OSDN Git Service

android-x86/bionic.git
9 years agoUpdate bionic to tzdata2014i.
Elliott Hughes [Wed, 22 Oct 2014 19:31:02 +0000 (12:31 -0700)]
Update bionic to tzdata2014i.

From the release notes:

   Changes affecting future time stamps

     Pacific/Fiji will observe DST from 2014-11-02 02:00 to
     2015-01-18 03:00.  (Thanks to Ken Rylander for the heads-up.)
     Guess that future years will use a similar pattern.

     A new Zone Pacific/Bougainville, for the part of Papua New
     Guinea that plans to switch from UTC+10 to UTC+11 on
     2014-12-28 at 02:00.  (Thanks to Kiley Walbom for the
     heads-up.)

   Changes affecting time zone abbreviations

     Since Belarus is not changing its clocks even though Moscow
     is, the time zone abbreviation in Europe/Minsk is changing
     from FET to its more-traditional value MSK on 2014-10-26 at
     01:00.  (Thanks to Alexander Bokovoy for the heads-up about
     Belarus.)

     The new abbreviation IDT stands for the pre-1976 use of UT+8
     in Indochina, to distinguish it better from ICT (UT+7).

   Changes affecting past time stamps

     Many time stamps have been corrected for Asia/Ho_Chi_Minh
     before 1976 (thanks to Trần Ngọc Quân for an indirect pointer
     to Trần Tiến Bình's authoritative book).  Asia/Ho_Chi_Minh has
     been added to zone1970.tab, to give tzselect users in Vietnam
     two choices, since north and south Vietnam disagreed after our
     1970 cutoff.

     Asia/Phnom_Penh and Asia/Vientiane have been turned into
     links, as they differed from existing zones only for older
     time stamps.  As usual, these changes affect pre-1970 time
     stamps only.  Their old contents have been moved to the
     'backzone' file.

Bug: 18085936
Change-Id: I89c065b4788b10ed7530cc4e8cfbc1b65c05c1b7

9 years agoMerge "Replace ambiguous cmp operator with cmpl."
Chih-Hung Hsieh [Wed, 22 Oct 2014 00:55:14 +0000 (00:55 +0000)]
Merge "Replace ambiguous cmp operator with cmpl."

9 years agoReplace ambiguous cmp operator with cmpl.
Chih-Hung Hsieh [Tue, 21 Oct 2014 23:25:35 +0000 (16:25 -0700)]
Replace ambiguous cmp operator with cmpl.

Clang assembler rejects ambiguous cmp instruction.

BUG: 17302482
Change-Id: Ie8dca7e4f46e0906c47076bba21c1941482ba6b3

9 years agoMerge "Rename library_offset to library_fd_offset"
Dmitriy Ivanov [Tue, 21 Oct 2014 21:06:18 +0000 (21:06 +0000)]
Merge "Rename library_offset to library_fd_offset"

9 years agoRename library_offset to library_fd_offset
Dmitriy Ivanov [Tue, 21 Oct 2014 19:09:18 +0000 (12:09 -0700)]
Rename library_offset to library_fd_offset

 replace lseek() and use pread() instead
 add test for library_fd_offset > file_size case

Bug: 17762003
Change-Id: I4555f0be635124efe849c1f226985bcba72ffcbd

9 years agoMerge "Never use ASAN."
Dan Albert [Tue, 21 Oct 2014 18:28:18 +0000 (18:28 +0000)]
Merge "Never use ASAN."

9 years agoNever use ASAN.
Dan Albert [Fri, 17 Oct 2014 17:33:25 +0000 (10:33 -0700)]
Never use ASAN.

Unless we completely redo how we build bionic (so that the object
files for libc.so get built separately from libc.a), we can't enable
ASAN here, as libc.a gets linked into static executables.

Change-Id: I2ce4f51248bd51c4213a555ff481b6faabbf53f8

9 years agoMerge "Fix generic __memcpy_chk implementation."
Christopher Ferris [Mon, 20 Oct 2014 18:21:21 +0000 (18:21 +0000)]
Merge "Fix generic __memcpy_chk implementation."

9 years agoMerge "Add -no-integrated-as at local level."
Chih-Hung Hsieh [Mon, 20 Oct 2014 16:51:05 +0000 (16:51 +0000)]
Merge "Add -no-integrated-as at local level."

