perf_counter: Add forward/backward attribute ABI compatibility

Provide for means of extending the perf_counter_attr in a 'natural' way.

We allow growing the structure by appending fields at the end by specifying
the full structure size inside it.

When a new kernel sees a smaller (old) structure, it will 0 pad the tail.
When an old kernel sees a larger (new) structure, it will verify the tail
consists of 0s, otherwise fail.

If we fail due to a size-mismatch, we return -E2BIG and write the kernel's
native attribe size back into the provided structure.

Furthermore, add some attribute verification, so that we'll fail counter
creation when unknown bits are present (PERF_SAMPLE, PERF_FORMAT, or in
the __reserved fields).

(This ABI detail is introduced while keeping the existing syscall ABI.)

Signed-off-by: Peter Zijlstra <[e-mail]>
Cc: Mike Galbraith <[e-mail]>
Cc: Paul Mackerras <[e-mail]>
Cc: Arnaldo Carvalho de Melo <[e-mail]>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <[e-mail]>

include/linux/perf_counter.h

···
 	PERF_SAMPLE_ID				= 1U << 6,
 	PERF_SAMPLE_CPU				= 1U << 7,
 	PERF_SAMPLE_PERIOD			= 1U << 8,
+
+	PERF_SAMPLE_MAX = 1U << 9,		/* non-ABI */
 };
 
 /*
···
 	PERF_FORMAT_TOTAL_TIME_ENABLED		= 1U << 0,
 	PERF_FORMAT_TOTAL_TIME_RUNNING		= 1U << 1,
 	PERF_FORMAT_ID				= 1U << 2,
+
+	PERF_FORMAT_MAX = 1U << 3, 		/* non-ABI */
 };
+
+#define PERF_ATTR_SIZE_VER0	64	/* sizeof first published struct */
 
 /*
  * Hardware event to monitor via a performance monitoring counter:
  */
 struct perf_counter_attr {
+
 	/*
 	 * Major type: hardware/software/tracepoint/etc.
 	 */
 	__u32			type;
-	__u32			__reserved_1;
+
+	/*
+	 * Size of the attr structure, for fwd/bwd compat.
+	 */
+	__u32			size;
 
 	/*
 	 * Type specific configuration information.
···
 				comm	       :  1, /* include comm data     */
 				freq           :  1, /* use freq, not period  */
 
-				__reserved_2   : 53;
+				__reserved_1   : 53;
 
 	__u32			wakeup_events;	/* wakeup every n events */
-	__u32			__reserved_3;
+	__u32			__reserved_2;
 
-	__u64			__reserved_4;
+	__u64			__reserved_3;
 };
 
 /*

include/linux/syscalls.h

···
 
 
 asmlinkage long sys_perf_counter_open(
-		const struct perf_counter_attr __user *attr_uptr,
+		struct perf_counter_attr __user *attr_uptr,
 		pid_t pid, int cpu, int group_fd, unsigned long flags);
 #endif

kernel/perf_counter.c

···
 	case PERF_TYPE_TRACEPOINT:
 		pmu = tp_perf_counter_init(counter);
 		break;
+
+	default:
+		break;
 	}
 done:
 	err = 0;
···
 	return counter;
 }
 
+static int perf_copy_attr(struct perf_counter_attr __user *uattr,
+			  struct perf_counter_attr *attr)
+{
+	int ret;
+	u32 size;
+
+	if (!access_ok(VERIFY_WRITE, uattr, PERF_ATTR_SIZE_VER0))
+		return -EFAULT;
+
+	/*
+	 * zero the full structure, so that a short copy will be nice.
+	 */
+	memset(attr, 0, sizeof(*attr));
+
+	ret = get_user(size, &uattr->size);
+	if (ret)
+		return ret;
+
+	if (size > PAGE_SIZE)	/* silly large */
+		goto err_size;
+
+	if (!size)		/* abi compat */
+		size = PERF_ATTR_SIZE_VER0;
+
+	if (size < PERF_ATTR_SIZE_VER0)
+		goto err_size;
+
+	/*
+	 * If we're handed a bigger struct than we know of,
+	 * ensure all the unknown bits are 0.
+	 */
+	if (size > sizeof(*attr)) {
+		unsigned long val;
+		unsigned long __user *addr;
+		unsigned long __user *end;
+
+		addr = PTR_ALIGN((void __user *)uattr + sizeof(*attr),
+				sizeof(unsigned long));
+		end  = PTR_ALIGN((void __user *)uattr + size,
+				sizeof(unsigned long));
+
+		for (; addr < end; addr += sizeof(unsigned long)) {
+			ret = get_user(val, addr);
+			if (ret)
+				return ret;
+			if (val)
+				goto err_size;
+		}
+	}
+
+	ret = copy_from_user(attr, uattr, size);
+	if (ret)
+		return -EFAULT;
+
+	/*
+	 * If the type exists, the corresponding creation will verify
+	 * the attr->config.
+	 */
+	if (attr->type >= PERF_TYPE_MAX)
+		return -EINVAL;
+
+	if (attr->__reserved_1 || attr->__reserved_2 || attr->__reserved_3)
+		return -EINVAL;
+
+	if (attr->sample_type & ~(PERF_SAMPLE_MAX-1))
+		return -EINVAL;
+
+	if (attr->read_format & ~(PERF_FORMAT_MAX-1))
+		return -EINVAL;
+
+out:
+	return ret;
+
+err_size:
+	put_user(sizeof(*attr), &uattr->size);
+	ret = -E2BIG;
+	goto out;
+}
+
 /**
  * sys_perf_counter_open - open a performance counter, associate it to a task/cpu
  *
···
  * @group_fd:		group leader counter fd
  */
 SYSCALL_DEFINE5(perf_counter_open,
-		const struct perf_counter_attr __user *, attr_uptr,
+		struct perf_counter_attr __user *, attr_uptr,
 		pid_t, pid, int, cpu, int, group_fd, unsigned long, flags)
 {
 	struct perf_counter *counter, *group_leader;
···
 	if (flags)
 		return -EINVAL;
 
-	if (copy_from_user(&attr, attr_uptr, sizeof(attr)) != 0)
-		return -EFAULT;
+	ret = perf_copy_attr(attr_uptr, &attr);
+	if (ret)
+		return ret;
 
 	if (!attr.exclude_kernel) {
 		if (perf_paranoid_kernel() && !capable(CAP_SYS_ADMIN))

tools/perf/perf.h

···
 	_min1 < _min2 ? _min1 : _min2; })
 
 static inline int
-sys_perf_counter_open(struct perf_counter_attr *attr_uptr,
+sys_perf_counter_open(struct perf_counter_attr *attr,
 		      pid_t pid, int cpu, int group_fd,
 		      unsigned long flags)
 {
-	return syscall(__NR_perf_counter_open, attr_uptr, pid, cpu,
+	attr->size = sizeof(*attr);
+	return syscall(__NR_perf_counter_open, attr, pid, cpu,
 		       group_fd, flags);
 }