MergeProcess: add fallback setcpv call
It's important that this metadata access happens in the parent process,
since closing of file descriptors in the subprocess can prevent access
to open database connections such as that used by the sqlite metadata
cache module.
diff --git a/pym/portage/dbapi/_MergeProcess.py b/pym/portage/dbapi/_MergeProcess.py
index ea7c3e6..afb2e08 100644
--- a/pym/portage/dbapi/_MergeProcess.py
+++ b/pym/portage/dbapi/_MergeProcess.py
@@ -30,6 +30,21 @@
'_buf', '_elog_keys')
def _start(self):
+ # Portage should always call setcpv prior to this
+ # point, but here we have a fallback as a convenience
+ # for external API consumers. It's important that
+ # this metadata access happens in the parent process,
+ # since closing of file descriptors in the subprocess
+ # can prevent access to open database connections such
+ # as that used by the sqlite metadata cache module.
+ cpv = "%s/%s" % (self.mycat, self.mypkg)
+ settings = self.settings
+ if cpv != settings.mycpv or \
+ "IUSE" not in settings.configdict["pkg"]:
+ settings.reload()
+ settings.reset()
+ settings.setcpv(cpv, mydb=self.mydbapi)
+
self._handle_self_reinstall()
super(MergeProcess, self)._start()
@@ -116,9 +131,10 @@
fcntl.fcntl(elog_reader_fd, fcntl.F_GETFL) | os.O_NONBLOCK)
blockers = None
if self.blockers is not None:
- # Query blockers in the main process, since metadata cache
- # queries may not work for some databases from within a
- # subprocess. For example, sqlite is known to misbehave.
+ # Query blockers in the main process, since closing
+ # of file descriptors in the subprocess can prevent
+ # access to open database connections such as that
+ # used by the sqlite metadata cache module.
blockers = self.blockers()
mylink = self.dblink(self.mycat, self.mypkg, settings=self.settings,
treetype=self.treetype, vartree=self.vartree,