9 years agoAdd -no-integrated-as at local level.
Chih-Hung Hsieh [Fri, 17 Oct 2014 23:27:03 +0000 (16:27 -0700)]
Add -no-integrated-as at local level.

Later we will enable integrated-as as default at the global level.

BUG: 17820427
BUG: 17302991
Change-Id: I233204ca04d7222b2cb1fbd0fa48a482f8b63f6f

9 years agoFix generic __memcpy_chk implementation.
Christopher Ferris [Fri, 17 Oct 2014 21:08:54 +0000 (14:08 -0700)]
Fix generic __memcpy_chk implementation.

- Clean up the labels (add .L to make them local).
- Change to using cfi directives.
- Fix unwinding of the __memcpy_chk fail path.

Bug: 18033671
Change-Id: I12845f10c7ce5e6699c15c558bda64c83f6a392a

9 years agoMerge "Correct way to specify additional dependencies"
Dmitriy Ivanov [Fri, 17 Oct 2014 20:19:18 +0000 (20:19 +0000)]
Merge "Correct way to specify additional dependencies"

9 years agoCorrect way to specify additional dependencies
Dmitriy Ivanov [Fri, 17 Oct 2014 18:47:18 +0000 (11:47 -0700)]
Correct way to specify additional dependencies

 Previous one was not covering all the targets

Bug: 17548097
Change-Id: I425e0f8d6e5a43feabd29dcb327049b82754d249

9 years agoMerge "Fix misleading comment."
Dmitriy Ivanov [Thu, 16 Oct 2014 17:54:02 +0000 (17:54 +0000)]
Merge "Fix misleading comment."

9 years agoMerge "Only use <atomic> for C++11 and newer."
Dan Albert [Thu, 16 Oct 2014 16:19:43 +0000 (16:19 +0000)]
Merge "Only use <atomic> for C++11 and newer."

9 years agoMerge "Don't use an STL for libm."
Dan Albert [Thu, 16 Oct 2014 15:56:31 +0000 (15:56 +0000)]
Merge "Don't use an STL for libm."

9 years agoOnly use <atomic> for C++11 and newer.
Dan Albert [Thu, 16 Oct 2014 14:52:51 +0000 (07:52 -0700)]
Only use <atomic> for C++11 and newer.

Any pre-C++11 clients of stdatomic.h that use libc++ are being forced
over to <atomic>, which they don't have the language support to use.

Change-Id: I62445c1f2541410a1569498c09433c7196635537

9 years agoDon't use an STL for libm.
Dan Albert [Thu, 16 Oct 2014 14:44:47 +0000 (07:44 -0700)]
Don't use an STL for libm.

The default is none right now, but this will be needed when the
default becomes libc++.

Change-Id: Ib022adac2b9dc08fbfae830908128043cfffb91e

9 years agoMerge "Ignore DT_BIND_NOW (0x18)"
Dmitriy Ivanov [Wed, 15 Oct 2014 23:46:26 +0000 (23:46 +0000)]
Merge "Ignore DT_BIND_NOW (0x18)"

9 years agoIgnore DT_BIND_NOW (0x18)
Dmitriy Ivanov [Wed, 15 Oct 2014 21:59:01 +0000 (14:59 -0700)]
Ignore DT_BIND_NOW (0x18)

Bug: 17552334
Change-Id: Ibe93fd3d6580f369f9a6ef8d235bf0df00fe7607

9 years agoFix misleading comment.
Robert Grosse [Wed, 15 Oct 2014 21:32:19 +0000 (14:32 -0700)]
Fix misleading comment.

Change-Id: I9e2aa6c2ba938c8f8d7c74eba32d313c866c77b2

9 years agoMerge "We don't actually need all of libc++ here."
Dan Albert [Wed, 15 Oct 2014 19:59:08 +0000 (19:59 +0000)]
Merge "We don't actually need all of libc++ here."

9 years agoWe don't actually need all of libc++ here.
Dan Albert [Tue, 14 Oct 2014 21:19:34 +0000 (14:19 -0700)]
We don't actually need all of libc++ here.

All we're actually interested in is the unwinder. Since that's now a
separate library, just use that.

Change-Id: If86071a0d850da961336a58147b70369ace7bd12

9 years agoMerge "Add clock_settime and clock_nanosleep."
Elliott Hughes [Tue, 14 Oct 2014 03:07:42 +0000 (03:07 +0000)]
Merge "Add clock_settime and clock_nanosleep."

