×

Notice

The forum is in read only mode.

Code_saturne 1.4b trouble [install]

More
15 years 7 months ago #2361 by Claus
Hi. Im trying to compile Code_Saturne again, now on ubuntu 8.10 and I've run into this error - Im not a programmer, so I don't know what this means - help :)

[code:1]gcc -I/opt/code_saturne/opt/bft-1.0.7/arch/Linux/include -I/opt/code_saturne/opt/fvm-0.11.0/arch/Linux/include -I/usr/include/libxml2 -D_POSIX_SOURCE -fPIC -D_CS_HAVE_SOCKET -D_CS_HAVE_XML -DNDEBUG -std=c99 -funsigned-char -pedantic -W -Wall -Wshadow -Wpointer-arith -Wcast-qual -Wcast-align -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -Wmissing-declarations -Wnested-externs -Wno-uninitialized -O -Wno-unused -c cs_proxy_comm.c
cs_proxy_comm.c: In function ‘_comm_sock_connect’:
cs_proxy_comm.c:313: error: ‘struct hostent’ has no member named ‘h_addr’
make[1]: *** [cs_proxy_comm.o] Error 1
make[1]: Leaving directory `/opt/code_saturne/Noyau/ncs-1.4.b/build/Linux/BASE'
make: *** [privcompil] Error 2
Directory: /opt/code_saturne/Noyau/ncs-1.4.b/build/Linux/BASE

Error : make lib_install MOD=BASE failed

Thank you for choosing Code_Saturne[/code:1]

I have the same 'libs and compilers as on ubuntu 8.04, but obviously something has changed.

Thanks in advance

Claus

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
More
15 years 7 months ago #2364 by David Monfort
Replied by David Monfort on topic Re:Code_saturne 1.4b trouble [install]
Hi Claus,

Some header files have changed in the new Ubuntu 8.10, that's why you get this error. You have to change the h_addr structure item in h_addr_list[0] at line 313 in order to compile Code_Saturne. Note that is is fixed in the (soon released) 1.4.c development version :whistle:

David
More
15 years 7 months ago #2366 by Claus
Replied by Claus on topic Re:Code_saturne 1.4b trouble [install]
I tried replacing with what you wrote, but this is the error I get when I try to run the case. I'll wait for 1.4c and see, it's not critical in any way right now.

Don't forget to post here when 1.4c is ready! I forget to check the code_saturne site :whistle:

(maybe you can use this for something :dry: )
[code:1] ********************************************
Calculation starting
********************************************

