- 20 Dec, 2007 24 commits
-
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Alexandre Julliard authored
Make the stubs return success.
-
Roy Shea authored
-
Alexandre Julliard authored
This allows keeping the default SIGTERM behavior of killing the whole process.
-
Alexandre Julliard authored
-
Dmitry Timoshkov authored
-
James Hawkins authored
-
Rob Shearman authored
ole32: Fix RPC_GetLocalClassObject to wait for 30 seconds, even if messages arrive while we are waiting.
-
Rob Shearman authored
safe_copy_from_buffer has the side-effect of incrementing the buffer and this still needs to be done when we point the memory into the buffer. (Thanks to Maarten Lankhorst for finding the mistake and suggesting a fix.)
-
Stefan Dösinger authored
-
Stefan Dösinger authored
-
Stefan Dösinger authored
-
Stefan Dösinger authored
-
Stefan Dösinger authored
-
Stefan Dösinger authored
-
Dmitry Timoshkov authored
-
Dmitry Timoshkov authored
-
James Hawkins authored
-
James Hawkins authored
-
Andrew Riedi authored
-
Rico Schüller authored
-
Andrew Talbot authored
-
- 19 Dec, 2007 16 commits
-
-
Alex Villacís Lasso authored
-
Rob Shearman authored
It is usually used during marshalling, where pStubMsg->BufferStart and pStubMsg->BufferEnd won't be valid. Replace it with a check using RpcMsg->Buffer and pStubMsg->BufferLength.
-
Rob Shearman authored
aligns contains padding, but the memory is marshalled in one block so call memset to avoid Valgrind warnings. padded and padded2 are marshalled as complex types so this is not required here. Initialise the dummy member of test_list_t to zero for the TL_NULL case. Change the type to the smallest available to not waste buffer space.
-
Rob Shearman authored
Also clear memory when FC_STRUCTPAD* is encountered.
-
Rob Shearman authored
-
Rob Shearman authored
Only do the clearing when marshalling a base type.
-
Rob Shearman authored
widl: The detection of types is highly dependent on the ordering of the various type detection functions. Therefore, needs_freeing would have to duplicate a lot of write_remoting_arg in order to get the detection right. Because of this, it is easier and will cause less problems in the future to simply move the logic of needs_freeing into write_remoting_arg in the appropriate handling code for the detected type.
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Alexandre Julliard authored
-
Roy Shea authored
-
Dmitry Timoshkov authored
-
Dmitry Timoshkov authored
-
Dmitry Timoshkov authored
-