[klibc] sample sleep command killed

[Phil Howard]

On Wed, Jul 14, 2010 at 12:22 PM, maximilian attems <max at stro.at> wrote:
> On Wed, Jul 14, 2010 at 07:34:46AM -0400, Phil Howard wrote:
>> I re-wrote the compile script to build klibc (cleaning up how I do it,
>> and how to find the source tarballs).
>
> why would there be a need for this. this seems just like a waste of
> time.

If you are meaning why do I do it by scripting, I do it because it is
a tremendous savings of time.  One big factor that wastes time doing
it manually is that errors can creep into the process way too easily.

That, and this script will eventually become a component of a larger
setup that builds a complete system from scratch.

Many years ago I tried out "Linux From Scratch".  I scripted it right
from the start.  That allowed me to do more rapid development, which
in that case included porting it to the Sparc architecture, which had
not yet been done, then.  I ended up producing bootable CDs that could
boot on both x86 and Sparc based on LFS and my own patches and
modifications.

But even that wasn't the first time I did so.  The first time I did a
full system build from scratch using a script was in 1980.  Obviously,
it wasn't Linux.


>> It seems to be compiling fine.
>> I tried several of the commands in usr/utils/static and most one.
>> However, the sleep command fails in a strange way (EINVAL on execve):
>
> why doesn't this surprise me!? works just fine here
> ./usr/utils/static/sleep 0.1

Other programs work.  I posted to ask for insight to investigate.  But
you haven't offered any that seems plausible.  I know it can't be
because the process is scripted.  It most likely has to be some option
entered into the script error, or read from this mailing list in
error.

Of course, once I find out what the error is, I can make the
appropriate change.  If it's how my script does the build, then I can
change it once and not have to worry about that mistake again.  But if
I do it manually, mistakes can happen each time ... new ones and old
ones.

You can see from answers to my past posts that just doing it as
described in the README file doesn't work.


> axe your wrong built and do it the recommended way.

Can you put the recommended way into some kind of complete document?