Re: [Qemu-devel] [PATCH 1/4] configure: keep track of Python version

[Cleber Rosa]

On 11/9/18 1:25 PM, Eduardo Habkost wrote:
> On Fri, Nov 09, 2018 at 11:39:32AM -0500, Cleber Rosa wrote:
>>
>>
>> On 11/9/18 10:49 AM, Eduardo Habkost wrote:
>>> On Fri, Nov 09, 2018 at 10:07:07AM -0500, Cleber Rosa wrote:
>>>> Some functionality is dependent on the Python version
>>>> detected/configured on configure.  While it's possible to run the
>>>> Python version later and check for the version, doing it once is
>>>> preferable.  Also, it's a relevant information to keep in build logs,
>>>> as the overall behavior of the build can be affected by it.
>>>>
>>>> Signed-off-by: Cleber Rosa <address@hidden>
>>>> ---
>>>>  configure | 6 +++++-
>>>>  1 file changed, 5 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/configure b/configure
>>>> index 74e313a810..67fff0290d 100755
>>>> --- a/configure
>>>> +++ b/configure
>>>> @@ -1740,6 +1740,9 @@ if ! $python -c 'import sys; 
>>>> sys.exit(sys.version_info < (2,7))'; then
>>>>        "Use --python=/path/to/python to specify a supported Python."
>>>>  fi
>>>>  
>>>> +# Preserve python version since some functionality is dependent on it
>>>> +python_version=$($python -V 2>&1 | sed -e 's/Python\ //')
>>>
>>> What about:
>>>   $($python -c 'import sys;print(sys.version)')
>>> ?
>>>
>>> It is very verbose, but I think that's a good thing.
>>>
>>
>> Well, something like:
>>
>> '3.7.1 (default, Oct 23 2018, 18:19:07) \n[GCC 8.2.1 20180801 (Red Hat
>> 8.2.1-2)]'
>>
>> Doesn't qualify as simply the Python *version*.  The documentation[1]
>> puts it clearly: "A string containing the version number of the Python
>> interpreter plus additional information on the build number and compiler
>> used. This string is displayed when the interactive interpreter is started."
>>
>> I can't see how the compiler used on the Python build, or the build
>> date, can be significant to someone debugging the type of Python code
>> that will be on QEMU.
> 
> No problem to me.
> 
>>
>>>> +
>>>>  # Suppress writing compiled files
>>>>  python="$python -B"
>>>>  
>>>> @@ -5918,7 +5921,7 @@ echo "LDFLAGS           $LDFLAGS"
>>>>  echo "QEMU_LDFLAGS      $QEMU_LDFLAGS"
>>>>  echo "make              $make"
>>>>  echo "install           $install"
>>>> -echo "python            $python"
>>>> +echo "python            $python ($python_version)"
>>>>  if test "$slirp" = "yes" ; then
>>>>      echo "smbd              $smbd"
>>>>  fi
>>>> @@ -6823,6 +6826,7 @@ echo "INSTALL_DATA=$install -c -m 0644" >> 
>>>> $config_host_mak
>>>>  echo "INSTALL_PROG=$install -c -m 0755" >> $config_host_mak
>>>>  echo "INSTALL_LIB=$install -c -m 0644" >> $config_host_mak
>>>>  echo "PYTHON=$python" >> $config_host_mak
>>>> +echo "PYTHON_VERSION=$python_version" >> $config_host_mak
>>>
>>> The output of "python -V" and "sys.version" seems to be meant for
>>> humans, not software.  If we really want something to be used in
>>> conditional makefile rules, I'd prefer to use sys.version_info.
>>> e.g.:
>>>
>>
>> "Python -V" is quite different from "sys.version".  Indeed it includes
>> the "Python " prefix, but that's all, everything else is the version number.
> 
> Is this a guarantee documented somewhere?
> 
>>
>>>   python_major_version="$($python -c 'import 
>>> sys;print(sys.version_info[0])')"
>>>   echo "PYTHON_MAJOR_VERSION=$python_major_version"
>>>
>>>
>>
>> No, I'm actually interested in the other version components, not just
>> the major version.  As I tried to explain in another thread, differences
>> from Python 3.0 to 3.4 are many, and from 3.4 to 3.6 and so on.
> 
> Do you have any examples in mind?  I really doubt we will need to
> look at the Python minor version number inside shell scripts or
> makefiles.
> 

I guess I failed to make myself clear, because it looks like you're
missing my point here.  The situation that I envision is a developer
writes a Python based test, runs it locally, gets a PASS, puts in a
patch, sends it upstream.  It gets executed on a number of different CI
environments.  It fails on one (or some).  What's causing the failure?
The Python *minor* version may be a significant difference.

Like I said before, printing that is the most important functionality at
this point.  It is indeed debatable whether we'll need to keep it
available for make/shell usage.

>>
>> Then, we can either introduce "PYTHON_MAJOR_VERSION",
>> "PYTHON_MINOR_VERSION", "PYTHON_RELEASE" of sorts, or just have a single
>> dot separated version string.
> 
> A dot separated version string would work for me, too.  I don't
> mind the format that much, because I expect the new variable to
> become unnecessary in the next year.  :)
> 

Cool.  Thanks for the feedback.  Just to get this right, what do I need
to change here?

- Cleber.