Skip to contentSkip to conversation

uprobes: switch to RCU Tasks Trace flavor for better performance

PATCH Applied Andrii Nakryiko 4 participants 4 messages Sep 10, 2024 – Oct 8, 2024

Subsystems (4)
lore
Share & tools
lore

Andrii Nakryiko

author maintainer
#0 lore raw

PATCH 1/1 uprobes: switch to RCU Tasks Trace flavor for better performance

This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which is optimized for more lightweight and quick readers (at the expense of slower writers, which for uprobes is a fine tradeof) and has better performance and scalability with number of CPUs.

Similarly to baseline vs SRCU, we've benchmarked SRCU-based implementation vs RCU Tasks Trace implementation.

SRCU
====

uprobe-nop      ( 1 cpus):    3.276 ± 0.005M/s  (  3.276M/s/cpu)
uprobe-nop      ( 2 cpus):    4.125 ± 0.002M/s  (  2.063M/s/cpu)
uprobe-nop      ( 4 cpus):    7.713 ± 0.002M/s  (  1.928M/s/cpu)
uprobe-nop      ( 8 cpus):    8.097 ± 0.006M/s  (  1.012M/s/cpu)
uprobe-nop      (16 cpus):    6.501 ± 0.056M/s  (  0.406M/s/cpu)
uprobe-nop      (32 cpus):    4.398 ± 0.084M/s  (  0.137M/s/cpu)
uprobe-nop      (64 cpus):    6.452 ± 0.000M/s  (  0.101M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.055 ± 0.001M/s  (  2.055M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.677 ± 0.000M/s  (  1.339M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.561 ± 0.003M/s  (  1.140M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.291 ± 0.002M/s  (  0.661M/s/cpu)
uretprobe-nop   (16 cpus):    5.065 ± 0.019M/s  (  0.317M/s/cpu)
uretprobe-nop   (32 cpus):    3.622 ± 0.003M/s  (  0.113M/s/cpu)
uretprobe-nop   (64 cpus):    3.723 ± 0.002M/s  (  0.058M/s/cpu)

RCU Tasks Trace
===============

uprobe-nop      ( 1 cpus):    3.396 ± 0.002M/s  (  3.396M/s/cpu)
uprobe-nop      ( 2 cpus):    4.271 ± 0.006M/s  (  2.135M/s/cpu)
uprobe-nop      ( 4 cpus):    8.499 ± 0.015M/s  (  2.125M/s/cpu)
uprobe-nop      ( 8 cpus):   10.355 ± 0.028M/s  (  1.294M/s/cpu)
uprobe-nop      (16 cpus):    7.615 ± 0.099M/s  (  0.476M/s/cpu)
uprobe-nop      (32 cpus):    4.430 ± 0.007M/s  (  0.138M/s/cpu)
uprobe-nop      (64 cpus):    6.887 ± 0.020M/s  (  0.108M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.174 ± 0.001M/s  (  2.174M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.853 ± 0.001M/s  (  1.426M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.913 ± 0.002M/s  (  1.228M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.883 ± 0.002M/s  (  0.735M/s/cpu)
uretprobe-nop   (16 cpus):    5.147 ± 0.001M/s  (  0.322M/s/cpu)
uretprobe-nop   (32 cpus):    3.738 ± 0.008M/s  (  0.117M/s/cpu)
uretprobe-nop   (64 cpus):    4.397 ± 0.002M/s  (  0.069M/s/cpu)

Peak throughput for uprobes increases from 8 mln/s to 10.3 mln/s (+28%!), and for uretprobes from 5.3 mln/s to 5.8 mln/s (+11%), as we have more work to do on uretprobes side.

Even single-thread (no contention) performance is slightly better: 3.276 mln/s to 3.396 mln/s (+3.5%) for uprobes, and 2.055 mln/s to 2.174 mln/s (+5.8%) for uretprobes.

We also select TASKS_TRACE_RCU for UPROBES in Kconfig due to the new dependency.

