Re: Current (from 11-aug-2011) fails booting - gcc 4.5 related

[thilo <thilo%nispuk.com@localhost>]

Hi All,

I narrowed the problem down to using build.sh (which in turn uses gcc
4.5) causes the kernel to fail.

When I manually compile the kernel it defaults to use the system gcc 4.1.
Using the old 4.1 compiler make the kernel work (asside from some 
compile glitches that I manually tweeked)

Anyone has any ideas as to why or what went wrong when using
build.sh/gcc 4.5 ?

regards thilo


On 8/12/2011 7:49 PM, thilo wrote:
> On 8/12/2011 2:49 PM, Andy Ruhl wrote:
>> On Fri, Aug 12, 2011 at 12:13 AM, thilo <thilo%nispuk.com@localhost> wrote:
>>> Thank's that would be an option if the kernel would get past the
>>> copyright message.
>> It's booting well past the copyright message. It's possible that it
> How can you tell? The last message it prints is "Loading ffs",
> the "Copyright" message neither shows up on the 9600bd serial nor on the
> console.
> (with DDB_ONPANIC it still reboots before printing the Copyright message
> - I verified that).
>> could panic at that point, which is why I asked you about ddb.
>>
>> Your suspicions about acpi could be correct. You might try booting a
>> kernel without it, or maybe disable some or all non essential hardware
> Doesn't boot -12  disable ACPI and SMP ?  - Shouldn't that have the same
> effect as compiling without support?
>
> My next try is to enable KGDB and see if I can attach to the system.
> Is it possible to already attach in the bootstrap loader, or do I have
> to wait for the kernel to setup the remote-gdb interface?
>
> thilo
>> in the bios.
>>
>> Andy

begin:vcard
fn:Thilo Jeremias
n:Jeremias;Thilo
email;internet:thilo%nispuk.com@localhost
tel;work:+49 15782492240
tel;home:+49 4243941633
version:2.1
end:vcard