[PATCH 0/5] y2038: Make ppoll, io_pgetevents and pselect y2038 safe

[PATCH 0/5] y2038: Make ppoll, io_pgetevents and pselect y2038 safe

[Deepa Dinamani]

The series transitions the ppoll, io_getevents, and pselect syscalls
to be y2038 safe.

This is part of the work proceeding for syscalls for y2038.
It is based on the series [1] from Arnd Bergmann.

The overview of the series is as below:
1. Refactor sigmask handling logic for the above syscalls.
2. Provide y2038 safe versions of syscalls for all ABIs.

[1] https://lkml.org/lkml/2018/8/27/651

Deepa Dinamani (5):
  signal: Add set_user_sigmask()
  signal: Add restore_user_sigmask()
  ppoll: use __kernel_timespec
  pselect6: use __kernel_timespec
  io_pgetevents: use __kernel_timespec

 fs/aio.c                 | 138 ++++++++++-----
 fs/eventpoll.c           |  52 +-----
 fs/select.c              | 367 +++++++++++++++++++++++----------------
 include/linux/compat.h   |  20 +++
 include/linux/signal.h   |   4 +
 include/linux/syscalls.h |  20 ++-
 kernel/signal.c          |  78 +++++++++
 7 files changed, 437 insertions(+), 242 deletions(-)

-- 
2.17.1

[PATCH 1/5] signal: Add set_user_sigmask()

[Deepa Dinamani]

Refactor reading sigset from userspace and updating sigmask
into an api.

This is useful for versions of syscalls that pass in the
sigmask and expect the current->sigmask to be changed during
the execution and restored after the execution of the syscall.

With the advent of new y2038 syscalls in the subsequent patches,
we add 2 more new versions of the syscalls(for pselect, ppoll
and io_pgetevents) in addition to the existing native and compat
versions. Adding such an api reduces the logic that would need to
be replicated otherwise.

Note that the calls to sigprocmask() ignored the return value
from the api as the function only returns an error on an invalid
first argument that is hardcoded at these call sites.
The updated logic uses set_current_blocked() instead.

Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
---
 fs/aio.c               | 23 ++++++-------------
 fs/eventpoll.c         | 22 +++++--------------
 fs/select.c            | 50 ++++++++++--------------------------------
 include/linux/compat.h |  4 ++++
 include/linux/signal.h |  2 ++
 kernel/signal.c        | 45 +++++++++++++++++++++++++++++++++++++
 6 files changed, 76 insertions(+), 70 deletions(-)

diff --git a/fs/aio.c b/fs/aio.c
index 2914e8c1b3d2..67e5b1f6fb0f 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -2105,14 +2105,10 @@ SYSCALL_DEFINE6(io_pgetevents,
 	if (usig && copy_from_user(&ksig, usig, sizeof(ksig)))
 		return -EFAULT;
 
-	if (ksig.sigmask) {
-		if (ksig.sigsetsize != sizeof(sigset_t))
-			return -EINVAL;
-		if (copy_from_user(&ksigmask, ksig.sigmask, sizeof(ksigmask)))
-			return -EFAULT;
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL) | sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+
+	ret = set_user_sigmask(ksig.sigmask, &ksigmask, &sigsaved, ksig.sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &ts : NULL);
 	if (signal_pending(current)) {
@@ -2175,14 +2171,9 @@ COMPAT_SYSCALL_DEFINE6(io_pgetevents,
 	if (usig && copy_from_user(&ksig, usig, sizeof(ksig)))
 		return -EFAULT;
 
-	if (ksig.sigmask) {
-		if (ksig.sigsetsize != sizeof(compat_sigset_t))
-			return -EINVAL;
-		if (get_compat_sigset(&ksigmask, ksig.sigmask))
-			return -EFAULT;
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL) | sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+	ret = set_compat_user_sigmask(ksig.sigmask, &ksigmask, &sigsaved, ksig.sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &t : NULL);
 	if (signal_pending(current)) {
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 42bbe6824b4b..2d86eeba837b 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -2223,14 +2223,9 @@ SYSCALL_DEFINE6(epoll_pwait, int, epfd, struct epoll_event __user *, events,
 	 * If the caller wants a certain signal mask to be set during the wait,
 	 * we apply it here.
 	 */
-	if (sigmask) {
-		if (sigsetsize != sizeof(sigset_t))
-			return -EINVAL;
-		if (copy_from_user(&ksigmask, sigmask, sizeof(ksigmask)))
-			return -EFAULT;
-		sigsaved = current->blocked;
-		set_current_blocked(&ksigmask);
-	}
+	error = set_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (error)
+		return error;
 
 	error = do_epoll_wait(epfd, events, maxevents, timeout);
 
@@ -2266,14 +2261,9 @@ COMPAT_SYSCALL_DEFINE6(epoll_pwait, int, epfd,
 	 * If the caller wants a certain signal mask to be set during the wait,
 	 * we apply it here.
 	 */
-	if (sigmask) {
-		if (sigsetsize != sizeof(compat_sigset_t))
-			return -EINVAL;
-		if (get_compat_sigset(&ksigmask, sigmask))
-			return -EFAULT;
-		sigsaved = current->blocked;
-		set_current_blocked(&ksigmask);
-	}
+	err = set_compat_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (err)
+		return err;
 
 	err = do_epoll_wait(epfd, events, maxevents, timeout);
 
diff --git a/fs/select.c b/fs/select.c
index 22b3bf89f051..65c78b4147a2 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -717,16 +717,9 @@ static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
 			return -EINVAL;
 	}
 
-	if (sigmask) {
-		/* XXX: Don't preclude handling different sized sigset_t's.  */
-		if (sigsetsize != sizeof(sigset_t))
-			return -EINVAL;
-		if (copy_from_user(&ksigmask, sigmask, sizeof(ksigmask)))
-			return -EFAULT;
-
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+	ret = set_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = core_sys_select(n, inp, outp, exp, to);
 	ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
@@ -1061,16 +1054,9 @@ SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds,
 			return -EINVAL;
 	}
 
-	if (sigmask) {
-		/* XXX: Don't preclude handling different sized sigset_t's.  */
-		if (sigsetsize != sizeof(sigset_t))
-			return -EINVAL;
-		if (copy_from_user(&ksigmask, sigmask, sizeof(ksigmask)))
-			return -EFAULT;
-
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+	ret = set_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = do_sys_poll(ufds, nfds, to);
 
@@ -1323,15 +1309,9 @@ static long do_compat_pselect(int n, compat_ulong_t __user *inp,
 			return -EINVAL;
 	}
 
