Replies: 6 comments 7 replies
-
|
Hello @mardar572, could you please provide additional details? How did you configure SU2 with meson? What is the output of meson? Which compiler do you use? |
Beta Was this translation helpful? Give feedback.
6 replies
-
|
The meson output shows that you are using a very old version of GCC (4.8.5), this might be the issue. Do you have the option to try a more recent version of GCC? |
Beta Was this translation helpful? Give feedback.
-
|
I have installed GCC (12.2) and try with it but it didn't solve the issue. Then I add the -Denable-tecio=false option to meson input. It helped the passed previous step but now I am having the following error :( As stated earlier I can successfully install SU2 751 without this option -Denable-tecio=false
|
Beta Was this translation helpful? Give feedback.
-
|
Could you double-check the meson output? Does meson actually use your GCC 12.2? |
Beta Was this translation helpful? Give feedback.
-
|
@jblueh thank you very much for your lead. The problem solved when I set CC env variable to /path_of_mpi/bin/mpicc |
Beta Was this translation helpful? Give feedback.
-
|
Yes, this can happen if there are multiple GCCs on the system, setting |
Beta Was this translation helpful? Give feedback.
-
|
I just tried to install the latest SU2 8.00 release to my HPC account, and
when building it using meson.py I got this message:
...
ninja executable not found. Building ...
bootstrapping ninja...
wrote build.ninja.
bootstrap complete. rebuilding...
[31/31] LINK ninja
Traceback (most recent call last):
File
"/scratch/677131/S800/SOURCE/SU2/externals/meson/mesonbuild/mesonmain.py",
line 146, in run
return options.run_func(options)
File
"/scratch/677131/S800/SOURCE/SU2/externals/meson/mesonbuild/msetup.py",
line 294, in run
app.generate()
File
"/scratch/677131/S800/SOURCE/SU2/externals/meson/mesonbuild/msetup.py",
line 184, in generate
with mesonlib.BuildDirLock(self.build_dir):
File
"/scratch/677131/S800/SOURCE/SU2/externals/meson/mesonbuild/mesonlib/posix.py",
line 32, in __enter__
fcntl.flock(self.lockfile, fcntl.LOCK_EX | fcntl.LOCK_NB)
OSError: [Errno 38] Function not implemented
ERROR: Unhandled python exception
This is a Meson bug and should be reported!
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
I have tested some python3 versions : 3.6, 3.97, 3.7.16.
For 3.6 version, it is old and continuation is rejected: Ok!
For the versions > 3.7, the message is the same shown above!
Kind regards,
Jairo
…On Fri, Sep 8, 2023 at 9:42 AM Johannes Blühdorn ***@***.***> wrote:
The meson output shows that you are using a very old version of GCC
(4.8.5), this might be the issue. Do you have the option to try a more
recent version of GCC?
—
Reply to this email directly, view it on GitHub
<#2128 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHGXZUX2MXD5LMZ7ELNBW3LXZMHEFANCNFSM6AAAAAA4QIQ23A>
.
You are receiving this because you are subscribed to this thread.Message
ID: ***@***.***>
|
Beta Was this translation helpful? Give feedback.
1 reply
-
|
Did this work for you with older versions? The final flock error could mean that on your cluster, some file locking functionality has been disabled. I see that you are compiling in /scratch. Do you have another filesystem where you can compile? For instance inside /home instead of /scratch? |
Beta Was this translation helpful? Give feedback.
-
|
Hello Nijso, Unfortunately, not so far: My home directory is
/scratch/677131. I will talk to HPC sysadmin to see whether there is or not
another directory available, at least for a test! /scratch is based on
cephfs/NFS.
…On Mon, Feb 26, 2024 at 7:10 PM Nijso ***@***.***> wrote:
Did this work for you with older versions? The final flock error could
mean that on your cluster, some file locking functionality has been
disabled. I see that you are compiling in /scratch. Do you have another
filesystem where you can compile? For instance inside /home instead of
/scratch?
—
Reply to this email directly, view it on GitHub
<#2128 (reply in thread)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHGXZUQ4ZFB7BWLICTOXQD3YVUB5BAVCNFSM6AAAAAA4QIQ23CVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DKOJYGE3TO>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
Beta Was this translation helpful? Give feedback.
0 replies
-
|
Supplementary info: Yes, for SU2 7.31 I was able to install it on another
cluster , in 2022.