---
 arch/Kconfig            |  1 +
 kernel/events/uprobes.c | 38 ++++++++++++++++----------------------
 2 files changed, 17 insertions(+), 22 deletions(-)
arch/Kconfig +1 -0
index 975dd22a2dbd..a0df3f3dc484 100644
@@ -126,6 +126,7 @@ config KPROBES_ON_FTRACE
 config UPROBES
 	def_bool n
 	depends on ARCH_SUPPORTS_UPROBES
+	select TASKS_TRACE_RCU
 	help
 	  Uprobes is the user-space counterpart to kprobes: they
 	  enable instrumentation applications (such as 'perf probe')
kernel/events/uprobes.c +16 -22
index 4b7e590dc428..a2e6a57f79f2 100644
@@ -26,6 +26,7 @@
 #include <linux/task_work.h>
 #include <linux/shmem_fs.h>
 #include <linux/khugepaged.h>
+#include <linux/rcupdate_trace.h>
 
 #include <linux/uprobes.h>
 
@@ -42,8 +43,6 @@ static struct rb_root uprobes_tree = RB_ROOT;
 static DEFINE_RWLOCK(uprobes_treelock);	/* serialize rbtree access */
 static seqcount_rwlock_t uprobes_seqcount = SEQCNT_RWLOCK_ZERO(uprobes_seqcount, &uprobes_treelock);
 
-DEFINE_STATIC_SRCU(uprobes_srcu);
-
 #define UPROBES_HASH_SZ	13
 /* serialize uprobe->pending_list */
 static struct mutex uprobes_mmap_mutex[UPROBES_HASH_SZ];
@@ -652,7 +651,7 @@ static void put_uprobe(struct uprobe *uprobe)
 	delayed_uprobe_remove(uprobe, NULL);
 	mutex_unlock(&delayed_uprobe_lock);
 
-	call_srcu(&uprobes_srcu, &uprobe->rcu, uprobe_free_rcu);
+	call_rcu_tasks_trace(&uprobe->rcu, uprobe_free_rcu);
 }
 
 static __always_inline
@@ -707,7 +706,7 @@ static struct uprobe *find_uprobe_rcu(struct inode *inode, loff_t offset)
 	struct rb_node *node;
 	unsigned int seq;
 
-	lockdep_assert(srcu_read_lock_held(&uprobes_srcu));
+	lockdep_assert(rcu_read_lock_trace_held());
 
 	do {
 		seq = read_seqcount_begin(&uprobes_seqcount);
@@ -935,8 +934,7 @@ static bool filter_chain(struct uprobe *uprobe, struct mm_struct *mm)
 	bool ret = false;
 
 	down_read(&uprobe->consumer_rwsem);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		ret = consumer_filter(uc, mm);
 		if (ret)
 			break;
@@ -1157,7 +1155,7 @@ void uprobe_unregister_sync(void)
 	 * unlucky enough caller can free consumer's memory and cause
 	 * handler_chain() or handle_uretprobe_chain() to do an use-after-free.
 	 */
-	synchronize_srcu(&uprobes_srcu);
+	synchronize_rcu_tasks_trace();
 }
 EXPORT_SYMBOL_GPL(uprobe_unregister_sync);
 
@@ -1241,19 +1239,18 @@ EXPORT_SYMBOL_GPL(uprobe_register);
 int uprobe_apply(struct uprobe *uprobe, struct uprobe_consumer *uc, bool add)
 {
 	struct uprobe_consumer *con;
-	int ret = -ENOENT, srcu_idx;
+	int ret = -ENOENT;
 
 	down_write(&uprobe->register_rwsem);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(con, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(con, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (con == uc) {
 			ret = register_for_each_vma(uprobe, add ? uc : NULL);
 			break;
 		}
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 
 	up_write(&uprobe->register_rwsem);
 
@@ -2123,8 +2120,7 @@ static void handler_chain(struct uprobe *uprobe, struct pt_regs *regs)
 
 	current->utask->auprobe = &uprobe->arch;
 
