Commit graph

43 commits

Author SHA1 Message Date
Andreas Gampe
2819c0b831 Base: Explicitly set abort message on FATAL
Currently bionic only stores the first invocation of
android_set_abort_message, libbase splits the logging invocations
into discrete lines, and liblog automatically calls bionic when
fatal severity is used. This leads to only the first line of
LOG(FATAL) being stored for tombstoned.

Eagerly set the abort message directly before logging when the
severity is FATAL. This ensures the complete message will be
available.

Bug: 120506942
Test: m
Test: manual
Change-Id: I104d6960a2b1f66f21f5ada383fb4ab0f35e96a9
2018-12-05 12:59:15 -08:00
Elliott Hughes
4679a39610 adb: rationalize fatal/error logging.
Let's use LOG(FATAL)/PLOG(FATAL) for actual fatal stuff.

Add a Windows error(3) and move folks who didn't really mean "abort"
fatal over to it. Also get rid of syntax_error which wasn't adding a
lot of value, and most of the places it was adding "usage: " didn't seem
entirely appropriate anyway.

In particular, we seemed to have confused fastdeploy.cpp into aborting
in most user error cases, and none of the reviewers noticed. Clearly
we'd all lost track of far too many options.

(I've also cleaned up a few random instances of fprintf(3) + exit(2).)

Bug: N/A
Test: manual
Change-Id: I3e8440848a24e30d928de9eded505916bc324786
2018-10-19 14:04:24 -07:00
Tomasz Wasilczyk
3b70eedfa0 Remove unused logging symbols.
Bug: 35361699
Test: flash sailfish, it boots
Change-Id: Ie545145789faa299e1228b6c91566e0767f7ea79
2018-07-31 16:33:38 -07:00
Elliott Hughes
11a64eb0e8 libbase: fix libbase_test.logging.StdioLogger.
I don't think the StdioLogger test ever passed. Move GetFileBasename
to where we can use it from StdioLogger, and undo the mix of anonymous
namespace vs static to consistently use static (which is the majority)
while we're touching those lines.

Bug: N/A
Test: ran tests
Change-Id: I95b3966cdb8af642bed71752bd7d4e3a86ac84ca
2018-06-06 12:54:41 -07:00
Elliott Hughes
1be0d1481b Add StdioLogger for command-line tools.
Bug: N/A
Test: ran tests
Change-Id: If366a4ea25aea1becdd3e443eba225e9bd52ebba
2018-05-23 10:06:20 -07:00
Mark Salyzyn
2507a048ab base: kernel logging get pre-opened /dev/kmsg from environment
If the environment has an instance of /dev/kmsg pre-opened, pick up.
This happens when an init script has 'file w /dev/kmsg' in the
daemon's recipe.

libcutils android_get_control_file operations are open-coded to
prevent a dependency on libcutils itself.  Also check that /dev/kmsg
file descriptor is valid and open'd for write only.

Test: check to make sure init() kernel logging still works
Bug: 77541952
Bug: 77661605
Change-Id: I007553acd594ef0815d23f32f8aa2867518570ba
2018-04-09 07:19:33 -07:00
Josh Gao
5791e21697 base: export GetThreadId.
and stop using the incorrect (on Mac) version in adb..

Test: mma
Change-Id: If28f6c8d73b03db863be13ed88df81422c3aef87
2018-03-16 15:35:52 -07:00
Andreas Gampe
1923e768ef Base: Add default tag manipulation
Allow the default tag (the program name) to be overwritten.

Bug: 34867873
Test: m
Test: logging_test
Test: manual
Change-Id: I4ef32bad413a7cc82e46ce16a2f26212925964b1
2018-03-06 14:45:11 -08:00
Andreas Gampe
af05f3b761 Base: Add an NDK-compiled version of libbase
To support building libraries for CTS.

Bug: 73206894
Bug: 73659009
Test: mmma system/core/base
Change-Id: I630be88f44cf5cefcb13be7959fc1a710686e29b
2018-02-21 01:59:24 +00:00
Tomasz Wasilczyk
c251600102 Use LOG_TAG instead of binary name as a tag.
If LOG_TAG was not defined, falling back to a default
behaviour (using binary name).