*** glibc detected *** /opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition: munmap_chunk(): invalid pointer: 0x09f1e4b8 ***
======= Backtrace: =========
/lib/tls/i686/cmov/libc.so.6[0x4a9933f4]
/opt/code_saturne/opt/bft-1.0.7/arch/Linux/lib/libbft.so.1[0xb7fc53f9]
/lib/tls/i686/cmov/libc.so.6(cfree+0x35)[0x4a9953f5]
/opt/code_saturne/opt/bft-1.0.7/arch/Linux/lib/libbft.so.1(bft_mem_free+0x30)[0xb7fc3400]
/opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition(main+0x21c4)[0x804f0f4]
/lib/tls/i686/cmov/libc.so.6(__libc_start_main+0xe5)[0x4a93a685]
/opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition[0x804c4f1]
======= Memory map: ========
08048000-08066000 r-xp 00000000 08:05 639005 /opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition
08066000-08067000 r--p 0001d000 08:05 639005 /opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition
08067000-08068000 rw-p 0001e000 08:05 639005 /opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition
09f17000-09f81000 rw-p 09f17000 00:00 0 [heap]
44595000-445af000 r-xp 00000000 08:05 843885 /lib/ld-2.8.90.so
445b0000-445b1000 r--p 0001a000 08:05 843885 /lib/ld-2.8.90.so
445b1000-445b2000 rw-p 0001b000 08:05 843885 /lib/ld-2.8.90.so
4a924000-4aa7c000 r-xp 00000000 08:05 860162 /lib/tls/i686/cmov/libc-2.8.90.so
4aa7c000-4aa7e000 r--p 00158000 08:05 860162 /lib/tls/i686/cmov/libc-2.8.90.so
4aa7e000-4aa7f000 rw-p 0015a000 08:05 860162 /lib/tls/i686/cmov/libc-2.8.90.so
4aa7f000-4aa82000 rw-p 4aa7f000 00:00 0
4aa8a000-4aaae000 r-xp 00000000 08:05 860382 /lib/tls/i686/cmov/libm-2.8.90.so
4aaae000-4aaaf000 r--p 00023000 08:05 860382 /lib/tls/i686/cmov/libm-2.8.90.so
4aaaf000-4aab0000 rw-p 00024000 08:05 860382 /lib/tls/i686/cmov/libm-2.8.90.so
4aacd000-4aae1000 r-xp 00000000 08:05 943291 /usr/lib/libz.so.1.2.3.3
4aae1000-4aae3000 rw-p 00013000 08:05 943291 /usr/lib/libz.so.1.2.3.3
4af36000-4af43000 r-xp 00000000 08:05 843788 /lib/libgcc_s.so.1
4af43000-4af44000 r--p 0000c000 08:05 843788 /lib/libgcc_s.so.1
4af44000-4af45000 rw-p 0000d000 08:05 843788 /lib/libgcc_s.so.1
b7e9a000-b7e9b000 rw-p b7e9a000 00:00 0
b7e9b000-b7eda000 r--p 00000000 08:05 843823 /usr/lib/locale/en_DK.utf8/LC_CTYPE
b7eda000-b7fbb000 r--p 00000000 08:05 843851 /usr/lib/locale/en_DK.utf8/LC_COLLATE
b7fbb000-b7fbd000 rw-p b7fbb000 00:00 0
b7fbd000-b7fc7000 r-xp 00000000 08:05 1187942 /opt/code_saturne/opt/bft-1.0.7/arch/Linux/lib/libbft.so.1.0.0
b7fc7000-b7fc8000 r--p 00009000 08:05 1187942 /opt/code_saturne/opt/bft-1.0.7/arch/Linux/lib/libbft.so.1.0.0
b7fc8000-b7fc9000 rw-p 0000a000 08:05 1187942 /opt/code_saturne/opt/bft-1.0.7/arch/Linux/lib/libbft.so.1.0.0
b7fd0000-b7fd1000 r--p 00000000 08:05 892930 /usr/lib/locale/en_DK.utf8/LC_NUMERIC
b7fd1000-b7fd2000 r--p 00000000 08:05 892931 /usr/lib/locale/en_DK.utf8/LC_TIME
b7fd2000-b7fd3000 r--p 00000000 08:05 892932 /usr/lib/locale/en_DK.utf8/LC_MONETARY
b7fd3000-b7fd4000 r--p 00000000 08:05 901358 /usr/lib/locale/en_DK.utf8/LC_MESSAGES/SYS_LC_MESSAGES
b7fd4000-b7fd5000 r--p 00000000 08:05 843865 /usr/lib/locale/en_DK.utf8/LC_PAPER
b7fd5000-b7fd6000 r--p 00000000 08:05 843890 /usr/lib/locale/en_DK.utf8/LC_NAME
b7fd6000-b7fd7000 r--p 00000000 08:05 892969 /usr/lib/locale/en_DK.utf8/LC_ADDRESS
b7fd7000-b7fd8000 r--p 00000000 08:05 892970 /usr/lib/locale/en_DK.utf8/LC_TELEPHONE
b7fd8000-b7fd9000 r--p 00000000 08:05 843907 /usr/lib/locale/en_DK.utf8/LC_MEASUREMENT
b7fd9000-b7fe0000 r--s 00000000 08:05 954221 /usr/lib/gconv/gconv-modules.cache
b7fe0000-b7fe1000 r--p 00000000 08:05 892971 /usr/lib/locale/en_DK.utf8/LC_IDENTIFICATION
b7fe1000-b7fe3000 rw-p b7fe1000 00:00 0
b7fe3000-b7fe4000 r-xp b7fe3000 00:00 0 [vdso]
bfacf000-bfae4000 rw-p bffeb000 00:00 0 [stack]
/home/claus/Stuff/CAD/Study_saturne-2beam/1/2/SCRIPTS/runcase: line 818: 15537 Aborted ${CSPP_PATH}/cs_partition $nproc_kernel > listpart 2>&1
Error running the partitioner.
Check partitioner log (listpart) for details.


********************************************
Error in partitioning stage.
********************************************
[/code:1]

Eagerly awaiting the 1.4c :silly:

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
More
15 years 7 months ago #2368 by David Monfort
Replied by David Monfort on topic Re:Code_saturne 1.4b trouble [install]
This error is not related to your compilation issue.
Nonetheless, difficult to say why you get this...

Could you post/attach some files so that I have a clearer view ? (listing, listpart, listenv, runcase, resume and error* if present)

Thanks.

ps: I'll post the announce when the new development version is released
More
15 years 7 months ago #2369 by Claus
Replied by Claus on topic Re:Code_saturne 1.4b trouble [install]
Hi - I've had a little time to test today, and it seems the partitioner only crashes when 2 cpus/cores are used. I have made sure that the macros.linux.mk have the proper entries for openmpi.

I have attached the output of the runcase here, I hope you can extract some info from it - otherwise tell me what to post :)


www.zshare.net/download/5123044444099e9d/

Code_Aster release : STA11.4 on OpenSUSE 12.3 64 bits - EDF/Intel version
More
15 years 7 months ago #2370 by David Monfort
Replied by David Monfort on topic Re:Code_saturne 1.4b trouble [install]
Thanks! It doesn't give me the answer, though.

Could you, when you have time, give me the output of the following command ? I have an idea that you're problem comes from the library our partitionning tool is using.
[code:1]ldd /opt/code_saturne/Enveloppe/ecs-1.4.a/arch/Linux/bin/cs_partition[/code:1]
I suspect that you link with the METIS wrapper of the SCOTCH library (and I did have some problems with that in the past).
If so, you can either install the METIS library in your opt directory and point to that while configuring the preprocessor (ecs), or disable the METIS fonctionnality (the partitionner will still work but the domain partitionning will be less efficient).
If not, I'll ask to send your whole test case directory so that I can restart it on my computer (if it is not confidentail of course ;-))
Moderators: catux
Time to create page: 0.155 seconds
Powered by Kunena Forum