home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.sys.att
- Path: sparky!uunet!uunet.ca!geac!torag!robohack!woods
- From: woods@robohack.UUCP (Greg A. Woods)
- Subject: 3b2 EPORTS device driver bugs
- Organization: Elegant Communications Inc.
- Summary: CLOCAL vs. DCD and proper modem control functionality
- Message-ID: <1993Jan4.062531.19906@robohack.UUCP>
- Date: Mon, 4 Jan 93 06:25:31 GMT
- Lines: 76
-
- I'm running (among other things) two modems on an EPORTS card plugged
- into a 3b2/400 with UNIX SysV Rel. 3.1, and version 1.3 of the EPORTS
- drivers.
-
- Before I lost my AT&T 2224 modem, and other than the traditional
- hardware flow control problems, I'd been quite happy with this setup.
-
- However, as some of you may recall, I posted a note about a problem
- which began to plague my Hayes 2400 and Telebit T-1000 some time ago.
-
- The symptoms were a hung modem, and on the Hayes, the appearance of
- RD, SD, TR, and MR being on continuously.
-
- Since I'm too poor to afford a fancy RS-232 protocol analyser, or even
- a nice logic analyser, and my multi-channel scope is dead, I had no
- idea what was actually happening to the ports.
-
- The only solution to this lockup is to either cycle the power on the
- affected modem, or to re-pump the firmware to the EPORTS card.
-
- Eventually I discovered that running the modems with DCD in what I
- call "interrupt" mode (i.e. S131=3 on the T1000, or AT&C3 on other
- conventional modems) effectively solved the problem. (The Hayes
- SmartModem 2400 couldn't do this, but since I only use it to dial out,
- I can use AT&C0 for it.)
-
- Recently I began experimenting with a new scheme for multi-class (link
- layer protocol) dialer control for modern modems using entries like
- these in the Devices file:
-
- ACU tty32,M - 19200 directM "" tb1init "" clocaloff "" tb1PEP "" tb1td \D
- ACU tty32,M - 2400 directM "" tb1init "" clocaloff "" tb124 "" tb1td \D
- ACUnC tty32,M - 2400 directM "" tb1init "" clocaloff "" tb124 "" tb1nC "" tb1td \D
- ACU tty32,M - 1200 directM "" tb1init "" clocaloff "" tb112 "" tb1td \D
- ACUkerm tty32,M - Any directM "" tb1init "" clocaloff "" tb1KERM "" tb1td \D
-
- And like these in the Dialers file:
-
- directM "" "" \M\c
- clocaloff "" "" \m\c
- tb1init "" "" A\pA\pATZ OK-A\pA\pATZ-OK ATQ0E1 OK-ATQ0E1-OK
- tb1td =W-, "" \EATDT\T\r\e\c CONNECT-\c-CONNECT-\c-CONNECT
- tb1pd =W-, "" \EATDP\T\r\e\c CONNECT-\c-CONNECT-\c-CONNECT
- tb112 "" "" ATS50=2 OK-ATS50=2-OK
- tb124 "" "" ATS50=3 OK-ATS50=3-OK
- tb1PEP "" "" ATS50=255 OK-AS50=255-OK
- tb1nMNP "" "" ATS95=0 OK-ATS95=0-OK
- tb1nC "" "" ATS96=0 OK-ATS96=0-OK
- tb1UUCP "" "" ATS111=30 OK-ATS111=30-OK
- tb1KERM "" "" ATS111=13 OK-ATS111=13-OK
-
- My original single-entry dialer had not used '\m', but since this is
- the secret to having cu(1) say "disconnected" when the modem drops
- DCD, I put it back in the dialer sequence.
-
- Since that time, every (successful?) call out results in a subsequent
- (but random?) series of locked up modems. Eventually the frequency of
- lock-ups is reduced. (So I've again removed the clocaloff invocation,
- and have gone back to ensuring I use '~.' to end calls).
-
- I should also note that if the board is re-pumped, it seems no
- subsequent lock-up's occur.
-
- [Seems to be a plot against proper modem control and use of the modem
- control features in HDB-UUCP (i.e. BNU)!]
-
- So, does anyone have any clues? Has anyone else experienced these
- problems? Does anyone have access to source for the EPORTS drivers
- and firmware, and perhaps the ability to get this fixed and release
- the fix? While you're at it, could the hardware flow control be fixed
- such that it can be permanently enabled at boot time for a given port?
- --
- Greg A. Woods
-
- woods@robohack.UUCP, woods@Elegant.COM VE3TCP UniForum Canada & ECI
- +1 416 443-1734 [home] +1 416 362-XRSA [work] Toronto, Ontario; CANADA
-