Bug: 35361699
Test: manual
Change-Id: I209a6ebaf0df882f98642f6d1831766cb296c951
2017-12-20 10:59:46 -08:00
Christopher N. Hesse
684b442b9f libcutils: Fix thread ID on macOS
macOS 10.12 / Xcode 8 actively prohibit the syscall(2)
usage.

Change-Id: I91ff16600a76e97d0b9b5db7e8905b969fb3482f
2017-10-22 22:58:41 +02:00
Yabin Cui
0c68953778 base: support logging before enter main function.
Bug: http://b/30974760
Test: run libbase_test.

Change-Id: I3f45fc21e8fe6301b5fef027728ea91f2188911d
2017-01-25 15:50:47 -08:00
Mark Salyzyn
cfd5b080af system/core: preparation to pull back interfaces from android/log.h
Point to log/log.h where necessary, define LOG_TAG where necessary.
Accept that private/android_logger.h is suitable replacement for
log/logger.h and android/log.h.

Correct liblog/README

Effectively a cleanup and controlled select revert of
'system/core: drop or replace log/logger.h' and
'system/core: Replace log/log.h with android/log.h'.

Test: compile
Bug: 30465923
Change-Id: Ic2ad157bad6f5efe2c6af293a73bb753300b17a2
2016-10-20 08:11:39 -07:00
Andreas Gampe
b4e32f3b33 Base: Hand complete log message to aborter
Undo zero-termination-substitution for linebreaks when logging.
This results in handing the complete log message to the aborter.
Add a test.

Bug: 31893081
Test: mmma system/core/base && $ANDROID_HOST_OUT/nativetest64/libbase_test/libbase_test64
Change-Id: I2ef6c6351db2fd494a02985f634f439104136227
2016-10-04 19:19:59 -07:00
Mark Salyzyn
ff2dcd9af9 system/core Replace log/log.h with android/log.h
Should use android/log.h instead of log/log.h as a good example
to all others.  Adjust header order to comply with Android Coding
standards.

Test: Compile
Bug: 26552300
Bug: 31289077
Change-Id: I33a8fb4e754d2dc4754d335660c450e0a67190fc
2016-09-30 12:47:05 -07:00
Andreas Gampe
1f5fb43047 Base: Remove LOG_S variants.
Partially reverts commit 436f5a031f.

Remove the variants taking a fully qualified LogSeverity. Instead
use a lambda with "using" statements to translate both qualified
and unqualified names into valid expressions.

Compile-time regression was measured as 0.1s for a thousand LOG
statements on a z840.

Update tests.

Bug: 31338270
Test: m
Test: mmma system/core/base && $ANDROID_HOST_OUT/nativetest64/libbase_test/libbase_test64
Change-Id: I36fdf30a9d535b19543307b85d1b3c19a97f20dd
2016-09-24 10:18:38 -07:00
Andreas Gampe
19ff8f1e57 Base: Check severity in destructor
Bug: 31338270
Test: m
Test: mmma system/core/base && $ANDROID_HOST_OUT/nativetest64/libbase_test/libbase_test64
Change-Id: Ic077822bcfd591bd11f4fa9d4b6e440b461c70c1
2016-09-23 13:31:52 -07:00
Pirama Arumuga Nainar
70f281ae36 Remove mutex.h and its uses.
mutex.h is now available as a part of MinGW itself.

Test: Successful build
Change-Id: I036ee37600a39130d6fbf293398585cfc58209ca
2016-09-16 15:58:00 -07:00
Josh Gao
63bdcb57aa base: extract mutex.h from logging.cpp.
Extract the Windows-specific mutex implementation from logging.cpp.

Bug: http://b/31468413
Change-Id: I0a895911ec6d815b8011b09d55209b64bbf9a70e
Test: mma
2016-09-13 15:59:46 -07:00
Andreas Gampe
2691e335fe Base: Add AbortFunction for logging
Add a hook to modify FATAL behavior. Add a test.

Bug: 31338270
Test: m
Test: mmma system/core/base && $ANDROID_HOST_OUT/nativetest64/libbase_test/libbase_test64
Change-Id: I966da319cda613738b3f2ccdac8c21900d6a5c72
2016-09-08 15:33:27 -07:00
Andreas Gampe
550829d84b Base: Add INTERNAL_FATAL to logging severity
Add the INTERNAL_FATAL level. It will print 'F' like FATAL, but
does not abort after logging.

Add a test to logging_test.

