I ran it without the --coverage flag and got the same qemu.cfg error
although the exit code changed from 2 to 1 which from what I read
amounts to about the same problem, syntax error, invalid path,
something like that.
error: qemu.cfg:81: execute failed: qemu-system-i386 -m 128 -boot b
-hda
/home/cpod/development/rtems/test/rtems-tools/tester/rtems/testing/coverage/rtems-boot.img
-no-reboot -serial null -serial mon:stdio -nographic -append
"--console=com1;boot;" -kernel
/home/cpod/development/rtems/pc386/i386-rtems4.12/c/pc386/testsuites/samples/unlimited/unlimited.exe
1: exit-code:1
warning: switched to dry run due to errors
Interestingly I ran rtems-test with --coverage again and just by
chance added the --coverage flag last. Now the qemu.cfg error
disappeared and all the sample tests took some time and actually ran
but all timed out now instead of all invalid, defaulting to dry-run.
Is this because the default time isn't long enough for my processor to
run or with them all timing out, does this indicate a bigger problem.
I suppose hello.exe test should barely take any time at all to run.
There's some post test run problems with the symbolSets.cfg file for
covoar that were there before that I'll take a look at now. Sample
test output below
cpod@cpod ~/coverage_test $
$HOME/development/rtems/test/rtems-tools/tester/rtems-test \
--log=log_pc386 \
--rtems-bsp=pc386 \
--rtems-tools=$HOME/development/rtems/4.12 \
--coverage \
$HOME/development/rtems/pc386/i386-rtems4.12/c/pc386/testsuites/samples
RTEMS Testing - Tester, 4.12 (b047c7737e9d modified)
Coverage analysis requested
Coverage environment prepared
[ 1/13] p:0 f:0 u:0 e:0 I:0 B:0 t:0 i:0 | i386/pc386: base_sp.exe
[ 3/13] p:0 f:0 u:0 e:0 I:0 B:0 t:0 i:0 | i386/pc386: cdtest.exe
[ 2/13] p:0 f:0 u:0 e:0 I:0 B:0 t:0 i:0 | i386/pc386: capture.exe
[ 4/13] p:0 f:0 u:0 e:0 I:0 B:0 t:0 i:0 | i386/pc386: fileio.exe
[ 6/13] p:0 f:0 u:0 e:0 I:0 B:0 t:2 i:0 | i386/pc386: cxx_iostream.exe
[ 5/13] p:0 f:0 u:0 e:0 I:0 B:0 t:2 i:0 | i386/pc386: hello.exe
[ 7/13] p:0 f:0 u:0 e:0 I:0 B:0 t:3 i:0 | i386/pc386: loopback.exe
[ 8/13] p:0 f:0 u:0 e:0 I:0 B:0 t:6 i:0 | i386/pc386: minimum.exe
[ 9/13] p:0 f:0 u:0 e:0 I:0 B:0 t:6 i:0 | i386/pc386: nsecs.exe
[10/13] p:0 f:0 u:0 e:0 I:0 B:0 t:6 i:0 | i386/pc386: paranoia.exe
[11/13] p:0 f:0 u:0 e:0 I:0 B:0 t:7 i:0 | i386/pc386: pppd.exe
[12/13] p:0 f:0 u:0 e:0 I:0 B:0 t:9 i:0 | i386/pc386: ticker.exe
[13/13] p:0 f:0 u:0 e:0 I:0 B:0 t:9 i:0 | i386/pc386: unlimited.exe
Passed: 0
Failed: 0
User Input: 0
Expected Fail: 0
Indeterminate: 0
Benchmark: 0
Timeout: 13
Invalid: 0
Total: 13
Average test time: 0:00:55.477863
Testing time : 0:12:01.212219
Linking executables to /home/cpod/coverage_test/coverage
Symlinks made
Invalid library path:
/home/development/rtems/pc386/i386-rtems4.12/c/pc386/cpukit/score/libscore.a
Invalid symbol set score. Skipping covoar run.
Invalid library path:
/home/development/rtems/pc386/i386-rtems4.12/c/pc386/cpukit/libfs/libdevfs.a
Invalid symbol set filesystems. Skipping covoar run.
Generating reports
Cleaning workspace up
Coverage analysis finished. You can find results in /home/cpod/coverage_test
On 31 May 2017 at 00:31, Gedare Bloom <ged...@rtems.org> wrote:
> Does it work without --coverage flag?
