Re: Loading initrd above 4G causes freeze on boot

[Mantas MikulÄnas]

On Sat, Aug 9, 2014 at 10:23 PM, Matt Fleming <matt@xxxxxxxxxxxxxxxxx> wrote:
>
> On Sat, 09 Aug, at 09:44:55AM, Yinghai Lu wrote:
> > On Sat, Aug 9, 2014 at 7:23 AM, Mantas MikulÄnas <grawity@xxxxxxxxx> wrote:
> > > As of commit 4bf7111f5016 ("x86/efi: Support initrd loaded above 4G"),
> > > the kernel freezes at the earliest possible moment when trying to boot
> > > via UEFI on my Asus laptop. (It still boots via BIOS.)
> > >
> > > If I revert that commit on current master (c309bfa9b481), it boots
> > > correctly again [although I see "setup_efi_pci() failed" being printed].
> > >
> > > (Seems like it freezes when handle_cmdline_files() attempts to read
> > > the last chunk of the initramfs -- the last call to efi_file_read()
> > > never returns. Figuring out why that happens is beyond me, though.)
> > >
> >
> > Matt found that efi_file_read()/firmware has problem that could not do
> > several reading. Please check if reverting that commit have any
> > difference.
> >
> > ---
> > I spent some time playing around with this bug and it appears to be
> > triggered by the read-a-chunk-at-a-time logic in handle_cmdline_files(),
> > (which is itself a bug workaround) introduced here,
> >
> >   commit 2d2da60fb40a
> >   Author: Maarten Lankhorst <m.b.lankhorst@xxxxxxxxx>
> >   Date:   Fri Dec 16 13:30:58 2011 +0100
> >
> >       x86, efi: Break up large initrd reads
>
> Right, I committed a patch to workaround this problem here,
>
>   http://git.kernel.org/cgit/linux/kernel/git/mfleming/efi.git/commit/?h=next&id=1c24a2bef39f041eb578189207240d0457ef0ac3
>
> since on my ASUS machine the combination of breaking up initrd reads
> into chunks and reading into a buffer above 4G actually triggers a whole
> new bug.
>
> Mantas, could you try out the patch referenced above and pass
> "efi=nochunk" on the command line to the EFI boot stub. I'm hoping that
> fixes your problem.

Thanks for the patch, I've just tested it but the problem remains; the
system freezes even when reading the entire initramfs at once.

[However... Whether it's five chunks, four (after making the initrd
smaller), or just one (with your efi=nochunk), I noticed that it
*always* hangs during the *last* read, and the allocated memory always
ends at roughly the same address (0x137fffXXX). I wonder if it's
selecting a memory location already in use, or something like that...]

(The laptop is Asus K52JT, AMI firmware v206, no updates available.)

-- 
Mantas MikulÄnas <grawity@xxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/