02-tests/task 01 Release 2016.10 - RC1 about release-specs HOT 22 CLOSED

Ouput for mutex_order:

main(): This is RIOT! (Version: 2017.01-devel-hedon-2016.10-branch)
Mutex order test
Please refer to the README.md for more information

T3 (prio 6): locking mutex now
T4 (prio 4): locking mutex now
T5 (prio 0): locking mutex now
T6 (prio 2): locking mutex now
T7 (prio 1): locking mutex now
T3 (prio 6): unlocking mutex now
T5 (prio 0): unlocking mutex now
T7 (prio 1): unlocking mutex now
T6 (prio 2): unlocking mutex now
T4 (prio 4): unlocking mutex now

Test END, check the order of priorities above

from release-specs.

Compile error for cpp11_condition_variable, cpp11_mutex and cpp11_thread:

/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.9/libstdc++.so when searching for -lstdc++
/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.9/libstdc++.a when searching for -lstdc++
/usr/bin/ld: cannot find -lstdc++
collect2: error: ld returned 1 exit status

Any idea? SOLVED!

from release-specs.

Compile error for cpp11_condition_variable, cpp11_mutex and cpp11_thread:

/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.9/libstdc++.so when searching for -lstdc++
/usr/bin/ld: skipping incompatible /usr/lib/gcc/x86_64-linux-gnu/4.9/libstdc++.a when searching for -lstdc++
/usr/bin/ld: cannot find -lstdc++
collect2: error: ld returned 1 exit status

Any idea?

You need to install that library for 32-bit. But I don't really remember how it is called in Ubuntu (or Arch). @josephnoir can you help?

About mutex_order I think the locking order isn't that important, but the unlocking order is (since that is where the priority comes into play). Judging from the order of the priorities in that test's main.c the order of the priorities on locking in your output make more sense than the ones in the README. @OlegHahm @kaspar030 what do you think?

from release-specs.

You need to install that library for 32-bit. But I don't really remember how it is called in Ubuntu (or Arch). @josephnoir can you help?

@miri64 , thanks for your suggestion. This is the error after I installed libstdc++-4.9-dev for the i386 architecture.
But I had to install lib32stdc++-4.9-dev for the amd64 architecture.
Multiarch is hilarious. :P

from release-specs.

The README of this test is definitely wrong. I prepare a fix.

from release-specs.

@A-Paul I can't reproduce the failed od test (I have pexpect==4.0.1 installed, older versions were always a bit iffy). Did you install from ubuntu repository or using pip? If the first is the case, please use the pip version.

from release-specs.

Regarding c++ problems, I have a similar error:

arm-none-eabi-g++ \
        -DRIOT_FILE_RELATIVE=\"tests/cpp11_condition_variable/main.cpp\" \
        -DRIOT_FILE_NOPATH=\"main.cpp\" \
        -Werror -mcpu=cortex-m3 -mlittle-endian -mthumb -mfloat-abi=soft -mno-thumb-interwork -ffunction-sections -fdata-sections -fno-builtin -fshort-enums -ggdb -g3 -Os -fno-delete-null-pointer-checks -Werror=strict-prototypes -Werror=old-style-definition -fno-common -Wall -include '/home/oleg/git/RIOT/tests/cpp11_condition_variable/bin/cc2538dk/riotbuild/riotbuild.h' -std=c++11 -isystem /usr/arm-none-eabi/include/nano -isystem /usr/arm-none-eabi/include -I/home/oleg/git/RIOT/core/include -I/home/oleg/git/RIOT/drivers/include -I/home/oleg/git/RIOT/sys/include -I/home/oleg/git/RIOT/cpu/cc2538/include -I/home/oleg/git/RIOT/boards/cc2538dk/include -I/home/oleg/git/RIOT/cpu/cortexm_common/include -I/home/oleg/git/RIOT/sys/libc/include -I/home/oleg/git/RIOT/sys/cpp11-compat/include  -MD -MP -c -o /home/oleg/git/RIOT/tests/cpp11_condition_variable/bin/cc2538dk/cpp11_condition_variable/main.o /home/oleg/git/RIOT/tests/cpp11_condition_variable/main.cpp
In file included from /usr/arm-none-eabi/include/c++/6.2.0/ext/string_conversions.h:41:0,
                 from /usr/arm-none-eabi/include/c++/6.2.0/bits/basic_string.h:5402,
                 from /usr/arm-none-eabi/include/c++/6.2.0/string:52,
                 from /home/oleg/git/RIOT/tests/cpp11_condition_variable/main.cpp:21:
/usr/arm-none-eabi/include/c++/6.2.0/cstdlib:75:25: fatal error: stdlib.h: No such file or directory
 #include_next <stdlib.h>
                         ^

However:

`# ls -lah /usr/arm-none-eabi/include/c++/6.2.0/stdlib.h
-rw-r--r-- 1 root root 2.2K Aug 31 19:52 /usr/arm-none-eabi/include/c++/6.2.0/stdlib.h

from release-specs.

Shouldn't this rather go to #26? These here are native tests ;-)

from release-specs.

Yepp.

from release-specs.

@A-Paul You solved your C++ problem, right?
@OlegHahm I can compile the cpp11_condition_variable test without issues for the samr21-xpro using compiler version 5.4.1. However, my stdlib.h file has a different location:

$ find . -name "stdlib.h"
arm-none-eabi/include/c++/5.4.1/tr1/stdlib.h
arm-none-eabi/include/machine/stdlib.h
arm-none-eabi/include/stdlib.h

Where did you get the 6.2.0 version?

from release-specs.

@josephnoir, let's continue in #28.

from release-specs.

Hi @josephnoir, thanks for stopping by...

@A-Paul You solved your C++ problem, right?

Yes. Turns out I didn't have a problem, I have been the problem. :-)

from release-specs.

@A-Paul I hope #27 (comment) did not get lost in the other comments ;-)

from release-specs.

@miri64,

Did you install from ubuntu repository or using pip?

Neither. Debian jessie (python-pexpect 3.2-1). But I declare you as trustworthy and take this statement:

@A-Paul I can't reproduce the failed od test

to set the test to success.

from release-specs.

You should be able to install pexpect 4 with pip (python-pip in debian repository ;-))

from release-specs.

xtimer_drift : FAILED stops after ~30-90 seconds (two different PCs)

confirmed.

from release-specs.

@A-Paul I hope #27 (comment) did not get lost in the other comments ;-)

see RIOT-OS/RIOT#6051

from release-specs.

@A-Paul I hope #27 (comment) did not get lost in the other comments ;-)

see RIOT-OS/RIOT#6051

Those two don't seem to be related :P

from release-specs.

see RIOT-OS/RIOT#6051

@OlegHahm output and README.md correlate now. The python script isn't very helpful now. But that's not an error and can be improved later.

from release-specs.

Discussing PR not in the PR is not very fruitful ;-).

from release-specs.

@miri64 my fault. I thought it is already merged and just wanted to give feedback.
I've left a comment in the PR now. But the main issue is resolved.

from release-specs.

superseded by #30.

from release-specs.