-	if (sigmask) {
-		if (sigsetsize != sizeof(compat_sigset_t))
-			return -EINVAL;
-		if (get_compat_sigset(&ksigmask, sigmask))
-			return -EFAULT;
-
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+	ret = set_compat_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = compat_core_sys_select(n, inp, outp, exp, to);
 	ret = compat_poll_select_copy_remaining(&end_time, tsp, 0, ret);
@@ -1389,15 +1369,9 @@ COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
 			return -EINVAL;
 	}
 
-	if (sigmask) {
-		if (sigsetsize != sizeof(compat_sigset_t))
-			return -EINVAL;
-		if (get_compat_sigset(&ksigmask, sigmask))
-			return -EFAULT;
-
-		sigdelsetmask(&ksigmask, sigmask(SIGKILL)|sigmask(SIGSTOP));
-		sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
-	}
+	ret = set_compat_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
+		return ret;
 
 	ret = do_sys_poll(ufds, nfds, to);
 
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 6fb5abdb87be..03d65c509eeb 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -166,6 +166,10 @@ typedef struct {
 	compat_sigset_word	sig[_COMPAT_NSIG_WORDS];
 } compat_sigset_t;
 
+int set_compat_user_sigmask(const compat_sigset_t __user *usigmask,
+			    sigset_t *set, sigset_t *oldset,
+			    size_t sigsetsize);
+
 struct compat_sigaction {
 #ifndef __ARCH_HAS_IRIX_SIGACTION
 	compat_uptr_t			sa_handler;
diff --git a/include/linux/signal.h b/include/linux/signal.h
index 3d4cd5db30a9..403e63d01bcf 100644
--- a/include/linux/signal.h
+++ b/include/linux/signal.h
@@ -263,6 +263,8 @@ extern int group_send_sig_info(int sig, struct siginfo *info,
 			       struct task_struct *p, enum pid_type type);
 extern int __group_send_sig_info(int, struct siginfo *, struct task_struct *);
 extern int sigprocmask(int, sigset_t *, sigset_t *);
+extern int set_user_sigmask(const sigset_t __user *usigmask, sigset_t *set,
+	sigset_t *oldset, size_t sigsetsize);
 extern void set_current_blocked(sigset_t *);
 extern void __set_current_blocked(const sigset_t *);
 extern int show_unhandled_signals;
diff --git a/kernel/signal.c b/kernel/signal.c
index 0831d56a731a..1d72dcddcaaf 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -2738,6 +2738,51 @@ int sigprocmask(int how, sigset_t *set, sigset_t *oldset)
 	return 0;
 }
 
+/*
+ * The api helps set app-provided sigmasks.
+ *
+ * This is useful for syscalls such as ppoll, pselect, io_pgetevents and
+ * epoll_pwait where a new sigmask is passed from userland for the syscalls.
+ */
+int set_user_sigmask(const sigset_t __user *usigmask, sigset_t *set,
+		     sigset_t *oldset, size_t sigsetsize)
+{
+	if (!usigmask)
+		return 0;
+
+	if (sigsetsize != sizeof(sigset_t))
+		return -EINVAL;
+	if (copy_from_user(set, usigmask, sizeof(sigset_t)))
+		return -EFAULT;
+
+	*oldset = current->blocked;
+	set_current_blocked(set);
+
+	return 0;
+}
+EXPORT_SYMBOL(set_user_sigmask);
+
+#ifdef CONFIG_COMPAT
+int set_compat_user_sigmask(const compat_sigset_t __user *usigmask,
+			    sigset_t *set, sigset_t *oldset,
+			    size_t sigsetsize)
+{
+	if (!usigmask)
+		return 0;
+
+	if (sigsetsize != sizeof(compat_sigset_t))
+		return -EINVAL;
+	if (get_compat_sigset(set, usigmask))
+		return -EFAULT;
+
+	*oldset = current->blocked;
+	set_current_blocked(set);
+
+	return 0;
+}
+EXPORT_SYMBOL(set_compat_user_sigmask);
+#endif
+
 /**
  *  sys_rt_sigprocmask - change the list of currently blocked signals
  *  @how: whether to add, remove, or set signals
-- 
2.17.1

[PATCH 2/5] signal: Add restore_user_sigmask()

[Deepa Dinamani]

Refactor the logic to restore the sigmask before the syscall
returns into an api.
This is useful for versions of syscalls that pass in the
sigmask and expect the current->sigmask to be changed during
the execution and restored after the execution of the syscall.

With the advent of new y2038 syscalls in the subsequent patches,
we add 2 more new versions of the syscalls(for pselect, ppoll
and io_pgetevents) in addition to the existing native and compat
versions. Adding such an api reduces the logic that would need to
be replicated otherwise.

Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
---
 fs/aio.c               | 29 +++++---------------
 fs/eventpoll.c         | 30 ++-------------------
 fs/select.c            | 60 ++++++------------------------------------
 include/linux/signal.h |  2 ++
 kernel/signal.c        | 33 +++++++++++++++++++++++
 5 files changed, 51 insertions(+), 103 deletions(-)

diff --git a/fs/aio.c b/fs/aio.c
index 67e5b1f6fb0f..b81c216534d6 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -2111,18 +2111,9 @@ SYSCALL_DEFINE6(io_pgetevents,
 		return ret;
 
 	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &ts : NULL);
-	if (signal_pending(current)) {
-		if (ksig.sigmask) {
-			current->saved_sigmask = sigsaved;
-			set_restore_sigmask();
-		}
-
-		if (!ret)
-			ret = -ERESTARTNOHAND;
-	} else {
-		if (ksig.sigmask)
-			sigprocmask(SIG_SETMASK, &sigsaved, NULL);
-	}
+	restore_user_sigmask(ksig.sigmask, &sigsaved);
+	if (signal_pending(current) && !ret)
+		ret = -ERESTARTNOHAND;
 
 	return ret;
 }
@@ -2176,17 +2167,9 @@ COMPAT_SYSCALL_DEFINE6(io_pgetevents,
 		return ret;
 
 	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &t : NULL);
-	if (signal_pending(current)) {
-		if (ksig.sigmask) {
-			current->saved_sigmask = sigsaved;
-			set_restore_sigmask();
-		}
-		if (!ret)
-			ret = -ERESTARTNOHAND;
-	} else {
-		if (ksig.sigmask)
-			sigprocmask(SIG_SETMASK, &sigsaved, NULL);
-	}
+	restore_user_sigmask(ksig.sigmask, &sigsaved);
+	if (signal_pending(current) && !ret)
+		ret = -ERESTARTNOHAND;
 
 	return ret;
 }
diff --git a/fs/eventpoll.c b/fs/eventpoll.c
index 2d86eeba837b..8a5a1010886b 100644
--- a/fs/eventpoll.c
+++ b/fs/eventpoll.c
@@ -2229,20 +2229,7 @@ SYSCALL_DEFINE6(epoll_pwait, int, epfd, struct epoll_event __user *, events,
 
 	error = do_epoll_wait(epfd, events, maxevents, timeout);
 
