superenv: handle formulae with runtime CPU detection

Some formulae are able to detect the features of the runtime CPU, and
execute code accordingly. This typically entails 1) the detection of
features of the build-time CPU in order to determine the targets that
the compiler can generate code for, and 2) generating code for the
targets that the compiler can support.

Our filtering of optimization flags can cause misdetection of compiler
features, leading to failed builds [1], and miscompilation even when the
build does not fail [2].

Let's try to fix this by allowing formulae to declare
`ENV.runtime_cpu_detection` which skips the filtering of `-march` and
related flags.

I've also skipped the filtering of the optimisation
level, since it seems to me that if upstream maintainers have gone to
the lengths of writing code that detects runtime hardware, they probably
also know better about appropriate `-O` flags to use.

This is a partial list of formulae that should make use of this feature:
1. apache-arrow
2. fftw
3. gromacs
4. open-mpi
5. openblas

Partially resolves Homebrew/homebrew-core#76537.

[1] open-mpi/ompi#8306 and linked issues/PRs
[2] Homebrew/homebrew-core#76537

Library/Homebrew/extend/ENV/super.rb

@@ -91,10 +91,11 @@ module Superenv
     # g - Enable "-stdlib=libc++" for clang.
     # h - Enable "-stdlib=libstdc++" for clang.
     # K - Don't strip -arch <arch>, -m32, or -m64
+    # d - Don't strip -march=<target>. Use only in formulae that
+    #     have runtime detection of CPU features.
     # w - Pass -no_weak_imports to the linker
     #
     # These flags will also be present:
-    # s - apply fix for sed's Unicode support
     # a - apply fix for apr-1-config path
   end
   alias generic_setup_build_environment setup_build_environment
@@ -314,6 +315,11 @@ module Superenv
     append_to_cccfg "K"
   end
 
+  sig { void }
+  def runtime_cpu_detection
+    append_to_cccfg "d"
+  end
+
   sig { void }
   def cxx11
     append_to_cccfg "x"

Library/Homebrew/shims/super/cc

@@ -173,10 +173,12 @@ class Cmd
 
     case arg
     when /^-g\d?$/, /^-gstabs\d+/, "-gstabs+", /^-ggdb\d?/,
-      /^-march=.+/, /^-mtune=.+/, /^-mcpu=.+/,
+      /^-mtune=.+/, /^-mcpu=.+/, /^-O[0-9zs]?$/,
-      /^-O[0-9zs]?$/, "-fast", "-no-cpp-precomp",
+      "-fast", "-no-cpp-precomp", "-pedantic",
-      "-pedantic", "-pedantic-errors", "-Wno-long-double",
+      "-pedantic-errors", "-Wno-long-double",
       "-Wno-unused-but-set-variable"
+    when /^-march=.+/
+      args << arg if runtime_cpu_detection?
     when "-fopenmp", "-lgomp", "-mno-fused-madd", "-fforce-addr", "-fno-defer-pop",
       "-mno-dynamic-no-pic", "-fearly-inlining", /^-f(?:no-)?inline-functions-called-once/,
       /^-finline-limit/, /^-f(?:no-)?check-new/, "-fno-delete-null-pointer-checks",
@@ -283,7 +285,7 @@ class Cmd
     args << "-pipe"
     args << "-w" unless configure?
     args << "-#{ENV["HOMEBREW_OPTIMIZATION_LEVEL"]}"
-    args.concat(optflags)
+    args.concat(optflags) unless runtime_cpu_detection?
     args.concat(archflags)
     args << "-std=#{@arg0}" if /c[89]9/.match?(@arg0)
     args
@@ -385,6 +387,10 @@ class Cmd
     config.include?("K")
   end
 
+  def runtime_cpu_detection?
+    config.include?("d")
+  end
+
   def no_weak_imports?
     config.include?("w")
   end