rtems-test error with coverage

Vijay Kumar Banerjee vijaykumar9597 at gmail.com
Sat Nov 24 07:55:57 UTC 2018


On Sat, 24 Nov 2018 at 04:59, Chris Johns <chrisj at rtems.org> wrote:

>
> On 24 Nov 2018, at 10:07 am, Joel Sherrill <joel at rtems.org> wrote:
>
> On Fri, Nov 23, 2018, 4:20 PM Chris Johns <chrisj at rtems.org wrote:
>
>> On 24/11/18 12:50 am, Vijay Kumar Banerjee wrote:
>> > On Fri, 23 Nov 2018 at 16:21, Chris Johns <chrisj at rtems.org
>> > <mailto:chrisj at rtems.org>> wrote:
>> >
>> >     On 23 Nov 2018, at 8:55 pm, Vijay Kumar Banerjee <
>> vijaykumar9597 at gmail.com
>> >     <mailto:vijaykumar9597 at gmail.com>> wrote:
>> >>     On Fri, 23 Nov 2018 at 15:09, Chris Johns <chrisj at rtems.org
>> >>     <mailto:chrisj at rtems.org>> wrote:
>> >>
>> >>         On 23/11/18 7:32 pm, Vijay Kumar Banerjee wrote:
>> >>         > Hi,
>> >>         > I was trying to run coverage with rtems-test on a fresh
>> install, and
>> >>         faced
>> >>         > the following errors. This is strange as everything was
>> working and
>> >>         I can't
>> >>         > figure out what went wrong.
>> >>         > Also, I don't see covoar executable in the covoar/ directory.
>> >>         >
>> >>         > I'm using Python 2.7.15
>> >>
>> >>         Are you sure? The latest version selects the python used.
>> >>
>> >>     `python --version` returns 2.7.15
>> >
>> >     The python command may not be used. It could python3.
>> >
>> >>         If you do ...
>> >>
>> >>          $ export RTEMS_PYTHON_OVERRIDE=python2
>> >>
>> >>         ... does it work?
>> >>
>> >>     Yes, this worked. Now I see
>> >>
>> >>     error: coverage: covoar not found
>> >
>> >     I am not sure about this error but it does mean there maybe some
>> issues with
>> >     python3.
>> >
>> > For some reason, the covoar executable is not installed in the
>> tester/covoar
>> > directory, where the the script is searching for it.
>>
>> OK.
>>
>> > The covoar is installed in the bin directory. Is there any reason that
>> script
>> > searches for covoar in the tester/ instead of the bin directory?
>>
>> Would anyone call 'covoar' directly?
>>
>> If the answer is No then I suggest the waf script is updated to install
>> covoar
>> into the tester directory. There should be examples in the script to do
>> this. I
>> am fine with this happening.
>>
>
> There is no reason why it wouldn't be if any other application or library
> used it.
>
>
> Yes and we can move it if there is a demand. It would need a lot of
> documentation to support it properly.
>
> But 5hey would have to use rtems-tester or a similar framework to drive
> it.
>
>
> Yeap. The tester is a nice simple interface.
>
> For now, feel free to put it in tester/bin
>
>
> OK, we just need a patch. :)
>
> this worked
========================
diff --git a/tester/covoar/wscript b/tester/covoar/wscript
index d447117..bf70bfb 100644
--- a/tester/covoar/wscript
+++ b/tester/covoar/wscript
@@ -125,6 +125,7 @@ def build(bld):
     bld.program(target = 'covoar',
                 source = ['covoar.cc'],
                 use = ['ccovoar'] + modules,
+                install_path = bld.srcnode.abspath() + '/tester/bin',
                 cflags = ['-O2', '-g'],
                 cxxflags = ['-std=c++11', '-O2', '-g'],
                 includes = ['.'] + rtl_includes)

========================

since there's a bin folder being created as a result, do we include bin in
.gitignore ?

> Chris
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/devel/attachments/20181124/7db7702c/attachment-0002.html>


More information about the devel mailing list