Re: Filed PR jit/66812 for the code generation issue

[David Malcolm <dmalcolm@redhat.com>]

On Wed, 2015-07-08 at 20:36 +0100, Dibyendu Majumdar wrote:
> On 8 July 2015 at 20:03, David Malcolm <dmalcolm@redhat.com> wrote:
> > On Wed, 2015-07-08 at 20:01 +0100, Dibyendu Majumdar wrote:
> >> On 8 July 2015 at 18:46, David Malcolm <dmalcolm@redhat.com> wrote:
> >> > Dibyendu: what Lua code generated the reproducer?  What is the code
> >> > meant to be doing?
> >> >
> >>
> >> Hi Dave - the Lua test is this:
> >>
> >> function x()
> >>   local IX
> >>   if ((10 or true) and false) then
> >>     IX = true
> >>   end;
> >>   return ((10 or true) and false)
> >> end
> >> assert(x() == false)
> >>
> >> In the original test IX is an upvalue - i.e. a variable in outer
> >> scope. This is my standalone version of the test. The original test is
> >> generated as part of the Lua test suite - its purpose is to test
> >> various permutations of boolean operators.
> >>
> >> The original test compares IX and the function return.
> >>
> >> The issue is that this test should return false - if you see the
> >> return statement. However when -O2 or -O3 is enabled it returns true.
> >>
> >> The if statement is indeed redundant in this cut down version as IX is
> >> a local variable. But the return statement is not redundant.
> >
> > Thanks.  What does this look like as bytecodes?
> >
> 
> Ok the bug is still there - I was running the test incorrectly. I have
> attached the standalone replication and the output.
> Note that if I remove the redundant if statement then it works correctly.

Thanks.


BTW, are Lua/Ravi constants truly constant?  If so, then I'd believe
you'd get a performance win by implementing LOADK by emitting code to
write the specific tt and value_ directly, rather than code that copies
a value from the table.

This would enable the optimizer to "know" the tt and value_, and
optimize accordingly.  For example, in this case, I believe it would
allow the function to be optimized away down to the equivalent of just a
"return false;".  Obviously won't help much for a function without a
loop, but if it saves instructions inside a loop, that's probably a win.

(...though maybe not before we track down this issue)