-	/*
-	 * If we changed the signal mask, we need to restore the original one.
-	 * In case we've got a signal while waiting, we do not restore the
-	 * signal mask yet, and we allow do_signal() to deliver the signal on
-	 * the way back to userspace, before the signal mask is restored.
-	 */
-	if (sigmask) {
-		if (error == -EINTR) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-			       sizeof(sigsaved));
-			set_restore_sigmask();
-		} else
-			set_current_blocked(&sigsaved);
-	}
+	restore_user_sigmask(sigmask, &sigsaved);
 
 	return error;
 }
@@ -2267,20 +2254,7 @@ COMPAT_SYSCALL_DEFINE6(epoll_pwait, int, epfd,
 
 	err = do_epoll_wait(epfd, events, maxevents, timeout);
 
-	/*
-	 * If we changed the signal mask, we need to restore the original one.
-	 * In case we've got a signal while waiting, we do not restore the
-	 * signal mask yet, and we allow do_signal() to deliver the signal on
-	 * the way back to userspace, before the signal mask is restored.
-	 */
-	if (sigmask) {
-		if (err == -EINTR) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-			       sizeof(sigsaved));
-			set_restore_sigmask();
-		} else
-			set_current_blocked(&sigsaved);
-	}
+	restore_user_sigmask(sigmask, &sigsaved);
 
 	return err;
 }
diff --git a/fs/select.c b/fs/select.c
index 65c78b4147a2..eb9132520197 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -724,19 +724,7 @@ static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
 	ret = core_sys_select(n, inp, outp, exp, to);
 	ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
 
-	if (ret == -ERESTARTNOHAND) {
-		/*
-		 * Don't restore the signal mask yet. Let do_signal() deliver
-		 * the signal on the way back to userspace, before the signal
-		 * mask is restored.
-		 */
-		if (sigmask) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-					sizeof(sigsaved));
-			set_restore_sigmask();
-		}
-	} else if (sigmask)
-		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+	restore_user_sigmask(sigmask, &sigsaved);
 
 	return ret;
 }
@@ -1060,21 +1048,11 @@ SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds,
 
 	ret = do_sys_poll(ufds, nfds, to);
 
+	restore_user_sigmask(sigmask, &sigsaved);
+
 	/* We can restart this syscall, usually */
-	if (ret == -EINTR) {
-		/*
-		 * Don't restore the signal mask yet. Let do_signal() deliver
-		 * the signal on the way back to userspace, before the signal
-		 * mask is restored.
-		 */
-		if (sigmask) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-					sizeof(sigsaved));
-			set_restore_sigmask();
-		}
+	if (ret == -EINTR)
 		ret = -ERESTARTNOHAND;
-	} else if (sigmask)
-		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
 
@@ -1316,19 +1294,7 @@ static long do_compat_pselect(int n, compat_ulong_t __user *inp,
 	ret = compat_core_sys_select(n, inp, outp, exp, to);
 	ret = compat_poll_select_copy_remaining(&end_time, tsp, 0, ret);
 
-	if (ret == -ERESTARTNOHAND) {
-		/*
-		 * Don't restore the signal mask yet. Let do_signal() deliver
-		 * the signal on the way back to userspace, before the signal
-		 * mask is restored.
-		 */
-		if (sigmask) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-					sizeof(sigsaved));
-			set_restore_sigmask();
-		}
-	} else if (sigmask)
-		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
+	restore_user_sigmask(sigmask, &sigsaved);
 
 	return ret;
 }
@@ -1375,21 +1341,11 @@ COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
 
 	ret = do_sys_poll(ufds, nfds, to);
 
+	restore_user_sigmask(sigmask, &sigsaved);
+
 	/* We can restart this syscall, usually */
-	if (ret == -EINTR) {
-		/*
-		 * Don't restore the signal mask yet. Let do_signal() deliver
-		 * the signal on the way back to userspace, before the signal
-		 * mask is restored.
-		 */
-		if (sigmask) {
-			memcpy(&current->saved_sigmask, &sigsaved,
-				sizeof(sigsaved));
-			set_restore_sigmask();
-		}
+	if (ret == -EINTR)
 		ret = -ERESTARTNOHAND;
-	} else if (sigmask)
-		sigprocmask(SIG_SETMASK, &sigsaved, NULL);
 
 	ret = compat_poll_select_copy_remaining(&end_time, tsp, 0, ret);
 
diff --git a/include/linux/signal.h b/include/linux/signal.h
index 403e63d01bcf..ed8be17afe89 100644
--- a/include/linux/signal.h
+++ b/include/linux/signal.h
@@ -265,6 +265,8 @@ extern int __group_send_sig_info(int, struct siginfo *, struct task_struct *);
 extern int sigprocmask(int, sigset_t *, sigset_t *);
 extern int set_user_sigmask(const sigset_t __user *usigmask, sigset_t *set,
 	sigset_t *oldset, size_t sigsetsize);
+extern void restore_user_sigmask(const void __user *usigmask,
+				 sigset_t *sigsaved);
 extern void set_current_blocked(sigset_t *);
 extern void __set_current_blocked(const sigset_t *);
 extern int show_unhandled_signals;
diff --git a/kernel/signal.c b/kernel/signal.c
index 1d72dcddcaaf..457d1abe62a4 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -2783,6 +2783,39 @@ int set_compat_user_sigmask(const compat_sigset_t __user *usigmask,
 EXPORT_SYMBOL(set_compat_user_sigmask);
 #endif
 
+/*
+ * restore_user_sigmask:
+ * usigmask: sigmask passed in from userland.
+ * sigsaved: saved sigmask when the syscall started and changed the sigmask to
+ *           usigmask.
+ *
+ * This is useful for syscalls such as ppoll, pselect, io_pgetevents and
+ * epoll_pwait where a new sigmask is passed in from userland for the syscalls.
+ */
+void restore_user_sigmask(const void __user *usigmask, sigset_t *sigsaved)
+{
+
+	if (!usigmask)
+		return;
+	/*
+	 * When signals are pending, do not restore them here.
+	 * Restoring sigmask here can lead to delivering signals that the above
+	 * syscalls are intended to block because of the sigmask passed in.
+	 */
+	if (signal_pending(current)) {
+		current->saved_sigmask = *sigsaved;
+		set_restore_sigmask();
+		return;
+	}
+
+	/*
+	 * This is needed because the fast syscall return path does not restore
+	 * saved_sigmask when signals are not pending.
+	 */
+	set_current_blocked(sigsaved);
+}
+EXPORT_SYMBOL(restore_user_sigmask);
+
 /**
  *  sys_rt_sigprocmask - change the list of currently blocked signals
  *  @how: whether to add, remove, or set signals
-- 
2.17.1

