Kernel-Package (Bug?) Error 127
 

After compiling a new kernel from 2.6.8-6 named 2.6.8-6i on Raven, i received error 127 before the actual final image stamp...
over and over.

I apt-get update && apt-get upgrade daily
I have two partitions with debian, Raven and Starfire with identical kernels.

This time i created on the bloated desktop partition, RAVEN instead of the minamalist kernel compiling and package testing compiling Starfire. IDENTICAL otherwise.

Raven is child of starfire. Anyway, i had not upgraded starfire last friday.

I needed a new kernel to get lirc working (serial must be modules) even though its not yet.

What was 2.6.8-6i got compiled in starfire as 2.6.8-6 (2.6.8-6i designating -6 debian and initrd already present starfire kernel as well)
----No problems. So I copied the source, remade symlink and unpacked fresh .deb. (Need the source for nvidia, ndiswrapper, captive, lirc ... modules)

Viola, no problems.
I went into starfire, and upgraded ... received kernel-package and two other small packages...tried it same error 127.

Well these funny errors always happen before a new stable pristine kernel release or final debianized kernel release :( or well at least oopses at about 40% of these release dates from 2.6.3 to the present.

Has anyone else experienced this phenomenon or even got an error 127.
I compile every new one.

Oh yes, I'm debian sid (UNSTABLE)

Next time I stop autocleaning so I can go back and replace Kernel Package. Compare version releases to signify bug over and move on. Yup, thats a what Immaaa gonna do.

Learned something.

Please Explain kernel-source updating by Debian, SID
 

As it turns, kernel-package was upgraded this morning on minimalist STARFIRE along with a new kernel-source-2.6.8-7...made new initrd kernel and appended -7i to it and voila! The same thing for RAVEN!

Yeah Debian!

Are the SID debianized kernel's re-releases (for instance 2.6.8-1 ==> 2.6.8-7) include among various patches, things like bug-fixes or rather are they shots of release candidates of the next kernel and kernel packages represent that trend?

Or am I just too curious and looking at it all wrong. Like part gets updated on Friday and continues on Monday, therefore missing something usually or quick fix/bug! ?

Anyway I'll also pay attention to weekends and release candidate dates.