- 16 Nov, 2015 2 commits
- 13 Nov, 2015 4 commits
-
-
Josh Gao authored
Change-Id: I84a4a3500f2a9a386e1ddb11405448644b2bad2c
-
Josh Gao authored
Bug: http://b/25648674 Change-Id: Iae021a611ad77cf721d36213d6b276a2023042b1
-
Josh Gao authored
Change-Id: I7213f3b47d12448fe1eeacf0743ee58d52fbfdf7
-
Josh Gao authored
Change-Id: I95bc77792ca6f2a60827aa36758709115da4c7df
-
- 11 Nov, 2015 1 commit
-
-
Josh Gao authored
Change-Id: Ibbc7f4ea185b60f4521254dc6b99130bc16d964f
-
- 16 Sep, 2015 5 commits
-
-
Josh Gao authored
os.normpath converts slashes to backslashes on windows, which gdb does not like. Change-Id: Iaba0ab1abda6a34d07f67b952bd259717decce9d
-
Josh Gao authored
Bug: http://b/23791565 Change-Id: I5312a1813bb7c9cbb601f7fc2bd039c25b6cdc4f
-
Josh Gao authored
Bug: http://b/23787206 Change-Id: I589dac158485cc3605586ec8f9598de8d5d29ba4
-
Josh Gao authored
Previously, ctrl-c would send SIGINT to all of the children spawned by ndk-gdb.py, which would result in everything immediately blowing up because gdbserver would get killed. Bug: http://b/23906844 Change-Id: Ic4b2a111759a95d6d82b6ffb185dd114b8ec0df0
-
Josh Gao authored
Remove another adb shell roundtrip. Bug: http://b/23793246 Change-Id: Idf1dccddb8c5ed8cb1327fadfffbfbf0822232e2
-
- 09 Sep, 2015 2 commits
-
-
Josh Gao authored
Change-Id: Ic7fec5ba367fcce939dde585826e274fcbd12d41
-
Josh Gao authored
Saves an adb roundtrip. Bug: http://b/23793246 Change-Id: Iffce325f7576102cdbf8d2ca7530c77c049b1503
-
- 08 Sep, 2015 1 commit
-
-
Josh Gao authored
Bug: http://b/23899431 Change-Id: I9ff4b38ae5c4c03467ecc61aebda5fdcc9ae8608
-
- 04 Sep, 2015 5 commits
-
-
Josh Gao authored
Change-Id: I4be46cc93159b28faf78d8231b24d918fa330788
-
Josh Gao authored
Saves about 25ms on my machine. Bug: http://b/23793246 Change-Id: I89484477c200825eeee1d2c9c3f3e1df3c225222
-
Josh Gao authored
Change-Id: I4d79ca1a314c8f8cb454be4bd2a58882b96e9126
-
Josh Gao authored
Change-Id: I693ffcbc66d74bcabea643e0ebc8a4d26a169414
-
Josh Gao authored
Bug: http://b/23790835 Change-Id: I5c4d145a99a30b48e78253d55cc2f58bec63d670
-
- 02 Sep, 2015 5 commits
-
-
Josh Gao authored
Change-Id: Ia2d616f81060241bc723784226b2ee2b073c3f03
-
Josh Gao authored
Bug: http://b/22128526 Change-Id: I64a2babfa1f3db91af742e02e73b9abae9a38573
-
Josh Gao authored
Change-Id: I8e800ea28a3a90493a19e4bf05d1125737d25c3b
-
Josh Gao authored
Change-Id: I8e9b5add36b6ad13ef1da3815b6323b1d2414deb
-
Josh Gao authored
Change-Id: Ia9c02a569ff4196674838bfd07d011ed7904aca0
-
- 22 Jan, 2015 1 commit
-
-
Alexander Ivchenko authored
1) Always try to use app_process32/app_process64 to avoid pulling symlinks (this does not work on win hosts). 2) Don't change the name of linker64 when pulling it from the device (otherwise symbols from linker64 are not being loaded in 64-bit case). 3) Always pull app_process32/app_process64 as app_process to avoid inconsistency. Change-Id: I1e4f066b77ee2fc37b879c1a1924b8396921539c Signed-off-by:
Alexander Ivchenko <alexander.ivchenko@intel.com>
-
- 19 Jan, 2015 1 commit
-
-
Alexander Ivchenko authored
For details, please, refer to f91d44c1 Signed-off-by:
Alexander Ivchenko <alexander.ivchenko@intel.com> Change-Id: Ic9af37f43adf4684e81555ba94692e9a084ff507
-
- 06 Jan, 2014 1 commit
-
-
Todd Fiala authored
This change fixes a list comprehension call that generates an exception when the PYPRPR_GNUSTDCXX_BASE directory does not exist. Change-Id: Icb5c9258b9960bc09cda167137f585225ba239db
-
- 10 Aug, 2013 2 commits
-
-
Ray Donnelly authored
It wasn't correct on 64bit. This meant that the gnumake found on those systems tended to be the system one. Now, sys.executable is used and if that fails, some more exhaustive checks of platform.platform() and sys.maxsize are used (in-case the user uses their own Python - note if this happens, arch differences between that Python and the NDK could still cause the system make to be used, but this is a small corner case that will usually work ok ..)
-
Ray Donnelly authored
The quotes in os.environ['PATH'] need to be removed otherwise executables are not found.
-
- 28 Jul, 2013 1 commit
-
-
Ray Donnelly authored
Made the description format use RawTextHelpFormatter to prevent line breaks in the filename on Windows. Also replaced \ with / in NDK so it looks more tidy.
-
- 14 Jun, 2013 1 commit
-
-
Ray Donnelly authored
This is the last commit needed to enable Python pretty-printers in ndk-gdb.py Valid options are 'auto', 'none', 'stlport', 'gnustdcxx' and 'gnustdcxx-$VER' where $VER is a foldername in: $NDK/prebuilt/$SYSTEM/share/pretty-printers/libstdcxx with 'gcc-' removed. 'auto' mode works by looking at the .so files for a libstdc++ library, and as such only works in the shared scenario. When linking statically to a libstdc++ library, the required printers must be specified. The default value of this option is 'none', if people feel 'auto' is preferable then I'll change it. Change-Id: Icdebd4f787387b99eaa60e0cf48058e9b8c484d4
-
- 05 Jun, 2013 1 commit
-
-
Ray Donnelly authored
Prevents early breakpoints from being missed. Original bug report from: https://code.google.com/p/android/issues/detail?id=41278 Original shell script version from: https://android-review.googlesource.com/#/c/48029/ Change-Id: I62809e73ee06ce962f4de3fc8ddc652744384c5b
-
- 04 Jun, 2013 1 commit
-
-
Ray Donnelly authored
-
- 09 Apr, 2013 1 commit
-
-
Andrew Hsieh authored
Instead of in the order of APP_ABIS, which will miss armeabi-v7a if APP_ABI := armeabi armeabi-v7a (note that armeabi appears first and get matched) ndk-gdb.py looks in the order of CPU_ABIS already. Fixed two bugs (unrelated nonetheless). See https://code.google.com/p/android/issues/detail?id=54033 Change-Id: I4c97e90a552a16598295e0b1597882f7d3c939b8
-
- 08 Jan, 2013 1 commit
-
-
Ray Donnelly authored
Pythonic replacements for the ndk-gdb shell script. Attempts to emulate the old behaviour, including verbose mode and logging and also the general layout of the shell script. Added a new command line option, --tui/-t to launch gdb in text user interface mode, if the gdb is built with support for this. Change-Id: Ic02feb18d70fe449bf54a531ffb973513dc82783
-