Return to BSD News archive
Received: by minnie.vk1xwt.ampr.org with NNTP id AA1081 ; Tue, 23 Feb 93 14:25:26 EST Newsgroups: comp.unix.bsd Path: sserve!manuel.anu.edu.au!munnari.oz.au!spool.mu.edu!decwrl!pacbell.com!UB.com!quack!dfox From: dfox@quack.sac.ca.us (David Fox) Subject: [386bsd] npx-0.1 causes lockup in npxprobe() - now what? Message-ID: <fZaiZXG@quack.sac.ca.us> Keywords: npx cyrix npxprobe ndp Organization: The Duck Pond public unix: +1 408 249 9630, log in as 'guest'. Date: 11 Feb 1993 08:38:16 UTC Lines: 60 Hi *, Sometime back, I complained (probably in e-mail) about spurious npxintr() panics on my machine after doing relatively heavy work in one or two terminal windows. Specifically, I was running rayshade in one terminal window, ghostscript was in another terminal window, sending a document to the printer. When done I was just at the command prompt doing simple things like cleaning up temp files etc., and bang! - a npxintr() panic. Following the complaint, Amancio Hasty sent me someone's npx-0.1 patches, which I applied. However, applying the patch to npx.c caused the system to lock up hard during the boot routine, just before the "npx0 at irq 13 on isa" confirmation message. Subsequent postings (which I replied to by email) suggested I could comment out the fp_divideby0() and/or the npxprobe() routine. I tried the first, didn't want to do the second, but did comment out a small section where it was looking for irq16 (this applies only for 486s which I don't have, it seems). The result after doing this was very good stability in the numerics department. I couldn't make my earlier heavy NDP use crash the system. However, a program 'fpetest' supplied with the patch crashed the system hard (only reset button worked) and in subsequent checks things like 'od -f ' on executable files caused the system also to crash hard. It seems that the npxprobe() stuff needs to be intact if certain programs like fpetest are going to work, so I can't exactly comment it out. Either that, or I haven't commented out the proper sections. Also, I am starting to recompile the kernel with gcc-2.33. Using the patches recently posted in comp.unix.bsd, there's one for npx.c which won't work with the npx-0.1 patched npx.c, mostly because the asm instructions are in #define macros rather than coded in-line. The npx.c compiles fine with gcc-2.33, but the patch (whether in-line or in macros) does look different for gcc-2.33 vs. gcc-1.39. I'm wondering if the asm macros need to be changed for gcc-2.33 in npx.c, essentially, and I don't understand the syntax for encoding asm directly in gcc at all, unfortunately. ------------------------------------------------------------------------ in real life: david fox email: dfox@quack.sac.ca.us 5479 Castle Manor Drive, San Jose, CA 95129 "Thanks for letting me change some USA magnetic patterns on your HD" -me 408-253-7992 "You've just won an absolutely FREE trip to the Isles of Langerhans" -me ------------------------------------------------------------------------ -- ------------------------------------------------------------------------ in real life: david fox email: dfox@quack.sac.ca.us 5479 Castle Manor Drive, San Jose, CA 95129 "Thanks for letting me change some USA magnetic patterns on your HD" -me 408-253-7992 "You've just won an absolutely FREE trip to the Isles of Langerhans" -me ------------------------------------------------------------------------