[PATCH 3/5] ppoll: use __kernel_timespec

[Deepa Dinamani]

struct timespec is not y2038 safe.
struct __kernel_timespec is the new y2038 safe structure for all
syscalls that are using struct timespec.
Update ppoll interfaces to use struct __kernel_timespec.

sigset_t also has different representations on 32 bit and 64 bit
architectures. Hence, we need to support the following different
syscalls:

New y2038 safe syscalls:
(Controlled by CONFIG_64BIT_TIME for 32 bit ABIs)

Native 64 bit(unchanged) and native 32 bit : sys_ppoll
Compat : compat_sys_ppoll_time64

Older y2038 unsafe syscalls:
(Controlled by CONFIG_32BIT_COMPAT_TIME for 32 bit ABIs)

Native 32 bit : ppoll_time32
Compat : compat_sys_ppoll

Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
---
 fs/select.c              | 171 ++++++++++++++++++++++++++-------------
 include/linux/compat.h   |   5 ++
 include/linux/syscalls.h |   5 +-
 3 files changed, 124 insertions(+), 57 deletions(-)

diff --git a/fs/select.c b/fs/select.c
index eb9132520197..653eed2add8e 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -287,12 +287,18 @@ int poll_select_set_timeout(struct timespec64 *to, time64_t sec, long nsec)
 	return 0;
 }
 
+enum poll_time_type {
+	PT_TIMEVAL = 0,
+	PT_OLD_TIMEVAL = 1,
+	PT_TIMESPEC = 2,
+	PT_OLD_TIMESPEC = 3,
+};
+
 static int poll_select_copy_remaining(struct timespec64 *end_time,
 				      void __user *p,
-				      int timeval, int ret)
+				      enum poll_time_type pt_type, int ret)
 {
 	struct timespec64 rts;
-	struct timeval rtv;
 
 	if (!p)
 		return ret;
@@ -310,18 +316,40 @@ static int poll_select_copy_remaining(struct timespec64 *end_time,
 		rts.tv_sec = rts.tv_nsec = 0;
 
 
-	if (timeval) {
-		if (sizeof(rtv) > sizeof(rtv.tv_sec) + sizeof(rtv.tv_usec))
-			memset(&rtv, 0, sizeof(rtv));
-		rtv.tv_sec = rts.tv_sec;
-		rtv.tv_usec = rts.tv_nsec / NSEC_PER_USEC;
+	switch (pt_type) {
+	case PT_TIMEVAL:
+		{
+			struct timeval rtv;
 
-		if (!copy_to_user(p, &rtv, sizeof(rtv)))
+			if (sizeof(rtv) > sizeof(rtv.tv_sec) + sizeof(rtv.tv_usec))
+				memset(&rtv, 0, sizeof(rtv));
+			rtv.tv_sec = rts.tv_sec;
+			rtv.tv_usec = rts.tv_nsec / NSEC_PER_USEC;
+			if (!copy_to_user(p, &rtv, sizeof(rtv)))
+				return ret;
+		}
+		break;
+	case PT_OLD_TIMEVAL:
+		{
+			struct old_timeval32 rtv;
+
+			rtv.tv_sec = rts.tv_sec;
+			rtv.tv_usec = rts.tv_nsec / NSEC_PER_USEC;
+			if (!copy_to_user(p, &rtv, sizeof(rtv)))
+				return ret;
+		}
+		break;
+	case PT_TIMESPEC:
+		if (!put_timespec64(&rts, p))
 			return ret;
-
-	} else if (!put_timespec64(&rts, p))
-		return ret;
-
+		break;
+	case PT_OLD_TIMESPEC:
+		if (!put_old_timespec32(&rts, p))
+			return ret;
+		break;
+	default:
+		BUG();
+	}
 	/*
 	 * If an application puts its timeval in read-only memory, we
 	 * don't want the Linux-specific update to the timeval to
@@ -689,7 +717,7 @@ static int kern_select(int n, fd_set __user *inp, fd_set __user *outp,
 	}
 
 	ret = core_sys_select(n, inp, outp, exp, to);
-	ret = poll_select_copy_remaining(&end_time, tvp, 1, ret);
+	ret = poll_select_copy_remaining(&end_time, tvp, PT_TIMEVAL, ret);
 
 	return ret;
 }
@@ -722,7 +750,7 @@ static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
 		return ret;
 
 	ret = core_sys_select(n, inp, outp, exp, to);
-	ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_TIMESPEC, ret);
 
 	restore_user_sigmask(sigmask, &sigsaved);
 
@@ -1026,7 +1054,7 @@ SYSCALL_DEFINE3(poll, struct pollfd __user *, ufds, unsigned int, nfds,
 }
 
 SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds,
-		struct timespec __user *, tsp, const sigset_t __user *, sigmask,
+		struct __kernel_timespec __user *, tsp, const sigset_t __user *, sigmask,
 		size_t, sigsetsize)
 {
 	sigset_t ksigmask, sigsaved;
@@ -1054,60 +1082,50 @@ SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds, unsigned int, nfds,
 	if (ret == -EINTR)
 		ret = -ERESTARTNOHAND;
 
-	ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_TIMESPEC, ret);
 
 	return ret;
 }
 
-#ifdef CONFIG_COMPAT
-#define __COMPAT_NFDBITS       (8 * sizeof(compat_ulong_t))
+#if defined(CONFIG_COMPAT_32BIT_TIME) && !defined(CONFIG_64BIT)
 
-static
-int compat_poll_select_copy_remaining(struct timespec64 *end_time, void __user *p,
-				      int timeval, int ret)
+SYSCALL_DEFINE5(ppoll_time32, struct pollfd __user *, ufds, unsigned int, nfds,
+		struct old_timespec32 __user *, tsp, const sigset_t __user *, sigmask,
+		size_t, sigsetsize)
 {
-	struct timespec64 ts;
+	sigset_t ksigmask, sigsaved;
+	struct timespec64 ts, end_time, *to = NULL;
+	int ret;
 
-	if (!p)
-		return ret;
+	if (tsp) {
+		if (get_old_timespec32(&ts, tsp))
+			return -EFAULT;
 
-	if (current->personality & STICKY_TIMEOUTS)
-		goto sticky;
+		to = &end_time;
+		if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
+			return -EINVAL;
+	}
 
-	/* No update for zero timeout */
-	if (!end_time->tv_sec && !end_time->tv_nsec)
+	ret = set_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
 		return ret;
 
-	ktime_get_ts64(&ts);
-	ts = timespec64_sub(*end_time, ts);
-	if (ts.tv_sec < 0)
-		ts.tv_sec = ts.tv_nsec = 0;
-
-	if (timeval) {
-		struct old_timeval32 rtv;
-
-		rtv.tv_sec = ts.tv_sec;
-		rtv.tv_usec = ts.tv_nsec / NSEC_PER_USEC;
+	ret = do_sys_poll(ufds, nfds, to);
 