9 years agoAdd clock_settime and clock_nanosleep.
Haruki Hasegawa [Sun, 12 Oct 2014 15:50:47 +0000 (00:50 +0900)]
Add clock_settime and clock_nanosleep.

Add the missing prototypes, fix the existing prototypes to use clockid_t
rather than int, fix clock_nanosleep's failure behavior, and add simple
tests.

Bug: 17644443
Bug: https://code.google.com/p/android/issues/detail?id=77372
Change-Id: I03fba369939403918abcabae9551a7123953d780
Signed-off-by: Haruki Hasegawa <h6a.h4i.0@gmail.com>
9 years agoMerge "Add glibc alias for NT_PRFPREG"
Dmitriy Ivanov [Mon, 13 Oct 2014 22:08:51 +0000 (22:08 +0000)]
Merge "Add glibc alias for NT_PRFPREG"

9 years agoAdd glibc alias for NT_PRFPREG
Dmitriy Ivanov [Mon, 13 Oct 2014 19:23:55 +0000 (12:23 -0700)]
Add glibc alias for NT_PRFPREG

 Also fix comment style for DF_1_ constants

Change-Id: I7c3ed1d834f0a2b85ee8cc8016c79020955084f8

9 years agoMerge "Return total footprint, not high water mark."
Christopher Ferris [Mon, 13 Oct 2014 19:48:35 +0000 (19:48 +0000)]
Merge "Return total footprint, not high water mark."

9 years agoReturn total footprint, not high water mark.
Christopher Ferris [Fri, 10 Oct 2014 15:40:21 +0000 (08:40 -0700)]
Return total footprint, not high water mark.

The mallinfo usmblks value returned by dlmalloc is a little misleading.
It's not the current max, it's the historical high water mark. This
leads to dumpsys meminfo producing native memory numbers that don't add up.
Change this to the real total footprint, not this high water mark.

Bug: 17265653

(cherry pick from commit f4ada9c9ce31c7e56146a4cb703747385bc043a5)

Change-Id: I2fba10285859dccfe8331063c9be14cc169f2d91

9 years agoMerge "Use LOCAL_CXX_STL to select libc++_static."
Dan Albert [Mon, 13 Oct 2014 19:08:59 +0000 (19:08 +0000)]
Merge "Use LOCAL_CXX_STL to select libc++_static."

9 years agoUse LOCAL_CXX_STL to select libc++_static.
Dan Albert [Mon, 13 Oct 2014 18:20:29 +0000 (11:20 -0700)]
Use LOCAL_CXX_STL to select libc++_static.

Directly adding the library to LOCAL_STATIC_LIBRARIES thwarts a few
build system checks.

Change-Id: Ib23c43badfa85de21c691f79ccdcc2cbe2a89b33

9 years agoMerge "Fix some formatting so GitHub displays it nicely."
Dan Albert [Sat, 11 Oct 2014 00:48:32 +0000 (00:48 +0000)]
Merge "Fix some formatting so GitHub displays it nicely."

9 years agoFix some formatting so GitHub displays it nicely.
Dan Albert [Sat, 11 Oct 2014 00:14:37 +0000 (17:14 -0700)]
Fix some formatting so GitHub displays it nicely.

Change-Id: Id6ce9a61bebd98544c85e5780147715c73d0e78b

9 years agoMerge "Improve documentation for testing and coverage."
Dan Albert [Fri, 10 Oct 2014 18:59:02 +0000 (18:59 +0000)]
Merge "Improve documentation for testing and coverage."

9 years agoImprove documentation for testing and coverage.
Dan Albert [Fri, 10 Oct 2014 05:57:49 +0000 (22:57 -0700)]
Improve documentation for testing and coverage.

Also rename HACKING.txt to README.md so it will display on the GitHub
mirror.

Change-Id: I70157a4ad262700212bf9afd87253d195c7013a9

9 years agoMerge "Catch up on upstream's round/roundf/roundl."
Elliott Hughes [Fri, 10 Oct 2014 17:29:51 +0000 (17:29 +0000)]
Merge "Catch up on upstream's round/roundf/roundl."

9 years agoCatch up on upstream's round/roundf/roundl.
Elliott Hughes [Fri, 10 Oct 2014 17:21:43 +0000 (10:21 -0700)]
Catch up on upstream's round/roundf/roundl.

