OSDN Git Service
Christopher Ferris [Tue, 26 Jan 2016 01:35:03 +0000 (01:35 +0000)]
Merge "Fix gcc test issues."
Christopher Ferris [Tue, 26 Jan 2016 01:23:38 +0000 (17:23 -0800)]
Fix gcc test issues.
Change-Id: I2b26b779f5198c7c9362a452b971260a8852aa7b
Elliott Hughes [Tue, 26 Jan 2016 00:32:38 +0000 (00:32 +0000)]
Merge "Revert "Make sem_wait able to return errno EINTR for sdk > 23.""
Dan Albert [Tue, 26 Jan 2016 00:20:06 +0000 (00:20 +0000)]
Revert "Make sem_wait able to return errno EINTR for sdk > 23."
Broke the build. There's no such file as bionic_sdk_version.h anywhere in the tree.
This reverts commit
892b61d3409e8cdf0e121c688737eb058d57f7ab.
Change-Id: Iec3f4588edfb1d1524bb5f16451fd05dc6ebe44a
Yabin Cui [Mon, 25 Jan 2016 23:39:29 +0000 (23:39 +0000)]
Merge "Make sem_wait able to return errno EINTR for sdk > 23."
Christopher Ferris [Mon, 25 Jan 2016 22:44:56 +0000 (22:44 +0000)]
Merge "Fix build for gcc."
Christopher Ferris [Mon, 25 Jan 2016 22:36:34 +0000 (14:36 -0800)]
Fix build for gcc.
Gcc doesn't like the brace object initialization, so make all of
the Feature objects explicit.
Also, no arguments to the error_log macros make gcc unhappy, so add
an option to turn these into warnings. These will be fixed when we
add the explicit _error and _warn log functions.
Change-Id: I35af834dabb5548923e893dd980a751fdebfa13a
Yabin Cui [Mon, 25 Jan 2016 19:37:16 +0000 (11:37 -0800)]
Make sem_wait able to return errno EINTR for sdk > 23.
Posix standards says sem_wait is interruptible by the delivery
of a signal. To keep compatiblity with old apps, only fix that
in newer sdk versions.
Bug:
26743454
Change-Id: I924cbb436658e3e0f397c922d866ece99b8241a3
Christopher Ferris [Mon, 25 Jan 2016 20:48:29 +0000 (20:48 +0000)]
Merge "Malloc debug rewrite."
Elliott Hughes [Mon, 25 Jan 2016 19:35:36 +0000 (19:35 +0000)]
Merge "Move internal FILE flags macros to "local.h" from <stdio.h>."
Chih-hung Hsieh [Mon, 25 Jan 2016 19:31:23 +0000 (19:31 +0000)]
Merge "Include libz for new libelf."
Chih-Hung Hsieh [Mon, 25 Jan 2016 18:28:26 +0000 (10:28 -0800)]
Include libz for new libelf.
New external/elfutils needs libz.
Change-Id: Ib2f2d56d4831809509a0fb3a849656f68adece7f
Elliott Hughes [Mon, 25 Jan 2016 19:10:47 +0000 (11:10 -0800)]
Move internal FILE flags macros to "local.h" from <stdio.h>.
Bug: http://b/
24807045
Change-Id: Ia42ec1907370b4af957ece94a856f71c7ee27cb7
Christopher Ferris [Tue, 17 Nov 2015 01:30:32 +0000 (17:30 -0800)]
Malloc debug rewrite.
The major components of the rewrite:
- Completely remove the qemu shared library code. Nobody was using it
and it appears to have broken at some point.
- Adds the ability to enable/disable different options independently.
- Adds a new option that can enable the backtrace on alloc/free when
a process gets a specific signal.
- Adds a new way to enable malloc debug. If a special property is
set, and the process has an environment variable set, then debug
malloc will be enabled. This allows something that might be
a derivative of app_process to be started with an environment variable
being enabled.
- get_malloc_leak_info() used to return one element for each pointer that
had the exact same backtrace. The new version returns information for
every one of the pointers with same backtrace. It turns out ddms already
automatically coalesces these, so the old method simply hid the fact
that there where multiple pointers with the same amount of backtrace.