On Tue, Feb 27, 2024 at 10:06 AM Jairo Cavalcante ***@***.***>
wrote:
… Hello Nijso, Unfortunately, not so far: My home directory is
/scratch/677131. I will talk to HPC sysadmin to see whether there is or not
another directory available, at least for a test! /scratch is based on
cephfs/NFS.
On Mon, Feb 26, 2024 at 7:10 PM Nijso ***@***.***> wrote:
> Did this work for you with older versions? The final flock error could
> mean that on your cluster, some file locking functionality has been
> disabled. I see that you are compiling in /scratch. Do you have another
> filesystem where you can compile? For instance inside /home instead of
> /scratch?
>
> —
> Reply to this email directly, view it on GitHub
> <#2128 (reply in thread)>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AHGXZUQ4ZFB7BWLICTOXQD3YVUB5BAVCNFSM6AAAAAA4QIQ23CVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DKOJYGE3TO>
> .
> You are receiving this because you commented.Message ID:
> ***@***.***>
>
|
Beta Was this translation helpful? Give feedback.
0 replies
-
|
News: I just tried install SU2V801 in another cluster, and no problems with
locking files on it!
Ninja has been aborted on the compilation:
...
from
../SU2_CFD/src/output/filewriter/CParaviewFileWriter.cpp:28:
../externals/codi/include/codi/tapes/data/chunk.hpp:107:32: warning: inline
function ‘virtual void codi::ChunkBase::deleteData()’ used but never defined
CODI_INLINE virtual void deleteData() = 0; ///< Delete
the allocated data.
^
../externals/codi/include/codi/tapes/data/chunk.hpp:106:32: warning: inline
function ‘virtual void codi::ChunkBase::allocateData()’ used but never
defined
CODI_INLINE virtual void allocateData() = 0; ///<
Allocated the data if it was deallocated before.
^
[538/1371] Compiling C++ object Common/src/libSU2Common.a.p/CConfig.cpp.o
ninja: build stopped: subcommand failed.
Info on compiler:
***@***.*** SU2]$ echo $CC
/opt/ohpc/pub/mpi/openmpi-gnu/1.10.7/bin/mpicc
***@***.*** SU2]$ echo $CXX
/opt/ohpc/pub/mpi/openmpi-gnu/1.10.7/bin/mpicxx
***@***.*** SU2]$ mpicc -v
Using built-in specs.
COLLECT_GCC=/opt/ohpc/pub/compiler/gcc/5.4.0/bin/gcc
COLLECT_LTO_WRAPPER=/opt/ohpc/pub/compiler/gcc/5.4.0/libexec/gcc/x86_64-unknown-linux-gnu/5.4.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../configure --disable-multilib
--enable-languages=c,c++,fortran --prefix=/opt/ohpc/pub/compiler/gcc/5.4.0
Thread model: posix
gcc version 5.4.0 (GCC)
***@***.*** SU2]$ mpicxx -v
Using built-in specs.
COLLECT_GCC=/opt/ohpc/pub/compiler/gcc/5.4.0/bin/g++
COLLECT_LTO_WRAPPER=/opt/ohpc/pub/compiler/gcc/5.4.0/libexec/gcc/x86_64-unknown-linux-gnu/5.4.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../configure --disable-multilib
--enable-languages=c,c++,fortran --prefix=/opt/ohpc/pub/compiler/gcc/5.4.0
Thread model: posix
On Tue, Feb 27, 2024 at 10:25 AM Jairo Cavalcante ***@***.***>
wrote:
… Supplementary info: Yes, for SU2 7.31 I was able to install it on another
cluster , in 2022.
On Tue, Feb 27, 2024 at 10:06 AM Jairo Cavalcante ***@***.***>
wrote:
> Hello Nijso, Unfortunately, not so far: My home directory is
> /scratch/677131. I will talk to HPC sysadmin to see whether there is or not
> another directory available, at least for a test! /scratch is based on
> cephfs/NFS.
>
> On Mon, Feb 26, 2024 at 7:10 PM Nijso ***@***.***> wrote:
>
>> Did this work for you with older versions? The final flock error could
>> mean that on your cluster, some file locking functionality has been
>> disabled. I see that you are compiling in /scratch. Do you have another
>> filesystem where you can compile? For instance inside /home instead of
>> /scratch?
>>
>> —
>> Reply to this email directly, view it on GitHub
>> <#2128 (reply in thread)>,
>> or unsubscribe
>> <https://github.com/notifications/unsubscribe-auth/AHGXZUQ4ZFB7BWLICTOXQD3YVUB5BAVCNFSM6AAAAAA4QIQ23CVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DKOJYGE3TO>
>> .
