summaryrefslogtreecommitdiff
path: root/bridges/source
AgeCommit message (Collapse)Author
2012-04-25WaE: Clang unused variableCatalin Iacob
2012-04-23Port gcc3_linux_powerpc fix to gcc3_macosx_powerpcStephan Bergmann
This ports 84dbc4fe2547f8fc341a46d7f000e721c81e63ee "do not let gcc use registers we are setting ourselves," in the hope that it is also an improvement for Mac OS X PPC. (But the patch was applied blindly, so please revert if it makes things worse instead.)
2012-04-23do not let gcc use registers we are setting ourselvesDavid Tardon
gcc uses a register for the function call--and it tried r9 here...
2012-04-23save register arguments firstDavid Tardon
The "mr" stores r11 to indeterminate register (r9 in my case) before storing it into the output variable, thus overwriting one of our input arguments... This later leads to nice segfault in testtools/source/bridgetest ... I suppose there is a better way to get the variable, but I do not know it .-)
2012-04-22WaE: variable 'stackptr' is uninitialized when usedTor Lillqvist
2012-04-15Fix crasher and generate PIC for iOS simulatorTor Lillqvist
Had an embarrassing thinko in the assembler generation for the iOS simulator, had forgotten the $ for a literal number. While at it, make also the simulator read-only code snippet data structures position-independent. Some minor comment improvements. Bridge seems to work now on simulator.
2012-04-06Replaced equalsAsciiL(RTL_CONSTASCII_STRINGPARAM(...)) with == operatorSzabolcs Dezsi
2012-04-04adjust gcc3_linux_intel bridge to work with clangLuboš Luňák
Otherwise exception throwing seems to be broken. This is a partial revert of b0515ea5fa6c29faebed616ae3e0213c72d24904.
2012-04-04Fixed include guard copy/paste errorStephan Bergmann
2012-04-04Adapt cpp_uno/gcc3_linux_intel to GCC 4.7Stephan Bergmann
...the same way cpp_uno/gcc3_linux_x86-64 was already adapted.
2012-03-30Hack around different __cxa_* declarations in different GCC versionsStephan Bergmann
2012-03-30Missing includesStephan Bergmann
2012-03-30Adapt cpp_uno/gcc3_linux_x86-64 to GCC 4.7 cxxabi.hStephan Bergmann
...the same way cpp_uno/mingw_intel was already adapted.
2012-03-30MinGW: callVirtualMethod needs same treatment as on gcc3_linux_x86_64Stephan Bergmann
...for reasons that are yet unclear to me.
2012-03-30iOS simulator fix and some cosmeticsTor Lillqvist
2012-03-29alloca(3) needs stdlib.h on *BSD because there is no alloca.hRobert Nagy
2012-03-29Mark all registered as clobbered that are not saved across callStephan Bergmann
2012-03-29Make cpp_uno/gcc3_linux_x86-64 bridge work with GCC 4.7Stephan Bergmann
See comment in callvirtualmethod.cxx for details.
2012-03-272nd param of __cxa_throw is void* in GCC <= 4.6Stephan Bergmann
2012-03-27MinGW: Adapted cpp_uno bridge to GCC 4.7Stephan Bergmann
...which incompatibly changed the calling convention of the third (dest) function pointer parameter of __cxa_throw to __thiscall.
2012-03-26Revert "Field IDs on Dalvik are more unique than in JVMs"Stephan Bergmann
This reverts commit b5ca34fa409bc4cbeb41a84b900612e302f5b0fb. 829a9b9c1041626efb688157658c1e0e2f6e8414 "In Java, css.uno.RuntimeException is not derived from css.uno.Exception" is a better fix.
2012-03-26In Java, css.uno.RuntimeException is not derived from css.uno.ExceptionStephan Bergmann
which lead to Bridge::map_to_java (jni_data.cxx) copying the Context member of RuntimeException twice, once with a fieldID for Exception, which fails on Dalvik.
2012-03-26Field IDs on Dalvik are more unique than in JVMsTor Lillqvist
2012-03-21Uniquify uno_initEnvironment and uno_ext_getMapping in the static linking caseTor Lillqvist
2012-03-16solaris/gcc: add missing standard includesJonathan Adams
2012-03-12WaE: 'exec' may be used uninitialized in this functionDavid Tardon
2012-03-07Fix compilation for simulatorTor Lillqvist
2012-03-05Make the iOS ARM code position-independentTor Lillqvist
2012-03-05Make the iOS device (ARM) stuff at least link, surely does not work yetTor Lillqvist
2012-02-21ARM bridge: fixes for passing double arguments on the stackJani Monoses
2012-02-17fix armel/armv4t buildRene Engelhard
2012-02-16ARM bridge: VFP ABI (armhf) supportJani Monoses
2012-02-15Remove unnecessary variableJani Monoses
2012-02-15Simplify code and use proper register names for linux armhfJani Monoses
2012-02-05Some cppcheck cleaningJulien Nabet
2012-01-26Code cleanup: ( () ) replaced by (())Alexander Bergmann
2012-01-21Removed some unused parameters; added SAL_UNUSED_PARAMETER.Stephan Bergmann
SAL_UNUSED_PARAMETER (expanding to __attribute__ ((unused)) for GCC) is used to annotate legitimately unused parameters, so that static analysis tools can tell legitimately unused parameters from truly unnecessary ones. To that end, some patches for external modules are also added, that are only applied when compiling with GCC and add necessary __attribute__ ((unused)) in headers.
2012-01-13Use cacheflush() on AndroidTor Lillqvist
2012-01-05WaE and build errors with OSL_DEBUG_LEVEL == 2Caolán McNamara
2011-12-27function component_canUnload is exportedDavid Tardon
2011-12-21only use posix_fallocate() if it's actually availableRobert Nagy
2011-12-19fix armhf build in bridgesMatthias Klose
Signed-off-by: Bjoern Michaelsen <bjoern.michaelsen@canonical.com> Signed-off-by: Caolán McNamara <caolanm@redhat.com>
2011-12-15Resolves: rhbz#767708 avoid SIGBUS writing to overcommitted mmaped diskspaceStephan Bergmann
2011-12-13Fix for fdo43460 Part IV getLength to isEmptyOlivier Hallot
Part IV Module basic (small fix per demand from Ivan Timofeev) binaryurp bridges
2011-12-11I don't see NO_BSYMBOLIC being used anywhereTor Lillqvist
2011-11-30catch by constant referenceTakeshi Abe
2011-11-27remove include of pch header in bridgeNorbert Thiebaud
2011-11-24Use RTLD_DEFAULT on AndroidTor Lillqvist
dlopen()ing NULL (the main program) and dlsym()ing on that handle does not work on Android as in GNU/Linux. So don't bother with that, just call dlsym(RTLD_DEFAULT, symbol) instead. But, it doesn't help, it won't find the type_infos anyway, as they are weak symbols and the dlsym() at least in Android 4.0 does not like weak symbols, "Symbol is not global." So the fallback branch that dynamically generates a type_info, which normally should get used only for inter-process UNO (and thus presumably never on Android) is taken instead. But, that didn't work either at least with NDK r7 thanks to a bug in its libgnustl_shared.so. But we now patch that at run-time in android_main() in sal/osl/android/jni/lo-bootstrap.c. Are we having fun yet?
2011-11-21Don't use USE_DOUBLE_MMAP on AndroidTor Lillqvist
2011-11-17Let's try using the gcc3_linux_arm code for Android, tooTor Lillqvist
generated by cgit (git 2.34.1) at 2025-03-05 02:07:00 +0000