[Mesa-users] Installation failed while testing crlibm module

[Kuldeep Verma]

Dear All,

This is to update you on the issue I described in my first email. With the
help of Rob Farmer (thanks again Rob), I could install mesa-r10108 with
mesasdk released on 2nd August 2017. The installation failed with the exact
same error message reported in my first email for the later versions of
mesasdk (installation went fine with the earlier versions).

This email is to help someone who faces the same issue. I'm closing this
thread now.

Best wishes,
Kuldeep

On Fri, Feb 2, 2018 at 7:51 PM, Kuldeep Verma <kuldeepv89 at gmail.com> wrote:

> Hi Rob,
>
> I wish to thank you very much for your time and help. Not surprisingly
> (after above experiments), the mesa-r10108 was installed successfully with
> the August 2017 version of the mesasdk. For now, I'm happy with what I have.
>
> Best wishes,
> Kuldeep
>
> On Fri, Feb 2, 2018 at 6:09 PM, Kuldeep Verma <kuldeepv89 at gmail.com>
> wrote:
>
>> Thank you for your email. The installation of version 8845 goes fine with
>> the mesasdk released on 2nd August 2017. The installation of mesa 8845
>> fails with the same error message as 10108 for all the mesasdk released
>> afterwards.
>>
>> Best wishes,
>> Kuldeep
>>
>> On Fri, Feb 2, 2018 at 5:26 PM, Rob Farmer <r.j.farmer at uva.nl> wrote:
>>
>>> Thanks every data point is useful, your tests shows its possible to get
>>> mesa working on your system so its unlikely its your system that is somehow
>>> misconfigured, though it may be a combination of some package version you
>>> have with the sdk.
>>>
>>> How about mesa 8845 with the latest sdk? Just trying to narrow down
>>> whether it was a change to mesa or the sdk that stopped things working.
>>>
>>> Rob
>>>
>>> On 2 February 2018 at 15:54, Kuldeep Verma <kuldeepv89 at gmail.com> wrote:
>>>
>>>> Latest mesa with previous mesasdk failed with the exact same error
>>>> message. However, I just dug out the mesa version that I successfully
>>>> installed in the past (8845), and tried installing it now with then
>>>> available mesasdk (January 2016 version), and that worked fine! I'm not
>>>> sure if this can help us with the installation of the latest mesa,
>>>> nevertheless I just thought to share this with you.
>>>>
>>>> ************************************************
>>>> ************************************************
>>>> ************************************************
>>>>
>>>> MESA installation was successful
>>>>
>>>> ************************************************
>>>> ************************************************
>>>> ************************************************
>>>>
>>>>
>>>>
>>>>
>>>> [kuldeep at s81n11 mesa-r8845]$
>>>>
>>>>
>>>> Best wishes,
>>>> Kuldeep
>>>>
>>>> On Fri, Feb 2, 2018 at 3:07 PM, Kuldeep Verma <kuldeepv89 at gmail.com>
>>>> wrote:
>>>>
>>>>> Hi Rob,
>>>>>
>>>>> Thank you for the email. Below is the output:
>>>>>
>>>>> [kuldeep at s81n11 ~]$ locale
>>>>> LANG=en_US.UTF-8
>>>>> LC_CTYPE="en_US.UTF-8"
>>>>> LC_NUMERIC="en_US.UTF-8"
>>>>> LC_TIME="en_US.UTF-8"
>>>>> LC_COLLATE="en_US.UTF-8"
