Re: Kernel tainted - no version for "loop_unregister_transfer"

Well spotted, thanks Jari:  tr  is just Busybox tr, and yes, it doesn't work as required for extracting non-numeric chars.

I remember now this also caused problems with bz2aespipe.   I even tried writing a Busybox-only version of bz2aepipe but as I recall I found that Busybox head reads from standard input differently to gnu head, so it would only work in one direction without a total rethink, so I gave up and installed tr and head.

Even Busybox systems often have sed (and/or perl), or an alternative might be:

echo abc123def | sed 's/[^0-9]//g'

I'll hack this in and see what happens first.

Jari Ruusu <[email protected]> wrote: Phil H wrote:
> but the Makefile's not working for me.  Get:
> 
> /bin/sh: line 0: [: VERSION = 2PATCHLEVEL = 60: integer expression expected
> 
> repeated a number of times, with no modules at the end.

Above sounds like 'tr' program is missing or misbehaving. Old Makefile did
not use 'tr' program by default, but new one does. It extracts version
information kernel kernel Makefile using these lines:

VERSION:=$(shell grep '^VERSION' $(LS)/Makefile | head -n 1 | tr -d -c '0-9')
PATCHLEVEL:=$(shell grep '^PATCHLEVEL' $(LS)/Makefile | head -n 1 | tr -d -c '0-9')
SUBLEVEL:=$(shell grep '^SUBLEVEL' $(LS)/Makefile | head -n 1 | tr -d -c '0-9')

That should evaluate to this:

VERSION:=2
PATCHLEVEL:=6
SUBLEVEL:=18

But in your case it appears to evaluate to this:

VERSION:=VERSION = 2
PATCHLEVEL:=PATCHLEVEL = 6
SUBLEVEL:=SUBLEVEL = 18

$(LS) is either detected kernel source directory, or LINUX_SOURCE= specified
kernel source directory. 'tr -d -c '0-9' command is supposed to delete all
non-number characters. Your 'tr' program doesn't do that.

Could you provide more information about your build system? Outputs of these
will help:

    tr --version
    make --version
    /bin/sh --version
    echo abc123def | tr -d -c '0-9'

What distro?
Any strange locale setting in your environment variables?

Next version of loop-AES will include that new Makefile plus some minor
adjustments because of normal kernel drift. Whatever is causing this problem
needs to be sorted out. If other loop-AES users want to test that new
Makefile, now is the perfect time to complain. Private replies are ok if
someone doesn't want to post to public mailing list.

-- 
Jari Ruusu  1024R/3A220F51 5B 4B F9 BB D3 3F 52 E9  DB 1D EB E3 24 0E A9 DD

-
Linux-crypto:  cryptography in and on the Linux system
Archive:       http://mail.nl.linux.org/linux-crypto/



       
---------------------------------
Be a better Globetrotter. Get better travel answers from someone who knows.
Yahoo! Answers - Check it out.