Merge tag 'perf-core-for-bpf-next' from tip tree

Stable tag for bpf-next's uprobe work.

Signed-off-by: Andrii Nakryiko <andrii@kernel.org>

2 files changed, 74 insertions, 6 deletions

diff --git a/include/linux/cpuhotplug.h b/include/linux/cpuhotplug.h
index 2361ed4d2b15..37a9afffb59e 100644
--- a/include/linux/cpuhotplug.h
+++ b/include/linux/cpuhotplug.h
@@ -208,7 +208,6 @@ enum cpuhp_state {
 	CPUHP_AP_PERF_X86_UNCORE_ONLINE,
 	CPUHP_AP_PERF_X86_AMD_UNCORE_ONLINE,
 	CPUHP_AP_PERF_X86_AMD_POWER_ONLINE,
-	CPUHP_AP_PERF_X86_RAPL_ONLINE,
 	CPUHP_AP_PERF_S390_CF_ONLINE,
 	CPUHP_AP_PERF_S390_SF_ONLINE,
 	CPUHP_AP_PERF_ARM_CCI_ONLINE,
diff --git a/include/linux/uprobes.h b/include/linux/uprobes.h
index 2b294bf1881f..7a051b5d2edd 100644
--- a/include/linux/uprobes.h
+++ b/include/linux/uprobes.h
@@ -15,6 +15,7 @@
 #include <linux/rbtree.h>
 #include <linux/types.h>
 #include <linux/wait.h>
+#include <linux/timer.h>
 
 struct uprobe;
 struct vm_area_struct;
@@ -23,8 +24,17 @@ struct inode;
 struct notifier_block;
 struct page;
 
+/*
+ * Allowed return values from uprobe consumer's handler callback
+ * with following meaning:
+ *
+ * UPROBE_HANDLER_REMOVE
+ * - Remove the uprobe breakpoint from current->mm.
+ * UPROBE_HANDLER_IGNORE
+ * - Ignore ret_handler callback for this consumer.
+ */
 #define UPROBE_HANDLER_REMOVE		1
-#define UPROBE_HANDLER_MASK		1
+#define UPROBE_HANDLER_IGNORE		2
 
 #define MAX_URETPROBE_DEPTH		64
 
@@ -37,13 +47,15 @@ struct uprobe_consumer {
 	 * for the current process. If filter() is omitted or returns true,
 	 * UPROBE_HANDLER_REMOVE is effectively ignored.
 	 */
-	int (*handler)(struct uprobe_consumer *self, struct pt_regs *regs);
+	int (*handler)(struct uprobe_consumer *self, struct pt_regs *regs, __u64 *data);
 	int (*ret_handler)(struct uprobe_consumer *self,
 				unsigned long func,
-				struct pt_regs *regs);
+				struct pt_regs *regs, __u64 *data);
 	bool (*filter)(struct uprobe_consumer *self, struct mm_struct *mm);
 
 	struct list_head cons_node;
+
+	__u64 id;	/* set when uprobe_consumer is registered */
 };
 
 #ifdef CONFIG_UPROBES
@@ -56,6 +68,53 @@ enum uprobe_task_state {
 	UTASK_SSTEP_TRAPPED,
 };
 
+/* The state of hybrid-lifetime uprobe inside struct return_instance */
+enum hprobe_state {
+	HPROBE_LEASED,		/* uretprobes_srcu-protected uprobe */
+	HPROBE_STABLE,		/* refcounted uprobe */
+	HPROBE_GONE,		/* NULL uprobe, SRCU expired, refcount failed */
+	HPROBE_CONSUMED,	/* uprobe "consumed" by uretprobe handler */
+};
+
+/*
+ * Hybrid lifetime uprobe. Represents a uprobe instance that could be either
+ * SRCU protected (with SRCU protection eventually potentially timing out),
+ * refcounted using uprobe->ref, or there could be no valid uprobe (NULL).
+ *
+ * hprobe's internal state is setup such that background timer thread can
+ * atomically "downgrade" temporarily RCU-protected uprobe into refcounted one
+ * (or no uprobe, if refcounting failed).
+ *
+ * *stable* pointer always point to the uprobe (or could be NULL if there is
+ * was no valid underlying uprobe to begin with).
+ *
+ * *leased* pointer is the key to achieving race-free atomic lifetime state
+ * transition and can have three possible states:
+ *   - either the same non-NULL value as *stable*, in which case uprobe is
+ *     SRCU-protected;
+ *   - NULL, in which case uprobe (if there is any) is refcounted;
+ *   - special __UPROBE_DEAD value, which represents an uprobe that was SRCU
+ *     protected initially, but SRCU period timed out and we attempted to
+ *     convert it to refcounted, but refcount_inc_not_zero() failed, because
+ *     uprobe effectively went away (the last consumer unsubscribed). In this
+ *     case it's important to know that *stable* pointer (which still has
+ *     non-NULL uprobe pointer) shouldn't be used, because lifetime of
+ *     underlying uprobe is not guaranteed anymore. __UPROBE_DEAD is just an
+ *     internal marker and is handled transparently by hprobe_fetch() helper.
+ *
+ * When uprobe is SRCU-protected, we also record srcu_idx value, necessary for
+ * SRCU unlocking.
+ *
+ * See hprobe_expire() and hprobe_fetch() for details of race-free uprobe
+ * state transitioning details. It all hinges on atomic xchg() over *leaded*
+ * pointer. *stable* pointer, once initially set, is not modified concurrently.
+ */
+struct hprobe {
+	enum hprobe_state state;
+	int srcu_idx;
+	struct uprobe *uprobe;
+};
+
 /*
  * uprobe_task: Metadata of a task while it singlesteps.
  */
@@ -75,6 +134,7 @@ struct uprobe_task {
 	};
 
 	struct uprobe			*active_uprobe;
+	struct timer_list		ri_timer;
 	unsigned long			xol_vaddr;
 
 	struct arch_uprobe              *auprobe;
@@ -83,15 +143,24 @@ struct uprobe_task {
 	unsigned int			depth;
 };
 
+struct return_consumer {
+	__u64	cookie;
+	__u64	id;
+};
+
 struct return_instance {
-	struct uprobe		*uprobe;
+	struct hprobe		hprobe;
 	unsigned long		func;
 	unsigned long		stack;		/* stack pointer */
 	unsigned long		orig_ret_vaddr; /* original return address */
 	bool			chained;	/* true, if instance is nested */
+	int			consumers_cnt;
 
 	struct return_instance	*next;		/* keep as stack */
-};
+	struct rcu_head		rcu;
+
+	struct return_consumer	consumers[] __counted_by(consumers_cnt);
+} ____cacheline_aligned;
 
 enum rp_check {
 	RP_CHECK_CALL,