Bug: 31338270
Test: m
Test: mmma system/core/base && adb sync && adb shell /data/nativetest/libbase_test/libbase_test32
Change-Id: Idf74c08e8516881efccaefc58fa3f41d57e56396
2016-09-07 18:19:42 -07:00
Elliott Hughes
7bc87a5a78 Move init's kernel logging into libbase.
This will let other stuff (such as adbd and recovery) use it too.

Bug: http://b/28467098
Change-Id: Idab51f438ed3edd6fe65a56c4b396eaff5f3833e
2016-08-04 16:09:39 -07:00
Elliott Hughes
e5dd71a290 libbase should use its own logging!
Not doing so led to us using a bogus log tag.

Bug: http://b/30281203
Change-Id: I3ac91758a1a043146c65f2ae0f36fcfbe372c30f
2016-07-28 15:15:28 -07:00
Elliott Hughes
4e5fd111d8 Add timestamps to libbase's stderr logging.
adb is the client that wants this, but I'm not aware of any client that
doesn't.

Bug: http://b/28467098
Test: tested manually with cat.
Change-Id: I82c3f7ddf466a67aedfbb9010e0939a44b71d3e5
2016-06-21 14:27:16 -07:00
Dan Willemsen
86cf941c48 Fix windows 64-bit builds
pid_t is 64-bit in 64-bit mingw, but the windows process/thread
functions return a DWORD(uint32_t). Instead of promoting to a pid_t and
fixing the format strings, just use a uint32_t to store the values.

android_thread_id also cannot be a 64-bit pointer, so for windows just
force it to be a uint32_t.

libutils/ProcessCallStack only works under Linux, since it makes heavy
use of /proc. Don't compile it under Windows or Darwin.

Bug: 26957718
Change-Id: I4e43e7cf18a96f22b3a9a08dbab8c3e960c12930
2016-02-05 16:10:10 -08:00
Elliott Hughes
4f71319df0 Track rename of base/ to android-base/.
Change-Id: Idf9444fece4aa89c93e15640de59a91f6e758ccf
2015-12-04 22:00:26 -08:00
Josh Gao
26f2e1fd4f Merge changes Ic124ecb9,I94de55d2
* changes:
  libbase: remove exit-time destructors.
  adb: remove exit-time destructors.
2015-11-12 22:50:03 +00:00
Josh Gao
7df6b5fc79 libbase: remove exit-time destructors.
Removed for the same reason as the adb exit-time destructors.

Bug: http://b/23384853
Change-Id: Ic124ecb9df132b850a3855e207baffec926dde29
2015-11-12 11:54:47 -08:00
Elliott Hughes
c1fd492ac5 Revert "Revert "adb/base: fix adb push of Unicode filenames on Win32""
This reverts commit cc8cd59456.

With the dependency on libcutils (for gettid for non-bionic) removed,
this no longer breaks the build.

Change-Id: I645bd6876e2502ddc1535b69af1e645c0df9d178
2015-11-11 18:23:00 -08:00
Spencer Low
363af568b8 adb/base: minor compiler portability improvements
I've been using these changes to compile with Visual Studio.

- GetFileBasename(): __FILE__ uses \ with Visual Studio.

- adb_trace.cpp: Apparently VS needs an ampersand before the function name.

- "expr1 ? : expr2" is a GCC extension.

- <algorithm> contains std::min().

- seekdir can't always be #define'd because some headers have members
  named seekdir.

- adb_utils.cpp: Not really a compiler issue, just a random fix:
  0x7F/DEL is not printable.

Change-Id: I0dfb634f1ba4ccbc0d1b9f71b00e838fbebb3b41
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
2015-11-09 10:22:13 -08:00
Spencer Low
765ae6bcfc base logging: fix errno restoring, severity conditionality, dangling ifs
Fix LOG() to properly save and restore errno. Test this properly.

Only do logging if severity is >= the minimum.

Fix dangling if statements in CHECK(), CHECK_STR{EQ,NE}(). Test this
properly.

Fix base logging tests on Windows. All libbase_tests now pass on
Windows.

Change place to lock, so the lock can protect logging of all data in
LogMessage.

Change-Id: I7ff531c67ae10a99ef0a2bbfe279aa77282d5ae9
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
2015-09-21 12:46:12 -07:00
Spencer Low
bdab59a861 libbase: logging fixes
Win32:

- getprogname(): call basename() which is available in mingw's crt.
  Don't potentially go recursive with DCHECK_GT().

