revert #1667 process_iter() new_only param

On a second thought I realized that process_iter() uses a global
variable, so it's not thread safe.
That means that if the are 2 threads using it, the first thread one calling the function (+ consume the iterator), will "steal" the processes of the second thread.

psutil.cpu_percent() has the same problem. That means we have a problem
can't solve with the current API and requires a lot of thinking on how
to solve it as it's not obvious.

4 files changed, 2 insertions, 34 deletions

diff --git a/HISTORY.rst b/HISTORY.rst
index 605a5648..1634e433 100644
--- a/HISTORY.rst
+++ b/HISTORY.rst
@@ -12,7 +12,6 @@ XXXX-XX-XX
   directory for additional data.  (patch by Javad Karabi)
 - 1652_: [Windows] dropped support for Windows XP and Windows Server 2003.
   Minimum supported Windows version now is Windows Vista.
-- 1667_: added process_iter(new_only=True) parameter.
 - 1671_: [FreeBSD] add CI testing/service for FreeBSD (Cirrus CI).
 - 1677_: [Windows] process exe() will succeed for all process PIDs (instead of
   raising AccessDenied).
diff --git a/docs/index.rst b/docs/index.rst
index 67719992..c5de1802 100644
--- a/docs/index.rst
+++ b/docs/index.rst
@@ -857,7 +857,7 @@ Functions
   .. versionchanged::
     5.6.0 PIDs are returned in sorted order
 
-.. function:: process_iter(attrs=None, ad_value=None, new_only=False)
+.. function:: process_iter(attrs=None, ad_value=None)
 
   Return an iterator yielding a :class:`Process` class instance for all running
   processes on the local machine.
@@ -873,8 +873,6 @@ Functions
   ``info`` attribute attached to the returned :class:`Process` instances.
   If *attrs* is an empty list it will retrieve all process info (slow).
 
-  If *new_only* is true this function will take into consideration only
-  new PIDs which appeared since the last time it was was called.
   Sorting order in which processes are returned is based on their PID.
 
   Example::
@@ -898,18 +896,9 @@ Functions
      3: {'name': 'ksoftirqd/0', 'username': 'root'},
      ...}
 
-  Get only new processes since last call::
-
-    >>> for proc in psutil.process_iter(['pid', 'name'], new_only=True):
-    ...     print(proc.info)
-    ...
-
   .. versionchanged::
     5.3.0 added "attrs" and "ad_value" parameters.
 
-  .. versionchanged::
-    5.7.0 added "new_only" parameter.
-
 .. function:: pid_exists(pid)
 
   Check whether the given PID exists in the current process list. This is
diff --git a/psutil/__init__.py b/psutil/__init__.py
index a58b452c..2f4e147f 100644
--- a/psutil/__init__.py
+++ b/psutil/__init__.py
@@ -1407,7 +1407,7 @@ _pmap = {}
 _lock = threading.Lock()
 
 
-def process_iter(attrs=None, ad_value=None, new_only=False):
+def process_iter(attrs=None, ad_value=None):
     """Return a generator yielding a Process instance for all
     running processes.
 
@@ -1428,8 +1428,6 @@ def process_iter(attrs=None, ad_value=None, new_only=False):
     If *attrs* is an empty list it will retrieve all process info
     (slow).
 
-    If *new_only* is true this function will take into consideration
-    only new PIDs which appeared since the last time it was called.
     """
     def add(pid):
         proc = Process(pid)
@@ -1452,8 +1450,6 @@ def process_iter(attrs=None, ad_value=None, new_only=False):
 
     with _lock:
         ls = list(dict.fromkeys(new_pids).items())
-        if not new_only:
-            ls += list(_pmap.items())
         ls.sort()
 
     for pid, proc in ls:
diff --git a/psutil/tests/test_system.py b/psutil/tests/test_system.py
index c32e8a73..3834209f 100755
--- a/psutil/tests/test_system.py
+++ b/psutil/tests/test_system.py
@@ -105,22 +105,6 @@ class TestProcessAPIs(unittest.TestCase):
                 self.assertGreaterEqual(p.info['pid'], 0)
             assert m.called
 
-    def test_process_iter_new_only(self):
-        ls1 = list(psutil.process_iter(attrs=['pid']))
-        ls2 = list(psutil.process_iter(attrs=['pid'], new_only=True))
-        self.assertGreater(len(ls1), len(ls2))
-        # assume no more than 3 new processes were created in the meantime
-        self.assertIn(len(ls2), [0, 1, 2, 3, 4, 5])
-
-        sproc = get_test_subprocess()
-        ls = list(psutil.process_iter(attrs=['pid'], new_only=True))
-        self.assertIn(len(ls2), [0, 1, 2, 3, 4, 5])
-        for p in ls:
-            if p.pid == sproc.pid:
-                break
-        else:
-            self.fail("subprocess not found")
-
     @unittest.skipIf(PYPY and WINDOWS,
                      "get_test_subprocess() unreliable on PYPY + WINDOWS")
     def test_wait_procs(self):