Return to BSD News archive
Path: sserve!newshost.anu.edu.au!munnari.oz.au!foxhound.dsto.gov.au!fang.dsto.gov.au!myall.awadi.com.au!blymn From: blymn@awadi.com.au (Brett Lymn) Newsgroups: comp.os.386bsd.questions Subject: Re: Slcompress -- enforce HW flow-control, patch Date: 17 Jul 1993 02:43:20 GMT Organization: AWA Defence Industries Lines: 34 Message-ID: <BLYMN.93Jul16204321@mallee.awadi.com.au> References: <Felipe.28.740578609@utopia.hacktic.nl> <C9I9Hs.ABv@news.cso.uiuc.edu> <hastyC9IFM3.F9u@netcom.com> <BLYMN.93Jul6170320@mallee.awadi.com.au> <21f6er$2v7@max.in-berlin.de> <21v57mINN7n@kralizec.zeta.org.au> NNTP-Posting-Host: mallee.awadi.com.au In-reply-to: bde@kralizec.zeta.org.au's message of 14 Jul 1993 06:16:54 +1000 >>>>> On 14 Jul 1993 06:16:54 +1000, bde@kralizec.zeta.org.au (Bruce Evans) said: Bruce> NNTP-Posting-Host: kralizec.zeta.org.au Bruce> In <21f6er$2v7@max.in-berlin.de> berry@max.IN-Berlin.DE (Stefan Behrens) writes: >In article <BLYMN.93Jul6170320@mallee.awadi.com.au> blymn@awadi.com.au (Brett Lymn) writes: [....] Bruce> I use this in /etc/netstart: Bruce> nohup slattach /dev/sio00 38400 >/dev/null Bruce> stty 115200 crtscts </dev/sio00 # correct for deficiencies of slattach Bruce> Add other stty flags as desired. It would be better for slattach to do Bruce> this internally so that it can restore the original settings when it Bruce> exits. A good way for it to do this would be to fork /bin/stty to change Bruce> the settings. I fixed slattach so that it did. I do not think that forking stty is that good an idea, slattach is already playing with the line (setting it into slipdisc, setting the speed...and so on) so logically it should perform the other bits of line frobbing itself rather than using up another process slot just to fix up the line. Bruce> I used to like magic bits in the minor number, but now I think that they Bruce> cause too many problems for device open/close. The vfs layer doesn't Bruce> doesn't know about the magic bits, so it calls the device close routine Bruce> too often. This leads to bugs like keyboard getting killed when something Bruce> opens and closes /dev/vga. Yup, also there is some question where to stop. I mean should we allocate a minor number for each baud rate? :-) (I am NOT serious!!!) -- Brett Lymn