Skip to content
Commits on Feb 8, 2016
  1. @migueldeicaza
  2. @migueldeicaza
  3. @lambdageek

    Merge pull request #2555 from lambdageek/monoerror-mono_runtime_invoke-2

    lambdageek committed
    Use MonoError for mono_runtime_invoke
  4. @akoeplinger

    [mcs] Use $(PLATFORM_PATH_SEPARATOR) instead of ':' for MONO_PATH to …

    akoeplinger committed
    …fix Windows tests
    
    ./ was added to MONO_PATH in b5c2b58, but the Unix path separator ':' doesn't work on Windows.
  5. @akoeplinger
  6. @lambdageek

    Merge pull request #2571 from lambdageek/monoerror-mono_assembly_get_…

    lambdageek committed
    …object
    
    Use MonoError in mono_assembly_get_object
  7. @akoeplinger

    [System.Runtime.Caching] Make TestExpiredGetValues test more reliable

    akoeplinger committed
    It failed twice on Jenkins over the weekend:
    
    ```
    Test Case Failures:
    1) MonoTests.System.Runtime.Caching.MemoryCacheTest.TestExpiredGetValues :   #EGV4
      Expected: 0
      But was:  9
    
    at MonoTests.System.Runtime.Caching.MemoryCacheTest.TestExpiredGetValues () [0x0012b] in /media/ssd/jenkins/workspace/test-mono-mainline/label/debian-armel/mcs/class/System.Runtime.Caching/Test/System.Runtime.Caching/MemoryCacheTest.cs:1304
    ```
    
    We're sleeping just as long as the expiry time of the cache items (1s) which might not be enough.
    Bumping the sleep time should fix this.
  8. @akoeplinger

    [readme] Clarify --enable-parallel-mark description

    akoeplinger committed
    It's Boehm only and not enabled by default on some platforms (e.g. it's disabled on OSX).
  9. @lambdageek

    [runtime] Switch to mono_runtime_invoke_checked and mono_runtime_try_…

    lambdageek committed
    …invoke
    
    Mark mono_runtime_invoke external only.
    Runtime should use mono_runtime_invoke_checked and mono_runtime_try_invoke.
  10. @ludovic-henry @lambdageek
  11. @lambdageek

    [reflection] Use MonoError in mono_assembly_get_object

    lambdageek committed
    Mark mono_assembly_get_object external only.
    Runtime should use mono_assembly_get_object_checked.
    
    Also propagated MonoError through a few static functions in reflection:
    mono_reflection_get_type_internal () and
    mono_reflection_get_type_internal_dynamic ()
  12. @kumpera

    Merge pull request #2574 from abrasive/gcc5_fix_gc

    kumpera committed
    libgc: use __builtin_frame_address in GC_approx_sp
  13. @kumpera

    Merge pull request #2575 from alexrp/master

    kumpera committed
    [profiler] Fix a rare crash when sending SIGPROF to a starting thread.
  14. @alexrp

    [profiler] Fix a rare crash when sending SIGPROF to a starting thread.

    alexrp committed
    When sample profiling is enabled and we send a SIGPROF signal to a thread that
    hasn't yet transitioned to a running state (e.g. STATE_RUNNING), the signal can
    sometimes be delivered at such an early point that libc has not even finished
    initializing TLS for that thread yet. This is problematic because we then go
    right on to crash in the SIGPROF signal handler when we try to save errno. We
    then further crash in the SIGSEGV handler because we can't access the jit_tls
    variable either.
    
    The fix is fairly simple: Only send SIGPROF to threads that have transitioned
    to a running state. Technically, TLS is initialized 'much' earlier than when
    this state transition happens, but since the worst thing that can happen is
    that we miss one sampling event (i.e. ~1sec by default) right at the beginning
    of the thread's lifetime, this simple fix is good enough.
    
    This crash seemingly only happened on OS X.
  15. @vargaz
  16. @vargaz
Commits on Feb 7, 2016
  1. libgc: use __builtin_frame_address in GC_approx_sp

    James Laird-Wah committed
    With GCC 5.x, the compiler optimises away the dummy load used to
    estimate the stack pointer address in GC_approx_sp; it returns
    zero. This leads to segfaults when embedding Mono.
    
    Instead, use __builtin_frame_address, which GCC's libgc uses for
    the same purpose.
  2. @vargaz
  3. @vargaz

    [llvm] When storing valuetypes received in registers to the stack, st…

    vargaz committed
    …ore only the value parts of the register, the upper bits might container garbage.
Commits on Feb 6, 2016
  1. @lambdageek

    Merge pull request #2567 from lambdageek/monoerror-mono_module_file_g…

    lambdageek committed
    …et_object
    
    [reflection] Use MonoError in mono_module_file_get_object
  2. @vargaz

    [jit] Disable to returning of trampolines from mini_jit_info_table_fi…

    vargaz committed
    …nd () to see whenever that fixes failures on wrench.
  3. @vargaz

    [llvmonly] Remove the mono_llvmonly_get_calling_assembly () workaroun…

    vargaz committed
    …d since stack walks work in llvmonly mode now.
Commits on Feb 5, 2016
  1. @lambdageek

    [reflection] Use MonoError in mono_module_file_get_object

    lambdageek committed
    Mark mono_module_file_get_object external only.
    Use mono_module_file_get_object_checked in runtime code.
  2. @vargaz

    Fix the build.

    vargaz committed
  3. @xmcclure

    Merge pull request #2569 from xmcclure/gshared-enable

    xmcclure committed
    Miscellaneous minor improvements around amd64 aot
  4. @tritao
  5. @rolfbjarne
  6. @akoeplinger

    Merge pull request #2572 from akoeplinger/fix-monodoc-language

    akoeplinger committed
    [monodoc] Fix literal formatting on systems where decimal separator is not a dot
  7. @vargaz

    [llvmonly] Various small fixes. Make mono_llvm_throw_exception etc. d…

    vargaz committed
    …irectly callable in static mode. Add support for directly calling pinvoke methods. Fix an assertion during EH.
  8. @migueldeicaza

    [Docs] More API docs

    migueldeicaza committed
  9. @vargaz

    [llvmonly] Don't set the invariant.load flag on loads from the GOT, t…

    vargaz committed
    …hey are initialized by a call to init_method () so they are not completely invariant.
  10. @vargaz

    Revert "[jit] Don't allocate memory for unused vreg"

    vargaz committed
    This reverts commit 137c099.
    
    Revert this as leads to a memory overwrite on x86:
    
    			defs [ins->dreg + 1] = NULL;
Commits on Feb 4, 2016
  1. @xmcclure
  2. @xmcclure
  3. @lambdageek

    Merge pull request #2566 from lambdageek/monoerror-mono_module_get_ob…

    lambdageek committed
    …ject
    
    [reflection] Use MonoError for mono_module_get_object
Something went wrong with that request. Please try again.