5.4-STABLE changes breaks IDE boot (was Re: 5.3 -> 5.4 breaks ATA (Intel ICH2))

Julian C. Dunn lists at aquezada.com
Mon Sep 26 21:35:15 PDT 2005


On Sun, 2005-09-25 at 15:36 -0500, Billy Newsom wrote:

> I would like to concur with this sighting. I will repeat it to emphasize 
> this for whoever can decipher this for bug fixes:
> 
> 1. In recent  5-Stable
> 2. The IDE controller, atapci0,
> 3. seems to be detectected, but without a certain memory allocation
> 4. namely, 0xfff0-0xffff
> 5. while the other memory and I/O segments are detected fine.
> 
> Furthermore,
> 6. It won't boot on the affected machines.
> 7. More than one controller is affected (ICH2, PIIX3, and possibly PIIX4)
> 8. Mine, at least, boots fine to a July 4th 5-Stable.
> 
> Here is an example I found from someone reporting a similar issue 
> (Julian Dunn)
> 
> Look at this from his dmesg:
> 
> *First* his good booting dmesg line:
> atapci0: <Intel PIIX4 UDMA33 controller> port 
> 0xfff0-0xffff,0x376,0x170-0x177,0x3f6,0x1f0-0x1f7 at device 2.1 on pci0
> 
> *Second* his non-booting dmesg line:
> atapci0: <Intel PIIX4 UDMA33 controller> port
>   0x376,0x170-0x177,0x3f6,0x1f0-0x1f
> atapci0: Lazy allocation of 0x10 bytes rid 0x20 type 4 at 0
> 7 at device 2.1 on pci0
> 
> Can I point out the fact that these two lines (first from the good, 
> second from the bad dmesg) seem to say that atapci0 is not getting its 
> memory segment detected -- the segment at 0xfff0-0xffff.

Here's a bit more information about the defect. I discovered this
accidentally while testing out a -STABLE kernel built today. When I
booted the -STABLE kernel, it went and autoloaded the
old /boot/kernel/acpi.ko from the -RELEASE kernel, and the system would
boot as normal. However, if I broke to the loader prompt and loaded the
-STABLE kernel and its corresponding acpi.ko, the machine would fail to
boot (as above, with the memory segment not being detected).

Would this help anyone in filing a PR about the issue?

- Julian

-- 
[ Julian C. Dunn <jdunn at aquezada.com>  *  "You can throw confetti,   ]
[ WWW: www.aquezada.com/staff/julian   *  but you're still going     ]
[ PGP: 91B3 7A9D 683C 7C16 715F        *  through the motions, baby" ]
[      442C 6065 D533 FDC2 05B9        *                - Aimee Mann ]


More information about the freebsd-stable mailing list