Re: [PATCH RESEND] initramfs: cleanup incomplete rootfs

[David Engraf <david.engraf@sysgo.com>]

On 12.02.19 at 11:43, Andy Shevchenko wrote:
> On Mon, Feb 11, 2019 at 2:40 PM David Engraf <david.engraf@sysgo.com> wrote:
>> On 11.02.19 at 12:40, Andy Shevchenko wrote:
>>> On Mon, Feb 11, 2019 at 10:49 AM David Engraf <david.engraf@sysgo.com> wrote:
>>>> On 11.02.19 at 08:56, David Engraf wrote:
>>>>> On 09.02.19 at 11:35, Andy Shevchenko wrote:
>>>>>> On Sat, Feb 9, 2019 at 12:08 AM Andrew Morton
>>>>>> <akpm@linux-foundation.org> wrote:
>>>>>>> On Fri, 8 Feb 2019 21:45:21 +0200 Andy Shevchenko
>>>>>>> <andy.shevchenko@gmail.com> wrote:
>>>>>>>> On Tue, Oct 30, 2018 at 5:22 PM David Engraf
>>>>>>>> <david.engraf@sysgo.com> wrote:
> 
>>>>>> In my case I have got "Junk in compressed archive". I don't know (I
>>>>>> would check if needed) which exact condition I got  since there are
>>>>>> three places with this message. The file itself smaller than the size
>>>>>> passed through bootparam. So, when decomression is finished
>>>>>> (successfully!) we still have a garbarge in the memory which is not
>>>>>> related to archive. Message per se is okay to have, though I consider
>>>>>> this non-fatal.
>>>>>
>>>>> I can reproduce this special case. The unpacking decompresses the whole
>>>>> size instead of checking the archive size. I will have a look how to get
>>>>> the real archive size.
>>>>
>>>> I did some checks and manually increased the initramfs size but I always
>>>> get the following kernel panic:
>>>
>>> We need to be on the same page here.
>>> There are two sizes of initramfs compressed archive:
>>>    1) actual file size;
>>>    2) what is declared by boot loader and provided via boot parameters.
>>>
>>> In my case I have the 2) bigger than the actual file size.
>>> Kernel decompresses the initramfs, prints an error that there is junk,
>>> which is understandable and continues to run init, etc.
>>
>> Ok got it. When the memory behind the actual file size is clear (0x0)
>> the decompression doesn't complain and just ignores the padding. Any
>> other data will be interpreted as a new archive and thus you'll see the
>> error message.
> 
> Correct.
> 
>> Is it possible for you to fill the padding after the actual file size
>> with 0x00 ?
> 
> Not sure. This is boot loader realm. Even if I patch U-Boot, not every
> boot loader will guarantee this.
> So, it's fragile to rely on data being 0x00 after actual archive.

The problem is that the kernel expects another archive if there are data 
left. If these data do not contain a valid magic the kernel prints an 
error message which is correct.

I could make this error not critical and keep the rootfs, but it's still 
an error and unexpected. You're using a modified bootloader which 
reports a size larger than the file itself. Other bootloader will use 
the file size and report the correct size to the kernel. So this 
workaround is required by your setup only.

@Andrew: What do you think about that? Shall I create a workaround for 
the special case?

Best regards
- David