MNT: Add support for osx-64 builds

[matthewfeickert]

• Add osx-64 builds.
• Change -fstack-protector FFLAGS flag to -no-fstack-protector for x86 osx builds to allow for the share/*.dat files to be loaded properly. osx-arm64 sets -no-fstack-protector, so this is mirroring that behavior.
  • c.f. https://stackoverflow.com/questions/10712972/what-is-the-use-of-fno-stack-protector Having stack-protector enabled can break the ability to compile against non-standard userspace libraries, and can be considered harmful.
• Bump build number.

Checklist

• Used a personal fork of the feedstock to propose changes
• Bumped the build number (if the version is unchanged)
• [N/A] Reset the build number to 0 (if the version changed)
• Re-rendered with the latest conda-smithy (Use the phrase @conda-forge-admin, please rerender in a comment in this PR for automated rerendering)
• Ensured the license file is being packaged.

[matthewfeickert]

@conda-forge-admin, please rerender

[conda-forge-admin]

Hi! This is the friendly automated conda-forge-linting service.

I just wanted to let you know that I linted all conda-recipes in your PR (recipe/meta.yaml) and found it was in an excellent condition.

I do have some suggestions for making it better though...

For recipe/meta.yaml:

• ℹ️ The recipe is not parsable by parser conda-recipe-manager. The recipe can only be automatically migrated to the new v1 format if it is parseable by conda-recipe-manager.

_{This message was generated by GitHub Actions workflow run https://github.com/conda-forge/conda-forge-webservices/actions/runs/12783159295. Examine the logs at this URL for more detail.}

[matthewfeickert]

I don't think Fortran 90 code (like trim) can be added to Fortran 77 .f files. So will need to get an equivalent to trim working.

[matthewfeickert]

@conda-forge-admin, please rerender

[matthewfeickert]

@conda-forge/staged-recipes I'm not sure if anyone with Fortran experience can offer advice here, but this recipe currently builds and runs as expected for linux-64, linux-aarch64, linux-ppc64le, and osx-arm64. To make this legacy code installable (the original author is dead and I do not develop Fortran — this is Fortran 77) file paths that the author expected individual users to hardcode for their machines get patched at recipe build time with sed to inject $PREFIX.

ff-feedstock/recipe/meta.yaml

Lines 33 to 34 in 7ad7719

	- sed -i "s|/user/gj/lib/|$PREFIX/share/ff/ |g" src/ff/ffinit.f
	- sed -i "s|/usr/local/ff/|$PREFIX/share/ff/ |g" src/ff/ffinit.f

However, when trying to add in support for osx-64 I run into a runtime issue where the open command fails to be able to read the path

$PREFIX/share/ff/fferr.dat

which does exist

ff-feedstock/recipe/meta.yaml

Lines 61 to 63 in 7ad7719

	- test -f $PREFIX/share/ff/fferr.dat
	- test -f $PREFIX/share/ff/ffperm5.dat
	- test -f $PREFIX/share/ff/ffwarn.dat

and instead reports failing to open the path

/Users/runner/minifferr.dat

with

open: error: could not open /Users/runner/minifferr.dat

where /Users/runner/mini are the first 19 characters of $PREFIX for the conda-build process for osx-64.

I originally assumed that this had to come down to some strange difference in FFLAGS being set by fortran-compiler across different platforms, given that both Linux and macOS use gfortran for fortran-compiler but after testing the build process outside of conda-build as a pixi project on a macOS x86 GitHub Actions runner and having everything pass at runtime as expected (https://github.com/matthewfeickert/debug-ff-macos-delete) it seems that this is not a compiler/linker flag issue but a problem between something in the Fortran 77 code and the conda-build $PREFIX path. I had attempted to work around issues with the $PREFIX path length earlier by patching in a longer character length for the relevant variable, but that doesn't seem to be the issue here.

Can you offer any advice on how to attempt to debug this?

