Subject: 5100 status: volunteer testers?
To: None <port-pmax@netbsd.org>
From: Jonathan Stone <jonathan@DSG.Stanford.EDU>
List: port-pmax
Date: 09/30/1999 11:48:23
Re the recent enquiries about Netbsd on a Decsystem 5100:

The major problem with running 1.4.1 on a 5100 is a glitch in the
cache-flush code. The 5100 and the 5400 use R2000-style cores with
external writebuffers. The write-drain signal is wired onto on CPU
condition pin 3, rather than 0.

The dec_mips1_wbflush() routine in locore_machdep.c does the right
thing for these machines (it loops on "bc3f 1b" instead of "bc0f 1b").
Taking 1.4.1 source, Adding the line

	/* set correct wbflush routine for this motherboard */
	mips_set_wbflush(kn230_wbflush);


to dec_5100_os_init() in sys/arch/pmax/pmax/dec_5100.c, *should* build
a kernel that boots on a 5100. Simon tried putting a full base.tgz,
network config, etc) on a diskless server, and booted a 5100 to
multi-user.

Unfortunately, console input doesn't seem to work, so we can't yet do
installs on a 5100. Is there anyone else with a 5100 who's keen to
help debug this?