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		int rc = 0;
 
 		if (uc->handler) {
@@ -2162,15 +2158,13 @@ handle_uretprobe_chain(struct return_instance *ri, struct pt_regs *regs)
 {
 	struct uprobe *uprobe = ri->uprobe;
 	struct uprobe_consumer *uc;
-	int srcu_idx;
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (uc->ret_handler)
 			uc->ret_handler(uc, ri->func, regs);
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 static struct return_instance *find_next_ret_chain(struct return_instance *ri)
@@ -2255,13 +2249,13 @@ static void handle_swbp(struct pt_regs *regs)
 {
 	struct uprobe *uprobe;
 	unsigned long bp_vaddr;
-	int is_swbp, srcu_idx;
+	int is_swbp;
 
 	bp_vaddr = uprobe_get_swbp_addr(regs);
 	if (bp_vaddr == uprobe_get_trampoline_vaddr())
 		return uprobe_handle_trampoline(regs);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
+	rcu_read_lock_trace();
 
 	uprobe = find_active_uprobe_rcu(bp_vaddr, &is_swbp);
 	if (!uprobe) {
@@ -2319,7 +2313,7 @@ static void handle_swbp(struct pt_regs *regs)
 
 out:
 	/* arch_uprobe_skip_sstep() succeeded, or restart if can't singlestep */
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 /*
Signature · 1 line
2.43.5
#1 lore raw
Quoted 56 lines · Andrii Nakryiko

This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which is optimized for more lightweight and quick readers (at the expense of slower writers, which for uprobes is a fine tradeof) and has better performance and scalability with number of CPUs.

Similarly to baseline vs SRCU, we've benchmarked SRCU-based implementation vs RCU Tasks Trace implementation.

SRCU
====

uprobe-nop      ( 1 cpus):    3.276 ± 0.005M/s  (  3.276M/s/cpu)
uprobe-nop      ( 2 cpus):    4.125 ± 0.002M/s  (  2.063M/s/cpu)
uprobe-nop      ( 4 cpus):    7.713 ± 0.002M/s  (  1.928M/s/cpu)
uprobe-nop      ( 8 cpus):    8.097 ± 0.006M/s  (  1.012M/s/cpu)
uprobe-nop      (16 cpus):    6.501 ± 0.056M/s  (  0.406M/s/cpu)
uprobe-nop      (32 cpus):    4.398 ± 0.084M/s  (  0.137M/s/cpu)
uprobe-nop      (64 cpus):    6.452 ± 0.000M/s  (  0.101M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.055 ± 0.001M/s  (  2.055M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.677 ± 0.000M/s  (  1.339M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.561 ± 0.003M/s  (  1.140M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.291 ± 0.002M/s  (  0.661M/s/cpu)
uretprobe-nop   (16 cpus):    5.065 ± 0.019M/s  (  0.317M/s/cpu)
uretprobe-nop   (32 cpus):    3.622 ± 0.003M/s  (  0.113M/s/cpu)
uretprobe-nop   (64 cpus):    3.723 ± 0.002M/s  (  0.058M/s/cpu)

RCU Tasks Trace
===============

uprobe-nop      ( 1 cpus):    3.396 ± 0.002M/s  (  3.396M/s/cpu)
uprobe-nop      ( 2 cpus):    4.271 ± 0.006M/s  (  2.135M/s/cpu)
uprobe-nop      ( 4 cpus):    8.499 ± 0.015M/s  (  2.125M/s/cpu)
uprobe-nop      ( 8 cpus):   10.355 ± 0.028M/s  (  1.294M/s/cpu)
uprobe-nop      (16 cpus):    7.615 ± 0.099M/s  (  0.476M/s/cpu)
uprobe-nop      (32 cpus):    4.430 ± 0.007M/s  (  0.138M/s/cpu)
uprobe-nop      (64 cpus):    6.887 ± 0.020M/s  (  0.108M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.174 ± 0.001M/s  (  2.174M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.853 ± 0.001M/s  (  1.426M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.913 ± 0.002M/s  (  1.228M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.883 ± 0.002M/s  (  0.735M/s/cpu)
uretprobe-nop   (16 cpus):    5.147 ± 0.001M/s  (  0.322M/s/cpu)
uretprobe-nop   (32 cpus):    3.738 ± 0.008M/s  (  0.117M/s/cpu)
uretprobe-nop   (64 cpus):    4.397 ± 0.002M/s  (  0.069M/s/cpu)

Peak throughput for uprobes increases from 8 mln/s to 10.3 mln/s (+28%!), and for uretprobes from 5.3 mln/s to 5.8 mln/s (+11%), as we have more work to do on uretprobes side.