-		if (!copy_to_user(p, &rtv, sizeof(rtv)))
-			return ret;
-	} else {
-		if (!put_old_timespec32(&ts, p))
-			return ret;
+	/* We can restart this syscall, usually */
+	if (ret == -EINTR) {
+		restore_user_sigmask(sigmask, &sigsaved);
+		ret = -ERESTARTNOHAND;
 	}
-	/*
-	 * If an application puts its timeval in read-only memory, we
-	 * don't want the Linux-specific update to the timeval to
-	 * cause a fault after the select has completed
-	 * successfully. However, because we're not updating the
-	 * timeval, we can't restart the system call.
-	 */
 
-sticky:
-	if (ret == -ERESTARTNOHAND)
-		ret = -EINTR;
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_OLD_TIMESPEC, ret);
+
 	return ret;
 }
+#endif
+
+#ifdef CONFIG_COMPAT
+#define __COMPAT_NFDBITS       (8 * sizeof(compat_ulong_t))
 
 /*
  * Ooo, nasty.  We need here to frob 32-bit unsigned longs to
@@ -1239,7 +1257,7 @@ static int do_compat_select(int n, compat_ulong_t __user *inp,
 	}
 
 	ret = compat_core_sys_select(n, inp, outp, exp, to);
-	ret = compat_poll_select_copy_remaining(&end_time, tvp, 1, ret);
+	ret = poll_select_copy_remaining(&end_time, tvp, PT_OLD_TIMEVAL, ret);
 
 	return ret;
 }
@@ -1292,7 +1310,7 @@ static long do_compat_pselect(int n, compat_ulong_t __user *inp,
 		return ret;
 
 	ret = compat_core_sys_select(n, inp, outp, exp, to);
-	ret = compat_poll_select_copy_remaining(&end_time, tsp, 0, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_OLD_TIMESPEC, ret);
 
 	restore_user_sigmask(sigmask, &sigsaved);
 
@@ -1318,6 +1336,7 @@ COMPAT_SYSCALL_DEFINE6(pselect6, int, n, compat_ulong_t __user *, inp,
 				 sigsetsize);
 }
 
+#if defined(CONFIG_COMPAT_32BIT_TIME)
 COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
 	unsigned int,  nfds, struct old_timespec32 __user *, tsp,
 	const compat_sigset_t __user *, sigmask, compat_size_t, sigsetsize)
@@ -1347,8 +1366,48 @@ COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
 	if (ret == -EINTR)
 		ret = -ERESTARTNOHAND;
 
-	ret = compat_poll_select_copy_remaining(&end_time, tsp, 0, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_OLD_TIMESPEC, ret);
+
+	return ret;
+}
+#endif
+
+/* New compat syscall for 64 bit time_t*/
+#if defined(CONFIG_64BIT_TIME)
+
+COMPAT_SYSCALL_DEFINE5(ppoll_time64, struct pollfd __user *, ufds,
+	unsigned int,  nfds, struct __kernel_timespec __user *, tsp,
+	const compat_sigset_t __user *, sigmask, compat_size_t, sigsetsize)
+{
+	sigset_t ksigmask, sigsaved;
+	struct timespec64 ts, end_time, *to = NULL;
+	int ret;
+
+	if (tsp) {
+		if (get_timespec64(&ts, tsp))
+			return -EFAULT;
+
+		to = &end_time;
+		if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
+			return -EINVAL;
+	}
+
+	ret = set_compat_user_sigmask(sigmask, &ksigmask, &sigsaved, sigsetsize);
+	if (ret)
+		return ret;
+
+	ret = do_sys_poll(ufds, nfds, to);
+
+	/* We can restart this syscall, usually */
+	if (ret == -EINTR) {
+		restore_user_sigmask(sigmask, &sigsaved);
+		ret = -ERESTARTNOHAND;
+	}
+
+	ret = poll_select_copy_remaining(&end_time, tsp, PT_TIMESPEC, ret);
 
 	return ret;
 }
+
+#endif
 #endif
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 03d65c509eeb..349a2d98e450 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -646,6 +646,11 @@ asmlinkage long compat_sys_ppoll(struct pollfd __user *ufds,
 				 struct old_timespec32 __user *tsp,
 				 const compat_sigset_t __user *sigmask,
 				 compat_size_t sigsetsize);
+asmlinkage long compat_sys_ppoll_time64(struct pollfd __user *ufds,
+				 unsigned int nfds,
+				 struct __kernel_timespec __user *tsp,
+				 const compat_sigset_t __user *sigmask,
+				 compat_size_t sigsetsize);
 
 /* fs/signalfd.c */
 asmlinkage long compat_sys_signalfd4(int ufd,
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index 82682b69435e..9755e70cfbb0 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -470,7 +470,10 @@ asmlinkage long sys_pselect6(int, fd_set __user *, fd_set __user *,
 			     fd_set __user *, struct timespec __user *,
 			     void __user *);
 asmlinkage long sys_ppoll(struct pollfd __user *, unsigned int,
-			  struct timespec __user *, const sigset_t __user *,
+			  struct __kernel_timespec __user *, const sigset_t __user *,
+			  size_t);
+asmlinkage long sys_ppoll_time32(struct pollfd __user *, unsigned int,
+			  struct old_timespec32 __user *, const sigset_t __user *,
 			  size_t);
 
 /* fs/signalfd.c */
-- 
2.17.1

Re: [PATCH 3/5] ppoll: use __kernel_timespec

[Arnd Bergmann]

On Sat, Sep 1, 2018 at 10:48 PM Deepa Dinamani <deepa.kernel@gmail.com> wrote:

> +
> +       /* We can restart this syscall, usually */
> +       if (ret == -EINTR) {
> +               restore_user_sigmask(sigmask, &sigsaved);
> +               ret = -ERESTARTNOHAND;
> +       }

I need a little help understanding this part: In the two existing copies of that
function, you have changed it to

        restore_user_sigmask(sigmask, &sigsaved);
        if (ret == -EINTR)
                ret = -ERESTARTNOHAND;

in the previous patch, but the new new copies you add here do

       if (ret == -EINTR) {
               restore_user_sigmask(sigmask, &sigsaved);
               ret = -ERESTARTNOHAND;
       }

Is that intentional, or did you accidently leave a version in here that
you had changed afterwards?

