OSDN Git Service
Elliott Hughes [Wed, 2 Aug 2017 19:43:58 +0000 (19:43 +0000)]
Merge "Add some <sched.h> tests."
am:
26c59f7082
Change-Id: I5423c0d614e7fee238f0f2980956878111604efe
Elliott Hughes [Wed, 2 Aug 2017 19:28:38 +0000 (19:28 +0000)]
Merge "Add some <sched.h> tests."
Elliott Hughes [Wed, 2 Aug 2017 06:02:48 +0000 (23:02 -0700)]
Add some <sched.h> tests.
I was intending to change our behavior to match POSIX, but reality convinced
me otherwise.
Also add missing argument names to the header file on the assumption they'll
be shown in Studio one day.
Bug: http://b/
26203902
Test: ran tests
Change-Id: I2aaea48a88d408a32925033fc8f17448fb63252e
Elliott Hughes [Wed, 2 Aug 2017 15:09:50 +0000 (15:09 +0000)]
Merge "Finish <stdio_ext.h>."
am:
316a44f33c
Change-Id: I5ec959f2a1f4090ca3852c2ec453759e5676d93c
Elliott Hughes [Wed, 2 Aug 2017 15:00:38 +0000 (15:00 +0000)]
Merge "Finish <stdio_ext.h>."
George Burgess IV [Wed, 2 Aug 2017 08:12:28 +0000 (08:12 +0000)]
Merge "libc fortify: make fcntl.h use diagnose_if"
am:
5ac5a7c083
Change-Id: I2db3f6d42cbb07c703f142665fd2a0c5059f78bc
Treehugger Robot [Wed, 2 Aug 2017 08:06:20 +0000 (08:06 +0000)]
Merge "libc fortify: make fcntl.h use diagnose_if"
George Burgess IV [Wed, 2 Aug 2017 07:43:56 +0000 (07:43 +0000)]
Merge "libc fortify: make socket.h and stdlib.h use diagnose_if"
am:
a821283d22
Change-Id: I818da183356943d72e96d1e2264747de6fe6a066
Treehugger Robot [Wed, 2 Aug 2017 07:37:46 +0000 (07:37 +0000)]
Merge "libc fortify: make socket.h and stdlib.h use diagnose_if"
George Burgess IV [Wed, 2 Aug 2017 07:33:55 +0000 (07:33 +0000)]
Merge "libc fortify: make stdio.h use diagnose_if"
am:
6e0de9da49
Change-Id: Ic847d12c7822f7f8e25ec5a40e6cf7143417dad9
Treehugger Robot [Wed, 2 Aug 2017 07:27:23 +0000 (07:27 +0000)]
Merge "libc fortify: make stdio.h use diagnose_if"
George Burgess IV [Wed, 2 Aug 2017 06:50:53 +0000 (06:50 +0000)]
Merge "libc fortify: make poll.h and stat.h use diagnose_if"
am:
cff14d9e3e
Change-Id: I4cd2863d8f9d859e8441da7cc54f2fc15294b42e
Treehugger Robot [Wed, 2 Aug 2017 06:44:41 +0000 (06:44 +0000)]
Merge "libc fortify: make poll.h and stat.h use diagnose_if"
George Burgess IV [Tue, 1 Aug 2017 04:21:10 +0000 (21:21 -0700)]
libc fortify: make socket.h and stdlib.h use diagnose_if
Since realpath no longer needs to be overloaded, we can restore the
upstream source to purity. We'll be able to do this with most of the
other functions when we pull a newer clang in.
Bug:
12231437
Test: m checkbuild on bionic internal master; CtsBionicTestCases show
no new failures.
Change-Id: I484221bba0b291273fece23d2be2f5f9fd713d2c
George Burgess IV [Tue, 1 Aug 2017 04:23:34 +0000 (21:23 -0700)]
libc fortify: make stdio.h use diagnose_if
This also throws in some reformatting to match the GCC FORTIFY style in
this file (e.g. 4 space standard indent).
Per b/
36984245, some function protos are unwrapped beyond 100 chars, as
well.
Bug:
12231437
Test: m checkbuild on bionic internal master; CtsBionicTestCases show
no new failures.
