Comment 31 for bug 1094218

Revision history for this message
Barry Warsaw (barry) wrote : Re: lsb_release crashed with IOError in getstatusoutput(): [Errno 10] No child processes

My first reaction is to ask whether the people experiencing this are using an lsb_release that has the proper -Es flags on its #! line. Looking at the versions reported on the errors.u.c page above, it looks like it is (i.e. precise-updates version 4.0-0ubuntu20.2 as the SRUd patch to the #! line). You can double check this by doing `head /usr/bin/lsb_release` and verifying that it's got the -Es flag.

However, from the original bug description, this is suspicious:

ProcCmdline: /usr/bin/python /usr/bin/lsb_release -a

There's no -Es there. If you were to call lsb_release explicitly like so: `/usr/bin/python /usr/bin/lsb_release -a` then you would be omitting the -Es flag and allowing environmental factors to potentially break it.

In https://bugs.launchpad.net/ubuntu/+source/lsb/+bug/1094218/comments/7 bdmurray posts the script function where lsb_release is called, and it looks like it's calling it via $PATH and "lsb_release", so assuming the -Es flag is in the #! line, that *should* be okay, unless whatever $PATH is involved is finding a different -- and broken -- lsb_release. One thing you could try there is replacing `lsb_release` with `/usr/bin/lsb_release` in that function and see if it makes a difference, assuming of course that your /usr/bin/lsb_release has -Es in its #! line.

I don't know for sure that this is the problem, but let's rule this out before digging in further.