       Arnd

Re: [PATCH 3/5] ppoll: use __kernel_timespec

[Deepa Dinamani]

On Fri, Sep 14, 2018 at 8:57 AM Arnd Bergmann <arnd@arndb.de> wrote:
>
> On Sat, Sep 1, 2018 at 10:48 PM Deepa Dinamani <deepa.kernel@gmail.com> wrote:
>
> > +
> > +       /* We can restart this syscall, usually */
> > +       if (ret == -EINTR) {
> > +               restore_user_sigmask(sigmask, &sigsaved);
> > +               ret = -ERESTARTNOHAND;
> > +       }
>
> I need a little help understanding this part: In the two existing copies of that
> function, you have changed it to
>
>         restore_user_sigmask(sigmask, &sigsaved);
>         if (ret == -EINTR)
>                 ret = -ERESTARTNOHAND;
>
> in the previous patch, but the new new copies you add here do
>
>        if (ret == -EINTR) {
>                restore_user_sigmask(sigmask, &sigsaved);
>                ret = -ERESTARTNOHAND;
>        }
>
> Is that intentional, or did you accidently leave a version in here that
> you had changed afterwards?

Oops, you are right. This wasn't intentional.
I will send out a v2.

Thanks,
Deepa

[PATCH 4/5] pselect6: use __kernel_timespec

[Deepa Dinamani]

struct timespec is not y2038 safe.
struct __kernel_timespec is the new y2038 safe structure for all
syscalls that are using struct timespec.
Update pselect interfaces to use struct __kernel_timespec.

sigset_t also has different representations on 32 bit and 64 bit
architectures. Hence, we need to support the following different
syscalls:

New y2038 safe syscalls:
(Controlled by CONFIG_64BIT_TIME for 32 bit ABIs)

Native 64 bit(unchanged) and native 32 bit : sys_pselect6
Compat : compat_sys_pselect6_time64

Older y2038 unsafe syscalls:
(Controlled by CONFIG_32BIT_COMPAT_TIME for 32 bit ABIs)

Native 32 bit : pselect6_time32
Compat : compat_sys_pselect6

Note that all other versions of select syscalls will not have
y2038 safe versions.

Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
---
 fs/select.c              | 98 ++++++++++++++++++++++++++++++++++------
 include/linux/compat.h   |  5 ++
 include/linux/syscalls.h |  5 +-
 3 files changed, 94 insertions(+), 14 deletions(-)

diff --git a/fs/select.c b/fs/select.c
index 653eed2add8e..b1a9f10b521a 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -729,16 +729,27 @@ SYSCALL_DEFINE5(select, int, n, fd_set __user *, inp, fd_set __user *, outp,
 }
 
 static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
-		       fd_set __user *exp, struct timespec __user *tsp,
-		       const sigset_t __user *sigmask, size_t sigsetsize)
+		       fd_set __user *exp, void __user *tsp,
+		       const sigset_t __user *sigmask, size_t sigsetsize,
+		       enum poll_time_type type)
 {
 	sigset_t ksigmask, sigsaved;
 	struct timespec64 ts, end_time, *to = NULL;
 	int ret;
 
 	if (tsp) {
-		if (get_timespec64(&ts, tsp))
-			return -EFAULT;
+		switch (type) {
+		case PT_TIMESPEC:
+			if (get_timespec64(&ts, tsp))
+				return -EFAULT;
+			break;
+		case PT_OLD_TIMESPEC:
+			if (get_old_timespec32(&ts, tsp))
+				return -EFAULT;
+			break;
+		default:
+			BUG();
+		}
 
 		to = &end_time;
 		if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
@@ -750,7 +761,7 @@ static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
 		return ret;
 
 	ret = core_sys_select(n, inp, outp, exp, to);
-	ret = poll_select_copy_remaining(&end_time, tsp, PT_TIMESPEC, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, type, ret);
 
 	restore_user_sigmask(sigmask, &sigsaved);
 
@@ -764,7 +775,27 @@ static long do_pselect(int n, fd_set __user *inp, fd_set __user *outp,
  * the sigset size.
  */
 SYSCALL_DEFINE6(pselect6, int, n, fd_set __user *, inp, fd_set __user *, outp,
-		fd_set __user *, exp, struct timespec __user *, tsp,
+		fd_set __user *, exp, struct __kernel_timespec __user *, tsp,
+		void __user *, sig)
+{
+	size_t sigsetsize = 0;
+	sigset_t __user *up = NULL;
+
+	if (sig) {
+		if (!access_ok(VERIFY_READ, sig, sizeof(void *)+sizeof(size_t))
+		    || __get_user(up, (sigset_t __user * __user *)sig)
+		    || __get_user(sigsetsize,
+				(size_t __user *)(sig+sizeof(void *))))
+			return -EFAULT;
+	}
+
+	return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize, PT_TIMESPEC);
+}
+
+#if defined(CONFIG_COMPAT_32BIT_TIME) && !defined(CONFIG_64BIT)
+
+SYSCALL_DEFINE6(pselect6_time32, int, n, fd_set __user *, inp, fd_set __user *, outp,
+		fd_set __user *, exp, struct old_timespec32 __user *, tsp,
 		void __user *, sig)
 {
 	size_t sigsetsize = 0;
@@ -778,9 +809,11 @@ SYSCALL_DEFINE6(pselect6, int, n, fd_set __user *, inp, fd_set __user *, outp,
 			return -EFAULT;
 	}
 
-	return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize);
+	return do_pselect(n, inp, outp, exp, tsp, up, sigsetsize, PT_OLD_TIMESPEC);
 }
 
+#endif
+
 #ifdef __ARCH_WANT_SYS_OLD_SELECT
 struct sel_arg_struct {
 	unsigned long n;
@@ -1289,16 +1322,26 @@ COMPAT_SYSCALL_DEFINE1(old_select, struct compat_sel_arg_struct __user *, arg)
 
 static long do_compat_pselect(int n, compat_ulong_t __user *inp,
 	compat_ulong_t __user *outp, compat_ulong_t __user *exp,
-	struct old_timespec32 __user *tsp, compat_sigset_t __user *sigmask,
-	compat_size_t sigsetsize)
+	void __user *tsp, compat_sigset_t __user *sigmask,
+	compat_size_t sigsetsize, enum poll_time_type type)
 {
 	sigset_t ksigmask, sigsaved;
 	struct timespec64 ts, end_time, *to = NULL;
 	int ret;
 
 	if (tsp) {
-		if (get_old_timespec32(&ts, tsp))
-			return -EFAULT;
+		switch (type) {
+		case PT_OLD_TIMESPEC:
+			if (get_old_timespec32(&ts, tsp))
+				return -EFAULT;
+			break;
+		case PT_TIMESPEC:
+			if (get_old_timespec32(&ts, tsp))
+				return -EFAULT;
+			break;
+		default:
+			BUG();
+		}
 
 		to = &end_time;
 		if (poll_select_set_timeout(to, ts.tv_sec, ts.tv_nsec))
@@ -1310,13 +1353,39 @@ static long do_compat_pselect(int n, compat_ulong_t __user *inp,
 		return ret;
 
