HP3000-L Archives

September 1997, Week 2

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Wayne, Roby" <[log in to unmask]>
Reply To:
Wayne, Roby
Date:
Wed, 10 Sep 1997 16:23:39 -0400
Content-Type:
text/plain
Parts/Attachments:
text/plain (160 lines)
Stan,

That explained the whole problem.  The lt_msg_enter was expecting
4-bytes and we had the "SHORT" declaration.

Seems to be working just fine now!!

Best regards,

Roby

> -----Original Message-----
> From: Stan Sieler [SMTP:[log in to unmask]]
> Sent: Wednesday, September 10, 1997 4:06 PM
> To:   [log in to unmask]
> Subject:      Re: [HP3000-L] TRAPS 68 Error
>
> Hi,
>
> > Here is the dump.  Alas, I am too green to understand what it means.
> If
> > I compile an FOPEN into the program, I get an abort with TRAPS 68.
> If I
> > comment the FOPEN statement out and recompile, it works just fine.
>
> The program aborted on a  "STWS" (Store Word, short) instruction.
> All word-oriented instructions require addresses that are multiples of
> 4 bytes (a word is 4 bytes).  In this instance, the store is being
> done
> to the address in register R21.  R21 has $41836182 (hex).  Since the
> value ends in 2 (and not 0, 4, 8, or $c), we know it isn't a multiple
> of 4.
>
> Possible cause:
>   You have some routine (LANenterCSF ?), which is compiled in a way
>   that says "I expect a 4-byte reference parameter".
>   The caller (lt_msg_enter) isn't taking any pains to ensure that the
>   variable passed into LANenterCSF is 4-byte aligned.
>   The compiler is emitting 2 extra bytes of local scratch storage if
>   there are any FOPEN calls ... this means that without the FOPEN
> call,
>   by pure chance the variable is allocated on a 4-byte boundary, and
>   with an FOPEN call, it gets pushed 2 bytes further in memory, and
>   isn't 4-byte aligned.
>
>   A $TABLES$ (?) directive may help by reporting where the variables
>   are allocated.
>
>   Also, you may be using a $HP3000_16$ directive, which says
> "generally
>   emit variables on 2-byte boundaries, not 4-byte boundaries".
>
> I hope this helps.  If not, you could send me the source for
> lt_msg_enter, and I'd look at it.
>
> BTW, I'm stepping out for an hour now.
>
> > 00152574  LANenterCSF+$d0     0ebc1280  STWS     28,0(0,21)
> > I have calls in to Starvision and HP.  Fun!
> >
> > Thanks, again!
> >
> > Roby
> >
> > ====> BEGIN CARCOMM
> > CARCOMM  Connect to remote workstation
> > CARCOMM  LT_BUFF=[Carousel Processor]
> > **** Data memory protection trap (TRAPS 68).
> >
> > ABORT: CARCOMM.CAR.MMV090
> >        PC=c38.00152574 LANenterCSF+$d0
> > NM* 0) SP=41836670 RP=c38.00152c2c lt_msg_enter+$28
> > NM  1) SP=418363f0 RP=c38.00152bf0 ?lt_msg_enter+$8
> >          export stub: ded.00005b70 carcomm+$368
> > NM  2) SP=418363b0 RP=ded.00000000
> >      (end of NM stack)
> >
> > R0 =00000000 41642200 00152572 82f55c00 R4 =d6effa88 00000001
> 00000000
> > 00000000
> > R8 =00000000 00000000 00000000 00000000 R12=00000000 00000000
> 00000000
> > 00000000
> > R16=00000000 00000000 00000000 0000005a R20=00000186 41836182
> 00000000
> > 4163c484
> > R24=00000000 418369cc 0000001e 4163a000 R28=00000000 82f55c00
> 41836670
> > 0000001e
> >
> > IPSW=0004000f=jthlnxbCvmrQPDI  PRIV=2   SAR=0019 PCQF=c38.152576
> > c38.15257a
> >
> > SR0=0000000a 00000000 00000000 00000000 SR4=00000c38 00000c74
> 0000000b
> > 0000000a
> > TR0=816df000 00000ff2 01136a03 d16e8390 TR4=ffffff88 80000000
> 000095c1
> > 00000006
> > PID1=04e8=0274(W) PID2=0000=0000(W)     PID3=0000=0000(W)
> > PID4=0000=0000(W)
> >
> > RCTR=00000000 ISR=00000c74 IOR=41836182 IIR=0ebc1280 IVA=00122000
> > ITMR=275e7911
> > EIEM=ffffffff EIRR=00000000 CCR=0000
> >
> > **** PROCESS ABORT INTERACTIVE DEBUG FACILITY ****
> >
> > $2 ($137) nmdebug >
> > $3 ($137) nmdebug > dc pc
> > USER $c38.152574
> > 00152574  LANenterCSF+$d0     0ebc1280  STWS     28,0(0,21)
> > $4 ($137) nmdebug > abort
> > :
> >
> > > -----Original Message-----
> > > From:     Stan Sieler [SMTP:[log in to unmask]]
> > > Sent:     Wednesday, September 10, 1997 3:52 PM
> > > To:       [log in to unmask]
> > > Cc:       [log in to unmask]
> > > Subject:  Re: [HP3000-L] TRAPS 68 Error
> > >
> > > Hi,
> > >
> > > Re:
> > > > **** Data memory protection trap (TRAPS 68).
> > > >
> > > > ABORT: CARCOMM.CAR.MMV090
> > > > NM USER  c38.00152574 LANenterCSF+$d0
> > > > NM PROG  339.00005c80 carcomm+$368
> > >
> > > Try:
> > >
> > >    :setdump
> > >    :run CARCOMM.CAR.MMV090
> > >
> > > when the abort happens, registers should be displayed, and you'll
> > > be left in Debug/iX.  Enter:
> > >
> > >    dc pc
> > > and
> > >    abort
> > >
> > > email/post and I'll look at it ... it might provide a clue as to
> > > what's going wrong.
> > >
> > > --
> > > Stan Sieler
> > > [log in to unmask]
> > >
> > > http://www.allegro.com/sieler.html
> >
>
>
> --
> Stan Sieler
> [log in to unmask]
>
> http://www.allegro.com/sieler.html

ATOM RSS1 RSS2