Bug 17661

Summary: 4.0-REL hangs after detecting ata1
Product: Base System Reporter: gandalf <gandalf>
Component: i386Assignee: Søren Schmidt <sos>
Status: Closed FIXED    
Severity: Affects Only Me    
Priority: Normal    
Version: 4.0-RELEASE   
Hardware: Any   
OS: Any   

Description gandalf 2000-03-29 14:10:01 UTC 
AMD K6-2/300, 128 MB Ram, ASUS Board (dont know what type)
Kernel detects atapci "Acer Labs Aladdin ATA 33 Controller", Port 0xb400-0xb40f IRQ 0  (???) dev 15.0 on pci0
ata0: at 0x1f0 irq 14 on atapci0
ata1: at 0x170 irq 15 on atapci0
And here the machine hangs. CTRL-ALT-DEL wont work.
(hardware connected: ata0: maxtor 20gig hdd, ata1: cdrom, disabling ata1 in bios doesnt help)

How-To-Repeat: Just boot from 4.0-REL CD (iso image) or floppy
Comment 1 Sheldon Hearn freebsd_committer freebsd_triage 2000-03-29 14:44:33 UTC 
Responsible Changed
From-To: freebsd-bugs->sos

Over to the ata maintainer.
Comment 2 Andre.Grosse.Bley 2000-03-30 12:59:45 UTC 
It's an ASUS P5A board. Funny: My father has an ASUS P5A, too. and on
his machine the boot does complete. It's even the same IDE Controller
revision - but he has 2 devices on both channels (as far i unterstand
the driver the controller cant use UDMA then). I suspect UDMA
initialisation responsible for this problem, i'll try to find out on
monday. Sorry, no time before...

    Andre
Comment 3 Andre.Grosse.Bley 2000-04-18 16:52:44 UTC 
3.4 runs fine on this, as long i dont enable DMA for wdc0/master. System
boots, but gives something like "dma error 5: dma active" on every disc access.
(sorry, forgot to note the correct error message)
Comment 4 Andre.Grosse.Bley 2000-05-15 10:25:39 UTC 
Ha!
Problem solved. After a long search i found out that a NE 2000 at 0x340
(and 0x300) hangs the system.
Win 2k had problems with 0x340, too, but works with 0x300. It looks for
me like that the PNP read probe
(Read_port) is misunderstood by the ne2k and blocks the isa bus. But i
cant imagine why it even in bootverbose hangs after ata1 ....

    Andre
Comment 5 Søren Schmidt freebsd_committer freebsd_triage 2000-11-14 08:21:20 UTC 
State Changed
From-To: open->closed

Hardware conflict.