[M3devel] narrow still failing..
mika at async.caltech.edu
mika at async.caltech.edu
Tue Jan 21 20:23:28 CET 2014
How do you get them into text form? They look binary...
I copied the whole directory bdd (and all subdirs) to
http://rover.gcapltd.com/~mika/bdd/
Antony Hosking writes:
>Please post .io / .mo in text form.
>
>Sent from my iPad
>
>> On Jan 21, 2014, at 1:08 PM, mika at async.caltech.edu wrote:
>>=20
>> The only thing bddtest uses from caltech_parser is Debug.Out. You can jus=
>t take out those calls or replace them with IO.Put.
>>=20
>> But again bddtest is not the program that fails... it's the "refany" progr=
>am (now in a sibling directory to bddtest)
>>=20
>> Mika
>>=20
>> "Rodney M. Bates" writes:
>>>=20
>>>=20
>>>> On 01/21/2014 11:11 AM, mika at async.caltech.edu wrote:
>>>> Did you try using my bdd.tgz ?
>>>>=20
>>>> I am getting this error on:
>>>>=20
>>>> ancient PM3 on FreeBSD4
>>>>=20
>>>> head CM3 on AMD64_LINUX
>>>=20
>>> Hmm. I can't get any errors compiling bdd and bddtest on my AMD64_LINUX s=
>ystem with
>>> a freshly updated head. Hacking around, I tried to use the binaries in b=
>dd.tgz,
>>> but my .M3SHIP wants a libbdd.so.5, not there. Some of the *.[im]o files=
> are
>>> slightly different in size when I recompiled, but I think there is an exp=
>lanation
>>> for that.
>>>=20
>>> I can't build bddtest with the release compiler, because it needs caltech=
>_parser,
>>> and that won't build in the release. This combination of [non]failures s=
>eems
>>> pretty strange. I am also wondering about a problem with type checks in t=
>he
>>> runtime.
>>>=20
>>>=20
>>>>=20
>>>> few months old CM3 on ARM_LINUX (Raspberry Pi)
>>>>=20
>>>> I doubt you will get it on LINUXLIBC6 if you don't get it on the others.=
>
>>>>=20
>>>> I already tried rearranging the files (the implementation and interfaces=
>
>>>> are split in a funny way across BDD.i3, BDDImpl.i3, BDD.m3, BDDImpl.m3
>>>> right now). With no effect. I don't know what it is about my code that=
>'s
>>>> tripping up the compiler.
>>>>=20
>>>> It's also a "decent" BDD package so someone (me I guess, but anyone
>>>> else who's interested is welcome to) might want to add it to CM3... :-)
>>>> Kind of a pain if you can't print out the debugging info though!
>>>> (Actually the workarounds are OK.)
>>>>=20
>>>> Mika
>>>>=20
>>>> "Rodney M. Bates" writes:
>>>>>=20
>>>>>=20
>>>>>> On 01/20/2014 10:18 PM, mika at async.caltech.edu wrote:
>>>>>> Were you able to get it to fail on any of your computers?
>>>>>=20
>>>>> No, the only times I got narrow failures were due to errors
>>>>> in my trial cases.
>>>>>=20
>>>>> I have only tried AMD64_LINUX, but should be able to get this onto a
>>>>> LINUXLIBC6 system.
>>>>>=20
>>>>>>=20
>>>>>> Your Z fails for me---CM3 and PM3 both.
>>>>>>=20
>>>>>> (579)truffles:~/bsd/refany/src>../AMD64_LINUX/refany
>>>>>>=20
>>>>>>=20
>>>>>> ***
>>>>>> *** runtime error:
>>>>>> *** An explicit or implicit NARROW() operation failed.
>>>>>> *** file "../src/Main.m3", line 30
>>>>>> ***
>>>>>=20
>>>>> Well that torpedos both my theories.
>>>>>=20
>>>>> But it is hard to imagine that bad code generated for an implicit narro=
>w
>>>>> on an assignment could ever have gone unnoticed so long. I thought it
>>>>> might be more believable if it only happened when narrowing an actual
>>>>> parameter before passing it.
