x86, bts: fix wrmsr and spinlock over kmalloc

Impact: fix sleeping-with-spinlock-held bugs/crashes

- Turn a wrmsr to write the DS_AREA MSR into a wrmsrl.
- Use irqsave variants of spinlocks.
- Do not allocate memory while holding spinlocks.

Reported-by: Stephane Eranian <[e-mail]>
Reported-by: Ingo Molnar <[e-mail]>
Signed-off-by: Markus Metzger <[e-mail]>
Signed-off-by: Ingo Molnar <[e-mail]>

arch/x86/kernel/ds.c

···
 static inline struct ds_context *ds_get_context(struct task_struct *task)
 {
 	struct ds_context *context;
+	unsigned long irq;
 
-	spin_lock(&ds_lock);
+	spin_lock_irqsave(&ds_lock, irq);
 
 	context = (task ? task->thread.ds_ctx : this_system_context);
 	if (context)
 		context->count++;
 
-	spin_unlock(&ds_lock);
+	spin_unlock_irqrestore(&ds_lock, irq);
 
 	return context;
 }
···
 /*
  * Same as ds_get_context, but allocates the context and it's DS
  * structure, if necessary; returns NULL; if out of memory.
- *
- * pre: requires ds_lock to be held
  */
 static inline struct ds_context *ds_alloc_context(struct task_struct *task)
 {
 	struct ds_context **p_context =
 		(task ? &task->thread.ds_ctx : &this_system_context);
 	struct ds_context *context = *p_context;
+	unsigned long irq;
 
 	if (!context) {
-		spin_unlock(&ds_lock);
-
 		context = kzalloc(sizeof(*context), GFP_KERNEL);
-
-		if (!context) {
-			spin_lock(&ds_lock);
+		if (!context)
 			return NULL;
-		}
 
 		context->ds = kzalloc(ds_cfg.sizeof_ds, GFP_KERNEL);
 		if (!context->ds) {
 			kfree(context);
-			spin_lock(&ds_lock);
 			return NULL;
 		}
 
-		spin_lock(&ds_lock);
-		/*
-		 * Check for race - another CPU could have allocated
-		 * it meanwhile:
-		 */
+		spin_lock_irqsave(&ds_lock, irq);
+
 		if (*p_context) {
 			kfree(context->ds);
 			kfree(context);
-			return *p_context;
+
+			context = *p_context;
+		} else {
+			*p_context = context;
+
+			context->this = p_context;
+			context->task = task;
+
+			if (task)
+				set_tsk_thread_flag(task, TIF_DS_AREA_MSR);
+
+			if (!task || (task == current))
+				wrmsrl(MSR_IA32_DS_AREA,
+				       (unsigned long)context->ds);
 		}
-
-		*p_context = context;
-
-		context->this = p_context;
-		context->task = task;
-
-		if (task)
-			set_tsk_thread_flag(task, TIF_DS_AREA_MSR);
-
-		if (!task || (task == current))
-			wrmsr(MSR_IA32_DS_AREA, (unsigned long)context->ds, 0);
-
-		get_tracer(task);
+		spin_unlock_irqrestore(&ds_lock, irq);
 	}
 
 	context->count++;
···
  */
 static inline void ds_put_context(struct ds_context *context)
 {
+	unsigned long irq;
+
 	if (!context)
 		return;
 
-	spin_lock(&ds_lock);
+	spin_lock_irqsave(&ds_lock, irq);
 
 	if (--context->count)
 		goto out;
···
 	kfree(context->ds);
 	kfree(context);
  out:
-	spin_unlock(&ds_lock);
+	spin_unlock_irqrestore(&ds_lock, irq);
 }
 
 
···
 	struct ds_context *context;
 	unsigned long buffer, adj;
 	const unsigned long alignment = (1 << 3);
+	unsigned long irq;
 	int error = 0;
 
 	if (!ds_cfg.sizeof_ds)
···
 		return -EOPNOTSUPP;
 
 
-	spin_lock(&ds_lock);
-
-	error = -ENOMEM;
 	context = ds_alloc_context(task);
 	if (!context)
-		goto out_unlock;
+		return -ENOMEM;
+
+	spin_lock_irqsave(&ds_lock, irq);
 
 	error = -EPERM;
 	if (!check_tracer(task))
 		goto out_unlock;
 
+	get_tracer(task);
+
 	error = -EALREADY;
 	if (context->owner[qual] == current)
-		goto out_unlock;
+		goto out_put_tracer;
 	error = -EPERM;
 	if (context->owner[qual] != NULL)
-		goto out_unlock;
+		goto out_put_tracer;
 	context->owner[qual] = current;
 
-	spin_unlock(&ds_lock);
+	spin_unlock_irqrestore(&ds_lock, irq);
 
 
 	error = -ENOMEM;
···
  out_release:
 	context->owner[qual] = NULL;
 	ds_put_context(context);
+	put_tracer(task);
+	return error;
+
+ out_put_tracer:
+	spin_unlock_irqrestore(&ds_lock, irq);
+	ds_put_context(context);
+	put_tracer(task);
 	return error;
 
  out_unlock:
-	spin_unlock(&ds_lock);
+	spin_unlock_irqrestore(&ds_lock, irq);
 	ds_put_context(context);
 	return error;
 }