Change-Id: I6dce60ec14d5cd703d9e724699634a197ca11023
George Burgess IV [Tue, 1 Aug 2017 04:16:05 +0000 (21:16 -0700)]
libc fortify: make poll.h and stat.h use diagnose_if
Bug:
12231437
Test: m checkbuild on bionic internal master; CtsBionicTestCases show
no new failures.
Change-Id: I7dbdeca12c46d1ee3804f13546b11c3f3b3a7596
George Burgess IV [Tue, 1 Aug 2017 04:39:33 +0000 (21:39 -0700)]
libc fortify: make fcntl.h use diagnose_if
Bug:
12231437
Test: m checkbuild on bionic internal master; CtsBionicTestCases show
no new failures.
Change-Id: I4a31d9e65ed5820a581261a7e8f63a83e44936c3
Dan Albert [Tue, 1 Aug 2017 23:19:17 +0000 (23:19 +0000)]
Merge "Mark new/delete as weak in the NDK stubs."
am:
aa8f766bf0
Change-Id: I1033bc0a963dd03476a77c0a54915b9c4353a5aa
George Burgess IV [Tue, 1 Aug 2017 23:18:59 +0000 (23:18 +0000)]
Merge "libc fortify: make unistd.h use diagnose_if"
am:
0ec52ce15b
Change-Id: I1361011fc3f466f681d651f85e8e828228c77b41
Treehugger Robot [Tue, 1 Aug 2017 23:11:12 +0000 (23:11 +0000)]
Merge "Mark new/delete as weak in the NDK stubs."
Treehugger Robot [Tue, 1 Aug 2017 23:06:09 +0000 (23:06 +0000)]
Merge "libc fortify: make unistd.h use diagnose_if"
Elliott Hughes [Tue, 1 Aug 2017 22:57:26 +0000 (22:57 +0000)]
Merge "Remove nullability specifications."
am:
fe3d58f49f
Change-Id: Id4fff7a1ba3f1ed76ae77ec42fdd0ca664dceeff
Elliott Hughes [Tue, 1 Aug 2017 22:46:40 +0000 (22:46 +0000)]
Merge "Remove nullability specifications."
Elliott Hughes [Tue, 1 Aug 2017 22:16:36 +0000 (15:16 -0700)]
Finish <stdio_ext.h>.
Implement __freading and __fwriting, and clarify the documentation that was
the cause of these not being implemented for years.
Bug: http://b/
17157253
Test: ran tests
Change-Id: I89542c8131b13889e2585417a024050ecf2abcb7
Elliott Hughes [Tue, 1 Aug 2017 20:57:43 +0000 (20:57 +0000)]
Merge "Clean up linker fatals slightly."
am:
1bac61f09f
Change-Id: Icf7d5b5b0def92ddb60c5c8a5c15e8867e76b2ad
George Burgess IV [Tue, 1 Aug 2017 04:30:47 +0000 (21:30 -0700)]
libc fortify: make unistd.h use diagnose_if
Also:
- remove a NULL check. The bug that's been working around has been
fixed since clang r298431, and our compiler is built from r300080, and
- unwrap some function prototypes, per b/
36984245.
If you're interested, here's what a new error versus an old one looks
like: https://drive.google.com/file/d/0B-8OBPboNJ_4MVdlYXdJbUFBYUk/view
Bug:
12231437
Test: m checkbuild on bullhead internal master; CtsBionicTestCases show
no new failures.
Change-Id: Ibafe61198988b4c8bee8391bcdf01bcbc04233de
Treehugger Robot [Tue, 1 Aug 2017 20:48:56 +0000 (20:48 +0000)]
Merge "Clean up linker fatals slightly."
Elliott Hughes [Tue, 1 Aug 2017 20:24:40 +0000 (13:24 -0700)]
Remove nullability specifications.
Bug: http://b/
64251432
Test: builds
Change-Id: I5b1613484783f7478d30b5e694007f77fa626659
Elliott Hughes [Tue, 1 Aug 2017 20:01:13 +0000 (20:01 +0000)]
Merge "Consistently say va_arg rather than __va_arg."
am:
3e523a6562
Change-Id: I3a01c16fdf736a5df9b9be4a6ba58706aadb9f6d
Treehugger Robot [Tue, 1 Aug 2017 19:51:32 +0000 (19:51 +0000)]
Merge "Consistently say va_arg rather than __va_arg."