- Moved all of the malloc debug specific code into the library.
Nothing related to the malloc debug data structures remains in libc.
- Removed the calls to the debug malloc cleanup routine. Instead, I
added an atexit call with the debug malloc cleanup routine. This gets
around most problems related to the timing of doing the cleanup.
The new properties and environment variables:
libc.debug.malloc.options
Set by option name (such as "backtrace"). Setting this to a bad value
will cause a usage statement to be printed to the log.
libc.debug.malloc.program
Same as before. If this is set, then only the program named will
be launched with malloc debug enabled. This is not a complete match,
but if any part of the property is in the program name, malloc debug is
enabled.
libc.debug.malloc.env_enabled
If set, then malloc debug is only enabled if the running process has the
environment variable LIBC_DEBUG_MALLOC_ENABLE set.
Bug:
19145921
Change-Id: I7b0e58cc85cc6d4118173fe1f8627a391b64c0d7
Elliott Hughes [Mon, 25 Jan 2016 18:50:41 +0000 (18:50 +0000)]
Merge "Add _seek64 to FILE."
Josh Gao [Mon, 25 Jan 2016 18:42:55 +0000 (18:42 +0000)]
Merge "Remove dlmalloc."
Elliott Hughes [Mon, 25 Jan 2016 18:31:35 +0000 (18:31 +0000)]
Merge "Add another stdio test."
Josh Gao [Thu, 7 Jan 2016 01:57:44 +0000 (17:57 -0800)]
Remove dlmalloc.
Bug: http://b/
17207577
Change-Id: Ie009badca6deb1f91b27a4340b70cdd6bedff893
Elliott Hughes [Fri, 22 Jan 2016 23:04:51 +0000 (15:04 -0800)]
Add _seek64 to FILE.
Move fdopen/fopen/freopen and change them to initialize _seek64 instead
of the legacy _seek. The in-memory streams can stick with _seek for now,
since you're not going to fit a > 4GiB in-memory stream on a 32-bit device
anyway.
Bug: http://b/
24807045
Change-Id: I09dcb426817b571415ce24d4d15f364cdda395b3
Elliott Hughes [Sat, 23 Jan 2016 04:50:40 +0000 (04:50 +0000)]
Merge "Put struct FILE back how NDK-built apps expect it."
Elliott Hughes [Sat, 23 Jan 2016 03:18:48 +0000 (19:18 -0800)]
Put struct FILE back how NDK-built apps expect it.
The first rule of stdio is you never change struct FILE. This broke all
NDK-built apps that used stdin/stdout/stderr. (Which is more than you
might think, given that those streams don't go anywhere useful. Svelte!)
I've added a big code comment because I knew when I removed the field that
doing so was a mistake, but I couldn't think why.
Bug: http://b/
24807045
Bug: http://b/
26747402
Change-Id: Ie1233586b223bb1cdf8e354c66d5ff23487a833a
Elliott Hughes [Sat, 23 Jan 2016 03:22:44 +0000 (19:22 -0800)]
Add another stdio test.
This test didn't catch anything, but it does ensure that we exercise
the "lots of files" case.
Bug: http://b/
26747402
Change-Id: I6c51c6436029572a49190d509f131eb93b808652
Dimitry Ivanov [Fri, 22 Jan 2016 18:38:08 +0000 (18:38 +0000)]
Merge "Revert "Temporary apply LIBC version to __pthread_gettid""
Elliott Hughes [Fri, 22 Jan 2016 03:19:20 +0000 (03:19 +0000)]
Merge "Loosen up sys_time.gettimeofday."
Dimitry Ivanov [Thu, 21 Jan 2016 22:25:32 +0000 (22:25 +0000)]
Revert "Temporary apply LIBC version to __pthread_gettid"
This reverts commit
0ef1d121b5e4845f4ef3b59ae9a1f99ceb531186.
