home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.sys.hp:15452 comp.windows.x:21678
- Path: sparky!uunet!opl.com!hri.com!spool.mu.edu!uwm.edu!linac!sunova!fremont.ssc.gov!bush
- From: bush@fremont.ssc.gov (Joe Bush)
- Newsgroups: comp.sys.hp,comp.windows.x
- Subject: Multi-headed hp9000/42S
- Keywords: vuewm multi-headed HPA98550
- Message-ID: <1993Jan27.230322.23897@sunova.ssc.gov>
- Date: 27 Jan 93 23:03:22 GMT
- References: <C056Aq.JAt@hpuerca.atl.hp.com>
- Sender: bush@NoHost.NoDomain (Joe Bush)
- Organization: Superconducting Super Collider Laboratory
- Lines: 127
- Nntp-Posting-Host: rosebud.ssc.gov
-
- Hello,
-
- I am having problems trying to configure a multi-headed HP9000/42S
- running HP-UX B.08.00. I have installed an HPA98550 as a second frame buffer
- and am now trying to get VUE/X11 to take advantage of it. Anybody have any
- experience with this? The good folks at HP have not been able to tell me how
- this should be done.
-
- Here are my results to date:
-
- (1) # uname -a
- HP-UX lepew B.08.00 A 9000/42S 08000924c893
-
- (2) I installed HP98550 Bit Mapped Display card in slot 0 (the
- middle slot) on the HP tower.
-
- (3) The dip switches were set to [msb-lsb] 1000 0100 to
- produce a select code of (decimal) 132 per page 4-135 of Installing
- Peripherals Vol 2.
-
- (4) The card is recognized at boot time.
-
- # dmesg
- Jan 25 13:23
- Internal HIL at 0x428000
- CONSOLE is ITE
- ITE + 1 port(s)
- MC68040 processor
- HP98620C DMA
- Card ID 37 at select code 5 ignored; interface driver not present
- Card ID 37 at select code 6 ignored; interface driver not present
- Internal HP-IB Interface - system controller at select code 7
- Parallel poll interrupts enabled.
- HP98644 Advanced RS-232C Serial Interface at select code 9
- With 16 byte rcv fifo, 16 byte xmit fifo, hardware handshake, and high speed clock.
- Card ID 6 at select code 12 ignored; interface driver not present
- HP98265A (SCSI Interface) 5 MB/s; parity enabled at select code 14
- HP98643 at select code 21
- ===> HP98550 Bit Mapped Display at select code 132
- SCSI Bus: hard reset (isc: 14)
- HPA1659A Bit-Mapped Display (revision 8.02/10) in SGC slot 0
- real mem = 33546240
- using 473 buffers containing 1937408 bytes of memory
- Root device major is 7, minor is 0xe0600, root site is 0
- Battery-backed real-time clock
- Swap device table: (start & size given in 512-byte units)
- entry 0 - auto-configured on root device; start = 1101184, size = 195328
- Warning: unable to configure dump device...using primary swap instead.
- core image of 8191 pages will be saved at block 615492 on device 0x70e0600
- avail mem = 28954624
- lockable mem = 28852224
-
- (5) I created a character special device /dev/crt2 to
- correspond with the device at the address of 132:
-
- # cd /dev
- # ls -la *crt*
- crw-rw-rw- 1 root root 12 0x000300 Jan 28 1992 crt
- crw-rw-rw- 1 root sys 12 0x000300 Mar 27 1992 crt1
- crw-rw-rw- 1 root sys 12 0x840200 Jan 21 15:36 crt2
- crw-rw-rw- 1 root other 12 0x000001 Apr 12 1991 ocrt
-
- The minor number of the /dev/crt2 device would appear correct
- and jives with the recommended number in table 4-40 (p. 4-138) of
- Installing Peripherals Vol 2.
-
- (6) I created the following entry in /usr/lib/X11/X0screens:
-
- /dev/crt
- /dev/crt2
-
- (7) I added the following line to the Vuesession file
- mentioned above:
-
- # cd /usr/lib/X11/app-defaults
- # more Vuesession
- Vuesession*wmStartupCommand: /usr/bin/X11/vuewm -multiscreen
-
- (8) I rebooted the machine...
-
- Initially, the login processes get started as usual:
-
- root 179 1 0 15:43:55 ? 0:00 /usr/lib/X11/vue/etc/vuelogin
- root 181 179 0 15:43:57 ? 0:00 /usr/lib/X11/vue/etc/vuelogin
- root 193 181 0 15:44:05 ? 0:01 vuegreet -display lepew:0
-
- After logging in, a vuewm is started in -multiscreen as
- expected. But the monitor connected to the new framebuffer shows
- nothing but random white lines.
-
- Interestingly, there is some evidence that X11 is trying
- to communicate with the second monitor device. As vuewm initializes
- the noise on the second monitor changes slightly (still noise but
- noticably different). Hence, there seems to be some communication with
- the second screen device though as yet things still are not working
- right...
-
- Also interesting is when I reset the DISPLAY variable in a
- window on the primary monitor to be HOSTNAME:0.1 (eg.,
- DISPLAY=lepew:0.1) and then start an xterm, there is again a
- corresponding change in the noise pattern on the second device. This
- all makes me think that I am getting close... That is, I can induce
- small changes in the noise pattern on the second monitor device from
- X11. Unfortunately, I can not get much more to happen.
-
- Is is possible that the /usr/lib/X11/X0screen file needs some
- options on the line corresponding to the /dev/crt2 device? The
- original /dev/crt has no such options and since both monitors are
- identical I assumed none were necessary.
-
- Note, both monitors are A1097C and both are in good working
- order (when I swap devices on the two controllers, they work the same
- [although the monitor on the second controller still does no work
- right, as mentioned above]).
-
- So, any kind soul have some suggestions as to what I am missing here?
- Any pointers would be greatly appreciated.
-
- Many thanks,
-
- - Joe
-
- ---------------------------------------------------------------
- Joe Bush bush@fremont.ssc.gov
- System Administrator Office: 214-708-3497
- Superconducting Super Collider Laboratory
- 2550 Beckleymeade Ave, MS 4005, Dallas, TX. 75237
-