I would be grateful if anyone can post the makefile using pathscale on Opteron machine.
Thanks in advance!!!
makefile using pathscale on Opteron Machine
Moderators: Global Moderator, Moderator
makefile using pathscale on Opteron Machine
Last edited by robycherian on Sun Jun 03, 2007 4:45 pm, edited 1 time in total.
-
- Newbie
- Posts: 3
- Joined: Tue Sep 13, 2005 7:10 pm
- License Nr.: jk35QGB7
makefile using pathscale on Opteron Machine
hello all,
I share the question.
In particular, when I compile with
./preprocess <whatever.F | /usr/bin/cpp -P -C -traditional > whatever.f90 -Dkind8 -DNGXhalf -DCACHE_SIZE=1000 -DRACCMU_DGEMV -DIFC
pathf95 -freeform -lowercase -fno-math-errno -mtune=opteron -mcpu=opteron -march=opteron -I/lustrefs/compiler/pathscale-3.0/include -freeform -O0 -c whatever.f90 ,
the binary is produced (serial), but on execution I receive POSCAR found : 2 types and 4 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: wrap around errors must be expected
FFT: planning ... 1
reading WAVECAR
**** Segmentation fault! Fault address: 0xfffffffee46d7ef0
Fault address is 4747067664 bytes below the nearest valid
mapping boundary, which is at 0xffffffffff600000.
This is likely to have been caused by a stack overflow.
Use your shell's ulimit or limit command to see if your
stack size limit is too small.
Aborted
btw: "ulimit" returns unlimited
Has anyone got any recommendations? TIA
I share the question.
In particular, when I compile with
./preprocess <whatever.F | /usr/bin/cpp -P -C -traditional > whatever.f90 -Dkind8 -DNGXhalf -DCACHE_SIZE=1000 -DRACCMU_DGEMV -DIFC
pathf95 -freeform -lowercase -fno-math-errno -mtune=opteron -mcpu=opteron -march=opteron -I/lustrefs/compiler/pathscale-3.0/include -freeform -O0 -c whatever.f90 ,
the binary is produced (serial), but on execution I receive POSCAR found : 2 types and 4 ions
LDA part: xc-table for Ceperly-Alder, standard interpolation
POSCAR, INCAR and KPOINTS ok, starting setup
WARNING: wrap around errors must be expected
FFT: planning ... 1
reading WAVECAR
**** Segmentation fault! Fault address: 0xfffffffee46d7ef0
Fault address is 4747067664 bytes below the nearest valid
mapping boundary, which is at 0xffffffffff600000.
This is likely to have been caused by a stack overflow.
Use your shell's ulimit or limit command to see if your
stack size limit is too small.
Aborted
btw: "ulimit" returns unlimited
Has anyone got any recommendations? TIA
Last edited by Oettking on Tue Jun 05, 2007 9:05 am, edited 1 time in total.