Bug: http://b/
26392296
Bug: http://b/
26391427
Change-Id: I7bbb555de3a43813e7623ff6ad4e17874d283eca
Dimitry Ivanov [Fri, 22 Jan 2016 01:41:27 +0000 (01:41 +0000)]
Merge "libc: hide __signalfd4 symbol"
Dimitry Ivanov [Fri, 22 Jan 2016 01:40:45 +0000 (01:40 +0000)]
Merge "libc: hide fake_gmtime_r and fake_localtime_r"
Elliott Hughes [Fri, 22 Jan 2016 00:43:43 +0000 (16:43 -0800)]
Loosen up sys_time.gettimeofday.
We've seen it take 1146us on Nexus 9 (which did have exceptionally slow
system calls).
Bug: http://b/
26724042
Change-Id: I263b7e1267d58fe4a6528403d03e5b245fdcd528
Dimitry Ivanov [Fri, 22 Jan 2016 00:05:23 +0000 (16:05 -0800)]
libc: hide fake_gmtime_r and fake_localtime_r
Bug: http://b/
26274444
Change-Id: Icf96a00142a287d8d3466d2545e213e605601b42
Dimitry Ivanov [Thu, 21 Jan 2016 23:23:49 +0000 (23:23 +0000)]
Merge "linker: align allocated blocks to 16 bytes"
Dimitry Ivanov [Thu, 21 Jan 2016 23:19:52 +0000 (15:19 -0800)]
libc: hide __signalfd4 symbol
Bug: http://b/
26274444
Change-Id: Iaa2df640cac07e2e407ca11b9701ee9e11729146
Dimitry Ivanov [Thu, 21 Jan 2016 18:55:40 +0000 (10:55 -0800)]
linker: align allocated blocks to 16 bytes
C/C++ requires the result of malloc/new to be
aligned for any primitive type.
Change-Id: I715b7679e738f34b3b409993fb3ef242e1321b7f
Elliott Hughes [Thu, 21 Jan 2016 18:35:18 +0000 (18:35 +0000)]
Merge "Simplify fseek/ftell."
Elliott Hughes [Thu, 21 Jan 2016 01:14:53 +0000 (17:14 -0800)]
Simplify fseek/ftell.
Another step towards _FILE_OFFSET_BITS=64 support.
Bug: http://b/
24807045
Change-Id: I00b83c81a7b108176c4d9437bc32611f73b7e967
Christopher Ferris [Thu, 21 Jan 2016 18:31:12 +0000 (18:31 +0000)]
Merge "Remove dependency on zipalign."
Dimitry Ivanov [Thu, 21 Jan 2016 05:30:31 +0000 (05:30 +0000)]
Merge "Do not unmap reserved region on dlclose"
Dimitry Ivanov [Fri, 15 Jan 2016 19:13:35 +0000 (11:13 -0800)]
Do not unmap reserved region on dlclose
dlclose used to unmap the part of the reserved region
for ANDROID_DLEXT_RESERVED_ADDRESS that was neccessary
to map PT_LOAD segments. With this change dlclose
replaces mapped PT_LOAD segments with a PROT_NONE,
MAP_ANONYMOUS | MAP_NORESERVE.
Previously caller was unmapping the reserved region after
the failed dlclose which led to race condition when someone
else reused the region freed by dlclose but before the unmap
by the chromium code.
Bug: http://code.google.com/p/chromium/issues/detail?id=568880
Change-Id: I0f5eaa2bf6641f83dde469b631c518482acc59a2
Christopher Ferris [Wed, 20 Jan 2016 04:32:37 +0000 (20:32 -0800)]
Remove dependency on zipalign.
Roll our own version of zipalign so that we can break the dependency
on the build tools zipalign. This breaks the transitive dependency
on androidfw so that building bionic unit tests in brillo works again.
Also modify the DlExtTest.ExtInfoUseFdWithOffset test so it dynamically
gets the offset of the shared library inside of the zip instead of
hard-coding the value.
Bug:
25446938
Change-Id: Idfb5d3089960a94eefa2c76e03da1ad2f4d7fb2f
Elliott Hughes [Wed, 20 Jan 2016 23:07:50 +0000 (23:07 +0000)]
Merge "Move stdio implementation details around a little."