- Use Win32 critical section instead of mutex.

Other:

- Change log_characters check to compile-time.

- Fix code that gets the basename of __FILE__. The previous code was not
  setting _file, so it didn't work.

- Save and restore errno for LOG calls. Inspired by similar Chromium code.

Change-Id: Ie7bb700918be726fa81d60177d1894d2daeff296
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
2015-08-13 14:08:31 -07:00
Spencer Low
ac3f7d9a78 adb / libbase: clean up NOGDI and evil ERROR macro
Instead of defining and undefining NOGDI:

1. Always #include "base/logging.h" after #include <windows.h>.
Unfortunately, I could not find an easy way to give the user a
warning/error if they include in the wrong order.

2. base/logging.h does #undef ERROR to undefine the evil ERROR macro
that is from another era and probably a bad idea to begin with.

Change-Id: I995d89620611e849af9d7ec47eb55fc0512377f2
Signed-off-by: Spencer Low <CompareAndSwap@gmail.com>
2015-05-20 12:25:19 -07:00
Dan Albert
ea41c213be Add additional shielding against the ERROR macro.
Including the logging header with ERROR defined will always be a
problem, so check for it explicitly.

Move the NOGDI define from logging.cpp to logging.h since the standard
library headers might be including windows.h, pulling in ERROR.

Change-Id: Ib426973d2f1840710c0bf0e0cfb2420d0d322e27
2015-05-04 16:40:38 -07:00
Dan Albert
5c19040b06 Support base::logging on Windows.
libc++ doesn't support std::mutex and friends for Windows yet, so we
just use a compatibility wrapper for now.

Change-Id: I2413d4c089e7d0fb232444043c6b772153035dab
2015-04-29 18:01:53 -07:00
Dan Albert
7a87d05974 Cleanup base/logging.
Some of this code was unused, most didn't need to be exposed, and we
can just use basename(3).

Also use a better default program name than "unknown".

Change-Id: I62d990f64e4fd0c16aa4b7e907dd06e4f26ddcdd
2015-04-03 16:49:55 -07:00
Dan Albert
b547c85b5b Support arbitrary loggers.
While the defaults (logd or stderr) make sense for most use cases,
there are places that can only log to the kernel, or need to log to a
file, etc.

Allow the user to pass in an arbitrary logging object, and provide
LogdLogger and StderrLogger as defaults.

Change-Id: I62368acc795ff313242bb205d65017404bf64e88
2015-04-03 16:22:39 -07:00
Dan Albert
0c055863eb Support logging to other log buffers.
LOGTO(dest, severity) and PLOGTO(dest, severity) log to other log
buffers. For example, `LOGTO(SYSTEM, FATAL) << "Foobar";`.

Change-Id: Id1ca1c8fdae72d69b73945ae9b006525d0be1582
2015-04-03 10:14:48 -07:00
Dan Albert
0cc351b1a7 Remove LogMessage::LogLineLowStack.
This was ported over from ART, but even they only use it in a single
place.

Change-Id: Ia34dc54b56c764ede79b2957bf3dc8df0f1323ee
2015-04-02 18:43:32 -07:00
Dan Albert
4b3f533be9 Include what we use.
Change-Id: If510431d3d445de94f781b4fdec2e1b445f7c294
2015-03-26 23:23:43 -07:00
Dan Albert
7dfb61dcdc Move gettid() into libcutils.
Change-Id: Ic8a15036833e6d129b7998d954b804be391de399
2015-03-23 13:01:21 -07:00
Dan Albert
47328c96d9 Update string Split API.
Return a new vector rather than appending to the parameter.

Delimiters are also a string rather than a character. Split on any
character in the string.

Change-Id: I039b332ace5578590df9e7ca0e8fa3db28db30a3
2015-03-23 10:23:11 -07:00
Dan Albert
58310b49fc Add google3 style logging to libbase.
ART already had a flavor of this, but it was specialized for their use
case a bit.

Note that the logging.* tests are currently disabled for the device
because there is no good way to capture the output of liblog. We can
make something that will execute logcat and then then scan the output,
but that's messy. Since we know it at least works on the host, we can
add better device tests later.

Change-Id: I47acd87a3312c0a5285b03f9c8dadef0c669f06a
2015-03-20 14:04:35 -07:00