syscall_user_dispatch: Add PR_SYS_DISPATCH_INCLUSIVE_ON

There are two possible scenarios for syscall filtering:
 - having a trusted/allowed range of PCs, and intercepting everything else
 - or the opposite: a single untrusted/intercepted range and allowing
   everything else (this is relevant for any kind of sandboxing scenario,
   or monitoring behavior of a single library)

The current API only allows the former use case due to allowed
range wrap-around check. Add PR_SYS_DISPATCH_INCLUSIVE_ON that
enables the second use case.

Add PR_SYS_DISPATCH_EXCLUSIVE_ON alias for PR_SYS_DISPATCH_ON
to make it clear how it's different from the new
PR_SYS_DISPATCH_INCLUSIVE_ON.

Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Link: https://lore.kernel.org/all/97947cc8e205ff49675826d7b0327ef2e2c66eea.1747839857.git.dvyukov@google.com

Documentation/admin-guide/syscall-user-dispatch.rst

@@ -53,20 +53,25 @@ following prctl:
 
   prctl(PR_SET_SYSCALL_USER_DISPATCH, <op>, <offset>, <length>, [selector])
 
-<op> is either PR_SYS_DISPATCH_ON or PR_SYS_DISPATCH_OFF, to enable and
+<op> is either PR_SYS_DISPATCH_EXCLUSIVE_ON/PR_SYS_DISPATCH_INCLUSIVE_ON
-disable the mechanism globally for that thread.  When
+or PR_SYS_DISPATCH_OFF, to enable and disable the mechanism globally for
-PR_SYS_DISPATCH_OFF is used, the other fields must be zero.
+that thread.  When PR_SYS_DISPATCH_OFF is used, the other fields must be zero.
 
-[<offset>, <offset>+<length>) delimit a memory region interval
+For PR_SYS_DISPATCH_EXCLUSIVE_ON [<offset>, <offset>+<length>) delimit
-from which syscalls are always executed directly, regardless of the
+a memory region interval from which syscalls are always executed directly,
-userspace selector.  This provides a fast path for the C library, which
+regardless of the userspace selector.  This provides a fast path for the
-includes the most common syscall dispatchers in the native code
+C library, which includes the most common syscall dispatchers in the native
-applications, and also provides a way for the signal handler to return
+code applications, and also provides a way for the signal handler to return
 without triggering a nested SIGSYS on (rt\_)sigreturn.  Users of this
 interface should make sure that at least the signal trampoline code is
 included in this region. In addition, for syscalls that implement the
 trampoline code on the vDSO, that trampoline is never intercepted.
 
+For PR_SYS_DISPATCH_INCLUSIVE_ON [<offset>, <offset>+<length>) delimit
+a memory region interval from which syscalls are dispatched based on
+the userspace selector. Syscalls from outside of the range are always
+executed directly.
+
 [selector] is a pointer to a char-sized region in the process memory
 region, that provides a quick way to enable disable syscall redirection
 thread-wide, without the need to invoke the kernel directly.  selector

include/uapi/linux/prctl.h

@@ -255,7 +255,12 @@ struct prctl_mm_map {
 /* Dispatch syscalls to a userspace handler */
 #define PR_SET_SYSCALL_USER_DISPATCH	59
 # define PR_SYS_DISPATCH_OFF		0
-# define PR_SYS_DISPATCH_ON		1
+/* Enable dispatch except for the specified range */
+# define PR_SYS_DISPATCH_EXCLUSIVE_ON	1
+/* Enable dispatch for the specified range */
+# define PR_SYS_DISPATCH_INCLUSIVE_ON	2
+/* Legacy name for backwards compatibility */
+# define PR_SYS_DISPATCH_ON		PR_SYS_DISPATCH_EXCLUSIVE_ON
 /* The control values for the user space selector when dispatch is enabled */
 # define SYSCALL_DISPATCH_FILTER_ALLOW	0
 # define SYSCALL_DISPATCH_FILTER_BLOCK	1

kernel/entry/syscall_user_dispatch.c

@@ -78,7 +78,7 @@ static int task_set_syscall_user_dispatch(struct task_struct *task, unsigned lon
 		if (offset || len || selector)
 			return -EINVAL;
 		break;
-	case PR_SYS_DISPATCH_ON:
+	case PR_SYS_DISPATCH_EXCLUSIVE_ON:
 		/*
 		 * Validate the direct dispatcher region just for basic
 		 * sanity against overflow and a 0-sized dispatcher
@@ -87,30 +87,40 @@ static int task_set_syscall_user_dispatch(struct task_struct *task, unsigned lon
 		 */
 		if (offset && offset + len <= offset)
 			return -EINVAL;
+		break;
+	case PR_SYS_DISPATCH_INCLUSIVE_ON:
+		if (len == 0 || offset + len <= offset)
+			return -EINVAL;
+		/*
+		 * Invert the range, the check in syscall_user_dispatch()
+		 * supports wrap-around.
+		 */
+		offset = offset + len;
+		len = -len;
+		break;
+	default:
+		return -EINVAL;
+	}
 
 	/*
 	 * access_ok() will clear memory tags for tagged addresses
 	 * if current has memory tagging enabled.
-
+	 *
 	 * To enable a tracer to set a tracees selector the
 	 * selector address must be untagged for access_ok(),
 	 * otherwise an untagged tracer will always fail to set a
 	 * tagged tracees selector.
 	 */
-		if (selector && !access_ok(untagged_addr(selector), sizeof(*selector)))
+	if (mode != PR_SYS_DISPATCH_OFF && selector &&
+		!access_ok(untagged_addr(selector), sizeof(*selector)))
 		return -EFAULT;
 
-		break;
-	default:
-		return -EINVAL;
-	}
-
 	task->syscall_dispatch.selector = selector;
 	task->syscall_dispatch.offset = offset;
 	task->syscall_dispatch.len = len;
 	task->syscall_dispatch.on_dispatch = false;
 
-	if (mode == PR_SYS_DISPATCH_ON)
+	if (mode != PR_SYS_DISPATCH_OFF)
 		set_task_syscall_work(task, SYSCALL_USER_DISPATCH);
 	else
 		clear_task_syscall_work(task, SYSCALL_USER_DISPATCH);

tools/include/uapi/linux/prctl.h

@@ -255,7 +255,12 @@ struct prctl_mm_map {
 /* Dispatch syscalls to a userspace handler */
 #define PR_SET_SYSCALL_USER_DISPATCH	59
 # define PR_SYS_DISPATCH_OFF		0
-# define PR_SYS_DISPATCH_ON		1
+/* Enable dispatch except for the specified range */
+# define PR_SYS_DISPATCH_EXCLUSIVE_ON	1
+/* Enable dispatch for the specified range */
+# define PR_SYS_DISPATCH_INCLUSIVE_ON	2
+/* Legacy name for backwards compatibility */
+# define PR_SYS_DISPATCH_ON		PR_SYS_DISPATCH_EXCLUSIVE_ON
 /* The control values for the user space selector when dispatch is enabled */
 # define SYSCALL_DISPATCH_FILTER_ALLOW	0
 # define SYSCALL_DISPATCH_FILTER_BLOCK	1