Colin Cross [Wed, 20 Jan 2016 21:10:34 +0000 (21:10 +0000)]
Merge "Update Android.bp to match Android.mk"
Colin Cross [Wed, 20 Jan 2016 20:01:19 +0000 (12:01 -0800)]
Update Android.bp to match Android.mk
Change-Id: Ia14a7af94d6220b3078b8d37b76990454b16703c
Colin Cross [Wed, 20 Jan 2016 19:58:16 +0000 (19:58 +0000)]
Merge "Update Android.bp to match Android.mk"
Elliott Hughes [Wed, 20 Jan 2016 00:28:15 +0000 (16:28 -0800)]
Move stdio implementation details around a little.
Change-Id: I24594426d5479bdd55cbef0ab1b7d76c249dbd0c
Elliott Hughes [Wed, 20 Jan 2016 00:13:45 +0000 (00:13 +0000)]
Merge "Make FILE*s less usable after fclose(3)."
Elliott Hughes [Tue, 19 Jan 2016 23:46:05 +0000 (15:46 -0800)]
Make FILE*s less usable after fclose(3).
BSD doesn't invalidate the fd stored in struct FILE, which can make
it possible (via fileno(3), for example), to perform operations on
an fd you didn't intend to (rather than just failing with EBADF).
Fixing this makes the code slightly simpler anyway, and might help
catch bad code before it ships.
Bug: http://stackoverflow.com/questions/
10816837/fclose-works-differently-on-android-and-linux
Change-Id: I9db74584038229499197a2695c70b58ed0372a87
Colin Cross [Tue, 19 Jan 2016 20:50:20 +0000 (12:50 -0800)]
Update Android.bp to match Android.mk
Change-Id: Ib6ab28c5b4569c9537c4647b3661166fdb82544a
Chih-hung Hsieh [Tue, 19 Jan 2016 20:04:47 +0000 (20:04 +0000)]
Merge "Switch libc default to clang."
Chih-Hung Hsieh [Tue, 19 Jan 2016 18:59:34 +0000 (10:59 -0800)]
Switch libc default to clang.
* Default to clang when USE_CLANG_PLATFORM_BUILD is not set
and the target has no clang bug.
BUG:
26102335
Change-Id: Ied6c9dc5593bfbadbb8d8b38e66ea237d649bae5
Elliott Hughes [Tue, 19 Jan 2016 17:48:28 +0000 (17:48 +0000)]
Merge "Fix GCC build breakage."
Elliott Hughes [Tue, 19 Jan 2016 17:47:47 +0000 (09:47 -0800)]
Fix GCC build breakage.
Change-Id: I0339f525c68af942fb97f4bffc771717e67e0e07
Elliott Hughes [Tue, 19 Jan 2016 17:25:01 +0000 (17:25 +0000)]
Merge "Add more fortify symbols to the expected differences from glibc."
Elliott Hughes [Tue, 19 Jan 2016 17:24:16 +0000 (17:24 +0000)]
Merge "Implement if_nameindex(3)/if_freenameindex(3)."
Elliott Hughes [Sat, 16 Jan 2016 05:02:56 +0000 (21:02 -0800)]
Implement if_nameindex(3)/if_freenameindex(3).
This is just a subset of the recently-implemented getifaddrs(3), though if
we want to handle interfaces (such as "rmnet_*") that don't have an address,
we need to either expose ifaddrs_storage and keep track of which interfaces
we've already seen (which is pretty messy), or refactor the netlink code so
we can reuse it and just extract the information we need for if_nameindex(3).
This patch goes the latter route.
Also clean up if_nametoindex(3) and if_indextoname(3).
Change-Id: I5ffc5df0bab62286cdda2e7af06f032c767119a8
Yi Kong [Sun, 17 Jan 2016 22:48:36 +0000 (22:48 +0000)]
Merge "Add more tests for getifaddrs(3)"
Elliott Hughes [Sat, 16 Jan 2016 03:54:31 +0000 (19:54 -0800)]
Add more fortify symbols to the expected differences from glibc.