Even single-thread (no contention) performance is slightly better: 3.276 mln/s to 3.396 mln/s (+3.5%) for uprobes, and 2.055 mln/s to 2.174 mln/s (+5.8%) for uretprobes.

We also select TASKS_TRACE_RCU for UPROBES in Kconfig due to the new dependency.

Quoted 7 lines
---
 arch/Kconfig            |  1 +
 kernel/events/uprobes.c | 38 ++++++++++++++++----------------------
 2 files changed, 17 insertions(+), 22 deletions(-)

Just in case this slipped through the cracks (and is not just waiting its turn to be applied), ping. It would be nice to have this patch with the rest of uprobe patches from the original patch set to go in together. Thanks!

Quoted 12 lines
arch/Kconfig +1 -0
index 975dd22a2dbd..a0df3f3dc484 100644
@@ -126,6 +126,7 @@ config KPROBES_ON_FTRACE
 config UPROBES
        def_bool n
        depends on ARCH_SUPPORTS_UPROBES
+       select TASKS_TRACE_RCU
        help
          Uprobes is the user-space counterpart to kprobes: they
          enable instrumentation applications (such as 'perf probe')

[...]

maintainer Reviewed-by on v1 1/1
#2 lore raw
In reply to Andrii Nakryiko
Quoted 48 lines · On Tue, 10 Sep 2024 10:43:12 -0700 Andrii Nakryiko <[email protected]> wrote:

This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which is optimized for more lightweight and quick readers (at the expense of slower writers, which for uprobes is a fine tradeof) and has better performance and scalability with number of CPUs.

Similarly to baseline vs SRCU, we've benchmarked SRCU-based implementation vs RCU Tasks Trace implementation.

SRCU
====

uprobe-nop      ( 1 cpus):    3.276 ± 0.005M/s  (  3.276M/s/cpu)
uprobe-nop      ( 2 cpus):    4.125 ± 0.002M/s  (  2.063M/s/cpu)
uprobe-nop      ( 4 cpus):    7.713 ± 0.002M/s  (  1.928M/s/cpu)
uprobe-nop      ( 8 cpus):    8.097 ± 0.006M/s  (  1.012M/s/cpu)
uprobe-nop      (16 cpus):    6.501 ± 0.056M/s  (  0.406M/s/cpu)
uprobe-nop      (32 cpus):    4.398 ± 0.084M/s  (  0.137M/s/cpu)
uprobe-nop      (64 cpus):    6.452 ± 0.000M/s  (  0.101M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.055 ± 0.001M/s  (  2.055M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.677 ± 0.000M/s  (  1.339M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.561 ± 0.003M/s  (  1.140M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.291 ± 0.002M/s  (  0.661M/s/cpu)
uretprobe-nop   (16 cpus):    5.065 ± 0.019M/s  (  0.317M/s/cpu)
uretprobe-nop   (32 cpus):    3.622 ± 0.003M/s  (  0.113M/s/cpu)
uretprobe-nop   (64 cpus):    3.723 ± 0.002M/s  (  0.058M/s/cpu)

RCU Tasks Trace
===============

uprobe-nop      ( 1 cpus):    3.396 ± 0.002M/s  (  3.396M/s/cpu)
uprobe-nop      ( 2 cpus):    4.271 ± 0.006M/s  (  2.135M/s/cpu)
uprobe-nop      ( 4 cpus):    8.499 ± 0.015M/s  (  2.125M/s/cpu)
uprobe-nop      ( 8 cpus):   10.355 ± 0.028M/s  (  1.294M/s/cpu)
uprobe-nop      (16 cpus):    7.615 ± 0.099M/s  (  0.476M/s/cpu)
uprobe-nop      (32 cpus):    4.430 ± 0.007M/s  (  0.138M/s/cpu)
uprobe-nop      (64 cpus):    6.887 ± 0.020M/s  (  0.108M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.174 ± 0.001M/s  (  2.174M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.853 ± 0.001M/s  (  1.426M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.913 ± 0.002M/s  (  1.228M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.883 ± 0.002M/s  (  0.735M/s/cpu)
uretprobe-nop   (16 cpus):    5.147 ± 0.001M/s  (  0.322M/s/cpu)
uretprobe-nop   (32 cpus):    3.738 ± 0.008M/s  (  0.117M/s/cpu)
uretprobe-nop   (64 cpus):    4.397 ± 0.002M/s  (  0.069M/s/cpu)

Peak throughput for uprobes increases from 8 mln/s to 10.3 mln/s (+28%!), and for uretprobes from 5.3 mln/s to 5.8 mln/s (+11%), as we have more work to do on uretprobes side.

