Date: Sun, 15 Feb 2004 15:50:13 -0500 From: Jesse Guardiani <jesse@wingnet.net> To: freebsd-current@freebsd.org Subject: Re: FreeBSD 5.2.1-RC2 available Message-ID: <c0om28$10b$1@sea.gmane.org> References: <402E3611.70804@freebsd.org> <c0mbfr$7kt$1@sea.gmane.org> <20040214234449.GA38808@xor.obsecurity.org>
next in thread | previous in thread | raw e-mail | index | archive | help
Kris Kennaway wrote:
> On Sat, Feb 14, 2004 at 06:37:30PM -0500, Jesse Guardiani wrote:
>
>> I'm having great difficulty with 5.2.1-RC2. I installed from an ISO just
>> this afternoon. Getting the installer running wasn't easy. I tried all of
>> the usual tricks that have worked for me in the past under 5.1-RELEASE
>> and 5.2-RELEASE:
>>
>> - set hw.pci.allow_unsupported_io_range=1
>> - disabling ACPI
>>
>> But none of that helped. Without ACPI, my kernel would panic before
>> starting sysinstall's device detection (somewhere before the keyboard
>> probe). And WITH ACPI it would get further - to the sysinstall device
>> detection screen, but promptly panic. Finally, I got it to install
>> 5.2.1-RC2 by selecting "safe mode" from the loader menu.
>
> We can't do much without at least a transcription of the panic.
>
>> hw.ata.atapi_dma=1
>>
>> Perhaps my hw.ata.atapi_dma line is breaking things?
>
> Certainly possible, because that's non-default and unsupported by a
> lot of hardware.
:) It's supported just fine on my hardware. It worked in 5.1-RELEASE and
5.2-RELEASE. And it's working now in 5.2.1-RC2.
I'm writing this on the 5.2.1-RC2 kernel. I finally got it booting by
manually executing these commands at the loader prompt:
set hw.pci.allow_unsupported_io_range=1
set hint.apic.0.disabled=1
I guess this is OK, but I never had to set hint.apic.0.disabled=1 before.
So I suppose it could be considered a regression.
In addition, APM suspend/resume seems to be even more broken than it was in 5.2-RELEASE.
I can't get the machine to come back from suspend no matter what I do now, whereas
I used to be able to get it to resume if I booted with `boot-vD` in 5.2-RELEASE, and
under 5.1-RELEASE I could get suspend/resume working with both APM *AND* ACPI.
Also, setting ACPI_LOAD="NO" in /boot/loader.conf seems to be broken in 5.2.1-RC2.
The loader will load acpi.ko regardless, unless explicitly told to not load it at
the boot prompt with:
unset ACPI_LOAD
set hint.acpi.0.disabled=1
Thanks.
--
Jesse Guardiani, Systems Administrator
WingNET Internet Services,
P.O. Box 2605 // Cleveland, TN 37320-2605
423-559-LINK (v) 423-559-5145 (f)
http://www.wingnet.net
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?c0om28$10b$1>