Elliott Hughes [Tue, 1 Aug 2017 18:14:00 +0000 (18:14 +0000)]
Merge "Clean up some <stdio.h> constants."
am:
7f86f1dd15
Change-Id: I917e76ed6786ce08fac63f887e84c04b20cd02b8
Elliott Hughes [Tue, 1 Aug 2017 18:09:02 +0000 (18:09 +0000)]
Merge "Clean up some <stdio.h> constants."
Elliott Hughes [Tue, 1 Aug 2017 17:52:30 +0000 (17:52 +0000)]
Merge "Remove restrict/__restrict."
am:
6616056f0f
Change-Id: I5343c191c3f424343d1c7617e054b2347d0dc6b5
Elliott Hughes [Tue, 1 Aug 2017 17:48:08 +0000 (10:48 -0700)]
Consistently say va_arg rather than __va_arg.
Bug: http://b/
64250392
Test: builds
Change-Id: I435242f620245d825b8a68365c71c99a60aba625
Elliott Hughes [Tue, 1 Aug 2017 17:44:19 +0000 (17:44 +0000)]
Merge "Remove restrict/__restrict."
Elliott Hughes [Mon, 31 Jul 2017 18:43:34 +0000 (11:43 -0700)]
Clean up linker fatals slightly.
Use __linker_cannot_link more consistently, and fix a comment.
Bug: http://b/
22798163
Test: ran tests
Change-Id: Id6d868f459997eaa67b8cbbf85b0b0f2749bf43d
Elliott Hughes [Tue, 1 Aug 2017 15:28:46 +0000 (08:28 -0700)]
Remove restrict/__restrict.
We've never really used __restrict: only <string.h> and <stdio.h> (which
are still very similar to upstream BSD headers) consistently have these
annotations. Neither clang nor GCC warns for trivial cases, and there's
little obvious documentation benefit.
Bug: http://b/
30833514
Test: builds
Change-Id: I3e4384281865475d0c55d764b546d8166419ee31
George Burgess IV [Tue, 1 Aug 2017 02:44:01 +0000 (02:44 +0000)]
Merge "libc: Add diagnose_if macros to cdefs."
am:
c2a10f7f1b
Change-Id: If9203f290902beb68cdc8d4986f74ba30dabd088
Treehugger Robot [Tue, 1 Aug 2017 02:32:33 +0000 (02:32 +0000)]
Merge "libc: Add diagnose_if macros to cdefs."
George Burgess IV [Tue, 1 Aug 2017 02:07:58 +0000 (02:07 +0000)]
Merge "libc: Split FORTIFY into its own headers"
am:
1db06faa38
Change-Id: I387fb3d28c9d3f827f0e0011718e72efb0949db7
Treehugger Robot [Tue, 1 Aug 2017 02:03:52 +0000 (02:03 +0000)]
Merge "libc: Split FORTIFY into its own headers"
Elliott Hughes [Tue, 1 Aug 2017 00:20:18 +0000 (17:20 -0700)]
Clean up some <stdio.h> constants.
And actually test an assertion rather than just state it in a comment.
Bug: N/A
Test: ran tests
Change-Id: I07699483aca4aac4e089d8b99123cb5bde9b3c63
George Burgess IV [Mon, 31 Jul 2017 23:57:15 +0000 (16:57 -0700)]
libc: Add diagnose_if macros to cdefs.
I plan on having one review per file for the enable_if->diagnose_if
FORTIFY migration. Having this in means that no one review is dependent
on another.
Bug:
12231437
Test: m.
Change-Id: Ic0b07d7f7f6782e371c8792eb1e40cdfa32d3e35
Dan Albert [Mon, 31 Jul 2017 16:01:14 +0000 (16:01 +0000)]
Merge "Allow calling v?dprintf for any target API level."
am:
4dd33f92d8
Change-Id: I740b3424c4e440350d63c3b7056c6e21acd77e60
Treehugger Robot [Mon, 31 Jul 2017 15:57:00 +0000 (15:57 +0000)]
Merge "Allow calling v?dprintf for any target API level."
