caf@omen.UUCP (Chuck Forsberg WA7KGX) (06/11/89)
Being about to take the plunge into Etnernet, TCP/IP etc., I would like to know which of the popular boards (3c01, wd8003, etc.) support the higher IRQ numbers on AT machines (IRQ8 and higher). The low number IRQs are used up already. Chuck Forsberg WA7KGX ...!tektronix!reed!omen!caf Author of YMODEM, ZMODEM, Professional-YAM, ZCOMM, and DSZ Omen Technology Inc "The High Reliability Software" 17505-V NW Sauvie IS RD Portland OR 97231 503-621-3406 TeleGodzilla:621-3746 FAX:621-3735 CIS:70007,2304 Genie:CAF
clements@bbn.com (Bob Clements) (06/12/89)
In article <782@omen.UUCP> caf@omen.UUCP (Chuck Forsberg WA7KGX) writes: >Being about to take the plunge into Etnernet, TCP/IP etc., I >would like to know which of the popular boards (3c01, wd8003, >etc.) support the higher IRQ numbers on AT machines (IRQ8 and >higher). The low number IRQs are used up already. Of the ones I know about: 3C501 IRQ2-7 Jumper select 3C503 IRQ2-5 Software select WD8003E IRQ2-7 Jumper select None of these will do IRQ8-15. /Rcc K1BC clements@bbn.com
jbvb@VAX.FTP.COM (James Van Bokkelen) (06/12/89)
The only AT-bus cards that hack interrupts greater than 7 are the 3Com 3C505 and the other "16-bit AT-bus" cards (Gateway, IMC and Tiara that I know of, maybe others). 3Com's PCS/TCP, our PC/TCP and NRC Fusion all have 3C505 drivers that I know of, and the other cards I mentioned all have Packet Drivers. The 3C505 is the only card that I know will run in either a 16-bit slot in 16-bit mode, or an 8-bit slot in 8-bit mode. The others come in 16-bit and 8-bit versions. Most Microchannel cards will work on some or all of the 2nd 8259's interrupts, but that wouldn't seem to help you... James B. VanBokkelen 26 Princess St., Wakefield, MA 01880 FTP Software Inc. voice: (617) 246-0900 fax: (617) 246-0901
dougm@ico.ISC.COM (Doug McCallum) (06/13/89)
In article <8906121543.AA25731@vax.ftp.com> jbvb@VAX.FTP.COM (James Van Bokkelen) writes: >The only AT-bus cards that hack interrupts greater than 7 are the 3Com >3C505 and the other "16-bit AT-bus" cards (Gateway, IMC and Tiara that >I know of, maybe others). 3Com's PCS/TCP, our PC/TCP and NRC Fusion While most don't support the signals for the second PIC, IRQ 2 (the IRQ the second PIC is chained off of) can be used and will give an interrupt in the upper range. This gets back an IRQ that people tend to think is missing on an AT. A lot of the 8-bit boards do support IRQ 2. This helped some of our 386/ix customers who thought they were out of IRQs.
acm@RELAY.PROTEON.COM (06/19/89)
Organization: Ontario Ministry of the Environment Message-Id:
<396@moegate.UUCP> Sender: pcip-request@twg.com
--------------------------------------------------------------------------------
I've seen this discussed before. In fact I'm using this hack on one
of my machines. It seems to work but I'm concerned that it may not be
reliable. Can anyone who really understands how the PIC's work
comment on the implications of strapping the hardware to IRQ2 and
calling it IRQ9 in software?
Also it doesn't do you any good on the DOS side of things if you're
using a commercial package, none of them try to do this.
This goes back to when the AT was introduced by IBM. What they did was
expand the PIC to 2 PIC's. In order to do this you must use a cascade of
the first to the second in hardware and in order to do this they took the
int2 connection to do it. In order for software to work on int2 they then
made int9 look like the old int2 of the PC. It is somewhat inefficient
but it does work on most stuff. Thus on an AT or compatible the int9 is
identical to int2. Automatic software discovery of the int connected will
usually get int9 but it is really coded like int2 in software, ie the
vector is at the location of int2.
Hope this helps understand what is going on. You are NOT in any problem
by using int9 in software with int2 in hardware (provided it works for you
in your particular situation as it should with all compatibles.)
BTW the changes are in the hardware bios and there is where the "fix" is
done.
-Alan Marshall, VP Proteon Inc.
2 Technology Drive CCMAIL: acm at proteonwebo
Westboro, MA 01581-5008 ARPANET: acm@Proteon.com
tel: (508)898-2120 MHS: acm @ ProteonWtwf@proteon.com ("Troy W. Frever") (06/19/89)
Having just played with this fairly recently it is still rather fresh in my mind. The interrupt line called IRQ-2 on an 8-bit AT-bus card will tug on the line connected (in hardware) to IRQ-9 on an AT. This is interrupt number 0x71, with vector located at 0x1C4. Since IRQs 9-F are indeed on PIC 2, and, yes the actual interrupt level 2 on PIC 1 is used to cascade the PICs together, you will need to ACK both PICs (just as you would when handling any IRQ 9-F). Pretty much business as usual. Now comes the tricky part. The BIOS installs an interrupt handler for interrupt 0x71 (IRQ-9) which ACKs PIC-2 and issues a software interrupt for int number 0x0A (IRQ-2). Voila! software rediriect - it looks just as though IRQ-2 was pulled, with the vector at location 0x28 being invoked and PIC-1 remaining to be ACKed. In short, the software may either field these interrupts as IRQ-9 (which they are) or let the BIOS installed redirect change them to IRQ-2 and field them that way. This information (although rather obscure) can be found in the IBM PCAT Technical Reference manual. Troy Frever twf@sonny.proteon.com
soley@moegate.UUCP (Norman S. Soley) (07/23/89)
In article <15853@vail.ICO.ISC.COM> dougm@ico.ISC.COM (Doug McCallum,violet,114,) writes: >In article <8906121543.AA25731@vax.ftp.com> jbvb@VAX.FTP.COM (James Van Bokkelen) writes: >>The only AT-bus cards that hack interrupts greater than 7 are the 3Com >>3C505 and the other "16-bit AT-bus" cards (Gateway, IMC and Tiara that >>I know of, maybe others). 3Com's PCS/TCP, our PC/TCP and NRC Fusion > >While most don't support the signals for the second PIC, IRQ 2 (the IRQ the >second PIC is chained off of) can be used and will give an interrupt in >the upper range. This gets back an IRQ that people tend to think is missing >on an AT. A lot of the 8-bit boards do support IRQ 2. This helped some >of our 386/ix customers who thought they were out of IRQs. I've seen this discussed before. In fact I'm using this hack on one of my machines. It seems to work but I'm concerned that it may not be reliable. Can anyone who really understands how the PIC's work comment on the implications of strapping the hardware to IRQ2 and calling it IRQ9 in software? Also it doesn't do you any good on the DOS side of things if you're using a commercial package, none of them try to do this. -- Norman Soley - The Communications Guy - Ontario Ministry of the Environment soley@moegate.UUCP or if you roll your own: uunet!attcan!ncrcan!moegate!soley The Minister speaks for the Ministry, I speak for myself. Got that! Good. Stay smart, go cool, be happy, it's the only way to get what you want