Date: Sat, 18 Jul 1998 15:41:22 +0200 From: Eivind Eklund <eivind@yes.no> To: hackers@FreeBSD.ORG Subject: Weird bug strikes in _NPX_ when adding PnP support to i4b Message-ID: <19980718154122.53467@follo.net>
next in thread | raw e-mail | index | archive | help
I'm slightly stuck here. I have done "draft code" (not nicely enough
integrated, but should be correct) to add PnP support to i4b. This
work fine for one card, but when I use two PnP cards, npx (of all
weird things) crash here:
* 16 works.
*/
printf("EE: npxprobe1 checks for exception 16\n");
control &= ~(1 << 2); /* enable divide by 0 trap */
fldcw(&control);
npx_traps_while_probing = npx_intrs_while_probing = 0;
printf("EE: npxprobe1 checks for divide by 0\n");
fp_divide_by_0();
if (npx_traps_while_probing != 0) {
/*
* Good, exception 16 works.
*/
printf("EE: exception 16 OK\n");
npx_ex16 = 1;
The first two printfs get through, before the third the machine
spontaneously reboot (and it sporadically reboot long before NPX, but
I assume this is related to the same root cause).
I don't have much of an i386-specific clue; I'd appreciate it if
anybody have ideas of what could cause it. I'll probably track it
down eventually, but would like to get it out of the world as fast as
possible...
Eivind.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19980718154122.53467>