On Nov 14, 2017, at 4:34 PM, Ben Coman <btc@openinworld.com> wrote:



On 15 November 2017 at 00:07, Todd Blanchard��<tblanchard@mac.com>��wrote:
I've got it loaded, have fixed up the library path and fixed the test for version (built in lib returns very different version string).

I have many tests green.�� However, as before, tests involving CXSourceRange that return CXSourceLocations return all zero'd data.�� The CXSourceRange struct looks sane.

The calls to��CXSourceLocation clang_getRangeStart(CXSourceRange range);

and its twin RangeEnd result in zero'd structs.

Any idea on where I can look to try to figure out where this is going wrong? ��

I have compiled little C programs and verified that both struct sizes agree in FFI and native code (24 bytes each).

Not sure what to try next.
��c
-Todd Blanchard

Compile your own debug-VM** and make a little C wrapper MyCXSourceRange��to forward to Clang's CXSourceRange, and set a gdb a breakpoint in MyCXSourceRange and compare 64 bit and 32 bit versions.����

That still leaves differences in the callout before your breakpoint in MyCXSourceRange is reached, so just do two callouts in a row and trace from the first to the second.��


cheers -ben

On 17 November 2017 at 14:57, Todd Blanchard <tblanchard@mac.com> wrote:
Right, so I have my own debug VM (64 bit Pharo Cog) and a debug version of libclang.��

I can attach with lldb and set a breakpoint on��clang_getRangeStart

Then I debug the test that gets diagnostics and I see that the CXSourceRange has data.�� Then I step into 'range start' and I land in my breakpoint for clang_getRangeStart and I can see that the argument to the function - supposedly a CXSourceRange passed by value - is all zero'd memory.

So it appears that in the 64bit VM, structs passed by value doesn't work right.

I've traced back up the stack and I confess the marshaling code is totally beyond my comprehension.�� This is what I was able to get:

Process 1126 stopped
* thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1 2.1
�� �� frame #0: 0x000000011e33136b libclang.dylib`::clang_getRangeStart(range=CXSourceRange @ 0x00007fff55f4e380) at CXSourceLocation.cpp:89
���� 86 �� ��
���� 87�� CXSourceLocation clang_getRangeStart(CXSourceRange range) {
���� 88�� �� // Special decoding for CXSourceLocations for CXLoadedDiagnostics.
-> 89�� �� if ((uintptr_t)range.ptr_data[0] & 0x1) {
���� 90�� �� �� CXSourceLocation Result = { { range.ptr_data[0], nullptr }, 0 };
���� 91�� �� �� return Result; �� ��
���� 92�� �� }
Target 0: (Pharo) stopped.
(lldb) p range
(CXSourceRange) $0 = {
�� ptr_data = ([0] = 0x0000000000000000, [1] = 0x0000000000000000)
�� begin_int_data = 0
�� end_int_data = 0
}
(lldb) bt
* thread #1, queue = 'com.apple.main-thread', stop reason = breakpoint 1.1 2.1
�� * frame #0: 0x000000011e33136b libclang.dylib`::clang_getRangeStart(range=CXSourceRange @ 0x00007fff55f4e380) at CXSourceLocation.cpp:89
�� �� frame #1: 0x0000000109e20b5a Pharo`primitiveCalloutWithArgs at X64SysVFFIPlugin.c:6041
�� �� frame #2: 0x0000000109cb964c Pharo`primitiveExternalCall at gcc3x-cointerp.c:77188
�� �� frame #3: 0x0000000109cb1629 Pharo`interpretMethodFromMachineCode at gcc3x-cointerp.c:19719
�� �� frame #4: 0x0000000109cb77fc Pharo`ceSendsupertonumArgs(selector=4713336760, superNormalBar=0, rcvr=4692466872, numArgs=1) at gcc3x-cointerp.c:17767
�� �� frame #5: 0x000000011774a135
�� �� frame #6: 0x0000000109c76227 Pharo`interpret at gcc3x-cointerp.c:2703
�� �� frame #7: 0x0000000109dbbf31 Pharo`-[sqSqueakMainApplication runSqueak](self=0x00006080000d7370, _cmd="runSqueak") at sqSqueakMainApplication.m:201
�� �� frame #8: 0x00007fff81e5a8dd Foundation`__NSFirePerformWithOrder + 368
�� �� frame #9: 0x00007fff803b1d37 CoreFoundation`__CFRUNLOOP_IS_CALLING_OUT_TO_AN_OBSERVER_CALLBACK_FUNCTION__ + 23
�� �� frame #10: 0x00007fff803b1ca7 CoreFoundation`__CFRunLoopDoObservers + 391
�� �� frame #11: 0x00007fff803926d9 CoreFoundation`__CFRunLoopRun + 873
�� �� frame #12: 0x00007fff80392114 CoreFoundation`CFRunLoopRunSpecific + 420
�� �� frame #13: 0x00007fff7f8f2ebc HIToolbox`RunCurrentEventLoopInMode + 240
�� �� frame #14: 0x00007fff7f8f2bf9 HIToolbox`ReceiveNextEventCommon + 184
�� �� frame #15: 0x00007fff7f8f2b26 HIToolbox`_BlockUntilNextEventMatchingListInModeWithFilter + 71
�� �� frame #16: 0x00007fff7de89a54 AppKit`_DPSNextEvent + 1120
�� �� frame #17: 0x00007fff7e6057ee AppKit`-[NSApplication(NSEvent) _nextEventMatchingEventMask:untilDate:inMode:dequeue:] + 2796
�� �� frame #18: 0x00007fff7de7e3db AppKit`-[NSApplication run] + 926
�� �� frame #19: 0x00007fff7de48e0e AppKit`NSApplicationMain + 1237
�� �� frame #20: 0x0000000109db5baf Pharo`main(argc=1, argv=0x00007fff55f8bca8, envp=0x00007fff55f8bcb8) at main.m:52
�� �� frame #21: 0x00007fff95b18235 libdyld.dylib`start + 1
�� �� frame #22: 0x00007fff95b18235 libdyld.dylib`start + 1
(lldb) up
frame #1: 0x0000000109e20b5a Pharo`primitiveCalloutWithArgs at X64SysVFFIPlugin.c:6041
���� 6038 result2 = floatObjectOf(floatRet1);
���� 6039 goto l14;
���� 6040 }
-> 6041 intRet1 = dispatchFunctionPointerwithwithwithwithwithwith(((SixteenByteReturn (*)(sqIntptr_t, sqIntptr_t, sqIntptr_t, sqIntptr_t, sqIntptr_t, sqIntptr_t)) (((void *) address2))), ((calloutState1->integerRegisters))[0], ((calloutState1->integerRegisters))[1], ((calloutState1->integerRegisters))[2], ((calloutState1->integerRegisters))[3], ((calloutState1->integerRegisters))[4], ((calloutState1->integerRegisters))[5]);
���� 6042 /* begin maybeOwnVM:threadIndex: */
���� 6043
���� 6044 #�� if COGMTVM

The struct that should have shown up appears in Pharo as��

"CXSourceRangeStruct (��
ptr_data1: (void*)@ 16r7FDF356C77B0
ptr_data2: (void*)@ 16r7FDF356D0C70
begin_int_data: 21
end_int_data: 28
)"

Open to ideas at this point but I think to progress I would require deep knowledge of Intel calling conventions and I don't really know where to go next.

Try to isolate the functions into a little dynamic lib and file a bug?

As you say, to minimize confounding factors the next step would be to try to reproduce the "structs passed by value" problem in a minimal library,
and also someone trying to reproduce it doesn't need to compile the whole libclang��

Also now is a good time to move discussion to vm-dev.

cheers -ben

��