Christopher Ferris [Sat, 29 Jul 2017 21:07:39 +0000 (21:07 +0000)]
Merge "Move back to the previous version of ion.h."
am:
f94ddd7a0e
Change-Id: I7a782bf91587547dfd4e2137dbdb529d1cd80040
Christopher Ferris [Sat, 29 Jul 2017 21:04:00 +0000 (21:04 +0000)]
Merge "Move back to the previous version of ion.h."
Christopher Ferris [Sat, 29 Jul 2017 20:42:28 +0000 (13:42 -0700)]
Move back to the previous version of ion.h.
The new kernel headers changed this and lots of parts of the tree
want the old definition, so restore the previous version.
Test: Builds.
Change-Id: I725e56bb7c51c9fcd23ebc4b67eac33d66e196e7
Christopher Ferris [Sat, 29 Jul 2017 19:06:47 +0000 (19:06 +0000)]
Merge "Remove workaround in <sys/epoll.h>."
am:
6c26e9ec6f
Change-Id: Ib77bb0a43f800061638e71bd572445820f4369b1
Christopher Ferris [Sat, 29 Jul 2017 19:06:41 +0000 (19:06 +0000)]
Merge "Update to kernel headers v4.12.3."
am:
c478c32d2f
Change-Id: If521c86d7895959f49e6933be0a8e9959105ad0c
Christopher Ferris [Sat, 29 Jul 2017 19:03:09 +0000 (19:03 +0000)]
Merge "Remove workaround in <sys/epoll.h>."
Christopher Ferris [Sat, 29 Jul 2017 19:03:09 +0000 (19:03 +0000)]
Merge "Update to kernel headers v4.12.3."
Josh Gao [Sat, 29 Jul 2017 16:44:23 +0000 (16:44 +0000)]
Merge "versioner: compile with `clang -include foo.h -`."
am:
b1bbf2bb24
Change-Id: Ia5cd7d937435a147b152b82433e483ab0bea3b57
Treehugger Robot [Sat, 29 Jul 2017 16:36:33 +0000 (16:36 +0000)]
Merge "versioner: compile with `clang -include foo.h -`."
Elliott Hughes [Sat, 29 Jul 2017 14:59:15 +0000 (14:59 +0000)]
Merge "POSIX getlogin_r/LOGIN_NAME_MAX/TTY_NAME_MAX."
am:
8aa4bc7c55
Change-Id: I578d3e5c096401b0447f9ed7223d3e20ffb5b19e
Elliott Hughes [Sat, 29 Jul 2017 14:56:02 +0000 (14:56 +0000)]
Merge "POSIX getlogin_r/LOGIN_NAME_MAX/TTY_NAME_MAX."
Elliott Hughes [Sat, 29 Jul 2017 07:07:47 +0000 (07:07 +0000)]
Merge "Fix the sysconf _SC_2_VERSION test."
am:
3e183a5d30
Change-Id: I384fc20676715a1385a182428e5bbd17d4e3d8e8
Treehugger Robot [Sat, 29 Jul 2017 07:01:32 +0000 (07:01 +0000)]
Merge "Fix the sysconf _SC_2_VERSION test."
Elliott Hughes [Fri, 28 Jul 2017 23:53:32 +0000 (16:53 -0700)]
Fix the sysconf _SC_2_VERSION test.
Not sure how this ended up broken. Too many branches!
Bug: N/A
Test: ran tests
Change-Id: Ieef1e8170c9aaf139d9e0664475ccb08b8fcdeba
Elliott Hughes [Fri, 28 Jul 2017 23:27:49 +0000 (16:27 -0700)]
POSIX getlogin_r/LOGIN_NAME_MAX/TTY_NAME_MAX.
Bug: N/A
Test: ran tests
Change-Id: I97c8ecc82fb09c3fa4bb5e6f000d46f04f3d1702
Elliott Hughes [Fri, 28 Jul 2017 21:50:26 +0000 (21:50 +0000)]
Merge "Implement wcwidth(3) in terms of icu4c."
am:
bafa1985ec
Change-Id: Ia7e74b6124206fe5f2c5bf372c03ee5246045462
Elliott Hughes [Fri, 28 Jul 2017 21:42:13 +0000 (21:42 +0000)]
Merge "Implement wcwidth(3) in terms of icu4c."