 	ret = compat_core_sys_select(n, inp, outp, exp, to);
-	ret = poll_select_copy_remaining(&end_time, tsp, PT_OLD_TIMESPEC, ret);
+	ret = poll_select_copy_remaining(&end_time, tsp, type, ret);
 
 	restore_user_sigmask(sigmask, &sigsaved);
 
 	return ret;
 }
 
+#if defined(CONFIG_64BIT_TIME)
+
+COMPAT_SYSCALL_DEFINE6(pselect6_time64, int, n, compat_ulong_t __user *, inp,
+	compat_ulong_t __user *, outp, compat_ulong_t __user *, exp,
+	struct __kernel_timespec __user *, tsp, void __user *, sig)
+{
+	compat_size_t sigsetsize = 0;
+	compat_uptr_t up = 0;
+
+	if (sig) {
+		if (!access_ok(VERIFY_READ, sig,
+				sizeof(compat_uptr_t)+sizeof(compat_size_t)) ||
+				__get_user(up, (compat_uptr_t __user *)sig) ||
+				__get_user(sigsetsize,
+				(compat_size_t __user *)(sig+sizeof(up))))
+			return -EFAULT;
+	}
+
+	return do_compat_pselect(n, inp, outp, exp, tsp, compat_ptr(up),
+				 sigsetsize, PT_TIMESPEC);
+}
+
+#endif
+
+#if defined(CONFIG_COMPAT_32BIT_TIME)
+
 COMPAT_SYSCALL_DEFINE6(pselect6, int, n, compat_ulong_t __user *, inp,
 	compat_ulong_t __user *, outp, compat_ulong_t __user *, exp,
 	struct old_timespec32 __user *, tsp, void __user *, sig)
@@ -1332,10 +1401,13 @@ COMPAT_SYSCALL_DEFINE6(pselect6, int, n, compat_ulong_t __user *, inp,
 				(compat_size_t __user *)(sig+sizeof(up))))
 			return -EFAULT;
 	}
+
 	return do_compat_pselect(n, inp, outp, exp, tsp, compat_ptr(up),
-				 sigsetsize);
+				 sigsetsize, PT_OLD_TIMESPEC);
 }
 
+#endif
+
 #if defined(CONFIG_COMPAT_32BIT_TIME)
 COMPAT_SYSCALL_DEFINE5(ppoll, struct pollfd __user *, ufds,
 	unsigned int,  nfds, struct old_timespec32 __user *, tsp,
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 349a2d98e450..6896e6e51c00 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -641,6 +641,11 @@ asmlinkage long compat_sys_pselect6(int n, compat_ulong_t __user *inp,
 				    compat_ulong_t __user *exp,
 				    struct old_timespec32 __user *tsp,
 				    void __user *sig);
+asmlinkage long compat_sys_pselect6_time64(int n, compat_ulong_t __user *inp,
+				    compat_ulong_t __user *outp,
+				    compat_ulong_t __user *exp,
+				    struct __kernel_timespec __user *tsp,
+				    void __user *sig);
 asmlinkage long compat_sys_ppoll(struct pollfd __user *ufds,
 				 unsigned int nfds,
 				 struct old_timespec32 __user *tsp,
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index 9755e70cfbb0..e9cd0409c3fe 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -467,7 +467,10 @@ asmlinkage long sys_sendfile64(int out_fd, int in_fd,
 
 /* fs/select.c */
 asmlinkage long sys_pselect6(int, fd_set __user *, fd_set __user *,
-			     fd_set __user *, struct timespec __user *,
+			     fd_set __user *, struct __kernel_timespec __user *,
+			     void __user *);
+asmlinkage long sys_pselect6_time32(int, fd_set __user *, fd_set __user *,
+			     fd_set __user *, struct old_timespec32 __user *,
 			     void __user *);
 asmlinkage long sys_ppoll(struct pollfd __user *, unsigned int,
 			  struct __kernel_timespec __user *, const sigset_t __user *,
-- 
2.17.1

[PATCH 5/5] io_pgetevents: use __kernel_timespec

[Deepa Dinamani]

struct timespec is not y2038 safe.
struct __kernel_timespec is the new y2038 safe structure for all
syscalls that are using struct timespec.
Update io_pgetevents interfaces to use struct __kernel_timespec.

sigset_t also has different representations on 32 bit and 64 bit
architectures. Hence, we need to support the following different
syscalls:

New y2038 safe syscalls:
(Controlled by CONFIG_64BIT_TIME for 32 bit ABIs)

Native 64 bit(unchanged) and native 32 bit : sys_io_pgetevents
Compat : compat_sys_io_pgetevents_time64

Older y2038 unsafe syscalls:
(Controlled by CONFIG_32BIT_COMPAT_TIME for 32 bit ABIs)

Native 32 bit : sys_io_pgetevents_time32
Compat : compat_sys_io_pgetevents

Note that io_getevents syscalls do not have a y2038 safe solution.

Signed-off-by: Deepa Dinamani <deepa.kernel@gmail.com>
---
 fs/aio.c                 | 88 ++++++++++++++++++++++++++++++++++++++--
 include/linux/compat.h   |  6 +++
 include/linux/syscalls.h | 10 ++++-
 3 files changed, 99 insertions(+), 5 deletions(-)

diff --git a/fs/aio.c b/fs/aio.c
index b81c216534d6..9560f147d7a5 100644
--- a/fs/aio.c
+++ b/fs/aio.c
@@ -2063,11 +2063,13 @@ static long do_io_getevents(aio_context_t ctx_id,
  *	specifies an infinite timeout. Note that the timeout pointed to by
  *	timeout is relative.  Will fail with -ENOSYS if not implemented.
  */
+#if !defined(CONFIG_64BIT_TIME) || defined(CONFIG_64BIT)
+
 SYSCALL_DEFINE5(io_getevents, aio_context_t, ctx_id,
 		long, min_nr,
 		long, nr,
 		struct io_event __user *, events,
-		struct timespec __user *, timeout)
+		struct __kernel_timespec __user *, timeout)
 {
 	struct timespec64	ts;
 	int			ret;
@@ -2081,6 +2083,8 @@ SYSCALL_DEFINE5(io_getevents, aio_context_t, ctx_id,
 	return ret;
 }
 
+#endif
+
 struct __aio_sigset {
 	const sigset_t __user	*sigmask;
 	size_t		sigsetsize;
@@ -2091,7 +2095,7 @@ SYSCALL_DEFINE6(io_pgetevents,
 		long, min_nr,
 		long, nr,
 		struct io_event __user *, events,
-		struct timespec __user *, timeout,
+		struct __kernel_timespec __user *, timeout,
 		const struct __aio_sigset __user *, usig)
 {
 	struct __aio_sigset	ksig = { NULL, };
@@ -2118,7 +2122,44 @@ SYSCALL_DEFINE6(io_pgetevents,
 	return ret;
 }
 