[matthewfeickert]

The (I assume) relevant bits of the sed injected patched src/ff/ffinit.f from @isuruf's cat commit (also thanks for taking a look at this!):

*	first try - my home directory
	path = '/Users/runner/miniforge3/conda-bld/ff_1736136218744/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeh/share/ff/ '
	fullname = path(1:index(path,' ')-1)//name
	open(ifile,file=fullname,status='OLD',err=30)
	return
   30	continue
*	second try - the system directory
	path = '/Users/runner/miniforge3/conda-bld/ff_1736136218744/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeh/share/ff/ '
	fullname = path(1:index(path,' ')-1)//name
	open(ifile,file=fullname,status='OLD',err=40)
	return
*	file could not be found
   40	continue
	print *,'ffopen: error: could not open ',fullname
	print *,'        adjust path in ffopen (ffinit.f)'
	ier = -1
*###] ffopen:

[matthewfeickert]

Maybe relevant environmental flags:

FFLAGS=-march=core2 -mtune=haswell -ftree-vectorize -fPIC -fstack-protector -O2 -pipe -isystem $PREFIX/include -fdebug-prefix-map=/Users/runner/miniforge3/conda-bld/ff_1736136218744/work=/usr/local/src/conda/ff-static-2.0.0 -fdebug-prefix-map=$PREFIX=/usr/local/src/conda-prefix

[matthewfeickert]

Looking at the full log for the osx-64 build from #2 (comment): osx-64_full_log.txt

I see now that the test program output

+ ./npointes
 ====================================================
   FF 2.0, a package to evaluate one-loop integrals
 written by G. J. van Oldenborgh, NIKHEF-H, Amsterdam
 ====================================================
 for the algorithms used see preprint NIKHEF-H 89/17,
 'New Algorithms for One-loop Integrals', by G.J. van
 Oldenborgh and J.A.M. Vermaseren, published in 
 Zeitschrift fuer Physik C46(1990)425.
 ====================================================
 ffini: precx =    4.4408920985006262E-016
 ffini: precc =    4.4408920985006262E-016
 ffini: xalogm =    4.9406564584124654E-324
 ffini: xclogm =    4.9406564584124654E-324
  0.0000000000000E+00 0.9156199386460E+06
 NPOIN: warning: D4 is not yet supported
 NPOIN: warning: B1' seems also not yet supported
 -0.3100623399361E+02-0.2054369006935E+03-0.5269961187649E-14-0.1218492318111E-08
  0.0000000000000E+00 0.1997630170032+305-0.1033542556010E+02-0.6793071440282E+02
 open: error: could not open ��/Users/runner/miniffwarn.dat                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                    
         adjust path in ffopen (ffinit.f)
ffwarn: warning cannot open ffwarn.dat with warning texts
 -0.0000000000000E+00 0.1739461728624E-08-0.0000000000000E+00-0.1286491875423E-08
  0.0000000000000E+00 0.4952559280190E-18-0.0000000000000E+00-0.4571732002955E-18
 ALIJ: error: not implemented
  0.0000000000000E+00
 open: error: could not open ��/Users/runner/minifferr.dat                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                     
         adjust path in ffopen (ffinit.f)
fferr:  warning cannot open fferr.dat with error texts
f
total number of errors and warnings
===================================
fferr: no errors
ffwarn:        1 times  18: ffwarn:  warning:   illegal value for ierr                                      
     (lost at most a factor     8.20    )
ffwarn:        1 times 129: ffwarn:  warning:   illegal value for ierr                                      
     (lost at most a factor     9.75    )
ffwarn:       10 times 163: ffwarn:  warning:   illegal value for ierr                                      
Note: The following floating-point exceptions are signalling: IEEE_UNDERFLOW_FLAG IEEE_DENORMAL
     (lost at most a factor     244.    )
ffwarn:        1 times 164: ffwarn:  warning:   illegal value for ierr                                      
     (lost at most a factor     8.33    )
f

has the error

 open: error: could not open ��/Users/runner/miniffwarn.dat 

so it seems that for x86 macOS only

	path = '/Users/runner/miniforge3/conda-bld/ff_1736136218744/_h_env_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placehold_placeh/share/ff/ '
	fullname = path(1:index(path,' ')-1)//name
	open(ifile,file=fullname,status='OLD',err=30)

is somehow getting null(?) charecters placed at the front of fullname?

[martin-g]

OSX sed command behaves differently than GNU sed when -i is used.

Another solution is to add sed as a build requirement to the recipe.

[matthewfeickert]

OSX sed command behaves differently than GNU sed when -i is used.

Another solution is to add sed as a build requirement to the recipe.

Thanks for your review, @martin-g. sed is already a build requirement of the recipe

ff-feedstock/recipe/meta.yaml

Lines 44 to 49 in 7ad7719

	build:
	- {{ stdlib('c') }}
	- {{ compiler('fortran') }}
	- make
	- libtool
	- sed

specifically to avoid the OS level pattern differences that you mention. I'm quite thankful that conda-forge sed is gsed so that people don't have to worry about these in differences in general!

[martin-g]

Ah, sorry! I didn't check before typing ... Then it must be something else