Quoted 7 lines

Even single-thread (no contention) performance is slightly better: 3.276 mln/s to 3.396 mln/s (+3.5%) for uprobes, and 2.055 mln/s to 2.174 mln/s (+5.8%) for uretprobes.

We also select TASKS_TRACE_RCU for UPROBES in Kconfig due to the new dependency.

Looks good to me. Peter, do you have any comment?

Quoted 158 lines
---
 arch/Kconfig            |  1 +
 kernel/events/uprobes.c | 38 ++++++++++++++++----------------------
 2 files changed, 17 insertions(+), 22 deletions(-)
arch/Kconfig +1 -0
index 975dd22a2dbd..a0df3f3dc484 100644
@@ -126,6 +126,7 @@ config KPROBES_ON_FTRACE
 config UPROBES
 	def_bool n
 	depends on ARCH_SUPPORTS_UPROBES
+	select TASKS_TRACE_RCU
 	help
 	  Uprobes is the user-space counterpart to kprobes: they
 	  enable instrumentation applications (such as 'perf probe')
kernel/events/uprobes.c +16 -22
index 4b7e590dc428..a2e6a57f79f2 100644
@@ -26,6 +26,7 @@
 #include <linux/task_work.h>
 #include <linux/shmem_fs.h>
 #include <linux/khugepaged.h>
+#include <linux/rcupdate_trace.h>
 
 #include <linux/uprobes.h>
 
@@ -42,8 +43,6 @@ static struct rb_root uprobes_tree = RB_ROOT;
 static DEFINE_RWLOCK(uprobes_treelock);	/* serialize rbtree access */
 static seqcount_rwlock_t uprobes_seqcount = SEQCNT_RWLOCK_ZERO(uprobes_seqcount, &uprobes_treelock);
 
-DEFINE_STATIC_SRCU(uprobes_srcu);
-
 #define UPROBES_HASH_SZ	13
 /* serialize uprobe->pending_list */
 static struct mutex uprobes_mmap_mutex[UPROBES_HASH_SZ];
@@ -652,7 +651,7 @@ static void put_uprobe(struct uprobe *uprobe)
 	delayed_uprobe_remove(uprobe, NULL);
 	mutex_unlock(&delayed_uprobe_lock);
 
-	call_srcu(&uprobes_srcu, &uprobe->rcu, uprobe_free_rcu);
+	call_rcu_tasks_trace(&uprobe->rcu, uprobe_free_rcu);
 }
 
 static __always_inline
@@ -707,7 +706,7 @@ static struct uprobe *find_uprobe_rcu(struct inode *inode, loff_t offset)
 	struct rb_node *node;
 	unsigned int seq;
 
-	lockdep_assert(srcu_read_lock_held(&uprobes_srcu));
+	lockdep_assert(rcu_read_lock_trace_held());
 
 	do {
 		seq = read_seqcount_begin(&uprobes_seqcount);
@@ -935,8 +934,7 @@ static bool filter_chain(struct uprobe *uprobe, struct mm_struct *mm)
 	bool ret = false;
 
 	down_read(&uprobe->consumer_rwsem);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		ret = consumer_filter(uc, mm);
 		if (ret)
 			break;
@@ -1157,7 +1155,7 @@ void uprobe_unregister_sync(void)
 	 * unlucky enough caller can free consumer's memory and cause
 	 * handler_chain() or handle_uretprobe_chain() to do an use-after-free.
 	 */
-	synchronize_srcu(&uprobes_srcu);
+	synchronize_rcu_tasks_trace();
 }
 EXPORT_SYMBOL_GPL(uprobe_unregister_sync);
 
