have preprocessing effects. This CL switches
the automotive to use the audio effects HAL configuration which
does not have preprocessing effects to match the HIDL configuration.
To be able to use a custom vendor config for the effects,
the effects configuration file is being removed from the audio
HAL VAPEX, thus it needs to be installed via 'device.mk' for
all CF targets.
Bug: 318423731
Test: run `atest audioeffect_tests` on cf_x86_64_auto-trunk_staging-userdebug
Change-Id: Ic0cf3a918f5b7f8a34dc55d7d43254ce8fac431b
|
|
non-gpu instances running QEMU should use swiftshader.
This carveout was originally added to aid bringing up riscv
but is no longer needed as qemu now works with swiftshader
This reverts commit 3a4bd964a2c203e680a1c885c6cce6e57c2061eb.
Bug: b/325329466
Change-Id: I212ba3a63da5af41b99da1e5d821459542e1b781
|
|
bootconfig to modules.options""""" into main
|
|
|
|
to modules.options""""
This reverts commit 93232aa398fea6f1b7d4f03877872955c5b12ff1.
Reason for revert: RC fixed with https://android-review.googlesource.com/2962479
Bug: 309328283
Change-Id: I436089329dbdae1d5a476e93970c707bb021332b
|
|
|
|
|
|
Bug: 315027339
Change-Id: I9380448ec27b15ac1c95903d9992e0aa99e21c5c
|
|
|
|
|
|
|
|
|
|
Bug: 318592126
Test: launch_cvd --enable_host_sandbox=true --resume=false
Change-Id: I4a68e2224aa7a961f6e9d62dd4bf5b72ab6425df
|
|
Triggers a power button event to:
- Wake up an instance when the instance is in S0ix/S3 power mode
- Switch the display mode when the instance is in S0 power mode
Test: Run `cvd powerbtn` to wake up a guest
Bug: 318540621
Change-Id: Ib4d1e1cede7a57739282974ab073f5c7c6da7526
|
|
main
|
|
There's an userspace API to wipe /data partition and reformat as EXT4.
This API is used to implement the 16K dev options feature. Since the
device's data partition might be reformatted w/o re-running launch_cvd
with a different --userdata_format flag, add ext4 fstab entries so that
device can always boot.
Test: th
Bug: 293313353
Change-Id: I6f7cbddfd12f902a6c12784ece83f162f318e26a
|
|
Bug: 325106413
Test: atest CtsNfcTestCases
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:f50bc38a0515653c6f8fcdc03053b1dee543e16b)
Merged-In: I4d3c70d4aa394fb35d3db31e18a1c48722cc5d69
Change-Id: I4d3c70d4aa394fb35d3db31e18a1c48722cc5d69
|
|
To being cvd_internal_(start|stop) again for the time being.
Bug: 325076463
Test: m && ls -l `which launch_cvd` `which stop_cvd`
Change-Id: I0b28d8295ae039cd5e4881ad8ca6abef5e6c4f6c
|
|
Currently cuttlefish runtime symlinks are created only
in case of succesfull boot. Acloud cannot pull them
when cvd isn't booted, so we cannot see launcher/kernel
logs on infra.
Fixing it by executing symlinks creating before cmd result
check.
Test: locally emulate vmm issue verified symlinks are created
Test: launcher log is pulled in case of vmm failure
https://android-build.corp.google.com/test_investigate/?invocationId=I49400010248251497
Bug: 325936323
Change-Id: I66d6aea50f011adde19fe175b460832ea281317c
Signed-off-by: Dmitrii Merkurev <dimorinny@google.com>
|
|
When uploading artifacts to the docker's host orchestrator, run for-
loop at server side to avoid repeated ssh connection / disconnection
while uploading each artifact file.
Test: run launch_cvd_arm64_server_docker.sh and succeeded to launch
instances
Change-Id: I39566c28a0827594b9545a61e319476e62649f92
|
|
CF Riscv uses mainline which has the wifi modules
built as gki modules. Reference to these gki modules
has been added for the unrelated 6.6 kernel update
so this workaround can now be removed.
This reverts commit 41a730a9cb8b4c3243fca4c8d2154c8b5275eced.
Bug: 309328283
Test: Treehugger
Change-Id: I1c784eb4b489870f019ac7b99df78e2a1242ef9d
|
|
|
|
Some early_hal depends on gralloc-mapper.
Bug: 325029533
Test: CF auto boots
Change-Id: I4a89c827c60bf5a44f3667664633c92a3eb2d99f
|
|
|
|
|
|
|
|
|
|
Automotive Dipslay service also needs to start after APEX gets mounted
on cuttlefish targets, to be able to use gralloc-mapper service.
Bug: 325029533
Test: atest CarEvsManagerTest
Change-Id: I2ab63d1d975a694d42b493734f2cbd5d5ff370f9
|