indirect register frames - current state and some notes
Leopold Toetsch
failing tests[1] currently.
I had to disable t/library/dumper.t, though. It doesn't setup registers
according to pdd03 and just assumes in a few places, that function
arguments will arrive at the caller's end. I think, that's simulating
some varariable arguments calling convention, but that violates pdd03.
We have to find something better for that.
Fixes for this are welcome. The same might be true for streams.t, which
at least in one function (setSource) had the same problem. Still 8
failures remaining.
The same problem is with extend.c:Parrot_call(), which isn't fully
compliant. But that code should wrap a normal Parrot function anyway and
not reside in extend.c. I'll fix that.
leo
[1]
Failed Test Stat Wstat Total Fail Failed List of Failed
-------------------------------------------------------------------------------
imcc/t/syn/pcc.t 1 256 37 1 2.70% 25
t/library/streams.t 8 2048 21 8 38.10% 10-12 14-17 20
t/op/gc.t 1 256 18 1 5.56% 13
t/src/extend.t 1 256 13 1 7.69% 12
3 tests and 68 subtests skipped.
Failed 4/123 test scripts, 96.75% okay. 11/1969 subtests failed, 99.44%
okay.