Change-Id: I6266b2e6cbea3c014b4ea30a7100186965e40cf1
Elliott Hughes [Sat, 16 Jan 2016 03:02:06 +0000 (03:02 +0000)]
Merge "Sync with upstream OpenBSD."
Dan Willemsen [Sat, 16 Jan 2016 00:47:39 +0000 (00:47 +0000)]
Merge "Align Soong builds closer to Make"
Elliott Hughes [Sat, 16 Jan 2016 00:30:18 +0000 (16:30 -0800)]
Sync with upstream OpenBSD.
Also some minor build cleanup.
Change-Id: Ibd20602d7ea45685f0c47fceb00ffd6c7ae35954
Dan Willemsen [Sat, 16 Jan 2016 00:00:57 +0000 (16:00 -0800)]
Align Soong builds closer to Make
This moves the Android.bp file to use wildcards in the same places that
they're using in the Android.mk file. It also fixes a file that was in a
different order.
Make and Soong still produce binaries with object files in different
orders, but that's due to Make reordering the object files based on
compile type. Soong keeps the original specified order. It's not
possible to emulate the reordered files in the Android.bp, since the
arch-specific files won't interleave properly.
Change-Id: I9052b3ed7c523c13df5cbe606f913c32c88c7c5f
Christopher Ferris [Thu, 14 Jan 2016 20:04:29 +0000 (20:04 +0000)]
Merge "Add comment for [p]select_smoke tests STDIN issue."
Christopher Ferris [Thu, 14 Jan 2016 19:12:38 +0000 (11:12 -0800)]
Add comment for [p]select_smoke tests STDIN issue.
Change-Id: I59cde8b5aaac3e27419ca86d16f85e5af568acf9
Yi Kong [Wed, 13 Jan 2016 11:28:14 +0000 (11:28 +0000)]
Add more tests for getifaddrs(3)
This adds the following two checks:
* getifaddrs sees the same list of interfaces as /sys/class/net.
* IPv4 addresses we get from netdevice(7) agrees with results from
getifaddrs.
Change-Id: I2f6d79d0b5cde6d98a0f671d1623b6b2bc75b60f
Dan Willemsen [Thu, 14 Jan 2016 08:16:02 +0000 (08:16 +0000)]
Merge "Fix Soong mips builds"
Dan Willemsen [Thu, 14 Jan 2016 07:22:22 +0000 (23:22 -0800)]
Fix Soong mips builds
Change-Id: I6ae8208e077d831969b08e7abd232deaa20ceca1
Christopher Ferris [Thu, 14 Jan 2016 07:04:53 +0000 (07:04 +0000)]
Merge "Fix select failures when STDIN is ready."
Christopher Ferris [Thu, 14 Jan 2016 06:38:09 +0000 (22:38 -0800)]
Fix select failures when STDIN is ready.
The select_smoke and pselect_smoke test can fail if STDIN has data ready
to be read. The easiest way to see the failure is to type on the command
line while running the tests.
To avoid this, allow the return value to be 2 or 3 and check which fds
are ready to be read.
Change-Id: Iafba332c5f3ed1943e3d34501f123dd45f06a8c4
Dan Albert [Thu, 14 Jan 2016 06:12:32 +0000 (06:12 +0000)]
Merge "Escape failure messages in XML test output."
Dan Albert [Thu, 14 Jan 2016 05:48:56 +0000 (21:48 -0800)]
Escape failure messages in XML test output.
The gtest XML format requires escaped HTML characters in the test
results.
Change-Id: Ieb9519a55cb52093dfb10a88e883b569bc372cdb
Christopher Ferris [Thu, 14 Jan 2016 05:16:33 +0000 (05:16 +0000)]
Merge "Add method to run fortified versions of funcs."
Christopher Ferris [Wed, 13 Jan 2016 21:47:58 +0000 (13:47 -0800)]
Add method to run fortified versions of funcs.
Add a way to turn fortify off for the files that test fortify functions.
This method involves simply compiling the same file with fortify off and
changing the test name slightly.