Elliott Hughes [Fri, 28 Jul 2017 20:25:46 +0000 (20:25 +0000)]
Merge "Actually turn on -Werror for crtbegin."
am:
07d2d327bf
Change-Id: Ia9be400c31c9815e2753bdbd30ef3f224dfda58f
Treehugger Robot [Fri, 28 Jul 2017 20:14:57 +0000 (20:14 +0000)]
Merge "Actually turn on -Werror for crtbegin."
Josh Gao [Fri, 28 Jul 2017 19:53:36 +0000 (12:53 -0700)]
versioner: compile with `clang -include foo.h -`.
At least one warning only triggers in files that are included, instead
of being passed directly. Switch to compiling with -include, and fix
the resulting warnings.
Bug: https://github.com/android-ndk/ndk/issues/474
Test: mma -j && versioner
Test: python tools/versioner/run_tests.py
Change-Id: I784698c18540c9cc30f372f279a1cec1d75721ea
Dan Albert [Fri, 28 Jul 2017 18:00:51 +0000 (11:00 -0700)]
Mark new/delete as weak in the NDK stubs.
Test: make ndk
Bug: None
Change-Id: I76eef4f0265e87ddbad84ac8c5d56225aa240239
Elliott Hughes [Fri, 28 Jul 2017 00:08:08 +0000 (17:08 -0700)]
Implement wcwidth(3) in terms of icu4c.
Based on com.google.i18n.CharWidth by the icu-team folks.
Bug: N/A
Test: ran tests
Change-Id: Ia9d818ec4ae60f2f3978533195330d00699397c5
George Burgess IV [Mon, 24 Jul 2017 22:05:05 +0000 (15:05 -0700)]
libc: Split FORTIFY into its own headers
This patch cleans up our standard headers by moving most of the FORTIFY
cruft out in to its own sandbox. In order to include the *_chk and
*_real declarations, you can either enable FORTIFY, or `#define
__BIONIC_DECLARE_FORTIFY_HELPERS`.
Both sys/select.h and strings.h are explicitly ignored by this patch.
Both of these files have very small __BIONIC_FORTIFY blocks, and don't
define any actual FORTIFY'ed functions (just macros, and 3 *_chk
functions).
This patch also makes the versioner ignore the FORTIFY implementation
headers, since we're guaranteed to pick the FORTIFY'ed headers up when
looking at the regular headers. (...Not to mention that making the
FORTIFY'ed headers freestanding would be annoying to do and maintain for
~no benefit).
We bake the knowledge of where FORTIFY headers live directly into the
versioner. We could go with a more general approach (e.g. adding an -X
IGNORED_FILE flag that tells the versioner to ignore
$HEADER_PATH/$IGNORED_FILE), but we'd then have to repeat that for every
test, every manual invocation of the versioner, etc. for no benefit
that's obvious to me.
Bug:
12231437
Test: m checkbuild on bullhead internal master + CtsBionicTestCases. no
new errors.
Change-Id: Iffc0cc609009b33d989cdaddde0a809282131a5b
Elliott Hughes [Thu, 27 Jul 2017 23:39:02 +0000 (16:39 -0700)]
Actually turn on -Werror for crtbegin.
And then -Wno-gcc-compat in the right place this time. (But leaving the
other one because I don't care in the non-crtbegin code either.)
Bug: N/A
Test: builds without warnings now
Change-Id: Id2cfa3392ee80c045644ebfdf92c6d341db90964
Dan Albert [Wed, 26 Jul 2017 21:09:45 +0000 (14:09 -0700)]
Allow calling v?dprintf for any target API level.
Old versions of Android called these fdprintf and vfdprintf out of
fears that the glibc names would collide with user debug printfs.
Allow users to just use dprintf and vfdprintf on any version by
renaming those calls to their legacy equivalents if needed.
Test: built trivial NDK module targeting android-14 and using dprintf
Test: make checkbuild
Bug: https://github.com/android-ndk/ndk/issues/72
Change-Id: I90de149278f931380418536abaef47c5cee5c195
Christopher Ferris [Wed, 26 Jul 2017 20:12:53 +0000 (13:12 -0700)]
Update to kernel headers v4.12.3.
Test: Built angler.