@@ -1241,19 +1239,18 @@ EXPORT_SYMBOL_GPL(uprobe_register);
 int uprobe_apply(struct uprobe *uprobe, struct uprobe_consumer *uc, bool add)
 {
 	struct uprobe_consumer *con;
-	int ret = -ENOENT, srcu_idx;
+	int ret = -ENOENT;
 
 	down_write(&uprobe->register_rwsem);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(con, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(con, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (con == uc) {
 			ret = register_for_each_vma(uprobe, add ? uc : NULL);
 			break;
 		}
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 
 	up_write(&uprobe->register_rwsem);
 
@@ -2123,8 +2120,7 @@ static void handler_chain(struct uprobe *uprobe, struct pt_regs *regs)
 
 	current->utask->auprobe = &uprobe->arch;
 
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		int rc = 0;
 
 		if (uc->handler) {
@@ -2162,15 +2158,13 @@ handle_uretprobe_chain(struct return_instance *ri, struct pt_regs *regs)
 {
 	struct uprobe *uprobe = ri->uprobe;
 	struct uprobe_consumer *uc;
-	int srcu_idx;
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (uc->ret_handler)
 			uc->ret_handler(uc, ri->func, regs);
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 static struct return_instance *find_next_ret_chain(struct return_instance *ri)
@@ -2255,13 +2249,13 @@ static void handle_swbp(struct pt_regs *regs)
 {
 	struct uprobe *uprobe;
 	unsigned long bp_vaddr;
-	int is_swbp, srcu_idx;
+	int is_swbp;
 
 	bp_vaddr = uprobe_get_swbp_addr(regs);
 	if (bp_vaddr == uprobe_get_trampoline_vaddr())
 		return uprobe_handle_trampoline(regs);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
+	rcu_read_lock_trace();
 
 	uprobe = find_active_uprobe_rcu(bp_vaddr, &is_swbp);
 	if (!uprobe) {
@@ -2319,7 +2313,7 @@ static void handle_swbp(struct pt_regs *regs)
 
 out:
 	/* arch_uprobe_skip_sstep() succeeded, or restart if can't singlestep */
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 /*
Signature · 1 line
2.43.5
Signature · 1 line
Masami Hiramatsu (Google) <[email protected]>
#3 lore raw
In reply to Andrii Nakryiko
The following commit has been merged into the perf/core branch of tip:

The following commit has been merged into the perf/core branch of tip:

Commit-ID:     87195a1ee332add27bd51448c6b54aad551a28f5
Gitweb:        https://git.kernel.org/tip/87195a1ee332add27bd51448c6b54aad551a28f5
Author:        Andrii Nakryiko <[email protected]>
AuthorDate:    Tue, 10 Sep 2024 10:43:12 -07:00
Committer:     Peter Zijlstra <[email protected]>
CommitterDate: Mon, 07 Oct 2024 09:28:42 +02:00

uprobes: switch to RCU Tasks Trace flavor for better performance

This patch switches uprobes SRCU usage to RCU Tasks Trace flavor, which is optimized for more lightweight and quick readers (at the expense of slower writers, which for uprobes is a fine tradeof) and has better performance and scalability with number of CPUs.

Similarly to baseline vs SRCU, we've benchmarked SRCU-based implementation vs RCU Tasks Trace implementation.

SRCU
====

uprobe-nop      ( 1 cpus):    3.276 ± 0.005M/s  (  3.276M/s/cpu)
uprobe-nop      ( 2 cpus):    4.125 ± 0.002M/s  (  2.063M/s/cpu)
uprobe-nop      ( 4 cpus):    7.713 ± 0.002M/s  (  1.928M/s/cpu)
uprobe-nop      ( 8 cpus):    8.097 ± 0.006M/s  (  1.012M/s/cpu)
uprobe-nop      (16 cpus):    6.501 ± 0.056M/s  (  0.406M/s/cpu)
uprobe-nop      (32 cpus):    4.398 ± 0.084M/s  (  0.137M/s/cpu)
uprobe-nop      (64 cpus):    6.452 ± 0.000M/s  (  0.101M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.055 ± 0.001M/s  (  2.055M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.677 ± 0.000M/s  (  1.339M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.561 ± 0.003M/s  (  1.140M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.291 ± 0.002M/s  (  0.661M/s/cpu)
uretprobe-nop   (16 cpus):    5.065 ± 0.019M/s  (  0.317M/s/cpu)
uretprobe-nop   (32 cpus):    3.622 ± 0.003M/s  (  0.113M/s/cpu)
uretprobe-nop   (64 cpus):    3.723 ± 0.002M/s  (  0.058M/s/cpu)

