Re: [RFC PATCH v10 for 4.15 01/14] Restartable sequences system call

----- On Nov 6, 2017, at 8:24 PM, Boqun Feng boqun.feng-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org wrote:

On Mon, Nov 06, 2017 at 03:56:31PM -0500, Mathieu Desnoyers wrote:
[...]

quoted

+
+/*
+ * struct rseq is aligned on 4 * 8 bytes to ensure it is always
+ * contained within a single cache-line.
+ *
+ * A single struct rseq per thread is allowed.
+ */
+struct rseq {
+	/*
+	 * Restartable sequences cpu_id_start field. Updated by the
+	 * kernel, and read by user-space with single-copy atomicity
+	 * semantics. Aligned on 32-bit. Always contain a value in the
+	 * range of possible CPUs, although the value may not be the
+	 * actual current CPU (e.g. if rseq is not initialized). This
+	 * CPU number value should always be confirmed against the value
+	 * of the cpu_id field.
+	 */
+	uint32_t cpu_id_start;
+	/*
+	 * Restartable sequences cpu_id field. Updated by the kernel,
+	 * and read by user-space with single-copy atomicity semantics.
+	 * Aligned on 32-bit. Values -1U and -2U have a special
+	 * semantic: -1U means "rseq uninitialized", and -2U means "rseq
+	 * initialization failed".
+	 */
+	uint32_t cpu_id;
+	/*
+	 * Restartable sequences rseq_cs field.
+	 *
+	 * Contains NULL when no critical section is active for the current
+	 * thread, or holds a pointer to the currently active struct rseq_cs.
+	 *
+	 * Updated by user-space at the beginning of assembly instruction
+	 * sequence block, and by the kernel when it restarts an assembly
+	 * instruction sequence block, and when the kernel detects that it
+	 * is preempting or delivering a signal outside of the range
+	 * targeted by the rseq_cs. Also needs to be cleared by user-space
+	 * before reclaiming memory that contains the targeted struct
+	 * rseq_cs.
+	 *
+	 * Read and set by the kernel with single-copy atomicity semantics.
+	 * Aligned on 64-bit.
+	 */
+	RSEQ_FIELD_u32_u64(rseq_cs);
+	/*
+	 * - RSEQ_DISABLE flag:
+	 *
+	 * Fallback fast-track flag for single-stepping.
+	 * Set by user-space if lack of progress is detected.
+	 * Cleared by user-space after rseq finish.
+	 * Read by the kernel.
+	 * - RSEQ_CS_FLAG_NO_RESTART_ON_PREEMPT
+	 *     Inhibit instruction sequence block restart and event
+	 *     counter increment on preemption for this thread.

Nit: "event counter" has been removed entirely ;-)

quoted

+	 * - RSEQ_CS_FLAG_NO_RESTART_ON_SIGNAL
+	 *     Inhibit instruction sequence block restart and event
+	 *     counter increment on signal delivery for this thread.
+	 * - RSEQ_CS_FLAG_NO_RESTART_ON_MIGRATE
+	 *     Inhibit instruction sequence block restart and event
+	 *     counter increment on migration for this thread.
+	 */
+	uint32_t flags;
+} __attribute__((aligned(4 * sizeof(uint64_t))));
+
+#endif /* _UAPI_LINUX_RSEQ_H */

[...]

quoted

+	} else {
+		/*
+		 * If there was no rseq previously registered,
+		 * we need to ensure the provided rseq is
+		 * properly aligned and valid.
+		 */
+		if (!IS_ALIGNED((unsigned long)rseq, __alignof__(*rseq))
+				|| rseq_len != sizeof(*rseq))
+			return -EINVAL;
+		if (!access_ok(VERIFY_WRITE, rseq, rseq_len))
+			return -EFAULT;
+		current->rseq = rseq;
+		current->rseq_len = rseq_len;
+		current->rseq_sig = sig;
+		/*
+		 * If rseq was previously inactive, and has just
+		 * been registered, ensure the cpu_id and
+		 * event_counter fields are updated before

s/event_counter/cpu_start_id/ ?

I guess you mean "cpu_id_start".

Good point. v11 will include that fix. Meanwhile, it is available
at https://git.kernel.org/pub/scm/linux/kernel/git/rseq/linux-rseq.git/log/?h=rseq/dev

Thanks,

Mathieu

Regards,
Boqun

quoted

+		 * returning to user-space.
+		 */
+		rseq_set_notify_resume(current);
+	}
+
+	return 0;
+}

[...]

-- 
Mathieu Desnoyers
EfficiOS Inc.
http://www.efficios.com