-#ifdef CONFIG_COMPAT
+#if defined(CONFIG_COMPAT_32BIT_TIME) && !defined(CONFIG_64BIT)
+
+SYSCALL_DEFINE6(io_pgetevents_time32,
+		aio_context_t, ctx_id,
+		long, min_nr,
+		long, nr,
+		struct io_event __user *, events,
+		struct old_timespec32 __user *, timeout,
+		const struct __aio_sigset __user *, usig)
+{
+	struct __aio_sigset	ksig = { NULL, };
+	sigset_t		ksigmask, sigsaved;
+	struct timespec64	ts;
+	int ret;
+
+	if (timeout && unlikely(get_old_timespec32(&ts, timeout)))
+		return -EFAULT;
+
+	if (usig && copy_from_user(&ksig, usig, sizeof(ksig)))
+		return -EFAULT;
+
+
+	ret = set_user_sigmask(ksig.sigmask, &ksigmask, &sigsaved, ksig.sigsetsize);
+	if (ret)
+		return ret;
+
+	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &ts : NULL);
+	restore_user_sigmask(ksig.sigmask, &sigsaved);
+	if (signal_pending(current) && !ret)
+		ret = -ERESTARTNOHAND;
+
+	return ret;
+}
+
+#endif
+
+#if defined(CONFIG_COMPAT_32BIT_TIME)
+
 COMPAT_SYSCALL_DEFINE5(io_getevents, compat_aio_context_t, ctx_id,
 		       compat_long_t, min_nr,
 		       compat_long_t, nr,
@@ -2137,12 +2178,17 @@ COMPAT_SYSCALL_DEFINE5(io_getevents, compat_aio_context_t, ctx_id,
 	return ret;
 }
 
+#endif
+
+#ifdef CONFIG_COMPAT
 
 struct __compat_aio_sigset {
 	compat_sigset_t __user	*sigmask;
 	compat_size_t		sigsetsize;
 };
 
+#if defined(CONFIG_COMPAT_32BIT_TIME)
+
 COMPAT_SYSCALL_DEFINE6(io_pgetevents,
 		compat_aio_context_t, ctx_id,
 		compat_long_t, min_nr,
@@ -2173,4 +2219,40 @@ COMPAT_SYSCALL_DEFINE6(io_pgetevents,
 
 	return ret;
 }
+
+#endif
+
+#if defined(CONFIG_64BIT_TIME)
+
+COMPAT_SYSCALL_DEFINE6(io_pgetevents_time64,
+		compat_aio_context_t, ctx_id,
+		compat_long_t, min_nr,
+		compat_long_t, nr,
+		struct io_event __user *, events,
+		struct __kernel_timespec __user *, timeout,
+		const struct __compat_aio_sigset __user *, usig)
+{
+	struct __compat_aio_sigset ksig = { NULL, };
+	sigset_t ksigmask, sigsaved;
+	struct timespec64 t;
+	int ret;
+
+	if (timeout && get_timespec64(&t, timeout))
+		return -EFAULT;
+
+	if (usig && copy_from_user(&ksig, usig, sizeof(ksig)))
+		return -EFAULT;
+
+	ret = set_compat_user_sigmask(ksig.sigmask, &ksigmask, &sigsaved, ksig.sigsetsize);
+	if (ret)
+		return ret;
+
+	ret = do_io_getevents(ctx_id, min_nr, nr, events, timeout ? &t : NULL);
+	restore_user_sigmask(ksig.sigmask, &sigsaved);
+	if (signal_pending(current) && !ret)
+		ret = -ERESTARTNOHAND;
+
+	return ret;
+}
+#endif
 #endif
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 6896e6e51c00..50cd0329c8bf 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -556,6 +556,12 @@ asmlinkage long compat_sys_io_pgetevents(compat_aio_context_t ctx_id,
 					struct io_event __user *events,
 					struct old_timespec32 __user *timeout,
 					const struct __compat_aio_sigset __user *usig);
+asmlinkage long compat_sys_io_pgetevents_time64(compat_aio_context_t ctx_id,
+					compat_long_t min_nr,
+					compat_long_t nr,
+					struct io_event __user *events,
+					struct __kernel_timespec __user *timeout,
+					const struct __compat_aio_sigset __user *usig);
 
 /* fs/cookies.c */
 asmlinkage long compat_sys_lookup_dcookie(u32, u32, char __user *, compat_size_t);
diff --git a/include/linux/syscalls.h b/include/linux/syscalls.h
index e9cd0409c3fe..3ff0e29c082c 100644
--- a/include/linux/syscalls.h
+++ b/include/linux/syscalls.h
@@ -297,12 +297,18 @@ asmlinkage long sys_io_getevents(aio_context_t ctx_id,
 				long min_nr,
 				long nr,
 				struct io_event __user *events,
-				struct timespec __user *timeout);
+				struct __kernel_timespec __user *timeout);
 asmlinkage long sys_io_pgetevents(aio_context_t ctx_id,
 				long min_nr,
 				long nr,
 				struct io_event __user *events,
-				struct timespec __user *timeout,
+				struct __kernel_timespec __user *timeout,
+				const struct __aio_sigset *sig);
+asmlinkage long sys_io_pgetevents_time32(aio_context_t ctx_id,
+				long min_nr,
+				long nr,
+				struct io_event __user *events,
+				struct old_timespec32 __user *timeout,
 				const struct __aio_sigset *sig);
 
 /* fs/xattr.c */
-- 
2.17.1

Re: [PATCH 0/5] y2038: Make ppoll, io_pgetevents and pselect y2038 safe

[Arnd Bergmann]

On Sat, Sep 1, 2018 at 10:48 PM Deepa Dinamani <deepa.kernel@gmail.com> wrote:
>
> The series transitions the ppoll, io_getevents, and pselect syscalls
> to be y2038 safe.
>
> This is part of the work proceeding for syscalls for y2038.
> It is based on the series [1] from Arnd Bergmann.
>
> The overview of the series is as below:
> 1. Refactor sigmask handling logic for the above syscalls.
> 2. Provide y2038 safe versions of syscalls for all ABIs.
>
> [1] https://lkml.org/lkml/2018/8/27/651

Hi Deepa,

This is a very nice series, sorry for taking way too long before
I had a good look at it.

The first two patches in particular are great. The third patch
looks like it has a small bug that I commented on, but given
how much more thought you have put into this than I did, it's
just as possible that I misunderstood it.

I've applied you series locally now, but have not pushed it out
yet for linux-next because of that one issue. I'll send a pull
request to Thomas for my other patches for now, and a
follow-up with these once we have clarified it.

        Arnd