GeezerVison is showing github.com — bpo-35189: Retry fnctl calls on EINTR by nierob · Pull Request #10413 · python/cpython
bpo-35189: Retry fnctl calls on EINTR
#10413
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and
privacy statement . We’ll occasionally send you account related emails.
Already on GitHub?
Sign in
to your account
Uh oh!
There was an error while loading. Please reload this page.
Show file tree
Hide file tree
Changes from all commits
Commits
File filter
Filter by extension
.c
(1)
.py
(1)
.rst
(1)
All 3 file types selected
Viewed files
Conversations
Failed to load comments.
Loading
Uh oh!
There was an error while loading. Please reload this page.
Jump to
Jump to file
Failed to load files.
Loading
Uh oh!
There was an error while loading. Please reload this page.
Diff view
Hide whitespace
Diff view
Hide whitespace
There are no files selected for viewing
Original file line number
Diff line number
Diff line change
Expand Up
@@ -10,6 +10,7 @@
import contextlib
import faulthandler
import fcntl
import os
import select
import signal
Expand Down
Expand Up
@@ -486,5 +487,43 @@ def test_devpoll(self):
self.assertGreaterEqual(dt, self.sleep_time)
class FNTLEINTRTest(EINTRBaseTest):
def _lock(self, lock_func, lock_name):
self.addCleanup(support.unlink, support.TESTFN)
code = '\n'.join((
"import fcntl, time",
"with open('%s', 'wb') as f:" % support.TESTFN,
" fcntl.%s(f, fcntl.LOCK_EX)" % lock_name,
" time.sleep(%s)" % self.sleep_time))
start_time = time.monotonic()
proc = self.subprocess(code)
with kill_on_error(proc):
with open(support.TESTFN, 'wb') as f:
while True: # synchronize the subprocess
dt = time.monotonic() - start_time
if dt > 60.0:
raise Exception("failed to sync child in %.1f sec" % dt)
try:
lock_func(f, fcntl.LOCK_EX | fcntl.LOCK_NB)
lock_func(f, fcntl.LOCK_UN)
time.sleep(0.01)
except BlockingIOError:
break
# the child locked the file just a moment ago for 'sleep_time' seconds
# that means that the lock below will block for 'sleep_time' minus some
# potential context switch delay
lock_func(f, fcntl.LOCK_EX)
Comment thread
Outdated
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more .
Here is a race condition that is kind of shared with all other tests in this file. We can not make sure that the function really blocked at least for some time. The race condition would be exposed under heavy load in which the main process was not given resources for at least self.sleep_time.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more .
Tests are run sequentially, so it's not an issue. Moreover, regrtest ensures that TESTFN is unique between two working processes running Python tests.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
dt = time.monotonic() - start_time
self.assertGreaterEqual(dt, self.sleep_time)
self.stop_alarm()
proc.wait()
def test_lockf(self):
self._lock(fcntl.lockf, "lockf")
def test_flock(self):
self._lock(fcntl.flock, "flock")
if __name__ == "__main__":
unittest.main()
Original file line number
Diff line number
Diff line change
@@ -0,0 +1,2 @@
Modify the following fnctl function to retry if interrupted by a signal
(EINTR): flock, lockf, fnctl
Original file line number
Diff line number
Diff line change
Expand Up
@@ -64,6 +64,7 @@ fcntl_fcntl_impl(PyObject *module, int fd, int code, PyObject *arg)
char *str;
Py_ssize_t len;
char buf[1024];
int async_err = 0;
if (arg != NULL) {
int parse_result;
Expand All
@@ -75,12 +76,13 @@ fcntl_fcntl_impl(PyObject *module, int fd, int code, PyObject *arg)
return NULL;
}
memcpy(buf, str, len);
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, code, buf);
Py_END_ALLOW_THREADS
do {
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, code, buf);
Py_END_ALLOW_THREADS
} while (ret == -1 && errno == EINTR && !(async_err = PyErr_CheckSignals()));
if (ret < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return NULL;
Comment thread
Outdated
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more .
That is something that I do not understand, why before result of PyErr_SetFromErrno(PyExc_OSError) was not returned. The new code is in sync with other modules, now.
/repeats
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
return !async_err ? PyErr_SetFromErrno(PyExc_OSError) : NULL;
}
return PyBytes_FromStringAndSize(buf, len);
}
Expand All
@@ -95,12 +97,13 @@ fcntl_fcntl_impl(PyObject *module, int fd, int code, PyObject *arg)
}
}
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, code, (int)int_arg);
Py_END_ALLOW_THREADS
do {
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, code, (int)int_arg);
Py_END_ALLOW_THREADS
} while (ret == -1 && errno == EINTR && !(async_err = PyErr_CheckSignals()));
if (ret < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return NULL;
return !async_err ? PyErr_SetFromErrno(PyExc_OSError) : NULL;
}
return PyLong_FromLong((long)ret);
}
Expand Down
Expand Up
@@ -283,11 +286,14 @@ fcntl_flock_impl(PyObject *module, int fd, int code)
/*[clinic end generated code: output=84059e2b37d2fc64 input=b70a0a41ca22a8a0]*/
{
int ret;
int async_err = 0;
#ifdef HAVE_FLOCK
Py_BEGIN_ALLOW_THREADS
ret = flock(fd, code);
Py_END_ALLOW_THREADS
do {
Py_BEGIN_ALLOW_THREADS
ret = flock(fd, code);
Py_END_ALLOW_THREADS
} while (ret == -1 && errno == EINTR && !(async_err = PyErr_CheckSignals()));
#else
#ifndef LOCK_SH
Expand All
@@ -310,14 +316,15 @@ fcntl_flock_impl(PyObject *module, int fd, int code)
return NULL;
}
l.l_whence = l.l_start = l.l_len = 0;
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, (code & LOCK_NB) ? F_SETLK : F_SETLKW, &l);
Py_END_ALLOW_THREADS
do {
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, (code & LOCK_NB) ? F_SETLK : F_SETLKW, &l);
Py_END_ALLOW_THREADS
} while (ret == -1 && errno == EINTR && !(async_err = PyErr_CheckSignals()));
}
#endif /* HAVE_FLOCK */
if (ret < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return NULL;
return !async_err ? PyErr_SetFromErrno(PyExc_OSError) : NULL;
}
Py_RETURN_NONE;
}
Expand Down
Expand Up
@@ -363,6 +370,7 @@ fcntl_lockf_impl(PyObject *module, int fd, int code, PyObject *lenobj,
/*[clinic end generated code: output=4985e7a172e7461a input=3a5dc01b04371f1a]*/
{
int ret;
int async_err = 0;
#ifndef LOCK_SH
#define LOCK_SH 1 /* shared lock */
Expand Down
Expand Up
@@ -407,13 +415,14 @@ fcntl_lockf_impl(PyObject *module, int fd, int code, PyObject *lenobj,
return NULL;
}
l.l_whence = whence;
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, (code & LOCK_NB) ? F_SETLK : F_SETLKW, &l);
Py_END_ALLOW_THREADS
do {
Py_BEGIN_ALLOW_THREADS
ret = fcntl(fd, (code & LOCK_NB) ? F_SETLK : F_SETLKW, &l);
Py_END_ALLOW_THREADS
} while (ret == -1 && errno == EINTR && !(async_err = PyErr_CheckSignals()));
}
if (ret < 0) {
PyErr_SetFromErrno(PyExc_OSError);
return NULL;
return !async_err ? PyErr_SetFromErrno(PyExc_OSError) : NULL;
}
Py_RETURN_NONE;
}
Expand Down
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.