Not sure how we missed these, but better late than never...

Change-Id: Ib08d1bb6e340a1907cbeb1cbe220e33f70642bdc

9 years agoMerge "Add back symbols to fix Greed for Glory franchise."
Christopher Ferris [Fri, 10 Oct 2014 05:25:25 +0000 (05:25 +0000)]
Merge "Add back symbols to fix Greed for Glory franchise."

9 years agoAdd back symbols to fix Greed for Glory franchise.
Christopher Ferris [Fri, 10 Oct 2014 01:31:01 +0000 (18:31 -0700)]
Add back symbols to fix Greed for Glory franchise.

Bug: 17813018

(cherry picked from commit c891e24073830e07ba7373dee554ff2c70e1d313)

Change-Id: I4e891c2dbad9ea00d2377c5175317d9ae3840153

9 years agoMerge "Sync libm with upstream."
Elliott Hughes [Fri, 10 Oct 2014 01:26:18 +0000 (01:26 +0000)]
Merge "Sync libm with upstream."

9 years agoSync libm with upstream.
Elliott Hughes [Thu, 9 Oct 2014 23:20:37 +0000 (16:20 -0700)]
Sync libm with upstream.

Change-Id: I3b4e2c9c6ce6c5934f270a51ce5eb9154c5805d5

9 years agoMerge "Fix pthread_attr_getstack under valgrind."
Elliott Hughes [Thu, 9 Oct 2014 21:17:40 +0000 (21:17 +0000)]
Merge "Fix pthread_attr_getstack under valgrind."

9 years agoFix pthread_attr_getstack under valgrind.
Elliott Hughes [Thu, 9 Oct 2014 21:01:47 +0000 (14:01 -0700)]
Fix pthread_attr_getstack under valgrind.

valgrind seems to mess with the stack enough that the kernel will
report "[stack:pid]" rather than "[stack]" in /proc/self/maps, so
switch to the task-specific file instead to force "[stack]". (There
are two conditions in the kernel code that decides which form to
output.)

Bug: 17897476
Change-Id: Iff85ceb6d52e8716251fab4e45d95a27184c5529

9 years agoMerge "Re-add dlmalloc for 32 bit."
Christopher Ferris [Thu, 9 Oct 2014 18:50:32 +0000 (18:50 +0000)]
Merge "Re-add dlmalloc for 32 bit."

9 years agoRe-add dlmalloc for 32 bit.
Christopher Ferris [Thu, 9 Oct 2014 05:48:20 +0000 (22:48 -0700)]
Re-add dlmalloc for 32 bit.

It turns out that appportable has a version that calls dlmalloc directly.
Re-add the dlmalloc symbol for 32 bit only as a compatibility shim that
calls malloc.

Bug: 17881362

(cherry pick commit from c9734d24d92f4737f5ab3808c77d816a1b084582)

Change-Id: Iee9a777f66a1edb407d7563a60792b767ac4f83a

9 years agoMerge "Use snprintf(3) instead of sprintf(3)."
Dan Albert [Thu, 9 Oct 2014 00:32:10 +0000 (00:32 +0000)]
Merge "Use snprintf(3) instead of sprintf(3)."

9 years agoUse snprintf(3) instead of sprintf(3).
Dan Albert [Thu, 9 Oct 2014 00:10:26 +0000 (17:10 -0700)]
Use snprintf(3) instead of sprintf(3).

This change should probably be made upstream as well, but they have a
note about not using it because it isn't available on all systems.

Change-Id: I6d8404c031bd2f486532ced55d94bbb4a4cd2e71

9 years agoMerge "Replace uses of sprintf(3) with snprintf(3)."
Dan Albert [Wed, 8 Oct 2014 23:44:48 +0000 (23:44 +0000)]
Merge "Replace uses of sprintf(3) with snprintf(3)."

9 years agoMerge "Build dlext_testzip as custom module."
Ying Wang [Wed, 8 Oct 2014 23:43:49 +0000 (23:43 +0000)]
Merge "Build dlext_testzip as custom module."

9 years agoBuild dlext_testzip as custom module.
Ying Wang [Wed, 8 Oct 2014 23:22:03 +0000 (16:22 -0700)]
Build dlext_testzip as custom module.

