Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 31 May 95 08:40 CDT
From:      uhclem%nemesis@fw.ast.com (Frank Durda IV)
To:        hackers@FreeBSD.ORG, current@FreeBSD.ORG
Subject:   Problems with 2.0.5-Alpha
Message-ID:  <m0sGo0D-0004w1C@nemesis.lonestar.org>

next in thread | raw e-mail | index | archive | help
I have two existing systems that were running SNAP-0412 without incident
that I have tried to install 2.0.5-Alpha on.  This has been a failure.

The system is a DEC 1027, also known as a GRid 433, also known as a
Tandy 4833, also, well you get the idea.

These are 486DX-33 systems, with external cache modules, 8Meg RAM,
a single 540 IDE drive.  No SCSI, CD-ROMs removed, WD8013 network
card (0x2, 0x280, 0xd400), standard COM1, COM2, LPT1 and nothing else.
OAK or WDC VGA adapters, depending on the machine.

With 8Meg present (what they have had for months), I note on the F2 screen
that dozens of processes are aborting with:

	Process 402 Killed by vm_pageout -- out of swap

The installation appears to be working although many help files appear
missing.  I expected some to be missing, but not all - the true error
may be a killed process rather than a missing file.

Anyway, the installation gets to "Making device files" and just
sits there forever.  I suspect it didn't actually make file systems
as that step reportedly took about one second.

I halted the system and added memory from an identical system, bringing
it up to 12 Meg thinking the additional RAM might avoid the swap 
altogether.  But now the system starts up:

	Booting the kernelel...done	(yes, it is displayed that way)
	BIOS basemem (639K) != RTC basemem (640K)

and it just hangs at this point.

So I tried 4Meg, as the system is supposed to work with just 4Meg.
The system also hangs after the BIOS BASEMEM message.  So I returned
to 8Meg memory size and it boots but by monitoring screen 2 it is obvious
it is killing processes left and right, starting with process 6.

Then I switched to 16Meg.  It also hangs.  But if I move the same
plug-in memory board loaded with the 16Meg on it to the second machine,
that system gets a trap when booting 2.0.5-Alpha:

	Booting the kernel
	BIOS basemem (639K) != RTC basemem (640K)
	Fatal trap 12: page fault while in kernel mode
	fault virtual address = 0x1c
	fault code	      = supervisor read, page not present
	instruction pointer   = 0x8:0xf0186235
	code segment	      = base 0x0, limit 0xfffff, type 0x1b, gran 1
	processor eflags      = resume, IOPL=0
	current process	      = 0 ()
	interrupt mask        = net tty bio
	panic: page fault
	hit reset please

This is 100% reproducible on one system and doesn't seem to happen
on the other.  The only difference was that the machine that paniced
had video shadowing disabled in the BIOS.

Again, both these systems ran 1.1.5.1, and all the SNAPs since February 
and have been used for numerous "make worlds" without incident.  The
system with the panic hasn't had its hard disk wiped yet, so after
the panic, I simply let the system boot the last snap from hard disk
and it ran fine with no hardware changes at all with the 16Meg present.

So with 2.0.5-Alpha, neither system runs at all unless exactly 8Meg
is present and it gets out of swap errors with 8Meg present.


Constructive suggestions are welcome and I'll be happy to try any
experiments to resolve this.

Fastest address is uhclem%nemesis@fw.ast.com

Frank Durda IV <uhclem@nemesis.lonestar.org>|"The Knights who say "LETNi"
or uhclem%nemesis@fw.ast.com (Fastest Route)| demand...  A SEGMENT REGISTER!!!"
...letni!rwsys!nemesis!uhclem               |"A what?"
...decvax!fw.ast.com!nemesis!uhclem         |"LETNi! LETNi! LETNi!"  - 1983




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?m0sGo0D-0004w1C>