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?
> On Nov 14, 2017, at 4:34 PM, Ben Coman <[email protected]> wrote:
>
>
>
> On 15 November 2017 at 00:07, Todd Blanchard <[email protected]
> <mailto:[email protected]>> 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.
>
> **
> https://github.com/OpenSmalltalk/opensmalltalk-vm/tree/Cog/build.linux64x64/pharo.cog.spur/build.debug
>
> <https://github.com/OpenSmalltalk/opensmalltalk-vm/tree/Cog/build.linux64x64/pharo.cog.spur/build.debug>
>
> cheers -ben