Use $(BUILD_SYSTEM)/base_rules to build it as custom module, so that
it's exposed to utilities like mm/mmma etc.

Bug: 17887283
Change-Id: Idd7b62d08e4d56331a3ff15da6aa46eae8c41898

9 years agoReplace uses of sprintf(3) with snprintf(3).
Dan Albert [Wed, 8 Oct 2014 04:08:41 +0000 (21:08 -0700)]
Replace uses of sprintf(3) with snprintf(3).

At -00, the compiler warns about sprintf(3), and it is promoted to an
error by -Werror.

Change-Id: Ibb380d27d7eb09dda8ac785be2478d656b379190

9 years agoMerge "Fix mma build: 'cd bionic && mma' case"
Dmitriy Ivanov [Wed, 8 Oct 2014 22:39:20 +0000 (22:39 +0000)]
Merge "Fix mma build: 'cd bionic && mma' case"

9 years agoFix mma build: 'cd bionic && mma' case
Dmitriy Ivanov [Wed, 8 Oct 2014 21:55:39 +0000 (14:55 -0700)]
Fix mma build: 'cd bionic && mma' case

Bug: 17887283
Change-Id: If2015f2361348eac8c53aed3f8715077b79ce2f5

9 years agoMerge "Don't silently return EINVAL for a NULL sem_t*."
Elliott Hughes [Wed, 8 Oct 2014 22:21:12 +0000 (22:21 +0000)]
Merge "Don't silently return EINVAL for a NULL sem_t*."

9 years agoMerge "Use ln(1) rather than cp(1) to set up host /system for testing."
Elliott Hughes [Wed, 8 Oct 2014 22:17:18 +0000 (22:17 +0000)]
Merge "Use ln(1) rather than cp(1) to set up host /system for testing."

9 years agoDon't silently return EINVAL for a NULL sem_t*.
Elliott Hughes [Wed, 8 Oct 2014 22:11:44 +0000 (15:11 -0700)]
Don't silently return EINVAL for a NULL sem_t*.

Bug: 17571197
Change-Id: Ied31d8de87fc10d20dc338be0457e9a4ddb9a0a5

9 years agoUse ln(1) rather than cp(1) to set up host /system for testing.
Elliott Hughes [Wed, 8 Oct 2014 21:56:27 +0000 (14:56 -0700)]
Use ln(1) rather than cp(1) to set up host /system for testing.

The trouble with cp is that if you're running a command directly (because
you're using gdb, say), you need to make sure you've copied the linker
and shell again if you've done a rebuild that caused a change there.

Change-Id: I312da30c0f92b515967e854e0faba272c961ecc7

9 years agoMerge "Fix compile-time tests for LP64."
Elliott Hughes [Wed, 8 Oct 2014 20:31:09 +0000 (20:31 +0000)]
Merge "Fix compile-time tests for LP64."

9 years agoFix compile-time tests for LP64.
Elliott Hughes [Wed, 8 Oct 2014 20:21:29 +0000 (13:21 -0700)]
Fix compile-time tests for LP64.

Change-Id: Ib9bec5c6c77c55679d64c2bc15ab70979c40e66c

9 years agoMerge "Add basic compile time tests for sn?printf."
Dan Albert [Wed, 8 Oct 2014 20:04:26 +0000 (20:04 +0000)]
Merge "Add basic compile time tests for sn?printf."

9 years agoAdd basic compile time tests for sn?printf.
Dan Albert [Wed, 8 Oct 2014 18:21:32 +0000 (11:21 -0700)]
Add basic compile time tests for sn?printf.

Bug: 17782583
Change-Id: I6443537c643d1cc6b4a8f2dab5c364f2fbc1faa3

9 years agoMerge "Revert "Work around a bug in Immersion's libImmEmulatorJ.so.""
Elliott Hughes [Wed, 8 Oct 2014 16:40:57 +0000 (16:40 +0000)]
Merge "Revert "Work around a bug in Immersion's libImmEmulatorJ.so.""

9 years agoMerge "Inline helpers need to be exported unmangled."
Dan Albert [Wed, 8 Oct 2014 00:01:08 +0000 (00:01 +0000)]
Merge "Inline helpers need to be exported unmangled."

9 years agoInline helpers need to be exported unmangled.
Dan Albert [Tue, 7 Oct 2014 18:10:36 +0000 (11:10 -0700)]
Inline helpers need to be exported unmangled.

