It was broken by the last change (because unlike `which adb`, `command
adb` does not search $PATH if there's already an entry in the hash).
No-one's noticed in almost a year, but this has caused (rare) confusion
in the past.
Bug: http://b/24473359
Change-Id: I6bbb207992449132ba8563a8bcf298256c599c0d
This gives us a good place to put scripts on the PATH instead of adding
more functions to envsetup.sh
Test: GENERATE_SOONG_DEBUG=true m nothing ; soongdbg ...
Change-Id: I3c1a80fc8aae8de548bfe9036dc86fa006fc58c6
We keep this as "trunk_staging" and "eng". However, we switch
from "aosp_arm" to "aosp_cf_x86_64_phone". This switches us
away from 32-bit development, and gives us a project (the
cuttlefish ("cf") emulator) at the end of the build which is
runnable with reasonable performance, and able to run all 32-bit
and 64-bit tests.
Bug: 318749634
Test: lunch
Change-Id: I920cdd413cbcecd4300bc46c719e5788244a924e
Otherwise it becomes a single line with an ugly long stretch of spaces
in the middle.
Test: hmm
Bug: 179779520
Change-Id: Ia095b75248952d9bee0d6417028d04460e74ad33
The same commands where we're setting things like
TARGET_BUILD_VARIANT to be explicitly empty, we do the
same with TARGET_RELEASE.
Bug: 307946156
Test: 'lunch' with no arguments
Change-Id: If78eacd62725122d47829e2e9079b182fdafb9b4
In the trunk stable builds, we're required to have a
TARGET_RELEASE set.
For now, we hard-code this to "trunk_staging" (and leave a TODO
to accept this as a function argument).
This will have us producing valid builds in places like git_main.
Bug: 307946156
Test: None
Change-Id: Ie418aee2fad192a38beca44cd7d168e0e1b45689
Instead of supporting both product-variant and
product-release-variant, we now require the release
type to be given to use.
Bug: 307946156
Test: 'lunch aosp_mokey-userdebug' (now) fails; 'lunch aosp_mokey-trunk_staging-userdebug' (still) works
Change-Id: Ica87b3969f950a57232615f33bfe5f4012a743d6
In the git_main trunk stable builds, we're required to have a
TARGET_RELEASE set.
For now, we hard-code this to "trunk_staging" (and leave a TODO
to accept this as a function argument).
This will have us producing valid builds under git_main.
Bug: 307975293
Test: 'tapas google-tradefed-all' now works on a tree that has go/aog/2806253
Change-Id: I70572fb4c7902350513e251694f3ddef91efcb13
This is because -eng provides faster local builds.
(also moving hints to where they are more likely to be seen)
Test: lunch cf_x86_64_phone-userdebug
Fixes: 307753064
Change-Id: I52cf9e9c399dbb3a52c7952f7204a9e567d15c22
In git_main, the correct target here is aosp_arm-trunk_staging-eng
(or "trunk_food" or "next", but we choose the most generic
"trunk_staging").
Bug: 307738446
Test: "lunch", hit return for the default, "echo $TARGET_RELEASE" to confirm
Change-Id: I4d4bcaf27553976e64133c7351db296953f88b08
Add an aninja function to envsetup.sh that can run ninja with the
appropriate ninja file for an Android build. Simplifies running
queries with ninja. For example:
aninja -t query droid
Test: aninja -t query droid
Change-Id: I7064984d55054cd4350800cfd75bf6aaa1740e9d
Add a script to help to create flag value building rules.
The script will only extract necessary information from the value file
but it will not validate the file. The validation will defer to building
system.
