From 69a309fde29aa6911522a987e7f4db11e07aeefb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E8=B5=B5=E5=9B=BD=E7=91=9E?= <18235787078@163.com>
Date: Sat, 21 Sep 2024 20:07:59 +0800
Subject: [PATCH] =?UTF-8?q?classify=E9=80=89=E9=A1=B9=E7=9B=B8=E5=85=B3?=
 =?UTF-8?q?=E6=96=87=E6=A1=A3?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 site/docs/doc/profiler.md    | 14 ++++++++++++++
 site/docs/en/doc/profiler.md | 13 +++++++++++++
 2 files changed, 27 insertions(+)

diff --git a/site/docs/doc/profiler.md b/site/docs/doc/profiler.md
index 0e0f2d22982..5007978f699 100644
--- a/site/docs/doc/profiler.md
+++ b/site/docs/doc/profiler.md
@@ -350,3 +350,17 @@ profiler start --loop 1h -f /var/log/profile-%t.jfr
 ## `--timeout` 选项
 
 这个选项指定 profiling 自动在多久后停止。该选项和 `--loop` 选项的格式一致，可以是时间点，也可以是一个时间间隔。这两个选项都是用于 `start` action 而不是 `collect` action 的。可参考 [async-profiler Github Discussions](https://github.com/async-profiler/async-profiler/discussions/789) 了解更多信息。
+
+
+## `--classify` 选项
+这个选项用于自动将执行样本分类到不同的类别中。
+这些类别包括垃圾回收、JIT 编译、类加载等。
+这个功能可以帮助你在分析火焰图时更容易地回答以下问题：
+* 总共花了多少时间在类加载上？
+* Java 代码在解释器和 C1/C2 编译方法中运行的时间百分比是多少？
+* Lambda 对启动时间的影响是什么？
+
+（推荐使用 --cstack dwarf 选项以提高准确性，可参考[async profiler pr#719](https://github.com/async-profiler/async-profiler/issues/719)）
+```bash
+profiler start -e cpu -f /tmp/profile.jfr --cstack dwarf
+```
diff --git a/site/docs/en/doc/profiler.md b/site/docs/en/doc/profiler.md
index fa7318a276e..abaf8c3dccb 100644
--- a/site/docs/en/doc/profiler.md
+++ b/site/docs/en/doc/profiler.md
@@ -352,3 +352,16 @@ profiler start --loop 1h -f /var/log/profile-%t.jfr
 This option specifies the time when profiling will automatically stop. The format is the same as in loop: it is either a wall clock time (12:34:56) or a relative time interval (2h).
 
 Both `--loop` and `--timeout` are used for `start` action but not for `collect` action, for further information refer to [async-profiler Github Discussions](https://github.com/async-profiler/async-profiler/discussions/789).
+
+## --classify Option
+This option is used to automatically classify execution samples into different categories.
+These categories include garbage collection, JIT compilation, class loading, etc.
+Specifically, this feature can help you more easily answer the following questions when analyzing flame graphs:
+* How much time was spent on class loading?
+* What percentage of Java code runs in the interpreter and C1/C2 compiled methods?
+* What is the impact of Lambda on startup time?
+
+(It is recommended to use the --cstack dwarf option to improve accuracy,, for further information refer to [async profiler pr#719](https://github.com/async-profiler/async-profiler/issues/719).)
+```bash
+profiler start -e cpu -f /tmp/profile.jfr --cstack dwarf
+```