Re: Broken initrd compression settings in 3.13

From: Jeff Layton
Date: Fri Dec 20 2013 - 22:11:27 EST


On Fri, 20 Dec 2013 18:15:29 -0800
Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:

> On Fri, Dec 20, 2013 at 5:39 PM, Andrew Morton
> <akpm@xxxxxxxxxxxxxxxxxxxx> wrote:
> > On Fri, 20 Dec 2013 16:41:43 -0800 Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
> >>
> >> (a) most sane people don't even have lz4 _installed_, so dracut won't
> >> actually succeed
> >>
> >> (b) there's no way to select the compression level (unlike the
> >> INITRAMFS_COMPRESSION thing that actually has a choice)
> >>
> >> (c) even if you *do* have lz4, it doesn't actually work, because
> >> while that causes the new F20 dracut to compress the initramfs with
> >> lz4, the end result is completely broken, because the F20 "lsinitrd"
> >> scripts don't understand the end result, so now the whole kernel
> >> install fails.
> >>
> >> (a) and (b) are very much kernel bugs.
> >
> > Jeff sent the below this morning. Will that fix (a)?
>
> Yes, it fixes (a), at least to some degree, in that at least
> defaulting to bzip2 is a lot more sane than defaulting to lz4. I
> suspect most everybody has bzip2 installed. And at least on my current
> F20 install, it looks like lsinitrd understands to use zcat, bzcat or
> xzcat on the resulting initrd image (and bzcat does that bzip2
> decoding).
>
> So I think Jeff's patch at least fixes the symptoms.
>
> That said, I think it does nothing *but* fix the symptoms, and we're
> actually still better off with the 3.12 behavior which was to never
> set INITRD_COMPRESS at all. Because quite frankly, there's currently
> no way for the kernel to know what the right compressor is. bz2 may
> well work, but can you guarantee it? I certainly can't..
>
> Now, if we asked the user, that would be a different thing. But right
> now we very much don't ask the user, and we just pick one at random.
>
> We're better off not picking a compression method at all, at which
> point the distro "installkernel" will do whatever the distro does.
>
> Linus

That works for me, you can drop my patch...

But, what's the point of setting INITRD_COMPRESS-y at all if you're not
going to export INITRD_COMPRESS? Would it be better to just remove that
entire block in the Makefile (i.e. just revert PJP's patch?)

--
Jeff Layton <jlayton@xxxxxxxxxx>
--
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/