It's not very pretty, and it assumes that only these few files test
functions that can be fortified.
Bug:
15195631
Change-Id: Iba9db1d508b7d28a1d6968019cb70fe08864827b
Dan Willemsen [Wed, 13 Jan 2016 21:45:14 +0000 (21:45 +0000)]
Merge "Fix Soong x86 builds"
Dan Willemsen [Wed, 13 Jan 2016 21:36:11 +0000 (13:36 -0800)]
Fix Soong x86 builds
Change-Id: Ida3f3584f832d61caf7200557ea57be799eaf38e
Dimitry Ivanov [Wed, 13 Jan 2016 20:35:25 +0000 (20:35 +0000)]
Merge "Revert "Move __system_property* to LIBC_PRIVATE for lp64 libc""
Dimitry Ivanov [Wed, 13 Jan 2016 20:32:12 +0000 (20:32 +0000)]
Revert "Move __system_property* to LIBC_PRIVATE for lp64 libc"
This reverts commit
ff14fb5a1f809d87125fa0d14f08d486b6bf1ea1.
Bug: http://b/
26497521
Change-Id: Id7131b37a31e936db0e72ef9969e1f13d2dd578f
Colin Cross [Wed, 13 Jan 2016 20:32:09 +0000 (20:32 +0000)]
Merge "Rearrange build files to support USE_SOONG=true"
Dan Willemsen [Wed, 13 Jan 2016 20:32:07 +0000 (20:32 +0000)]
Merge "Fix Soong x86 atom builds"
Colin Cross [Wed, 13 Jan 2016 20:31:39 +0000 (20:31 +0000)]
Merge "Update Android.bp for changes to Android.mk"
Dan Willemsen [Wed, 13 Jan 2016 03:35:40 +0000 (19:35 -0800)]
Fix Soong x86 atom builds
Change-Id: Ie92e605c499f128335a8da7cfefa7e6ae367a223
Elliott Hughes [Tue, 12 Jan 2016 23:16:00 +0000 (23:16 +0000)]
Merge "Fix <sys/sysmacros.h>."
Elliott Hughes [Tue, 12 Jan 2016 22:43:27 +0000 (14:43 -0800)]
Fix <sys/sysmacros.h>.
1. The definitions were wrong.
2. The definitions were inline functions.
3. The definitions were polluting the namespace even for code that doesn't
want BSD cruft.
Note that everybody will still get these by default, because you still get
all the BSD stuff by default.
Bug: http://b/
12706131
Change-Id: I062ecd09feef7a6e8ba1922d465b96a9c4bf4f4e
Colin Cross [Mon, 11 Jan 2016 20:28:17 +0000 (12:28 -0800)]
Rearrange build files to support USE_SOONG=true
Building with USE_SOONG=true will build with an Android.bp file if it is
present in the directory, otherwise an Android.mk file. Only a few of
the bionic directories compile with soong, so include all of them from
the top level Android.mk file and remove the top level Android.bp file.
Individual subdirectories with Android.bp files will use soong with
USE_SOONG=true.
Change-Id: Idf8d7977ea4668fa646be25b543bf9d3773de615
Elliott Hughes [Tue, 12 Jan 2016 20:57:48 +0000 (20:57 +0000)]
Merge "Add a getifaddrs(3) debugging tool."
Chih-hung Hsieh [Tue, 12 Jan 2016 19:41:46 +0000 (19:41 +0000)]
Merge "Explicit binary op order for both gcc and clang/llvm."
Chih-Hung Hsieh [Tue, 12 Jan 2016 17:48:09 +0000 (09:48 -0800)]
Explicit binary op order for both gcc and clang/llvm.
* gcc assembler parses (1 << n - 1) as ((1 << n) - 1)
* llvm assembler parses (1 << n - 1) as (1 << (n - 1))
BUG:
26289334
Change-Id: Iaaa3d07f59c0f87cb1b80b9a1c4e765b66846791
Elliott Hughes [Tue, 12 Jan 2016 05:30:07 +0000 (21:30 -0800)]
Add a getifaddrs(3) debugging tool.
