| Commit message (Expand) | Author | Age | Files | Lines |
* | Make tracing state exception safe.object_ownership | Jason Madden | 2021-11-02 | 2 | -32/+77 |
* | Cleanups for non-native thread state creation. | Jason Madden | 2021-11-02 | 3 | -25/+25 |
* | Update comment. We're leaking less than 1.1.2 | Jason Madden | 2021-11-02 | 2 | -246/+51 |
* | Always test C++. | Jason Madden | 2021-11-02 | 5 | -42/+39 |
* | Clean up argument passing. | Jason Madden | 2021-11-02 | 3 | -155/+239 |
* | Cleanup debugging code. | Jason Madden | 2021-10-30 | 7 | -249/+215 |
* | Remove set_terminate | Jason Madden | 2021-10-29 | 1 | -3/+0 |
* | Neither disabling faulthandler nor enabling optimizations changed the SEH depth. | Jason Madden | 2021-10-29 | 6 | -44/+37 |
* | Nope, both are needed. So what about if faulthandler is off and we optimize (... | Jason Madden | 2021-10-29 | 5 | -24/+22 |
* | That worked, but possibly by accident: We might have grabbed the SEH handler ... | Jason Madden | 2021-10-29 | 1 | -1/+6 |
* | Entering the try block doesn't change the SEH HEAD, entering the function con... | Jason Madden | 2021-10-29 | 1 | -5/+12 |
* | It seems that now we really are calling terminate() | Jason Madden | 2021-10-29 | 1 | -2/+12 |
* | Make the part of g_initialstub that never returns its own function. | Jason Madden | 2021-10-29 | 4 | -124/+163 |
* | Use a better testrunner. | Jason Madden | 2021-10-29 | 1 | -2/+2 |
* | Debugging. | Jason Madden | 2021-10-29 | 3 | -1/+15 |
* | Hmm, the build shows failed, but the log cuts off in the middle with no error... | Jason Madden | 2021-10-29 | 2 | -25/+19 |
* | Hot dog, that fixed the tracing tests. Did it fix all the tests? | Jason Madden | 2021-10-29 | 1 | -1/+1 |
* | Helps if you actually return it. | Jason Madden | 2021-10-29 | 1 | -11/+13 |
* | We seem to need to keep the bottom three handlers around, perhaps? | Jason Madden | 2021-10-29 | 3 | -6/+20 |
* | More debugging. | Jason Madden | 2021-10-29 | 2 | -1/+12 |
* | The constructor of the thread state is too late to capture, because it has it... | Jason Madden | 2021-10-29 | 1 | -2/+9 |
* | Must explicitly convert. | Jason Madden | 2021-10-29 | 1 | -1/+1 |
* | Save and restore the SEH state from the main greenlet into new greenlets. | Jason Madden | 2021-10-29 | 3 | -2/+38 |
* | It looks like unwinding through shared stack frames is corrupting the state. | Jason Madden | 2021-10-29 | 3 | -7/+21 |
* | Showing before replacement but after stack switch makes no sense. | Jason Madden | 2021-10-29 | 1 | -2/+4 |
* | No need for set_terminate either. | Jason Madden | 2021-10-29 | 1 | -17/+1 |
* | Maybe ignoring access violation wasn't a good idea. | Jason Madden | 2021-10-29 | 2 | -8/+2 |
* | Guard the SEH walking with SEH handling. This probably doesn't work, but lets... | Jason Madden | 2021-10-29 | 3 | -6/+25 |
* | Try to traverse SEH chain manually. | Jason Madden | 2021-10-29 | 3 | -4/+38 |
* | We do seem to get a vectored exception each time we throw a C++ exception. | Jason Madden | 2021-10-29 | 2 | -15/+35 |
* | Debugging: Temporarily add a vectored exception handler. | Jason Madden | 2021-10-29 | 1 | -1/+51 |
* | Can't use /MDd because we would need the Python debug libs. | Jason Madden | 2021-10-28 | 1 | -5/+6 |
* | Enable some debug options on MSVC. | Jason Madden | 2021-10-28 | 3 | -27/+16 |
* | One more try with /EHa. | Jason Madden | 2021-10-28 | 1 | -1/+1 |
* | More debugging. | Jason Madden | 2021-10-28 | 1 | -1/+15 |
* | That got us a lot further; now more debugging. | Jason Madden | 2021-10-28 | 2 | -5/+7 |
* | One level was not enough. Try two. | Jason Madden | 2021-10-28 | 1 | -17/+13 |
* | Silly compile error. | Jason Madden | 2021-10-28 | 1 | -1/+1 |
* | Ok, I give up on 32-bit windows. See how far we need to rewind to use excepti... | Jason Madden | 2021-10-28 | 2 | -0/+48 |
* | Quite possibly we weren't saving the seh in a place it would actually be saved. | Jason Madden | 2021-10-28 | 1 | -2/+8 |
* | Sprinkle in some more noexcept. | Jason Madden | 2021-10-28 | 3 | -6/+6 |
* | Too much | Jason Madden | 2021-10-28 | 1 | -1/+1 |
* | Some extra info lives on the stack of functions that use exceptions. Perhaps ... | Jason Madden | 2021-10-28 | 2 | -45/+42 |
* | Force g_switch_finish to be called virtually. | Jason Madden | 2021-10-28 | 1 | -40/+44 |
* | More debugging. | Jason Madden | 2021-10-28 | 1 | -1/+13 |
* | MSVC is picky about extern C matching. | Jason Madden | 2021-10-28 | 1 | -0/+4 |
* | Debugging. | Jason Madden | 2021-10-28 | 2 | -7/+10 |
* | Take the c off the /EHsc. | Jason Madden | 2021-10-28 | 2 | -2/+3 |
* | Try using the compiler intrinsics to access the SEH info | Jason Madden | 2021-10-28 | 3 | -16/+13 |
* | It looks like 32-bit Windows compilers don't like catch(...), no matter how n... | Jason Madden | 2021-10-28 | 2 | -1/+8 |