>
> On Tue, May 30, 2017 at 5:23 PM, Cillian O'Donnell
> <cpodonne...@gmail.com> wrote:
>> Ah nevermind.. just realised its already printing them out.
>>
>> On 30 May 2017 9:50 pm, "Cillian O'Donnell" <cpodonne...@gmail.com> wrote:
>>>
>>> Hi Tanu,
>>>
>>> Thanks for the help, I had been grepping through the files and
>>> backtracking through the definitions. I was just wondering was there a quick
>>> way to just print it out that I didn't know about.
>>>
>>> I also couldn't find the recipe for the bsp but the previous work I'm
>>> continuing on from worked for pc386 and leon2, the command I'm using is the
>>> standard command with an extra --coverage flag thats added in the work I'm
>>> continuing, it definitely worked in 2015.
>>>
>>> $HOME/development/rtems/test/rtems-tools/tester/rtems-test
>>> --rtems-bsp=pc386 \
>>> --log=log_pc386 --coverage \
>>> --rtems-tools=$HOME/development/rtems/4.12 \
>>> $HOME/development/rtems/pc386/i386-rtems4.12/c/pc386/testsuites/
>>>
>>>
>>> On 30 May 2017 at 20:34, Tanu Hari Dixit <tokencol...@gmail.com> wrote:
>>>>
>>>> Hi Cillian,
>>>>
>>>> I usually do a git grep -i "<string>" to find out where possibly a
>>>> macro or variable is used elsewhere in the code. Not the best way, but
>>>> works mostly.
>>>> Also %{qemu_cmd} has been defined here in the same file
>>>>
>>>> (https://devel.rtems.org/browser/rtems-tools/tester/rtems/testing/qemu.cfg#L60).
>>>> I also sometimes use winpdb to see what value a macro has but this is
>>>> a long process to step into each line.
>>>> I wanted to ask what command did you use exactly with rtems-test? I
>>>> couldn't find a recipe for qemu simulation for pc386 in
>>>> rtems-tools.git/tester/rtems/testing/bsps.
>>>>
>>>> Regards,
>>>> Tanu Hari Dixit.
>>>>
>>>> On Wed, May 31, 2017 at 12:42 AM, Cillian O'Donnell
>>>> <cpodonne...@gmail.com> wrote:
>>>> > Hi,
>>>> >
>>>> > I'm trying to run rtems-test for pc386 with the added --coverage tag
>>>> > for
>>>> > couverture-qemu. All tests default to dry-run because of errors in
>>>> > qemu.cfg
>>>> >
>>>> > error: qemu.cfg:81: execute failed: qemu-system-i386 -m 128 -boot b
>>>> > -hda
>>>> >
>>>> > /home/cpod/development/rtems/test/rtems-tools/tester/rtems/testing/coverage/rtems-boot.img
>>>> > -no-reboot -serial null -serial mon:stdio -nographic -append
>>>> > "--console=com1;boot;" -kernel
>>>> >
>>>> > /home/cpod/development/rtems/pc386/i386-rtems4.12/c/pc386/testsuites/samples/unlimited/unlimited.exe
>>>> > -exec-trace coverage/unlimited.exe.cov: exit-code:2
>>>> > warning: switched to dry run due to errors
>>>> >
>>>> > The exit code I believe refers to 'no such file or directory'. In
>>>> > qemu.cfg
>>>> > the line is:
>>>> >
>>>> > %execute %{qemu_cmd} %{qemu_opts} -kernel %{test_executable}
>>>> > %{coverage_arg}
>>>> >
>>>> > It's a bit tricky to trace those macros back to different files and see
>>>> > what
>>>> > they have. Is there any way to just print those out when it runs so I
>>>> > can
>>>> > take a look at them.
>>>> >
>>>> > I tried %print %{qemu_cmd} and %echo %{qemu_cmd}. Is there any way to
>>>> > do
>>>> > this in a config file?
>>>> >
>>>> > Thanks,
>>>> >
>>>> > Cillian.
>>>> >
>>>> > _______________________________________________
>>>> > devel mailing list
>>>> > devel@rtems.org
>>>> > http://lists.rtems.org/mailman/listinfo/devel
>>>
>>>
>>
>> _______________________________________________
>> devel mailing list
>> devel@rtems.org
>> http://lists.rtems.org/mailman/listinfo/devel
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
