Re: Rant about Debian reproducibility environment
- Date: Thu, 1 Mar 2018 18:04:13 +0000
- From: Simon McVittie <smcv@xxxxxxxxxx>
- Subject: Re: Rant about Debian reproducibility environment
On Thu, 01 Mar 2018 at 18:17:20 +0100, Steffen Nurpmeso wrote:
> And, of course, if there is a different kernel version, or
> a different uname(1) output as such, then how could a dumb end
> producer[consumer, the author] like S-nail deal with that? We
> hardwire those attributes into the binary, like many other
> programs do, e.g., "mutt(1) -v" output.
Why is the kernel version on the machine where s-nail was compiled useful
to you? If you're looking for more information about a bug report from
your users, for example if you are concerned that a syscall might have
changed behaviour, the kernel version on the machine where s-nail was
*used* seems far more useful - but you can't know that at compile time,
only at runtime (via uname(2) which is the same system call that uname(1)
Similarly, on 32-bit x86 and ARM systems, the architecture reported by
uname is (unfortunately) variable, because it's more specific than just
the machine architecture (i686 or i586 or armv5te or armv7hl, not just
i386 or arm). Again, if you pick this up at build-time and bake it into
your binary, it's misleading: a Debian binary built on an i686 autobuilder
could be used on an i586 machine, or vice versa. If you want this
information for debugging, the architecture of the machine where s-nail
is running seems a lot more interesting than the architecture where it
happens to have been compiled.
If you want "s-nail --version" or similar to give information about the
machine for debugging purposes, consider using uname(2) instead.
That's what mutt -v does, at least on Debian: you can see this by running
setarch x86_64 --uname-2.6 mutt -v
and noting that the kernel version that was reported changes.
 Actually it can't any more in current Debian, because the oldest
supported x86 instruction set is now i686, but in older releases
this was true.