Re: Assert in VM

[Roberto Ierusalimschy <roberto@...>]

> Please correct me swiftly if this is wrong. If this is true, it should
> have been pointed out in the manual, and is a severe limitation/bug in
> Lua.

It seems that that assertion is wrong. Of course when we return to a
function it does not need to be running on the same C frame it was
before.  That is the whole point of coroutines in Lua: That you *can*
"call a C function, which calls Lua to create a coroutine and then
resume your coroutine (once), yield and return to C, then resume the
coroutine through C from a different point."

Besides the assertion, the initialization of that field may be wrong also.
The line

> callentry:  /* entry point when calling new functions */
-  L->ci->u.l.pc = &pc;

should be here:

> retentry:  /* entry point when returning to old functions */
+  L->ci->u.l.pc = &pc;

It is tricky to get this bug, because this field (`pc') is used only for
debug/error report. The attached code shows the bug in Lua5.0.  BTW,
maybe this explains your previous problems with linehooks in coroutines!

I am quite busy until next week, so I cannot test all these now. But as
soon as I have more time I will double check all this stuff.

-- Roberto

function g(x)
    coroutine.yield(x)
end

function f (i)
  debug.sethook(print, "l")
  for j=1,1000 do
    g(i+j)
  end
end

co = coroutine.wrap(f)
co(10)
pcall(co)
pcall(co)