Skip to contentSkip to conversation

selftests/bpf: extend uprobe/uretprobe triggering benchmarks

PATCH Applied Andrii Nakryiko 3 participants 3 messages Mar 1, 2024 – Mar 4, 2024

Subsystems (3)
lore
Share & tools
lore

Andrii Nakryiko

author maintainer
#0 lore raw

PATCH 1/1 selftests/bpf: extend uprobe/uretprobe triggering benchmarks

Settle on three "flavors" of uprobe/uretprobe, installed on different kinds of instruction: nop, push, and ret. All three are testing different internal code paths emulating or single-stepping instructions, so are interesting to compare and benchmark separately.

To ensure push rbp instruction we ensure that uprobe_target_push() is not a leaf function by calling (global __weak) noop function and returning something afterwards (if we don't do that, compiler will just do a tail call optimization).

Also, we need to make sure that compiler isn't skipping frame pointer generation, so let's add -fno-omit-frame-pointers to Makefile.

Just to give an idea of where we currently stand in terms of relative performance of different uprobe/uretprobe cases vs a cheap syscall (getpgid()) baseline, here are results from my local machine:

$ benchs/run_bench_uprobes.sh
base           :    1.561 ± 0.020M/s
uprobe-nop     :    0.947 ± 0.007M/s
uprobe-push    :    0.951 ± 0.004M/s
uprobe-ret     :    0.443 ± 0.007M/s
uretprobe-nop  :    0.471 ± 0.013M/s
uretprobe-push :    0.483 ± 0.004M/s
uretprobe-ret  :    0.306 ± 0.007M/s
---
 tools/testing/selftests/bpf/Makefile          |   2 +-
 tools/testing/selftests/bpf/bench.c           |  20 +--
 .../selftests/bpf/benchs/bench_trigger.c      | 118 ++++++++++++------
 .../selftests/bpf/benchs/run_bench_uprobes.sh |   9 ++
 4 files changed, 103 insertions(+), 46 deletions(-)
 create mode 100755 tools/testing/selftests/bpf/benchs/run_bench_uprobes.sh
tools/testing/selftests/bpf/Makefile +1 -1
index a38a3001527c..eccba9f3a77d 100644
@@ -34,7 +34,7 @@ LIBELF_CFLAGS	:= $(shell $(PKG_CONFIG) libelf --cflags 2>/dev/null)
 LIBELF_LIBS	:= $(shell $(PKG_CONFIG) libelf --libs 2>/dev/null || echo -lelf)
 
 CFLAGS += -g $(OPT_FLAGS) -rdynamic					\
-	  -Wall -Werror 						\
+	  -Wall -Werror -fno-omit-frame-pointer				\
 	  $(GENFLAGS) $(SAN_CFLAGS) $(LIBELF_CFLAGS)			\
 	  -I$(CURDIR) -I$(INCLUDE_DIR) -I$(GENDIR) -I$(LIBDIR)		\
 	  -I$(TOOLSINCDIR) -I$(APIDIR) -I$(OUTPUT)
tools/testing/selftests/bpf/bench.c +12 -8
index 73ce11b0547d..c632811f9171 100644
@@ -499,10 +499,12 @@ extern const struct bench bench_trig_fentry;
 extern const struct bench bench_trig_fentry_sleep;
 extern const struct bench bench_trig_fmodret;
 extern const struct bench bench_trig_uprobe_base;
-extern const struct bench bench_trig_uprobe_with_nop;
-extern const struct bench bench_trig_uretprobe_with_nop;
-extern const struct bench bench_trig_uprobe_without_nop;
-extern const struct bench bench_trig_uretprobe_without_nop;
+extern const struct bench bench_trig_uprobe_nop;
+extern const struct bench bench_trig_uretprobe_nop;
+extern const struct bench bench_trig_uprobe_push;
+extern const struct bench bench_trig_uretprobe_push;
+extern const struct bench bench_trig_uprobe_ret;
+extern const struct bench bench_trig_uretprobe_ret;
 extern const struct bench bench_rb_libbpf;
 extern const struct bench bench_rb_custom;
 extern const struct bench bench_pb_libbpf;
@@ -541,10 +543,12 @@ static const struct bench *benchs[] = {
 	&bench_trig_fentry_sleep,
 	&bench_trig_fmodret,
 	&bench_trig_uprobe_base,
-	&bench_trig_uprobe_with_nop,
-	&bench_trig_uretprobe_with_nop,
-	&bench_trig_uprobe_without_nop,
-	&bench_trig_uretprobe_without_nop,
+	&bench_trig_uprobe_nop,
+	&bench_trig_uretprobe_nop,
+	&bench_trig_uprobe_push,
+	&bench_trig_uretprobe_push,
+	&bench_trig_uprobe_ret,
+	&bench_trig_uretprobe_ret,
 	&bench_rb_libbpf,
 	&bench_rb_custom,
 	&bench_pb_libbpf,
tools/testing/selftests/bpf/benchs/bench_trigger.c +81 -37

192 lines not on this page · raw message

tools/testing/selftests/bpf/benchs/run_bench_uprobes.sh new file +9 -0
new file mode 100755
index 000000000000..9bdcc74e03a4
@@ -0,0 +1,9 @@
+#!/bin/bash
+
+set -eufo pipefail
+
+for i in base {uprobe,uretprobe}-{nop,push,ret}
+do
+	summary=$(sudo ./bench -w2 -d5 -a trig-$i | tail -n1 | cut -d'(' -f1 | cut -d' ' -f3-)
+	printf "%-15s: %s\n" $i "$summary"
+done
Signature · 1 line
2.43.0
#1 lore raw
This patch was applied to bpf/bpf-next.git (master) by Daniel Borkmann <[email protected]>:

Hello:

This patch was applied to bpf/bpf-next.git (master) by Daniel Borkmann <[email protected]>:

Quoted 11 lines · On Fri, 1 Mar 2024 13:45:51 -0800 you wrote:

Settle on three "flavors" of uprobe/uretprobe, installed on different kinds of instruction: nop, push, and ret. All three are testing different internal code paths emulating or single-stepping instructions, so are interesting to compare and benchmark separately.

To ensure push rbp instruction we ensure that uprobe_target_push() is not a leaf function by calling (global __weak) noop function and returning something afterwards (if we don't do that, compiler will just do a tail call optimization).

[...]

Here is the summary with links:
- [bpf-next] selftests/bpf: extend uprobe/uretprobe triggering benchmarks

    https://git.kernel.org/bpf/bpf-next/c/8f79870ec8a9

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
maintainer
#2 lore raw
In reply to Andrii Nakryiko

On 3/1/24 10:45 PM, Andrii Nakryiko wrote:
[...]

Quoted 8 lines
tools/testing/selftests/bpf/benchs/bench_trigger.c +0 -0
index dbd362771d6a..a8ed8afff0b3 100644
@@ -113,12 +113,25 @@ static void trigger_fmodret_setup(void)
   * GCC doesn't generate stack setup preample for these functions due to them
   * having no input arguments and doing nothing in the body.
   */

tools/testing/selftests/bpf/benchs/bench_trigger.c static void trigger_fmodret_setup(void) · line 123

Quoted 10 lines
tools/testing/selftests/bpf/benchs/bench_trigger.c +4 -2
-__weak void uprobe_target_with_nop(void)
+__weak void uprobe_target_nop(void)
  {
  	asm volatile ("nop");
  }
 
-__weak void uprobe_target_without_nop(void)
+__weak void opaque_noop_func()
+{
+}

Nice work and super useful to have! I fixed up the void while applying.

Thanks,
Daniel