>>>>> LC_MONETARY="en_US.UTF-8"
>>>>> LC_MESSAGES="en_US.UTF-8"
>>>>> LC_PAPER="en_US.UTF-8"
>>>>> LC_NAME="en_US.UTF-8"
>>>>> LC_ADDRESS="en_US.UTF-8"
>>>>> LC_TELEPHONE="en_US.UTF-8"
>>>>> LC_MEASUREMENT="en_US.UTF-8"
>>>>> LC_IDENTIFICATION="en_US.UTF-8"
>>>>> LC_ALL=
>>>>>
>>>>> >>Also just to make sure there isn't something silly we've missed can
>>>>> you try downloading mesa again and also try a different sdk version as well
>>>>> and see if they work.
>>>>> I have already tried downloading mesa several times (both the ZIP file
>>>>> as well as through svn), but ended up with the exact same error message.
>>>>> I'm downloading a fresh copy of mesa again and will try to install it with
>>>>> the previous version of mesasdk. I'll let you know how it goes.
>>>>>
>>>>> Best wishes,
>>>>> Kuldeep
>>>>>
>>>>>
>>>>>
>>>>> On Fri, Feb 2, 2018 at 2:14 PM, Rob Farmer <r.j.farmer at uva.nl> wrote:
>>>>>
>>>>>> Thanks, still looking for the root cause. Can you run:
>>>>>>
>>>>>> locale
>>>>>> in a terminal and send the output
>>>>>>
>>>>>> Also just to make sure there isn't something silly we've missed can
>>>>>> you try downloading mesa again and also try a different sdk version as well
>>>>>> and see if they work.
>>>>>>
>>>>>> Thanks
>>>>>> Rob
>>>>>>
>>>>>> On 1 February 2018 at 17:41, Kuldeep Verma <kuldeepv89 at gmail.com>
>>>>>> wrote:
>>>>>>
>>>>>>> Hi Rob,
>>>>>>>
>>>>>>> Thank you so much for your time. Please find the attached file.
>>>>>>>
>>>>>>> Best wishes,
>>>>>>> Kuldeep
>>>>>>>
>>>>>>> On Thu, Feb 1, 2018 at 5:02 PM, Rob Farmer <r.j.farmer at uva.nl>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> Hi
>>>>>>>>
>>>>>>>> Thanks, can you try this file and repeat the procedure? This isn't
>>>>>>>> a fix, i'm just trying to get some more info on what is happening.
>>>>>>>>
>>>>>>>> The short answer is that mesa is reading in a number as a string,
>>>>>>>> converting that to a numeric value then checking that they are exactly the
>>>>>>>> same, for some reason that is failing for you.
>>>>>>>>
>>>>>>>> Rob
>>>>>>>>
>>>>>>>>
>>>>>>>> On 1 February 2018 at 15:51, Kuldeep Verma <kuldeepv89 at gmail.com>
>>>>>>>> wrote:
>>>>>>>>
>>>>>>>>> Hi Rob,
>>>>>>>>>
>>>>>>>>> Many thanks for looking into it. Please find the attached run.log
>>>>>>>>> file.
>>>>>>>>>
>>>>>>>>> [kuldeep at s81n11 ~]$ uname -a
>>>>>>>>> Linux s81n11.grendel.cscaa.dk 3.10.0-514.2.2.el7.x86_64 #1 SMP
>>>>>>>>> Tue Dec 6 23:06:41 UTC 2016 x86_64 x86_64 x86_64 GNU/Linux
>>>>>>>>>
>>>>>>>>> [kuldeep at s81n11 ~]$ ldd --version
>>>>>>>>> ldd (GNU libc) 2.17
>>>>>>>>> Copyright (C) 2012 Free Software Foundation, Inc.
>>>>>>>>> This is free software; see the source for copying conditions.
>>>>>>>>> There is NO
>>>>>>>>> warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR
>>>>>>>>> PURPOSE.
>>>>>>>>> Written by Roland McGrath and Ulrich Drepper.
>>>>>>>>>
>>>>>>>>> Best wishes,
>>>>>>>>> Kuldeep
>>>>>>>>>
>>>>>>>>> On Thu, Feb 1, 2018 at 1:58 PM, Rob Farmer <r.j.farmer at uva.nl>
>>>>>>>>> wrote:
>>>>>>>>>
>>>>>>>>>> Hi
>>>>>>>>>>
>>>>>>>>>> Thanks for the report, could you put this file into
