adam@ncifcrf.gov (Adam W. Feigin) (09/20/90)
I just upgraded our Iris 4d machines to Irix 3.3.1, and I've got this
bizzare bug/interaction with telnet and terminal characteristics, and
I'm wondering where the problem is, and if anyone has seen this
behavior before.
From an Iris 4D machine, if I telnet into a Cray (X/MP, Unicos
5.1.10), my terminal characteristics get all screwed up:
csh[30]:stty
speed 9600 baud; -parity hupcl
intr = b; quit = b; erase = ^b; kill = ^b; eof = ^b; swtch = ^`;
-inpck icrnl onlcr tab3
echo echoe echok
They were fine before I telnett'd in; Now, after I log out of the
Cray, and get back to the local system (Iris):
csh[49]:stty
speed 38400 baud; -parity
line = 1; intr = -b; quit = -b; erase = -^B; kill = -^B; eof = -^B; swtch = -B;
lnext = -^B; werase = -^B; rprnt = -^B; flush = -^B; stop = -^B; start = -^B;
brkint -inpck icrnl ixoff onlcr
echo echoe echok
And my terminal characteristics are screwed up on the local machine,
even though they were fine before I telnet'd to the Cray.
Rsh seems to work without screwing up the terminal characteristics.
I'm certain this is a bug, but who is responsible for it ? Has anybody
seen this behavior with their Iris's and Crays ?
Thanks in advance for any insight.
AWF
-------------------------------------------------------------------------------
Internet: adam@ncifcrf.gov Adam W. Feigin
UUCP: {backbonz}!ncifcrf!adam Senior Systems Manager
Mail: P.O. Box B, Bldg 430 National Cancer Institute-Superconmputer Center
Frederick, MD 21702 Frederick Cancer Research & Development Facilityarc@thyme.wpd.sgi.com (Andrew Cherenson) (09/29/90)
In article <1882@fcs280s.ncifcrf.gov> adam@ncifcrf.gov (Adam W. Feigin) writes: >I just upgraded our Iris 4d machines to Irix 3.3.1, and I've got this >bizzare bug/interaction with telnet and terminal characteristics, and >I'm wondering where the problem is, and if anyone has seen this >behavior before. > >From an Iris 4D machine, if I telnet into a Cray (X/MP, Unicos >5.1.10), my terminal characteristics get all screwed up: [Stuff deleteted] > AWF >------------------------------------------------------------------------------- >Internet: adam@ncifcrf.gov Adam W. Feigin >UUCP: {backbonz}!ncifcrf!adam Senior Systems Manager >Mail: P.O. Box B, Bldg 430 National Cancer Institute-Superconmputer Center > Frederick, MD 21702 Frederick Cancer Research & Development Facility There's a bug in the UNICOS telnetd. Here's the details: > From: dab@berserkly.cray.com (David Borman) > Date: Fri, 28 Sep 90 11:10:50 -0500 > > The problem that is being seen with telnetd on Unicos 5.1.10 is due to > a bug in the generation compiler, /usr/gen/bin/cc. There are two ways > to solve this problem: > 1) re-build telnetd with either /bin/scc or /bin/cc > (slc.c is the affected module) > 2) get mod e19534tcpa to change telnetd to work around > the compiler bug. This is a critical bugfix mod, that > should be applied to all 5.1.10 sites. The mod is real > simple: > > *f usr/etc/telnetd/slc.c > *d 518849tcpab.3 > register cc_t val; > *d 518849tcpab.8 > register cc_t val; > *d 518849tcpab.10 > register cc_t val; > > This changes the argument declarations in add_slc(), process_slc() > and change_slc() from non-register to register. The bug is that > if you have a function that takes three character arguments, and > the first two are declared as register and the third not as a > register, the value for the third argument will become the same > as the value for the second argument. > > -David Borman, dab@cray.com