[MLton] Patch ready

[Wesley W. Terpstra]

On Feb 11, 2007, at 11:26 PM, Matthew Fluet wrote:
> Wesley W. Terpstra wrote:
>> More worrying, however, is that the executable is bugged! It runs  
>> three times in a row on my system. All three runs are successful,  
>> and terminate at test43, but it should only be running once!
> I don't get this behavior at all.

I'm going to assume you mean it doesn't happen for you, as opposed to  
you don't understand it.

After doing a 'svn update; make clean; make' with the mlton-unicode- 
v3.patch applied, i run (on an intel machine):
> terpstra at dakar:/tmp/mlton/mlton/regression$ ../build/bin/mlton  
> widechar.sml
> terpstra at dakar:/tmp/mlton/mlton/regression$ ./widechar
>
> File widechar.sml: Testing structure WideChar...
> test4           OK
> test5           OK
> test6           OK
> test18          OK
> test19          OK
> test20          OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> test31          OK
> test32          OK
> test33          OK
> test34a         OK
> test34b         OK
> test35a         OK
> test35b         OK
> test36          OK
> test37          OK
> test38          OK
> test39          WRONG
> test40          OK
> test41          OK
> test42          OK
> test43          WRONG
>
> File bytechar.sml: Testing structure WideChar...
> test4           OK
> test5           OK
> test6           OK
> test18          OK
> test19          OK
> test20          OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> test31          OK
> test32          OK
> test33          OK
> test34a         OK
> test34b         OK
> test35a         OK
> test35b         OK
> test36          OK
> test37          OK
> test38          OK
> test39          WRONG
> test40          OK
> test41          OK
> test42          OK
> test43          WRONG
>
> File bytechar.sml: Testing structure WideChar...
> test4           OK
> test5           OK
> test6           OK
> test18          OK
> test19          OK
> test20          OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> checkset        OK
> test31          OK
> test32          OK
> test33          OK
> test34a         OK
> test34b         OK
> test35a         OK
> test35b         OK
> test36          OK
> test37          OK
> test38          OK
> test39          WRONG
> test40          OK
> test41          OK
> test42          OK
> test43          WRONG
> terpstra at dakar:/tmp/mlton/mlton/regression$

The same thing happens if I use the C codegen. I'm not sure how I can  
reproduce this for you? I can send you the intermediate assembler  
files and/or all the output passes?