Yep, the original code was about "cwd". Also, I think now that the logging messages really help, what if we turn them on with the lowest priority possible ?

Cheers,
Alex

On Tue, Jul 14, 2015 at 2:06 PM, Ed Bartosh <ed.bartosh@linux.intel.com> wrote:
Hi Michael,

Thanks for the review.

...

> While you're in localhostbecontroller
>
> We could also get rid of
>
>     def _shellcmd(self, command, cwd = None):
>         if cwd is None:
>             cwd = self.be.sourcedir
>
>         #logger.debug("lbc_shellcmmd: (%s) %s" % (cwd, command))
>         p = subprocess.Popen(command, cwd = cwd, shell=True,
> stdout=subprocess.PIPE, stderr=subprocess.PIPE)
>         (out,err) = p.communicate()
>         p.wait()
>         if p.returncode:
>             if len(err) == 0:
>                 err = "command: %s \n%s" % (command, out)
>             else:
>                 err = "command: %s \n%s" % (command, err)
>             #logger.warn("localhostbecontroller: shellcmd error %s" % err)
>             raise ShellCmdException(err)
>         else:
>             #logger.debug("localhostbecontroller: shellcmd success")
>             return out
>
>
> I believe with:
>
> subprocess.check_output
>
Yep, I've noticed this too. I think the reason for not using
check_output was that it doesn't have cwd parameter.

I'm planning to generalize build controllers, so I'll make much more
changes to this code anyway. Hopefully it will be for good :)
Just discussed this with Alex and he seems to be ok with this.

--
Regards,
Ed
--
_______________________________________________
toaster mailing list
toaster@yoctoproject.org
https://lists.yoctoproject.org/listinfo/toaster



--
Alex Damian
Yocto Project
SSG / OTC