>>>>>>>>>> $MESA_DIR/crlibm/public/ (overwriting the one that exists there now) then:
>>>>>>>>>>
>>>>>>>>>> cd $MESA_DIR/crlibm/test
>>>>>>>>>> ./clean && ./mk && ./tester | tee run.log
>>>>>>>>>>
>>>>>>>>>> and send the run.log file back that gets made.
>>>>>>>>>>
>>>>>>>>>> Also you say you are running centos 5.x/6.x, a 5.x version would
>>>>>>>>>> be very old, could you check which version you have with:
>>>>>>>>>>
>>>>>>>>>> uname -a
>>>>>>>>>> and also send the output of
>>>>>>>>>> ldd --version
>>>>>>>>>>
>>>>>>>>>> Thanks
>>>>>>>>>> Rob
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>> On 1 February 2018 at 13:28, Kuldeep Verma via Mesa-users <
>>>>>>>>>> mesa-users at lists.mesastar.org> wrote:
>>>>>>>>>>
>>>>>>>>>>> Dear All,
>>>>>>>>>>>
>>>>>>>>>>> I was trying to install the latest version of mesa (10108) with
>>>>>>>>>>> the latest version of mesasdk (January 2018) on a cluster (OS: CentOS
>>>>>>>>>>> 5.x/6.x Linux 64 bit, on which I had successfully installed an older
>>>>>>>>>>> version of mesa some time back). I could install mesasdk successfully (see
>>>>>>>>>>> below the outputs of few commands).
>>>>>>>>>>>
>>>>>>>>>>> [kuldeep at s81n11 ~]$ source $MESASDK_ROOT/bin/mesasdk_init.sh
>>>>>>>>>>> mesasdk_init.sh: checking architecture
>>>>>>>>>>> mesasdk_init.sh: checking prerequisites
>>>>>>>>>>> mesasdk_init.sh: regenerating headers
>>>>>>>>>>>
>>>>>>>>>>> [kuldeep at s81n11 ~]$ gfortran --version
>>>>>>>>>>> GNU Fortran (GCC) 7.2.0
>>>>>>>>>>> Copyright (C) 2017 Free Software Foundation, Inc.
>>>>>>>>>>> This is free software; see the source for copying conditions.
>>>>>>>>>>> There is NO
>>>>>>>>>>> warranty; not even for MERCHANTABILITY or FITNESS FOR A
>>>>>>>>>>> PARTICULAR PURPOSE.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> The ./clean followed ./install command in the the directory
>>>>>>>>>>> mesa-r10108 failed with the following error message:
>>>>>>>>>>>
>>>>>>>>>>> make[2]: Entering directory `/home/kuldeep/mesa-r10108/crl
>>>>>>>>>>> ibm/crlibm/tests'
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> test_val.o -MD -MP -MF .deps/test_val.Tpo -c -o test_val.o test_val.c
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> test_common.o -MD -MP -MF .deps/test_common.Tpo -c -o test_common.o
>>>>>>>>>>> test_common.c
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> soak_test.o -MD -MP -MF .deps/soak_test.Tpo -c -o soak_test.o soak_test.c
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> test_perf.o -MD -MP -MF .deps/test_perf.Tpo -c -o test_perf.o test_perf.c
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> blind_test.o -MD -MP -MF .deps/blind_test.Tpo -c -o blind_test.o
>>>>>>>>>>> blind_test.c
>>>>>>>>>>> gcc -DHAVE_CONFIG_H -I. -I.. -I..   -O0 -std=gnu99   -MT
>>>>>>>>>>> generate_test_vectors.o -MD -MP -MF .deps/generate_test_vectors.Tpo
>>>>>>>>>>> -c -o generate_test_vectors.o generate_test_vectors.c
>>>>>>>>>>> mv -f .deps/generate_test_vectors.Tpo
>>>>>>>>>>> .deps/generate_test_vectors.Po
>>>>>>>>>>> mv -f .deps/soak_test.Tpo .deps/soak_test.Po
>>>>>>>>>>> mv -f .deps/blind_test.Tpo .deps/blind_test.Po