>>>>>=20
>>>>>>=20
>>>>>> Abort
>>>>>> (580)truffles:~/bsd/refany/src>cat Mai
>>>>>> Main.m3~ Main.m3
>>>>>> (580)truffles:~/bsd/refany/src>cat -n Main.m3
>>>>>> 1 MODULE Main;
>>>>>> 2 IMPORT BDD, IO;
>>>>>> 3
>>>>>> 4 PROCEDURE P() : REFANY =3D
>>>>>> 5 BEGIN
>>>>>> 6 VAR x : REFANY;
>>>>>> 7 BEGIN
>>>>>> 8 x :=3D BDD.New("a");
>>>>>> 9 Z(x);
>>>>>> 10 RETURN x
>>>>>> 11 END
>>>>>> 12 END P;
>>>>>> 13
>>>>>> 14 PROCEDURE Q() =3D
>>>>>> 15 BEGIN
>>>>>> 16 IO.Put(P() & "\n")
>>>>>> 17 END Q;
>>>>>> 18
>>>>>> 19 (*
>>>>>> 20 PROCEDURE Z(VAR x : REFANY) =3D
>>>>>> 21 BEGIN
>>>>>> 22 TYPECASE x OF
>>>>>> 23 BDD.T(b) =3D> x :=3D BDD.Format(b)
>>>>>> 24 END;
>>>>>> 25 END Z;
>>>>>> 26 *)
>>>>>> 27 PROCEDURE Z(VAR x : REFANY) =3D
>>>>>> 28 VAR b: BDD.T;
>>>>>> 29 BEGIN
>>>>>> 30 b :=3D x;
>>>>>> 31 x :=3D BDD.Format(b)
>>>>>> 32 END Z;
>>>>>> 33
>>>>>> 34
>>>>>> 35
>>>>>> 36 BEGIN
>>>>>> 37 Q()
>>>>>> 38 END Main.
>>>>>> (581)truffles:~/bsd/refany/src>
>>>>>> "Rodney M. Bates" writes:
>>>>>>> What does this do?
>>>>>>>=20
>>>>>>> PROCEDURE Z(VAR x : REFANY) =3D
>>>>>>> VAR b: BDD.T;
>>>>>>> BEGIN
>>>>>>> b :=3D x;
>>>>>>> x :=3D BDD.Format(b)
>>>>>>> END Z;
>>>>>>>=20
>>>>>>> The common property of the failing cases seems to be the necessity
>>>>>>> of a runtime narrow check when passing x to BDD.Format.
>>>>>>>=20
>>>>>>> So far, I haven't been able to reproduce the failure on AMD64_LINUX,
>>>>>>> with even more irrelevant (one would think) stuff pared out. I moved=
>
>>>>>>> T, Root, and Format into main, removed all but one fields of T and Ro=
>ot,
>>>>>>> replaced BDD.New by NEW(Root) and removed Symtab and most of the body=
>
>>>>>>> from Format.
>>>>>>>=20
>>>>>>> Looking at your code, I agree it should work.
>>>>>>>=20
>>>>>>>=20
>>>>>>>> On 01/19/2014 04:01 PM, mika at async.caltech.edu wrote:
>>>>>>>> As in the ISTYPE/TYPECASE example, the following works:
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> PROCEDURE Z(VAR x : REFANY) =3D
>>>>>>>> BEGIN
>>>>>>>> TYPECASE x OF
>>>>>>>> BDD.T(b) =3D> x :=3D BDD.Format(b)
>>>>>>>> END;
>>>>>>>> END Z;
>>>>>>>>=20
>>>>>>>> (244)rover:~/refany/src>../FreeBSD4/refany
>>>>>>>> a
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> This does NOT work:
>>>>>>>>=20
>>>>>>>> PROCEDURE Z(VAR x : REFANY) =3D
>>>>>>>> BEGIN
>>>>>>>> TYPECASE x OF
>>>>>>>> BDD.T(b) =3D> x :=3D BDD.Format(x)
>>>>>>>> END;
>>>>>>>> END Z;
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> (242)rover:~/refany/src>../FreeBSD4/refany
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> ***
>>>>>>>> *** runtime error:
>>>>>>>> *** NARROW failed
>>>>>>>> *** file "/big/home/mika/refany/src/Main.m3", line 22
>>>>>>>> ***
>>>>>>>>=20
>>>>>>>> use option @M3stackdump to get a stack trace
>>>>>>>> Abort
>>>>>>>>=20
>>>>>>>>=20
>>>>>>>> Same result PM3 / CM3.
>>>>>>>>=20
>>>>>>>> Mika
>>>>=20
More information about the M3devel
mailing list