fix(compat.glx-runtime): ln -sf was last-wins, and nothing checked the ABI (mcpp#352)

On Fedora 44 this package produced

    libGLX.so.0 -> /usr/lib/libGLX.so.0        # 32-bit
    libOpenGL.so.0 -> /usr/lib64/libOpenGL.so.0

and the application died with `libGLX.so.0: wrong ELF class: ELFCLASS32`,
surfacing as a silent exit code 255 with no output at all.

THE OBVIOUS DIAGNOSIS IS NOT THE BUG. The issue reads this as "generated
assuming the Debian layout", but `/usr/lib64` is already ahead of `/usr/lib` in
`candidate_dirs()`. Two other things were wrong:

1. **`ln -sf` overwrites.** The loop reached /usr/lib64 first and linked the
   correct file, then reached /usr/lib and replaced it. Last-wins, not
   first-wins. `libOpenGL.so.0` survived as 64-bit only because that host's
   32-bit glvnd does not ship it -- which is why exactly one link in the bug
   report was right, and that is evidence rather than coincidence.

2. **No ABI check anywhere**, including in `required`, which asserted that
   libGLX.so.0 and libGL.so.1 EXIST. Both existed. Both were 32-bit. So the
   package reported success.

There is no layout to assume, which is why a better ordering is not the fix: the
FHS biarch clause makes /usr/lib 32-bit (Fedora/RHEL/SUSE), Debian explicitly
declined that clause and uses /usr/lib/<triplet> so its /usr/lib is 64-bit, and
Arch is a third answer again. An ABI check makes the order stop mattering.

So: enumerate and decide per file (first-wins, keyed by soname), read
e_ident[EI_CLASS] out of the file itself -- five bytes, no external tool, because
`file`/`readelf` may not be installed when a hook runs -- and make `required`
assert 64-bit rather than merely present. Both failure paths now name the cause
and what to install.

Verified against a forged biarch host (32-bit libGLX.so.0 in /usr/lib, the real
one in /usr/lib64): both required links resolve to /usr/lib64 and both are
ELF64. The same fixture under the old loop lands on /usr/lib.

The same three rules -- ask the loader, check the ABI, first hit wins -- are now
one shared module in the xlings index (libs/hostlib.lua), where four call sites
had four answers and three were wrong.

Refs: mcpp-community/mcpp#352
Design: openxlings/xlings .agents/docs/2026-08-07-graphics-experience-industry-survey-and-plan.md §8.1, §9.2

Author: Sunrisepeak

pkgs/c/compat.glx-runtime.lua

@@ -104,25 +104,90 @@ local required = {
     ["libGL.so.1"] = false,
 }
 
+-- Is FILE a 64-bit ELF? e_ident[EI_CLASS] == ELFCLASS64.
+--
+-- Five bytes read directly. `file`/`readelf`/`patchelf` would each answer this
+-- and each may be absent when a hook runs, and a probe that answers "cannot
+-- tell" by assuming "fine" is the bug below.
+local function is_elf64(file)
+    local f = io.open(file, "rb")
+    if not f then return false end
+    local head = f:read(5)
+    f:close()
+    return head ~= nil and #head == 5
+       and head:sub(1, 4) == "\127ELF" and head:byte(5) == 2
+end
+
+-- Link the host's GL runtime into one directory, FIRST HIT WINS, 64-bit only.
+--
+-- openxlings/xlings' mcpp#352: on Fedora 44 this produced
+--     libGLX.so.0 -> /usr/lib/libGLX.so.0
+-- a 32-bit library, and the application died with
+--     libGLX.so.0: wrong ELF class: ELFCLASS32
+--
+-- TWO BUGS, and the obvious diagnosis ("the candidate order assumes Debian") is
+-- not either of them -- `/usr/lib64` is already ahead of `/usr/lib` in the list:
+--
+--   1. `ln -sf` OVERWRITES. The loop reached /usr/lib64 first and linked the
+--      correct file, then reached /usr/lib and replaced it. Last-wins, not
+--      first-wins. `libOpenGL.so.0` survived as 64-bit purely because that host's
+--      32-bit glvnd does not ship it -- which is why exactly one link in the bug
+--      report was right.
+--   2. NO ABI CHECK ANYWHERE, including in `required` below, which asserted that
+--      libGLX.so.0 and libGL.so.1 EXIST. Both existed. Both were 32-bit.
+--
+-- There is no directory layout to assume: the FHS biarch clause makes /usr/lib
+-- 32-bit (Fedora/RHEL/SUSE), Debian explicitly declined that clause and uses
+-- /usr/lib/<triplet> so its /usr/lib is 64-bit, and Arch is a third answer
+-- again. So the fix cannot be a better ordering -- it has to be an ABI check,
+-- which makes the order stop mattering.
 local function link_runtime_libs(outdir)
     os.mkdir(outdir)
+    local claimed = {}
     for _, dir in ipairs(candidate_dirs()) do
         for _, pattern in ipairs(host_gl_patterns) do
-            os.exec(
-                "for lib in " .. sh_quote(dir) .. "/" .. pattern ..
-                "; do [ -e \"$lib\" ] || continue; " ..
-                "ln -sf \"$lib\" " .. sh_quote(outdir) .. "/\"$(basename \"$lib\")\"; " ..
-                "done"
-            )
+            -- Enumerate, then decide per file, instead of letting the shell
+            -- link them: the decision needs the ELF class and "have I already
+            -- taken this name", neither of which a `ln -sf` loop can express.
+            local pipe = io.popen("ls -1 " .. sh_quote(dir) .. "/" .. pattern
+                                  .. " 2>/dev/null")
+            if pipe then
+                for line in pipe:lines() do
+                    local lib = line:gsub("[\r\n]+$", "")
+                    local name = lib:match("[^/]+$")
+                    if lib ~= "" and name and not claimed[name]
+                       and is_elf64(lib) then
+                        claimed[name] = lib
+                        os.exec("ln -sf " .. sh_quote(lib) .. " "
+                                .. sh_quote(path.join(outdir, name)))
+                    end
+                end
+                pipe:close()
+            end
         end
     end
 
     for name, _ in pairs(required) do
-        if not os.isfile(path.join(outdir, name)) then
+        local link = path.join(outdir, name)
+        -- Existence AND ABI. Existence alone passed on the Fedora host with
+        -- both links 32-bit, which is how a broken package reported success and
+        -- the failure surfaced as a silent exit code 255 from the application.
+        if not os.isfile(link) then
             log.error("required host GL runtime library not found: %s", name)
+            log.error("  searched: %s", table.concat(candidate_dirs(), " "))
+            log.error("  install your distro's GL runtime (mesa / libglvnd)")
+            return false
+        end
+        if not is_elf64(link) then
+            log.error("host %s is not 64-bit (%s)", name, claimed[name] or link)
+            log.error("  a 32-bit library here fails at dlopen with")
+            log.error("  `wrong ELF class: ELFCLASS32` and the application")
+            log.error("  exits without output. Install the 64-bit GL runtime.")
             return false
         end
     end
+    log.info("glx-runtime: linked %d host GL libraries (64-bit)",
+             (function() local n = 0 for _ in pairs(claimed) do n = n + 1 end return n end)())
     return true
 end