Compatibility fix with EL 8

The 'capture_output=True' was introduced in Python 3.7, while Enterprise
Linux 8 is Python 3.6.

Per report in Bodhi update:
https://bodhi.fedoraproject.org/updates/FEDORA-EPEL-2023-45ace77fca

mock/py/mockbuild/podman.py

@@ -88,7 +88,9 @@ def mounted_image(self):
         logger = getLog()
         cmd_mount = [self.podman_binary, "image", "mount", self.image]
         cmd_umount = [self.podman_binary, "image", "umount", self.image]
-        result = subprocess.run(cmd_mount, capture_output=True, check=False, encoding="utf8")
+        result = subprocess.run(cmd_mount, stdout=subprocess.PIPE,
+                                stderr=subprocess.PIPE, check=False,
+                                encoding="utf8")
         if result.returncode:
             message = "Podman mount failed: " + result.stderr
             raise BootstrapError(message)
@@ -101,7 +103,8 @@ def mounted_image(self):
         finally:
             logger.info("umounting image %s (%s) with podman image umount",
                         self.image, mountpoint)
-            subprocess.run(cmd_umount, capture_output=True, check=True)
+            subprocess.run(cmd_umount, stdout=subprocess.PIPE,
+                           stderr=subprocess.PIPE, check=True)
 
     @traceLog()
     def cp(self, destination, tar_cmd):

releng/release-notes-next/epel-8-compat.md

@@ -0,0 +1,4 @@
+Fixes added in Mock 5.1 introduced an EL 8 Python incompatibility (dependency on
+Python `capture_output=True` feature in `subprocess` which has been added in
+Python 3.7 - while EL 8 is on Python 3.6).  This has been fixed in Mock (using
+`stdout=subprocess.PIPE` instead).