Booting UHCI with FILO

[Joseph Smith]

Hello,
Trying to boot Fedora 12 Live USB. FILO loads the Kernel just fine and 
then Loads the initrd just fine. Then:

Jumping to entry point...

Out of memory while allocating output buffer

-- System halted

Any Ideas?

[Patrick Georgi]

Am 08.06.2010 04:22, schrieb Joseph Smith:
> Hello,
> Trying to boot Fedora 12 Live USB. FILO loads the Kernel just fine and
> then Loads the initrd just fine. Then:
>
> Jumping to entry point...
>
> Out of memory while allocating output buffer
>
> -- System halted
>
> Any Ideas?

That error message is in the linux decompression code and appears when
it can't malloc the necessary memory (32kb), or if it's asked to
decompress to address 0.

I guess that this either means that linux doesn't find usable memory
tables, or that some parts of the linux bootblock data structure are not
filled in correctly.
The latter can happen when the kernel version is too new for the used
linux loader.

Patrick

[Joseph Smith]

On Tue, 08 Jun 2010 08:06:19 +0200, Patrick Georgi 
wrote:
> Am 08.06.2010 04:22, schrieb Joseph Smith:
> > Hello,
> > Trying to boot Fedora 12 Live USB. FILO loads the Kernel just fine and
> > then Loads the initrd just fine. Then:
> >
> > Jumping to entry point...
> >
> > Out of memory while allocating output buffer
> >
> > -- System halted
> >
> > Any Ideas?
>
> That error message is in the linux decompression code and appears when
> it can't malloc the necessary memory (32kb), or if it's asked to
> decompress to address 0.
>
> I guess that this either means that linux doesn't find usable memory
> tables, or that some parts of the linux bootblock data structure are not
> filled in correctly.
> The latter can happen when the kernel version is too new for the used
> linux loader.

Thanks Patrick.
Hmm by linux loader you mean FILO? I was booting this directly from the
FILO prompt. FYI the same OS boots fine from regular hard drive, and the
Live USB image boots fine on another PC.

[Patrick Georgi]

Am 08.06.2010 12:47, schrieb Joseph Smith:
> Thanks Patrick.
> Hmm by linux loader you mean FILO?

The linux loader in FILO, yes.
> FYI the same OS boots fine from regular hard drive, and the
> Live USB image boots fine on another PC.

These approaches have different code taking care of the tables that are
setup.

Patrick

[Joseph Smith]

On Tue, 08 Jun 2010 13:11:08 +0200, Patrick Georgi 
wrote:
> Am 08.06.2010 12:47, schrieb Joseph Smith:
> > Thanks Patrick.
> > Hmm by linux loader you mean FILO?
>
> The linux loader in FILO, yes.

So does this mean there is a bug in FILO?
> > FYI the same OS boots fine from regular hard drive, and the
> > Live USB image boots fine on another PC.
>
> These approaches have different code taking care of the tables that are
> setup.

Yes booting from hard drive uses the FILO/Grub and booting from Live USB
uses syslinux.

[Stefan Reinauer]

On 6/8/10 12:47 PM, Joseph Smith wrote:
> > That error message is in the linux decompression code and appears when
> > it can't malloc the necessary memory (32kb), or if it's asked to
> > decompress to address 0.
> >
> > I guess that this either means that linux doesn't find usable memory
> > tables, or that some parts of the linux bootblock data structure are not
> > filled in correctly.
> > The latter can happen when the kernel version is too new for the used
> > linux loader.
>
> Thanks Patrick.
> Hmm by linux loader you mean FILO? I was booting this directly from the
> FILO prompt. FYI the same OS boots fine from regular hard drive, and the
> Live USB image boots fine on another PC.

Is this the latest svn libpayload and filo?

Stefan

[Joseph Smith]

On Tue, 08 Jun 2010 13:23:31 +0200, Stefan Reinauer
 wrote:
> On 6/8/10 12:47 PM, Joseph Smith wrote:
> > > That error message is in the linux decompression code and appears when
> > > it can't malloc the necessary memory (32kb), or if it's asked to
> > > decompress to address 0.
> > >
> > > I guess that this either means that linux doesn't find usable memory
> > > tables, or that some parts of the linux bootblock data structure are

not
> > > filled in correctly.
> > > The latter can happen when the kernel version is too new for the used
> > > linux loader.
> >
> > Thanks Patrick.
> > Hmm by linux loader you mean FILO? I was booting this directly from the
> > FILO prompt. FYI the same OS boots fine from regular hard drive, and the
> > Live USB image boots fine on another PC.
>
> Is this the latest svn libpayload and filo?

This was from about a week ago. I know Patrick just made some changes to
UHCI stack, so maybe I will try new version. I was kind of wondering if
that would fix it though? Or, if this is completly different problem.

[Patrick Georgi]

Am 08.06.2010 13:35, schrieb Joseph Smith:
> This was from about a week ago. I know Patrick just made some changes to
> UHCI stack, so maybe I will try new version. I was kind of wondering if
> that would fix it though? Or, if this is completly different problem.

Your problem has nothing to do with UHCI.

And my recent changes shouldn't change behaviour anyway.

Patrick

[Joseph Smith]

On Tue, 08 Jun 2010 14:10:51 +0200, Patrick Georgi
 wrote:
> Am 08.06.2010 13:35, schrieb Joseph Smith:
> > This was from about a week ago. I know Patrick just made some changes to
> > UHCI stack, so maybe I will try new version. I was kind of wondering if
> > that would fix it though? Or, if this is completly different problem.
>
> Your problem has nothing to do with UHCI.
>
> And my recent changes shouldn't change behaviour anyway.

Hmm, then I guess FILO has a bug.