Linux kernel mirror (for testing) git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git
kernel os linux
1
fork

Configure Feed

Select the types of activity you want to include in your feed.

Merge tag 'for-linus-2019-11-21' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux

Pull pidfd fixlet from Christian Brauner:
"This contains a simple fix for the pidfd poll method. In the original
patchset pidfd_poll() was made to return an unsigned int. However, the
poll method is defined to return a __poll_t. While the unsigned int is
not a huge deal it's just nicer to return a __poll_t.

I've decided to send it right before the 5.4 release mainly so that
stable doesn't need to backport it to both 5.4 and 5.3"

* tag 'for-linus-2019-11-21' of git://git.kernel.org/pub/scm/linux/kernel/git/brauner/linux:
fork: fix pidfd_poll()'s return type

+3 -3
+3 -3
kernel/fork.c
··· 1708 1708 /* 1709 1709 * Poll support for process exit notification. 1710 1710 */ 1711 - static unsigned int pidfd_poll(struct file *file, struct poll_table_struct *pts) 1711 + static __poll_t pidfd_poll(struct file *file, struct poll_table_struct *pts) 1712 1712 { 1713 1713 struct task_struct *task; 1714 1714 struct pid *pid = file->private_data; 1715 - int poll_flags = 0; 1715 + __poll_t poll_flags = 0; 1716 1716 1717 1717 poll_wait(file, &pid->wait_pidfd, pts); 1718 1718 ··· 1724 1724 * group, then poll(2) should block, similar to the wait(2) family. 1725 1725 */ 1726 1726 if (!task || (task->exit_state && thread_group_empty(task))) 1727 - poll_flags = POLLIN | POLLRDNORM; 1727 + poll_flags = EPOLLIN | EPOLLRDNORM; 1728 1728 rcu_read_unlock(); 1729 1729 1730 1730 return poll_flags;