__open_2() is used by the fortify implementation of open(2) in
fcntl.h, and as such needs an unmangled C name. For some reason
(inlining?), this doesn't cause problems at the default optimization
level, but does for -O0.

The rest of these didn't cause build failures, but they look suspect
and probably will, we just haven't caught them yet.

Bug: 17784968
Change-Id: I7391a7a8999ee204eaf6abd14a3d5373ea419d5b

9 years agoRevert "Work around a bug in Immersion's libImmEmulatorJ.so."
Elliott Hughes [Tue, 7 Oct 2014 23:31:06 +0000 (23:31 +0000)]
Revert "Work around a bug in Immersion's libImmEmulatorJ.so."

This reverts commit 7d3f553f989f830976efa92ddc3c84661d4d42aa.

Change-Id: I8909b6aa1d97e9a61dbe95a2d91b9fbe336b58f0

9 years agoMerge "Work around a bug in Immersion's libImmEmulatorJ.so."
Elliott Hughes [Tue, 7 Oct 2014 23:08:11 +0000 (23:08 +0000)]
Merge "Work around a bug in Immersion's libImmEmulatorJ.so."

9 years agoWork around a bug in Immersion's libImmEmulatorJ.so.
Elliott Hughes [Tue, 7 Oct 2014 23:02:11 +0000 (16:02 -0700)]
Work around a bug in Immersion's libImmEmulatorJ.so.

This library calls pthread_mutex_lock and pthread_mutex_unlock with a NULL
pthread_mutex_t*. This gives them (and their users) one release to fix things.

Bug: 17443936
Change-Id: I3b63c9a3dd63db0833f21073e323b3236a13b47a

9 years agoMerge "Use snprintf instead of sprintf."
Dan Albert [Tue, 7 Oct 2014 18:38:11 +0000 (18:38 +0000)]
Merge "Use snprintf instead of sprintf."

9 years agoUse snprintf instead of sprintf.
Dan Albert [Tue, 7 Oct 2014 18:07:53 +0000 (11:07 -0700)]
Use snprintf instead of sprintf.

At -O0, the attribute warning on sprintf is actually triggered (why
doesn't this happen with -Os?!) and promoted to an error by -Werror.
asctime64_r() is a non-standard function, but the IBM docs state that
the buffer is assumed to be at least 26 characters wide, and the
format string does limit to that (assuming a 4 digit year, also
defined by the IBM docs).

http://www-01.ibm.com/support/knowledgecenter/SSLTBW_2.1.0/com.ibm.zos.v2r1.bpxbd00/asctimer.htm

Change-Id: I1c884474a769aa16c53e985c3d8d694c478c1189

9 years agoMerge "Fix 2nd arch prefix application"
Dmitriy Ivanov [Tue, 7 Oct 2014 17:04:23 +0000 (17:04 +0000)]
Merge "Fix 2nd arch prefix application"

9 years agoFix 2nd arch prefix application
Dmitriy Ivanov [Tue, 7 Oct 2014 16:46:46 +0000 (09:46 -0700)]
Fix 2nd arch prefix application

Bug: 17887283
Change-Id: I7b3b2332212fe416d6b5f5c2e5c1e07605ba09aa

9 years agoMerge "Force export symbols on all x86 variants in libc."
Christopher Ferris [Mon, 6 Oct 2014 23:45:02 +0000 (23:45 +0000)]
Merge "Force export symbols on all x86 variants in libc."

9 years agoForce export symbols on all x86 variants in libc.
Christopher Ferris [Mon, 6 Oct 2014 22:11:28 +0000 (15:11 -0700)]
Force export symbols on all x86 variants in libc.

For silvermont, the __popcountsi2 symbol does not get exported by libc.
But for atom, this symbol is exported. Since we already exported this symbol
for previous releases, it's better to just follow through and force
the export, but only for 32 bit. x86 64 bit will not export this symbol.

Bug: 17681440

(cherry picked from commit d11eac3455a059a092ceee92eda9905e1d661e94)

Change-Id: I93704c721d98d569922f606f214069bda24872ba

9 years agoMerge "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"
Nick Kralevich [Mon, 6 Oct 2014 23:21:03 +0000 (23:21 +0000)]
Merge "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"

9 years agocdefs.h: add artificial attribute to FORTIFY_SOURCE functions
Nick Kralevich [Mon, 6 Oct 2014 21:49:00 +0000 (14:49 -0700)]
cdefs.h: add artificial attribute to FORTIFY_SOURCE functions

Otherwise the gcc compiler warning doesn't show up.

