[NTLUG:Discuss] ubuntu 6.10 start up question

[Leroy Tennison]

Robert Citek wrote:
> Robert Citek wrote:
>   
>> Here's a script
>> that summarizes how I found the cheatcode:
>>
>> ( tmpdir=$(mktemp -d)
>>   cd $tmpdir
>>   zcat /boot/initrd.img-2.6.15-27-386 |
>>    cpio -i
>>   grep -R cmdline .
>>   rm -rf $tmpdir
>> )
>>
>> As you can see, from the output, /init is not the only script that
>> parses /proc/cmdline.
>>     
>
> Woops.  Forgot to paste the output:
>
> 31417 blocks
> Binary file ./bin/kinit matches
> Binary file ./bin/sh matches
> Binary file ./bin/busybox matches
> Binary file ./lib/libselinux.so.1 matches
> Binary file ./sbin/vgchange matches
> ./scripts/init-top/framebuffer:for x in $(cat /proc/cmdline); do
> ./scripts/init-top/usplash:for x in $(cat /proc/cmdline); do
> ./init:for x in $(cat /proc/cmdline); do
>
> BTW, the above was run on Ubuntu 6.06 with the 2.6.15-27-386 kernel.
> I'm assuming if you adjust the code you can find similar results for
> Ubuntu 6.10.
>
> Regards,
> - Robert
>
> _______________________________________________
> http://www.ntlug.org/mailman/listinfo/discuss
>
>   
If we are looking for 'break' in /proc/cmdline then it's technically a 
kernel parameter although it may not be processed by the kernel or init 
per se (I just tested passing the parameter  'bogus' on the command line 
and it was ignored - /var/log/messages showed it but no error messages 
were produced).  If boot scripts in initrd act on it's presence then 
this is distribution specific.  Did you by chance confirm that it was 
the initrd boot scripts acting on 'break'?