Change-Id: Icbcf4fac2334de8409b049ed7a3b4c24b4e98ce9
Elliott Hughes [Thu, 27 Jul 2017 22:14:52 +0000 (15:14 -0700)]
Remove workaround in <sys/epoll.h>.
Linux 4.12 has a better fix: https://lkml.org/lkml/2017/2/23/416 and
https://lkml.org/lkml/2017/2/23/417.
Bug: N/A
Test: builds
Change-Id: I9cbf0c60817332142c48297899062f1a25f2b082
Elliott Hughes [Thu, 27 Jul 2017 15:42:52 +0000 (15:42 +0000)]
Merge "Use O_APPEND for stdio append mode."
am:
38bcf2b704
Change-Id: I4e1274d93adafffc6c4795cbcb5755735533c4de
Elliott Hughes [Thu, 27 Jul 2017 15:33:30 +0000 (15:33 +0000)]
Merge "Use O_APPEND for stdio append mode."
Dan Albert [Thu, 27 Jul 2017 02:48:38 +0000 (02:48 +0000)]
Merge "Fix error condition on strtof inline."
am:
9faa95c9eb
Change-Id: Ic09c01e387ad5cf701368972aee6673e78822603
Treehugger Robot [Thu, 27 Jul 2017 02:40:40 +0000 (02:40 +0000)]
Merge "Fix error condition on strtof inline."
Dan Albert [Wed, 26 Jul 2017 21:59:24 +0000 (14:59 -0700)]
Fix error condition on strtof inline.
Test: make checkbuild
Test: pull into ndk, ./run_tests.py
Bug: https://github.com/android-ndk/ndk/issues/415
Change-Id: I1575f799c9e849d66456c785208a36709a7b8122
Dan Albert [Thu, 27 Jul 2017 00:17:13 +0000 (00:17 +0000)]
Merge "Add __ANDROID_NDK__ to identify using the NDK."
am:
d3e656abeb
Change-Id: I8c590998e93937c6702bcb91b9dcadb509f341ca
Treehugger Robot [Wed, 26 Jul 2017 23:47:38 +0000 (23:47 +0000)]
Merge "Add __ANDROID_NDK__ to identify using the NDK."
Dan Albert [Wed, 26 Jul 2017 21:27:43 +0000 (14:27 -0700)]
Add __ANDROID_NDK__ to identify using the NDK.
I've added some things like __ANDROID_MAJOR__ to an ndk-version.h, but
that is only in the NDK itself and so doesn't help the platform. Add
__ANDROID_NDK__ to identify that you're building for the NDK and not
the platform.
Test: make checkbuild
Bug: https://github.com/android-ndk/ndk/issues/407
Change-Id: I2d1f1c28e3764e4e658cf675b290b7a17253ee33
George Burgess IV [Wed, 26 Jul 2017 02:55:44 +0000 (02:55 +0000)]
Merge "libc: Move FORTIFY into one file; make style fixups"
am:
4f36cabd8d
Change-Id: I17f125e6100e0af30c083df7a6a7a19a90696dc3
Treehugger Robot [Wed, 26 Jul 2017 02:44:21 +0000 (02:44 +0000)]
Merge "libc: Move FORTIFY into one file; make style fixups"
Elliott Hughes [Wed, 26 Jul 2017 01:06:46 +0000 (18:06 -0700)]
Use O_APPEND for stdio append mode.
"Although not explicitly required by this volume of POSIX.1-2008, a good
implementation of append (a) mode would cause the O_APPEND flag to be set."
Yeah, about that...
Bug: N/A
Test: ran tests
Change-Id: I23c4bc5c1ebc92e0cb44025d2d313f321f9ffa68
George Burgess IV [Tue, 25 Jul 2017 18:43:39 +0000 (11:43 -0700)]
libc: Move FORTIFY into one file; make style fixups
This addresses post-commit feedback from
I88c39ca166bacde0b692aa3063e743bb046a5d2f. With this, our FORTIFY impl
now sits in one file.
Bug:
12231437
Test: mma; no new CtsBionicTestCases failures on bullhead internal
master.