RCU Tasks Trace
===============

uprobe-nop      ( 1 cpus):    3.396 ± 0.002M/s  (  3.396M/s/cpu)
uprobe-nop      ( 2 cpus):    4.271 ± 0.006M/s  (  2.135M/s/cpu)
uprobe-nop      ( 4 cpus):    8.499 ± 0.015M/s  (  2.125M/s/cpu)
uprobe-nop      ( 8 cpus):   10.355 ± 0.028M/s  (  1.294M/s/cpu)
uprobe-nop      (16 cpus):    7.615 ± 0.099M/s  (  0.476M/s/cpu)
uprobe-nop      (32 cpus):    4.430 ± 0.007M/s  (  0.138M/s/cpu)
uprobe-nop      (64 cpus):    6.887 ± 0.020M/s  (  0.108M/s/cpu)
uretprobe-nop   ( 1 cpus):    2.174 ± 0.001M/s  (  2.174M/s/cpu)
uretprobe-nop   ( 2 cpus):    2.853 ± 0.001M/s  (  1.426M/s/cpu)
uretprobe-nop   ( 4 cpus):    4.913 ± 0.002M/s  (  1.228M/s/cpu)
uretprobe-nop   ( 8 cpus):    5.883 ± 0.002M/s  (  0.735M/s/cpu)
uretprobe-nop   (16 cpus):    5.147 ± 0.001M/s  (  0.322M/s/cpu)
uretprobe-nop   (32 cpus):    3.738 ± 0.008M/s  (  0.117M/s/cpu)
uretprobe-nop   (64 cpus):    4.397 ± 0.002M/s  (  0.069M/s/cpu)

Peak throughput for uprobes increases from 8 mln/s to 10.3 mln/s (+28%!), and for uretprobes from 5.3 mln/s to 5.8 mln/s (+11%), as we have more work to do on uretprobes side.

Even single-thread (no contention) performance is slightly better: 3.276 mln/s to 3.396 mln/s (+3.5%) for uprobes, and 2.055 mln/s to 2.174 mln/s (+5.8%) for uretprobes.

We also select TASKS_TRACE_RCU for UPROBES in Kconfig due to the new dependency.

---
 arch/Kconfig            |  1 +
 kernel/events/uprobes.c | 38 ++++++++++++++++----------------------
 2 files changed, 17 insertions(+), 22 deletions(-)
arch/Kconfig +1 -0
index 8af374e..5b5c216 100644
@@ -135,6 +135,7 @@ config KPROBES_ON_FTRACE
 config UPROBES
 	def_bool n
 	depends on ARCH_SUPPORTS_UPROBES
+	select TASKS_TRACE_RCU
 	help
 	  Uprobes is the user-space counterpart to kprobes: they
 	  enable instrumentation applications (such as 'perf probe')
kernel/events/uprobes.c +16 -22
index 4b52cb2..5106dc1 100644
@@ -26,6 +26,7 @@
 #include <linux/task_work.h>
 #include <linux/shmem_fs.h>
 #include <linux/khugepaged.h>
+#include <linux/rcupdate_trace.h>
 
 #include <linux/uprobes.h>
 
@@ -42,8 +43,6 @@ static struct rb_root uprobes_tree = RB_ROOT;
 static DEFINE_RWLOCK(uprobes_treelock);	/* serialize rbtree access */
 static seqcount_rwlock_t uprobes_seqcount = SEQCNT_RWLOCK_ZERO(uprobes_seqcount, &uprobes_treelock);
 
-DEFINE_STATIC_SRCU(uprobes_srcu);
-
 #define UPROBES_HASH_SZ	13
 /* serialize uprobe->pending_list */
 static struct mutex uprobes_mmap_mutex[UPROBES_HASH_SZ];
@@ -651,7 +650,7 @@ static void put_uprobe(struct uprobe *uprobe)
 	delayed_uprobe_remove(uprobe, NULL);
 	mutex_unlock(&delayed_uprobe_lock);
 
