segmentation fault

[sbshewale@gmail_com]

hello,I am trying to run cam5 in serial mode on linux platform using ifort compiler (earlier i did  with pgf90 compiler).it is showing error as 
forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source             
cam                00000000010EC6EB  Unknown               Unknown  Unknown
cam                00000000010DE42F  Unknown               Unknown  Unknown
cam                00000000010D87E1  Unknown               Unknown  Unknown
cam                0000000000D988BC  Unknown               Unknown  Unknown
cam                0000000000D84CED  Unknown               Unknown  Unknown
cam                000000000135D860  Unknown               Unknown  Unknown
cam                0000000000C934FD  Unknown               Unknown  Unknown
cam                00000000005331D0  Unknown               Unknown  Unknown
cam                000000000050D55D  Unknown               Unknown  Unknown
cam                00000000005865D5  Unknown               Unknown  Unknown
cam                0000000000404462  Unknown               Unknown  Unknown
libc.so.6          0000003BF8A1D974  Unknown               Unknown  Unknown
cam                00000000004043A9  Unknown               Unknown  Unknown


so if anybody knows, will you please help me in this regard.
Thank You

 

[fischer]

Hi,
Going to need more information.  How did you set up the your run?  What release version of cam5 are you using?
You can also try your run with debug turned on, this should give you more information about the segmentation fault.

Chris
 

 

[sbshewale@gmail_com]

hello,Thank You for your suggestionI tried it, but again facing another problem.let me give you the details of model set upI am using cam5 (CESM1_0) using ifort compiler. I configure the model as follows./configure -dyn fv -hgrid 1.9*2.5 -fc ifort -cc icc -debug -testgmake -j2 (compile)./build-namelist./camI want to run model for 1 day only (test run)but when I give run, it executes endlessly, with following lines repeatedlyforrtl: warning (402): fort: (1): In call to NUCLEATI, an array temporary was created for argument #9 forrtl: warning (402): fort: (1): In call to NUCLEATI, an array temporary was created for argument #8 forrtl: warning (402): fort: (1): In call to NUCLEATI, an array temporary was created for argument #9 forrtl: warning (402): fort: (1): In call to NUCLEATI, an array temporary was created for argument #8          If you can suggest anything for above, you are welcomeThanks 

 

[santos]

Hi swati,The lines you mention are not errors; they are just warnings. But they make it difficult to see the real problem, so in recent versions of CAM, we have disabled this warning.Please try again with this configure option for a more readable log:-fflags '-check noarg_temp_created'

 

[sbshewale@gmail_com]

Hello,Thanks for suggestion.I tried your option at the time of configure that is -fflags '-check noarg_temp_created'.but now, I think it takes too much time to complete single nstep.It executes nstep with following warningprint_memusage iam   0 stepon after physics and dynamics. -1 in the next line means unavailableprint_memusage: size, rss, share, text, datastack=  755283  700670   14578   24198       0 nstep, te        1   0.33202427330084662E+10   0.33202487031513062E+10   0.33084636104768956E-03   0.98532611477194529E+05 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=104 Min. mixing ratio violated at    1 points.  Reset to  1.0E-36 Worst =-1.0E-12 at i,k=   6 30 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -1.5354E-06, lchnk = 110, i =    6, same as indices lat =  84, lon =  86 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -1.3289E-06, lchnk = 249, i =    6, same as indices lat =  83, lon =  81 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -2.0096E-06, lchnk = 393, i =    6, same as indices lat =  82, lon =  81 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -6.8510E-07, lchnk = 396, i =    6, same as indices lat =  82, lon =  84 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -2.1394E-06, lchnk = 397, i =    6, same as indices lat =  82, lon =  85 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -1.2732E-07, lchnk = 402, i =    6, same as indices lat =  82, lon =  90 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -1.7180E-07, lchnk = 536, i =    6, same as indices lat =  81, lon =  80 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -2.3049E-06, lchnk = 537, i =    6, same as indices lat =  81, lon =  81 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -2.4098E-06, lchnk = 539, i =    6, same as indices lat =  81, lon =  83 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -6.9158E-07, lchnk = 544, i =    6, same as indices lat =  81, lon =  88 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=680 Min. mixing ratio violated at    3 points.  Reset to  1.0E-36 Worst =-5.4E-12 at i,k=   4 30 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=681 Min. mixing ratio violated at    3 points.  Reset to  1.0E-36 Worst =-3.8E-12 at i,k=   4 30 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=689 Min. mixing ratio violated at    1 points.  Reset to  1.0E-36 Worst =-9.8E-12 at i,k=   4 30 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -6.4270E-07, lchnk = 829, i =    4, same as indices lat =  85, lon =  85 QNEG4 WARNING from TPHYSAC  Max possible LH flx exceeded at    1 points. , Worst excess =  -2.5430E-07, lchnk = 830, i =    4, same as indices lat =  85, lon =  86 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=835 Min. mixing ratio violated at    3 points.  Reset to  1.0E-36 Worst =-2.0E-12 at i,k=   4 30 QNEG3 from vertical diffusion/SO2:m=  8 lat/lchnk=836 Min. mixing ratio violated at    3 points.  Reset to  1.0E-36 Worst =-1.1E-12 at i,k=   4 30  nstep, te        2   0.33202240651212697E+10   0.33202251525474677E+10   0.60261795353279058E-04   0.98532467876740251E+05If you know any idea, Please let me know or should I go for another version of cam5 (suggest me version)Thanks 

 

[santos]

Hi,There are at least two reasons why it is slow. One is the "-debug" option. We only use that option to get information about problems, like in your first message. If you can figure out what the original problem was, you can remove that option.The other reason is that you are running in serial mode on a 1.9x2.5 grid, which is probably too much work for one processor on your system. You can use MPI with multiple processors to speed things up. You can also try running in single column mode ("SCAM") instead of a 1.9x2.5 grid, but I don't use that myself, so I don't know as much about it.-Sean

 

[moguo_sun@nasa_gov]

Dear Sir/Madam,
      I try to run CAM5 on the same platform as yours and run into the same 'segmentation fault' problem. I saw you solved the problem. Could you tell me how you solved the problem? Your help is very much appreciated. Thanks.
Moguo 

 

[sbshewale@gmail_com]

Hello,I did the same thing which they have suggested me in my earlier post.put their suggested  flags and -debug option at the time of configuration.after doing this my problem was solved.if possible, try to do this and also as I have just 3-4 months back started to run cam, so unable to give any suggestions on top of above. 

 

[moguo_sun@nasa_gov]

Hi, Swati,
      Thank you for your reply. In fact, I did put -debug as suggested, but I still got the same segmentation fault. $camcfg/configure -fc ifort -dyn fv -hgrid 10x15 -nospmd -nosmp -debug -test       But Thank you anyway.