Created attachment 187059 [details] Update patch New Features from http://www.dhondt.de/new_calc.htm : Procedures Introduced CETOL on the *VISCO card in order to control the viscous strain increase within a time increment Changed the requirements on initial values for networks: initial values are required now at all open end and all crossing points in the network. Improved the initial value generation for gas networks Improved code for laminar incompressible calculations. Solvers Installed dgmres for the iterative solution of nonsymmetrical structural matrix systems (active for SOLVER=ITERATIVE SCALING) Elements Coded an element type C3D10T = C3D10 with linear temperature interpolation (improves thermal strains in structural calculations). Keyword cards Introduced SHIFTX and SHIFTY on the *AMPLITUDE card Constraints Improved the pre-tension formulation (cut faces have to stay parallel). User subroutines propertynet.f: for user-defined changes of the properties of networks defined in the input deck. Extended the parameter list in the interfaces of uboun.f, utemp.f, film.f and others. Coded a network user element (cf. documentation: Fluid Section Types: Gases). Coded a network user MPC interface (keyword: *NETWORK MPC). Coded a general user element (cf. documentation: Element Types) Output For networks the sum of the incoming flows is assigned to the end nodes (instead of the outcoming flow) The body heating due to electromagnetic effects can be stored in the .dat file (key: EBHE)
I should mention I am getting this issue when using the latest gcc6 package: % ccx /lib/libgcc_s.so.1: version GCC_4.6.0 required by /usr/local/lib/gcc6/libgfortran.so.3 not found However, I see it on the previous version as well, so it's either local or a GCC6 wide issue.
Created attachment 187562 [details] Updated patch It seems the distfile has been re-rolled. New patch file contains updated distinfo.
(In reply to Pedro F. Giffuni from comment #1) I see the same problem. Do you have a fix for it?
(In reply to Thomas Zander from comment #3) As I mentioned the issue applies to existing port as well. There is now a wiki entry for the issue: https://wiki.freebsd.org/libgcc%20problem and I email gerald@ about it but the only "clean" way to fix it may be to move fortran ports to flang.
(In reply to Thomas Zander from comment #2) Thanks .. upstream likes to do that :(.
Created attachment 187570 [details] Updated patch to link with LDFLAGS I figured this one out by adding LDFLAGS when linking with the fortran compiler.
A commit references this bug: Author: riggs Date: Mon Oct 30 17:15:56 UTC 2017 New revision: 453199 URL: https://svnweb.freebsd.org/changeset/ports/453199 Log: Update to upstream version 2.13 Detailed log: - Introduction of new features, see http://www.dhondt.de/new_calc.htm https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=222914 PR: 222914 Submitted by: pfg (maintainer) Changes: head/cad/calculix-ccx/Makefile head/cad/calculix-ccx/distinfo head/cad/calculix-ccx/files/patch-CalculiX.h head/cad/calculix-ccx/files/patch-Makefile head/cad/calculix-ccx/pkg-plist