Change-Id: I6f9ff81c3e86cf9d6a0efa650eb5765f1e2fa09c
Elliott Hughes [Tue, 25 Jul 2017 21:12:22 +0000 (21:12 +0000)]
Merge "Build crtbegin with clang too."
am:
9306889509
Change-Id: Ic36b2f3a9c25fe2567917cc6f9bbca9c104c189a
Treehugger Robot [Tue, 25 Jul 2017 21:03:26 +0000 (21:03 +0000)]
Merge "Build crtbegin with clang too."
Steve Muckle [Tue, 25 Jul 2017 20:26:21 +0000 (20:26 +0000)]
Merge "Create global seccomp policy."
am:
33f9043f48
Change-Id: I4ee6138d1bbcb2e80360601a7af4b425ca44ac06
Treehugger Robot [Tue, 25 Jul 2017 20:18:51 +0000 (20:18 +0000)]
Merge "Create global seccomp policy."
Elliott Hughes [Mon, 24 Jul 2017 23:53:11 +0000 (16:53 -0700)]
Build crtbegin with clang too.
If _start_main is static, clang optimizes it out and then generates an
undefined reference from the inline assembler, so let's explicitly tell
the compiler it's used.
Bug: N/A
Test: ran tests
Change-Id: Iba3d9c63abae9a0dc32f236b01783515f0190956
George Burgess IV [Tue, 25 Jul 2017 18:13:10 +0000 (18:13 +0000)]
Merge "Split our FORTIFY implementation into libc_fortify"
am:
086b1b5a22
Change-Id: I6e05d09bc00135657a35420fe102a37ccf8c59fc
Treehugger Robot [Tue, 25 Jul 2017 18:05:42 +0000 (18:05 +0000)]
Merge "Split our FORTIFY implementation into libc_fortify"
Elliott Hughes [Tue, 25 Jul 2017 15:00:25 +0000 (15:00 +0000)]
Merge "tests: clean up bionic-unit-tests-static linking"
am:
b8115bdf4f
Change-Id: I851293e85617365eacdc0aa811473d5a6ebfbb1d
Elliott Hughes [Tue, 25 Jul 2017 14:51:32 +0000 (14:51 +0000)]
Merge "tests: clean up bionic-unit-tests-static linking"
Elliott Hughes [Mon, 24 Jul 2017 23:22:17 +0000 (16:22 -0700)]
tests: clean up bionic-unit-tests-static linking
Adding libc++_static explicitly is not needed when it is specified
via stl:, and was the real reason we needed --allow-multiple-definition.
Bug: N/A
Test: ran static tests (32- and 64-bit)
Change-Id: I254a4b8c4f7955a17c970d31b427b05c1a2e0acf
Signed-off-by: Kevin Brodsky <kevin.brodsky@arm.com>
George Burgess IV [Fri, 21 Jul 2017 20:28:42 +0000 (13:28 -0700)]
Split our FORTIFY implementation into libc_fortify
As requested in the bug. This also rips __memcpy_chk out of memcpy.S,
which lets us cut down on copypasta (all of the implementations look
identical).
Bug:
12231437
Test: mma on aosp_{arm,arm64,mips,x86,x86_64} internal master;
checkbuild on bullhead internal master; CtsBionicTestCases on bullhead.
No new failures.
Change-Id: I88c39ca166bacde0b692aa3063e743bb046a5d2f
Elliott Hughes [Mon, 24 Jul 2017 21:09:43 +0000 (21:09 +0000)]
Merge "Remove the HASLB and FREEUB macros."
am:
4fce647a5b
Change-Id: Ic7316b86c5c1063144d3ac3706963e4bc3484a09
Treehugger Robot [Mon, 24 Jul 2017 21:05:37 +0000 (21:05 +0000)]
Merge "Remove the HASLB and FREEUB macros."
Elliott Hughes [Mon, 24 Jul 2017 20:55:11 +0000 (20:55 +0000)]
Merge "Make xfail passes not failures."
am:
4dc80181cd
Change-Id: I38cfd064c28994372d6f39a6f7e1e93a3ee58779
Elliott Hughes [Mon, 24 Jul 2017 20:51:18 +0000 (20:51 +0000)]
Merge "Make xfail passes not failures."
Elliott Hughes [Mon, 24 Jul 2017 20:39:09 +0000 (20:39 +0000)]
Merge "Replace fmemopen."
am:
bff8da308f
Change-Id: I67a84a5bc4d879a51871778fd4cea5646e54c34f