-	call_srcu(&uprobes_srcu, &uprobe->rcu, uprobe_free_rcu);
+	call_rcu_tasks_trace(&uprobe->rcu, uprobe_free_rcu);
 }
 
 static __always_inline
@@ -706,7 +705,7 @@ static struct uprobe *find_uprobe_rcu(struct inode *inode, loff_t offset)
 	struct rb_node *node;
 	unsigned int seq;
 
-	lockdep_assert(srcu_read_lock_held(&uprobes_srcu));
+	lockdep_assert(rcu_read_lock_trace_held());
 
 	do {
 		seq = read_seqcount_begin(&uprobes_seqcount);
@@ -934,8 +933,7 @@ static bool filter_chain(struct uprobe *uprobe, struct mm_struct *mm)
 	bool ret = false;
 
 	down_read(&uprobe->consumer_rwsem);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		ret = consumer_filter(uc, mm);
 		if (ret)
 			break;
@@ -1156,7 +1154,7 @@ void uprobe_unregister_sync(void)
 	 * unlucky enough caller can free consumer's memory and cause
 	 * handler_chain() or handle_uretprobe_chain() to do an use-after-free.
 	 */
-	synchronize_srcu(&uprobes_srcu);
+	synchronize_rcu_tasks_trace();
 }
 EXPORT_SYMBOL_GPL(uprobe_unregister_sync);
 
@@ -1240,19 +1238,18 @@ EXPORT_SYMBOL_GPL(uprobe_register);
 int uprobe_apply(struct uprobe *uprobe, struct uprobe_consumer *uc, bool add)
 {
 	struct uprobe_consumer *con;
-	int ret = -ENOENT, srcu_idx;
+	int ret = -ENOENT;
 
 	down_write(&uprobe->register_rwsem);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(con, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(con, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (con == uc) {
 			ret = register_for_each_vma(uprobe, add ? uc : NULL);
 			break;
 		}
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 
 	up_write(&uprobe->register_rwsem);
 
@@ -2134,8 +2131,7 @@ static void handler_chain(struct uprobe *uprobe, struct pt_regs *regs)
 
 	current->utask->auprobe = &uprobe->arch;
 
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		int rc = 0;
 
 		if (uc->handler) {
@@ -2173,15 +2169,13 @@ handle_uretprobe_chain(struct return_instance *ri, struct pt_regs *regs)
 {
 	struct uprobe *uprobe = ri->uprobe;
 	struct uprobe_consumer *uc;
-	int srcu_idx;
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
-	list_for_each_entry_srcu(uc, &uprobe->consumers, cons_node,
-				 srcu_read_lock_held(&uprobes_srcu)) {
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(uc, &uprobe->consumers, cons_node, rcu_read_lock_trace_held()) {
 		if (uc->ret_handler)
 			uc->ret_handler(uc, ri->func, regs);
 	}
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 static struct return_instance *find_next_ret_chain(struct return_instance *ri)
@@ -2266,13 +2260,13 @@ static void handle_swbp(struct pt_regs *regs)
 {
 	struct uprobe *uprobe;
 	unsigned long bp_vaddr;
-	int is_swbp, srcu_idx;
+	int is_swbp;
 
 	bp_vaddr = uprobe_get_swbp_addr(regs);
 	if (bp_vaddr == uprobe_get_trampoline_vaddr())
 		return uprobe_handle_trampoline(regs);
 
-	srcu_idx = srcu_read_lock(&uprobes_srcu);
+	rcu_read_lock_trace();
 
 	uprobe = find_active_uprobe_rcu(bp_vaddr, &is_swbp);
 	if (!uprobe) {
@@ -2330,7 +2324,7 @@ static void handle_swbp(struct pt_regs *regs)
 
 out:
 	/* arch_uprobe_skip_sstep() succeeded, or restart if can't singlestep */
-	srcu_read_unlock(&uprobes_srcu, srcu_idx);
+	rcu_read_unlock_trace();
 }
 
 /*