C-Kermit progress - Protocols
This is a discussion on C-Kermit progress - Protocols ; Looks like the new build worked fine on AIX 5.3 and SunOS 4.1.4, but I'm
running into errors compiling it on NetBSD 2.0.2 and OpenBSD (current)
on the VAX platform (actually the VAX emulators running under SIMH for
windows). The ...
-
C-Kermit progress
Looks like the new build worked fine on AIX 5.3 and SunOS 4.1.4, but I'm
running into errors compiling it on NetBSD 2.0.2 and OpenBSD (current)
on the VAX platform (actually the VAX emulators running under SIMH for
windows). The errors I get are like this on both:
O" "LIBS= -lcurses -lutil -lm"
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckcmai.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckclib.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckucmd.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuusr.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus2.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus3.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus4.c
cc -DBSD44 -DCK_CURSES -DCK_NEWTERM -DTCPSOCKET -DOPENBSD -DUSE_UU_LOCK -DFNFLO
AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus5.c
cc: Internal compiler error: program cc1 got fatal signal 11
*** Error code 1
Stop in /tmp.
*** Error code 1
Stop in /tmp (line 1773 of makefile).
-
Re: C-Kermit progress
On 2005-09-28, Robert Rodriguez wrote:
: Looks like the new build worked fine on AIX 5.3 and SunOS 4.1.4, but I'm
: running into errors compiling it on NetBSD 2.0.2 and OpenBSD (current)
: on the VAX platform (actually the VAX emulators running under SIMH for
: windows). The errors I get are like this on both:
:
: ...AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus5.c
: cc: Internal compiler error: program cc1 got fatal signal 11
: *** Error code 1
:
Well, it builds OK on real NetBSD 2.0, and although I haven't tried it on
the latest OpenBSD yet, I'd be surprised at any surprises, since OpenBSD has
been one of the most reliable (i.e. unchanging) platforms. It looks like
some kind of resource (virtual swap space? /tmp space?) was exhausted. Does
it always happen in ckuus5.c?
What happens if you run the make again? In this case it would start with
ckuus5.c, perhaps with less resources allocated.
- Frank
-
Re: C-Kermit progress
Frank da Cruz wrote:
> On 2005-09-28, Robert Rodriguez wrote:
> : Looks like the new build worked fine on AIX 5.3 and SunOS 4.1.4, but I'm
> : running into errors compiling it on NetBSD 2.0.2 and OpenBSD (current)
> : on the VAX platform (actually the VAX emulators running under SIMH for
> : windows). The errors I get are like this on both:
> :
> : ...AT -DUSE_STRERROR -O -DKTARGET=\"openbsd\" -c ckuus5.c
> : cc: Internal compiler error: program cc1 got fatal signal 11
> : *** Error code 1
> :
> Well, it builds OK on real NetBSD 2.0, and although I haven't tried it on
> the latest OpenBSD yet, I'd be surprised at any surprises, since OpenBSD has
> been one of the most reliable (i.e. unchanging) platforms. It looks like
> some kind of resource (virtual swap space? /tmp space?) was exhausted. Does
> it always happen in ckuus5.c?
> What happens if you run the make again? In this case it would start with
> ckuus5.c, perhaps with less resources allocated.
> - Frank
It always fails on the ckuus5.c, even after a reboot. I tried
increasing the emulator virtual memory size to 256meg, but it didn't
help.
Bob Rodriguez
Also I tried building it on MVS Openedition with the POSIX option, but
looks like there is no param.h or utmp.h header file, and it got
errors looking for a timezone struct. Are there some other options I need
to add?
-
Re: C-Kermit progress
On 2005-09-29, Robert Rodriguez wrote:
: It always fails on the ckuus5.c, even after a reboot. I tried
: increasing the emulator virtual memory size to 256meg, but it didn't
: help.
:
Then it's a "hard" resource limit that doesn't exit on real NetBSD. I
don't know if the limit is in the emulated VAX or what. Anyway, you can
find a lot of hints in documents like:
http://www.columbia.edu/kermit/ckccfg.html
for trimming the size by trimming features and/or size of things. ckuus5.c
is mainly taken up by the script language implementation. I'll bet if you
did:
make netbsd KFLAGS=-DNOSPL
it would build. But it also wouldn't include the scripting features.
: Also I tried building it on MVS Openedition with the POSIX option, but
: looks like there is no param.h or utmp.h header file, and it got
: errors looking for a timezone struct. Are there some other options I need
: to add?
:
If I knew the answer to that I would have a makefile target. Usually the
way this works is, somebody gives me access to the platform in question,
I spend a few hours figuring it out, and then I put together the needed
targed, along with any necessary source-code changes.
- Frank