Add -Wno-error to fortify related tests. Fortify related tests
are expected to be examples of bad programs, and in many
cases shouldn't compile cleanly. Rewriting them to compile
cleanly isn't feasible nor desirable.

Bug: 17784968
Change-Id: I93bececa7444d965f18c7c27d46e7abce5c49a02

9 years agoMerge "Add compile time test infrastructure for bionic."
Dan Albert [Mon, 6 Oct 2014 19:45:31 +0000 (19:45 +0000)]
Merge "Add compile time test infrastructure for bionic."

9 years agoMerge "Resolve "unused DT entry" warnings for x86_64"
Dmitriy Ivanov [Mon, 6 Oct 2014 19:07:57 +0000 (19:07 +0000)]
Merge "Resolve "unused DT entry" warnings for x86_64"

9 years agoAdd compile time test infrastructure for bionic.
Dan Albert [Fri, 3 Oct 2014 20:46:42 +0000 (13:46 -0700)]
Add compile time test infrastructure for bionic.

Actual tests to follow...

Bug: 17782583
Change-Id: I22857f19899c3b39dbc2e5473ad3be1a8a291139

9 years agoResolve "unused DT entry" warnings for x86_64
Dmitriy Ivanov [Mon, 6 Oct 2014 18:30:43 +0000 (11:30 -0700)]
Resolve "unused DT entry" warnings for x86_64

Change-Id: I3a2c87bf6e6259313cd9203389623f13b7405ef3

9 years agoMerge "Add file_offset parameter to android_extinfo"
Dmitriy Ivanov [Mon, 6 Oct 2014 17:19:38 +0000 (17:19 +0000)]
Merge "Add file_offset parameter to android_extinfo"

9 years agoAdd file_offset parameter to android_extinfo
Dmitriy Ivanov [Sat, 4 Oct 2014 00:52:44 +0000 (17:52 -0700)]
Add file_offset parameter to android_extinfo

Bug: 17762003
Change-Id: I8fb267a3155acef3dba534038cf5d1ef00d7154b

9 years agoMerge "Revert "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions""
Nick Kralevich [Mon, 6 Oct 2014 00:40:07 +0000 (00:40 +0000)]
Merge "Revert "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions""

9 years agoRevert "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"
Nick Kralevich [Mon, 6 Oct 2014 00:39:47 +0000 (00:39 +0000)]
Revert "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"

Broke the build.

In file included from frameworks/rs/cpu_ref/rsCpuCore.cpp:36:
system/core/include/cutils/properties.h:118:1: error: unknown attribute '__artificial__' ignored [-Werror,-Wunknown-attributes]
__BIONIC_FORTIFY_INLINE
^
bionic/libc/include/sys/cdefs.h:537:110: note: expanded from macro '__BIONIC_FORTIFY_INLINE'
#define __BIONIC_FORTIFY_INLINE extern __inline__ __always_inline __attribute__((gnu_inline)) __attribute__((__artificial__))
                                                                                                             ^
1 error generated.
make: *** [out/target/product/generic/obj/SHARED_LIBRARIES/libRSCpuRef_intermediates/rsCpuCore.o] Error 1
This reverts commit 9b543ffeac216189cc8125f7624da9a8cbcbe2e4.

Change-Id: I6a1198747505dcb402b722887c1bfbc3a628a8b8

9 years agoMerge "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"
Nick Kralevich [Mon, 6 Oct 2014 00:10:47 +0000 (00:10 +0000)]
Merge "cdefs.h: add artificial attribute to FORTIFY_SOURCE functions"

9 years agoMerge "further cleanup memcpy fortify implementation"
Nick Kralevich [Sun, 5 Oct 2014 16:43:24 +0000 (16:43 +0000)]
Merge "further cleanup memcpy fortify implementation"

9 years agofurther cleanup memcpy fortify implementation
Nick Kralevich [Sun, 5 Oct 2014 13:52:24 +0000 (06:52 -0700)]
further cleanup memcpy fortify implementation

Change-Id: I0380f3c8ca8f9bb8a206accd839c32c9daf13c56

9 years agocdefs.h: add artificial attribute to FORTIFY_SOURCE functions
Nick Kralevich [Sat, 4 Oct 2014 22:26:29 +0000 (15:26 -0700)]
cdefs.h: add artificial attribute to FORTIFY_SOURCE functions

Otherwise the gcc compiler warning doesn't show up.