For input
```
flag_value {
package: "com.android.aconfig.test"
name: "disabled_ro"
state: DISABLED
permission: READ_ONLY
}
flag_value {
package: "com.android.demoapp.test"
name: "enabled_ro"
state: DISABLED
permission: READ_WRITE
}
flag_value {
package: "com.android.aconfig.test"
name: "enabled_rw"
state: ENABLED
permission: READ_WRITE
}
```
The output Android.bp file will be like
```
VALUE_LIST_LIST = [
aconfig-local-com.android.demoapp.test,
aconfig-local-com.android.aconfig.test
]
aconfig_values {
name: "aconfig-local-com.android.demoapp.test",
package: "com.android.demoapp.test",
srcs: [
"override_values.textproto",
]
}
aconfig_values {
name: "aconfig-local-com.android.aconfig.test",
package: "com.android.aconfig.test",
srcs: [
"override_values.textproto",
]
}
```
Bug: 298692416
Test: source build/envsetup.sh && overrideflags, and then check the
output
Change-Id: Ife4063e079811e0b29046be7a6bc127cad668ed0
This script has the following functionality:
1. Sets up the `out` symlink to point to a default location on disk
2. Copies RBE binaries outside Cog (temporary fix for build stall problem) - b/289391270
3. Exports ANDROID_BUILD_ENVIRONMENT_CONFIG variable to point to googler-cog.json file. This file will be available in aosp-main-with-phones and goog/main and downstream branches.
TESTED=ran the script within and outside the Cog workspace to test it.
Bug: b/295053716
Change-Id: I59cd6d2d5f954be6f7e4ecd7d64de1e63e51b72e
If adb is not found on PATH, which will silently fail and evaluate to
the empty string. This will cause the remaining arguments to be
interpreted as a command to run in the shell, which is generally
undesirable. (Consider, for example, "adb reboot" -> "reboot".)
Bug: 24473359
Test: Just run adb before lunch so it's not on PATH
Change-Id: I6b20722add6c67d1d2627f963dc66095502ab816
Signed-off-by: Saagar Jha <saagarjha@google.com>
Similar to
https://android-review.googlesource.com/c/platform/build/+/2403120.
The directory structure of //development/python-packages/gdbrunner
needs to change to be able to export type info to importers, and this
needs to adapt to keep gdbclient working.
Bug: None
Test: gdbclient.py --help
Change-Id: I8127ec7beae2a7840283be25ee236a8e2af48cae
This replaces support for the unused
TARGET_PLATFORM_VERSION variable.
Now, if you pass three - separated
items the first is product, the
second is release and the third
is variant. If you only pass two
they're still product-variant
and the build system will choose
a reasonable default for release.
Test: run lunch with two and three items, confirmed values in the build banner
Change-Id: I128177d96ffe81b79b6945a24ebf37861c3b25fc
adb root sometimes takes a second, and the next command
will fail.
Bug: N/A
Test: syswrite on device which isn't already rooted
Change-Id: I91754c31c74bf0589ea8cbae77c10e5a120bdce5
Importers of the adb python package like lldbclient.py and gdbrunner
are broken following the restructure that enabled adb to export type
info to importers.
Bug: http://b/266688086
Test: adb.__file__ is no longer None after envsetup.sh/lunch
Change-Id: I7bf27601e800043adee8971c093347ca9b2979a9
Since RBE is default for dev builds, and we no longer set RBE config
through env variables, this error message is incorrect.
Bug: b/265326161
Test: Ran a build and make sure this error message wasn't printed
anymore.
Change-Id: I128616ce37763400407e099b52ef284905239eb4
This is useful for searching Rust config files in external.
Test: cd external/rust/crates; tomlgrep foo
Change-Id: I1ff5118a8eee1a95f1bc6fa44c922673a03ebd8b
If the shell is set to exit on errors, the extraneous (incorrect syntax)
call to lunch.py causes an early exit.
Change-Id: I0176d5c9e7e42640514fd4988d82fbfdd825631d
This gets b (and other scripts that are coming soon) into the path without running lunch.
Also gets rid of the make dumpvar ANDROID_BUILD_PATHS variable, used to set a few more paths,
so that the setting is all in one place, now that the get_abs_build_var is cached.
Test: Lots of manual checks of before and after combinations
Change-Id: I745168c43ccf2fef4cdd22657fd087f1ad07e510
Generate a all_modules.txt file on every build that
can be used for autocompleting.
Because this adds time to every non-incremental build,
I profiled it by using `$(shell date +"%s.%3N")` before
and after, and disovered it only adds about 0.01 seconds
to the build. And of course it won't be rerun if there
weren't build changes.
Bug: 255867758
Test: Manually
Change-Id: I1d3f1008611fd542884322b27b19ea2fa1d341fe
