FPU detection weirdness

Nathan Bryant (nathan@falcon.destru.com)
Fri, 15 Aug 1997 21:25:46 -0400 (EDT)


Interesting output of "grep 386/387 /var/log/messages":

Aug 10 13:43:48 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 10 15:52:49 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 11 20:31:22 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 12 20:13:53 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 13 20:11:12 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 13 21:39:23 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 13 21:41:34 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 15 19:50:17 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.
Aug 15 22:07:26 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using old IRQ13 error reporting
Aug 15 22:17:16 DeathStar kernel: Checking 386/387 coupling... Ok, fpu using exception 16 error reporting.

Most of the early entries are from davem's pre-2.0.31-2, the later ones
on August 13th are from pre-2.0.31-5, and the last two entries (on the
15th) are both from kernel pre-2.0.31-6. The machine is a uniprocessor
ppro-200.