Delete some unittests. These unittests no longer compile cleanly
using -Wall -Werror, and rewriting them to compile cleanly
isn't feasible.

Bug: 17784968
Change-Id: I9bbdc7b6a1c2ac75754f5d0f90782e0dfae66721

9 years agoMerge "string.h: remove unused variable"
Nick Kralevich [Sat, 4 Oct 2014 22:38:08 +0000 (22:38 +0000)]
Merge "string.h: remove unused variable"

9 years agostring.h: remove unused variable
Nick Kralevich [Sat, 4 Oct 2014 22:20:00 +0000 (15:20 -0700)]
string.h: remove unused variable

Change-Id: I08b7dab1b374ad427fcbee4183f2c7d93d0199a9

9 years agoMerge "Reimplement clock(3) using clock_gettime(3)"
Elliott Hughes [Sat, 4 Oct 2014 01:57:42 +0000 (01:57 +0000)]
Merge "Reimplement clock(3) using clock_gettime(3)"

9 years agoReimplement clock(3) using clock_gettime(3)
Alex Van Brunt [Fri, 26 Sep 2014 20:32:47 +0000 (13:32 -0700)]
Reimplement clock(3) using clock_gettime(3)

Unlike times(), clock_gettime() is implemented as a vDSO on many architectures.
So, using clock_gettime() will return a more accurate time and do so with less
overhead because it does have the overhead of calling into the kernel.

It is also significantly more accurate because it measures the actual time in
nanoseconds rather than the number of ticks (typically 1 millisecond or more).

Bug: 17814435
Change-Id: Id4945d9f387330518f78669809639952e9227ed9

9 years agoMerge "Add RTLD_NODELETE flag support"
Dmitriy Ivanov [Thu, 2 Oct 2014 21:47:29 +0000 (21:47 +0000)]
Merge "Add RTLD_NODELETE flag support"

9 years agoMerge "Make memory_order_acquire visible in global namespace"
Hans Boehm [Wed, 1 Oct 2014 19:16:49 +0000 (19:16 +0000)]
Merge "Make memory_order_acquire visible in global namespace"

9 years agoMerge "Add stpcpy assembler version."
Christopher Ferris [Wed, 1 Oct 2014 18:16:14 +0000 (18:16 +0000)]
Merge "Add stpcpy assembler version."

9 years agoAdd RTLD_NODELETE flag support
Dmitriy Ivanov [Mon, 19 May 2014 22:06:58 +0000 (15:06 -0700)]
Add RTLD_NODELETE flag support

Bug: https://code.google.com/p/android/issues/detail?id=64069
Change-Id: Ie5f90482feae86391172be4b32d6cb7d76f446fb

9 years agoMerge "Update bionic to tzdata 2014h."
Elliott Hughes [Wed, 1 Oct 2014 17:20:33 +0000 (17:20 +0000)]
Merge "Update bionic to tzdata 2014h."

9 years agoMerge "Fix update-tzdata.py to rebuild icu4c .dat file."
Elliott Hughes [Wed, 1 Oct 2014 17:00:23 +0000 (17:00 +0000)]
Merge "Fix update-tzdata.py to rebuild icu4c .dat file."

9 years agoAdd stpcpy assembler version.
Christopher Ferris [Tue, 30 Sep 2014 18:53:13 +0000 (11:53 -0700)]
Add stpcpy assembler version.

For generic, continue to use the C version of the code.

Bug: 13746695
Change-Id: I77426a70b06131f2373bb51265bea1240bb3f101

9 years agoMake memory_order_acquire visible in global namespace
Hans Boehm [Wed, 1 Oct 2014 01:31:04 +0000 (18:31 -0700)]
Make memory_order_acquire visible in global namespace

We were missing that using directive when including <atomic>.

Bug:17736764
Change-Id: Ie8ca92a952749415567bcd5fa21d56629a364660

9 years agoMerge "Add coverage flag for libm."
Dan Albert [Wed, 1 Oct 2014 01:28:24 +0000 (01:28 +0000)]
Merge "Add coverage flag for libm."