...disguised as a test, because that's the easiest way to keep an eye on it.
Change-Id: Iec2531183790d12aec561d27645e5d400f69efbf
Colin Cross [Mon, 11 Jan 2016 21:20:55 +0000 (13:20 -0800)]
Update Android.bp for changes to Android.mk
Change-Id: Ie01e326e2776c699237b156155610d08801c0cd4
Elliott Hughes [Mon, 11 Jan 2016 21:43:58 +0000 (21:43 +0000)]
Merge "Add basic tests for <net/if.h>."
Elliott Hughes [Mon, 11 Jan 2016 21:18:20 +0000 (13:18 -0800)]
Add basic tests for <net/if.h>.
Bug: http://b/
26447711
Change-Id: I9bbcdffc07e99c98e367a63de861ddc7b274d579
Elliott Hughes [Mon, 11 Jan 2016 21:16:42 +0000 (21:16 +0000)]
Merge "Handle AF_PACKET in getifaddr(3)."
Yi Kong [Tue, 22 Dec 2015 17:07:23 +0000 (17:07 +0000)]
Handle AF_PACKET in getifaddr(3).
Also fix a bug where we were mutating the address/broadcast address
of an existing entry rather than the new entry, and use 'const' to
ensure we don't make that mistake again.
Change-Id: I31c127a5d21879b52c85cd0f7ed2e66554a21e39
Elliott Hughes [Fri, 8 Jan 2016 18:54:01 +0000 (18:54 +0000)]
Merge "Defend against -fstack-protector in libc startup."
Dimitry Ivanov [Fri, 8 Jan 2016 18:50:13 +0000 (18:50 +0000)]
Merge "Move __aeabi* symbols to LIBC_N version"
Dimitry Ivanov [Fri, 8 Jan 2016 18:06:44 +0000 (10:06 -0800)]
Move __aeabi* symbols to LIBC_N version
We had these symbols incorrectly versioned as LIBC_PRIVATE
in M release. This change moves __aeabi* symbols from LIBC
to LIBC_N and adds __gnu_Unwind_Find_exidx to the list
Bug: https://github.com/android-ndk/ndk/issues/1
Change-Id: I0b353012adeacb00ae29ea10c63b9d1cf1cadbe7
Elliott Hughes [Thu, 7 Jan 2016 03:51:43 +0000 (19:51 -0800)]
Defend against -fstack-protector in libc startup.
Exactly which functions get a stack protector is up to the compiler, so
let's separate the code that sets up the environment stack protection
requires and explicitly build it with -fno-stack-protector.
Bug: http://b/
26276517
Change-Id: I8719e23ead1f1e81715c32c1335da868f68369b5
Josh Gao [Wed, 6 Jan 2016 22:49:30 +0000 (22:49 +0000)]
Merge "Kill BOARD_MALLOC_ALIGNMENT."
Josh Gao [Wed, 6 Jan 2016 22:06:55 +0000 (14:06 -0800)]
Kill BOARD_MALLOC_ALIGNMENT.
Bug: http://b/
26403338
Change-Id: Iba4006277fdb7c7ca60542c517634ae86c847b4f
Josh Gao [Wed, 6 Jan 2016 20:08:13 +0000 (20:08 +0000)]
Merge "Replace MALLOC_IMPL with MALLOC_SVELTE."
Elliott Hughes [Wed, 6 Jan 2016 19:52:43 +0000 (19:52 +0000)]
Merge "Temporary hack to avoid stack protector crashes."
Tom Cherry [Wed, 6 Jan 2016 19:16:31 +0000 (19:16 +0000)]
Merge "Revert "system_properties.cpp: special case ro.* properties""
Tom Cherry [Wed, 6 Jan 2016 18:52:52 +0000 (18:52 +0000)]
Revert "system_properties.cpp: special case ro.* properties"
This reverts commit
c5fd81ab2524a06be907d9c5234e79346bbbbd7c.
Bug:
26416032
Change-Id: Id2d6761fdf55efa28c0b08b597daaa5cd381d758