Hi Enrique,

Thanks for reaching out. All the Mac’s I’ve tried are Intel Macs ranging from 
2013 to 2020 models, so I’m interested to hear you got Seisan to compile.

Yes, I had to do the same allowances for Seisan executables under Security & 
Privacy.

I wonder if I could possibly get a tarred/zipped copy of your Intel-compiled 
PRO and LIB directories to try? And did you have to symlink any dynamic 
libraries to get those executables to run on Intel Mac, or was that only on M1 
Mac?

Still, if there is someone in the Seisan community that can document successful 
steps to get Seisan to compile on MacOSX today, that would be even better. The 
instructions in the Seisan manual and in the extra file in the INF directory 
aren’t getting me there.

Cheers,

Glenn

From: Enrique Rodolfo Chon <[email protected]>
Date: Thursday, April 1, 2021 at 2:59 PM
To: Thompson, Glenn <[email protected]>
Cc: [email protected] <[email protected]>
Subject: Re: [seisan] Problems compiling Seisan on MacOSX Big Sur
Hi Glenn,

I recently set SEISAN up on a mac mini with the M1 chip/running Big Sur. I 
faced similar issues compiling on the machine directly, so I eventually 
resorted to compiling on an Intel mac and just copying the master directory 
over. Although the machine architecture is totally different, the programs work 
fine because of the Rosetta 2 interpreter. The only caveat was that I had to 
track down a few x86 libraries and put them in the correct place (seems like 
you've anticipated this already with your symlink workaround). By default, the 
security features under Big Sur have tried to prevent me from running the 
executables, so you must authorize them once in System Pref>Security and 
Privacy (there's an 'allow anyway' choice after the first attempt at running 
each program).

Best,

Enrique

On Thu, Apr 1, 2021 at 12:06 PM Thompson, Glenn 
<[email protected]<mailto:[email protected]>> wrote:
Hi Seisan community,

I’m trying to compile Seisan 11.0 on MacOSX (Big Sur) and running into similar 
issues reported by Rod Stewart from Aug 2019: 
https://www.mail-archive.com/[email protected]/msg01829.html<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwww.mail-archive.com%2Fseisan%40geo.uib.no%2Fmsg01829.html&data=04%7C01%7Cthompsong%40usf.edu%7C38f5c8526df04c8d38ba08d8f5403cc8%7C741bf7dee2e546df8d6782607df9deaa%7C0%7C0%7C637529003556832364%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=2ejo1X8%2BGb7K92FLj1lQEoKxEVwHOGy3X5ePaTH0k2M%3D&reserved=0>

Yesterday I tried this on a clean MacOS re-install: it just had Xcode, homebrew 
(+gcc), XQuartz, and Miniconda/ObsPy. Today I am trying it on a different Mac 
which already had many other applications, but I’m getting the same results.

At first I tried the pre-compiled programs (I am mainly using dirf, mulplt, 
eev), but I was getting failures on various libraries. I symlinked some newer 
versions of these but this resulted in segmentation faults.

I’ve tried (in PRO) make clean, make all, make mulplt. And in LIB, make clean, 
make all, make ./seisan.a

For example:


(base) thompsong@Glenns-MacBook-Pro PRO % make mulplt

make: *** No rule to make target `../LIB/./seisan.a', needed by `mulplt'.  Stop.



(base) thompsong@Glenns-MacBook-Pro LIB % make ./seisan.a

gfortran -m64 -c -I../INC -fdollar-ok -g -fno-automatic -o abstim.o abstim.for

gfortran -m64 -c -I../INC -fdollar-ok -g -fno-automatic -o auto_tr.o auto_tr.for



(… lots of make output hidden by Glenn …)


Error: Actual argument contains too few elements for dummy argument 'v' 
(150/200) at (1)
hyposub6.for:166:29:

  166 |           call ttcal(n_layer,v,z,d,vr,grad,z_hypo,p,phi,dist,
      |                             1
Error: Actual argument contains too few elements for dummy argument 'v' 
(150/200) at (1)
hyposub6.for:174:26:

  174 |        call ttcal(n_layer,v,z,d,vr,grad,z_hypo,p,phi,dist,tmin,n_hypo,
      |                          1
Error: Actual argument contains too few elements for dummy argument 'v' 
(150/200) at (1)
make: *** [hyposub6.o] Error 1
Edited LIB Makefile to skip hyposub*
(base) thompsong@Glenns-MacBook-Pro LIB % make ./seisan.a
gfortran -m64 -c -I../INC -fdollar-ok -g -fno-automatic -o comp_linux.o 
comp_linux.for
gcc -m64 -c -I../INC -I/usr/X11R6/include -g -o xlib.o xlib.c
xlib.c:651:14: error: implicit declaration of function 'get_udp_msg' is invalid 
in C99 [-Werror,-Wimplicit-function-declaration]
         if (get_udp_msg(in_msg)) {
             ^
xlib.c:784:12: warning: using the result of an assignment as a condition 
without parentheses [-Wparentheses]
      if (i=1)
          ~^~

(… lots of make output hidden by Glenn …)
^
xlib.c:1865:37: warning: array index 255 is past the end of the array (which 
contains 255 elements) [-Warray-bounds]
seis[255].red=171;seis[255].green=0;seis[255].blue=0;
                                    ^    ~~~
xlib.c:110:1: note: array 'seis' declared here
rgbcol seis[255];
^
13 warnings and 1 error generated.
make: *** [xlib.o] Error 1

I think I had similar issues trying to compile Seisan a few weeks ago on the 
previous version of MacOSX (Catalina) and I gave up, but the situation is a 
higher priority now as I’m setting this up for a graduate student to do her 
project.

Any help or tips appreciated,

Thanks,

Glenn Thompson
_______________________________________________
seisan mailing list
[email protected]<mailto:[email protected]>
https://mailman.uib.no/listinfo/seisan<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fmailman.uib.no%2Flistinfo%2Fseisan&data=04%7C01%7Cthompsong%40usf.edu%7C38f5c8526df04c8d38ba08d8f5403cc8%7C741bf7dee2e546df8d6782607df9deaa%7C0%7C0%7C637529003556832364%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=DFAuhpqI87fK%2FZR7LaT2umhRX6rawEEQzkHFYLlf8ZQ%3D&reserved=0>

[EXTERNAL EMAIL] DO NOT CLICK links or attachments unless you recognize the 
sender and know the content is safe.
_______________________________________________
seisan mailing list
[email protected]
https://mailman.uib.no/listinfo/seisan

Reply via email to