>>>>>>>>>>> mv -f .deps/test_val.Tpo .deps/test_val.Po
>>>>>>>>>>> mv -f .deps/test_common.Tpo .deps/test_common.Po
>>>>>>>>>>> gcc -O0 -std=gnu99     -o crlibm_testval test_val.o
>>>>>>>>>>> test_common.o ../libcrlibm.a -lm
>>>>>>>>>>> gcc -O0 -std=gnu99     -o crlibm_soaktest soak_test.o
>>>>>>>>>>> test_common.o ../libcrlibm.a -lm
>>>>>>>>>>> gcc -O0 -std=gnu99     -o crlibm_blindtest blind_test.o
>>>>>>>>>>> test_common.o ../libcrlibm.a -lm
>>>>>>>>>>> gcc -O0 -std=gnu99     -o crlibm_generate_test_vectors
>>>>>>>>>>> generate_test_vectors.o test_common.o ../libcrlibm.a -lm
>>>>>>>>>>> mv -f .deps/test_perf.Tpo .deps/test_perf.Po
>>>>>>>>>>> gcc -O0 -std=gnu99     -o crlibm_testperf test_perf.o
>>>>>>>>>>> test_common.o ../libcrlibm.a -lm
>>>>>>>>>>> make[2]: Leaving directory `/home/kuldeep/mesa-r10108/crl
>>>>>>>>>>> ibm/crlibm/tests'
>>>>>>>>>>> make[1]: Leaving directory `/home/kuldeep/mesa-r10108/crl
>>>>>>>>>>> ibm/crlibm'
>>>>>>>>>>> gcc -O3 -w -fno-common -fexceptions -c -I../../include
>>>>>>>>>>> ../private/crlibm_c_fortran.c -o crlibm_c_fortran.o
>>>>>>>>>>> gfortran -Wno-uninitialized -fno-range-check -fmax-errors=12
>>>>>>>>>>> -fprotect-parens -fno-sign-zero -fbacktrace -ggdb -finit-real=snan
>>>>>>>>>>> -fopenmp  -std=f2008 -Wno-error=tabs -I../public -I../private
>>>>>>>>>>> -I../../include -I/home/kuldeep/mesasdk/include -Wunused-value
>>>>>>>>>>> -Werror -W -Wno-compare-reals -Wno-unused-parameter -fimplicit-none   -O2
>>>>>>>>>>> -c -ffree-form -x f95-cpp-input ../public/crlibm_lib.f
>>>>>>>>>>> ar crs libf2crlibm.a crlibm_c_fortran.o crlibm_lib.o
>>>>>>>>>>> make: `libf2crlibm.a' is up to date.
>>>>>>>>>>> gfortran -Wno-uninitialized -fno-range-check -fmax-errors=12
>>>>>>>>>>> -fprotect-parens -fno-sign-zero -fbacktrace -ggdb -finit-real=snan -fopenmp
>>>>>>>>>>> -I../../make -I../../public -I../../../include -fbounds-check
>>>>>>>>>>> -Wuninitialized -Warray-bounds  -c -ffree-form -x f95-cpp-input
>>>>>>>>>>> ../src/test_crlibm.f
>>>>>>>>>>> gfortran -fopenmp -o ../tester test_crlibm.o \
>>>>>>>>>>> -L../../make -lf2crlibm -lcrlibm -L../../../lib -lutils -lconst
>>>>>>>>>>>
>>>>>>>>>>> FAILED
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> /home/kuldeep/mesa-r10108/crlibm/test
>>>>>>>>>>> ./build_and_test FAILED
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> /home/kuldeep/mesa-r10108/crlibm
>>>>>>>>>>> ./build_and_test FAILED
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> You can find the files mesa-r10108/crlibm/crlibm/config.log and
>>>>>>>>>>> mesa-r10108/crlibm/test/tmp.txt attached with the email. Your
>>>>>>>>>>> pointer towards the issue would be highly appreciated (I apologize in
>>>>>>>>>>> advance if I missed to find an existing solution of this problem on the
>>>>>>>>>>> mailing list).
>>>>>>>>>>>
>>>>>>>>>>> Best wishes,
>>>>>>>>>>> Kuldeep
>>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>>
>>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>
>>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.mesastar.org/pipermail/mesa-users/attachments/20180205/af0a6fb5/attachment.html>