>> You are receiving this because you commented.Message ID:
>> ***@***.***>
>>
>
|
Beta Was this translation helpful? Give feedback.
0 replies
-
|
FYI installation worked when I updated to GCC 8.8.3, but when I star the
flow solver, I got the following error: SU2_CFD: error while loading shared
libraries: libmpi.so.12: cannot open shared object file: No such file or
directory!
On Wed, Feb 28, 2024 at 11:33 AM Jairo Cavalcante ***@***.***>
wrote:
… News: I just tried install SU2V801 in another cluster, and no problems
with locking files on it!
Ninja has been aborted on the compilation:
...
from
../SU2_CFD/src/output/filewriter/CParaviewFileWriter.cpp:28:
../externals/codi/include/codi/tapes/data/chunk.hpp:107:32: warning:
inline function ‘virtual void codi::ChunkBase::deleteData()’ used but never
defined
CODI_INLINE virtual void deleteData() = 0; ///< Delete
the allocated data.
^
../externals/codi/include/codi/tapes/data/chunk.hpp:106:32: warning:
inline function ‘virtual void codi::ChunkBase::allocateData()’ used but
never defined
CODI_INLINE virtual void allocateData() = 0; ///<
Allocated the data if it was deallocated before.
^
[538/1371] Compiling C++ object Common/src/libSU2Common.a.p/CConfig.cpp.o
ninja: build stopped: subcommand failed.
Info on compiler:
***@***.*** SU2]$ echo $CC
/opt/ohpc/pub/mpi/openmpi-gnu/1.10.7/bin/mpicc
***@***.*** SU2]$ echo $CXX
/opt/ohpc/pub/mpi/openmpi-gnu/1.10.7/bin/mpicxx
***@***.*** SU2]$ mpicc -v
Using built-in specs.
COLLECT_GCC=/opt/ohpc/pub/compiler/gcc/5.4.0/bin/gcc
COLLECT_LTO_WRAPPER=/opt/ohpc/pub/compiler/gcc/5.4.0/libexec/gcc/x86_64-unknown-linux-gnu/5.4.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../configure --disable-multilib
--enable-languages=c,c++,fortran --prefix=/opt/ohpc/pub/compiler/gcc/5.4.0
Thread model: posix
gcc version 5.4.0 (GCC)
***@***.*** SU2]$ mpicxx -v
Using built-in specs.
COLLECT_GCC=/opt/ohpc/pub/compiler/gcc/5.4.0/bin/g++
COLLECT_LTO_WRAPPER=/opt/ohpc/pub/compiler/gcc/5.4.0/libexec/gcc/x86_64-unknown-linux-gnu/5.4.0/lto-wrapper
Target: x86_64-unknown-linux-gnu
Configured with: ../configure --disable-multilib
--enable-languages=c,c++,fortran --prefix=/opt/ohpc/pub/compiler/gcc/5.4.0
Thread model: posix
On Tue, Feb 27, 2024 at 10:25 AM Jairo Cavalcante ***@***.***>
wrote:
> Supplementary info: Yes, for SU2 7.31 I was able to install it on another
> cluster , in 2022.
>
> On Tue, Feb 27, 2024 at 10:06 AM Jairo Cavalcante <
> ***@***.***> wrote:
>
>> Hello Nijso, Unfortunately, not so far: My home directory is
>> /scratch/677131. I will talk to HPC sysadmin to see whether there is or not
>> another directory available, at least for a test! /scratch is based on
>> cephfs/NFS.
>>
>> On Mon, Feb 26, 2024 at 7:10 PM Nijso ***@***.***> wrote:
>>
>>> Did this work for you with older versions? The final flock error could
>>> mean that on your cluster, some file locking functionality has been
>>> disabled. I see that you are compiling in /scratch. Do you have another
>>> filesystem where you can compile? For instance inside /home instead of
>>> /scratch?
>>>
>>> —
>>> Reply to this email directly, view it on GitHub
>>> <#2128 (reply in thread)>,
>>> or unsubscribe
>>> <https://github.com/notifications/unsubscribe-auth/AHGXZUQ4ZFB7BWLICTOXQD3YVUB5BAVCNFSM6AAAAAA4QIQ23CVHI2DSMVQWIX3LMV43SRDJONRXK43TNFXW4Q3PNVWWK3TUHM4DKOJYGE3TO>
>>> .
>>> You are receiving this because you commented.Message ID:
>>> ***@***.***>
>>>
>>
|
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
How can I solve the following problem?
Beta Was this translation helpful? Give feedback.
All reactions