Commit 6e1fac41 authored by 4ast's avatar 4ast Committed by GitHub

Merge pull request #963 from goldshtn/utools-php

u* tools: PHP support
parents 0ea2ffa7 cfb5ee7c
......@@ -2,28 +2,29 @@
.SH NAME
ucalls \- Summarize method calls from high-level languages and Linux syscalls.
.SH SYNOPSIS
.B ucalls [-l {java,python,ruby}] [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
.B ucalls [-l {java,python,ruby,php}] [-h] [-T TOP] [-L] [-S] [-v] [-m] pid [interval]
.SH DESCRIPTION
This tool summarizes method calls from high-level languages such as Python,
Java, and Ruby. It can also trace Linux system calls. Whenever a method is
Java, Ruby, and PHP. It can also trace Linux system calls. Whenever a method is
invoked, ucalls records the call count and optionally the method's execution
time (latency) and displays a summary.
This uses in-kernel eBPF maps to store per process summaries for efficiency.
This tool relies on USDT probes embedded in many high-level languages, such as
Node, Java, Python, and Ruby. It requires a runtime instrumented with these
Java, Python, Ruby, and PHP. It requires a runtime instrumented with these
probes, which in some cases requires building from source with a USDT-specific
flag, such as "--enable-dtrace" or "--with-dtrace". For Java, method probes are
not enabled by default, and can be turned on by running the Java process with
the "-XX:+ExtendedDTraceProbes" flag.
the "-XX:+ExtendedDTraceProbes" flag. For PHP processes, the environment
variable USE_ZEND_DTRACE must be set to 1.
Since this uses BPF, only the root user can use this tool.
.SH REQUIREMENTS
CONFIG_BPF and bcc.
.SH OPTIONS
.TP
\-l {java,python,ruby,node}
\-l {java,python,ruby,php}
The language to trace. If not provided, only syscalls are traced (when the \-S
option is used).
.TP
......
......@@ -2,16 +2,17 @@
.SH NAME
uflow \- Print a flow graph of method calls in high-level languages.
.SH SYNOPSIS
.B uflow [-h] [-M METHOD] [-C CLAZZ] [-v] {java,python,ruby} pid
.B uflow [-h] [-M METHOD] [-C CLAZZ] [-v] {java,python,ruby,php} pid
.SH DESCRIPTION
uflow traces method calls and prints them in a flow graph that can facilitate
debugging and diagnostics by following the program's execution (method flow).
This tool relies on USDT probes embedded in many high-level languages, such as
Node, Java, Python, and Ruby. It requires a runtime instrumented with these
Java, Python, Ruby, and PHP. It requires a runtime instrumented with these
probes, which in some cases requires building from source with a USDT-specific
flag, such as "--enable-dtrace" or "--with-dtrace". For Java processes, the
startup flag "-XX:+ExtendedDTraceProbes" is required.
startup flag "-XX:+ExtendedDTraceProbes" is required. For PHP processes, the
environment variable USE_ZEND_DTRACE must be set to 1.
Since this uses BPF, only the root user can use this tool.
.SH REQUIREMENTS
......@@ -29,7 +30,7 @@ name interpretation strongly depends on the language. For example, in Java use
\-v
Print the resulting BPF program, for debugging purposes.
.TP
{java,python,ruby}
{java,python,ruby,php}
The language to trace.
.TP
pid
......
......@@ -2,7 +2,7 @@
.SH NAME
ustat \- Activity stats from high-level languages.
.SH SYNOPSIS
.B ustat [-l {java,python,ruby,node}] [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
.B ustat [-l {java,python,ruby,node,php}] [-C] [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d] [interval [count]]
.SH DESCRIPTION
This is "top" for high-level language events, such as garbage collections,
exceptions, thread creations, object allocations, method calls, and more. The
......@@ -12,11 +12,12 @@ can be sorted by various fields.
This uses in-kernel eBPF maps to store per process summaries for efficiency.
This tool relies on USDT probes embedded in many high-level languages, such as
Node, Java, Python, and Ruby. It requires a runtime instrumented with these
Node, Java, Python, Ruby, and PHP. It requires a runtime instrumented with these
probes, which in some cases requires building from source with a USDT-specific
flag, such as "--enable-dtrace" or "--with-dtrace". For Java, some probes are
not enabled by default, and can be turned on by running the Java process with
the "-XX:+ExtendedDTraceProbes" flag.
the "-XX:+ExtendedDTraceProbes" flag. For PHP processes, the environment
variable USE_ZEND_DTRACE must be set to 1.
Newly-created processes will only be traced at the next interval. If you run
this tool with a short interval (say, 1-5 seconds), this should be virtually
......@@ -28,7 +29,7 @@ Since this uses BPF, only the root user can use this tool.
CONFIG_BPF and bcc.
.SH OPTIONS
.TP
\-l {java,python,ruby,node}
\-l {java,python,ruby,node,php}
The language to trace. By default, all languages are traced.
.TP
\-C
......
......@@ -4,7 +4,7 @@
# ucalls Summarize method calls in high-level languages and/or system calls.
# For Linux, uses BCC, eBPF.
#
# USAGE: ucalls [-l {java,python,ruby}] [-h] [-T TOP] [-L] [-S] [-v] [-m]
# USAGE: ucalls [-l {java,python,ruby,php}] [-h] [-T TOP] [-L] [-S] [-v] [-m]
# pid [interval]
#
# Copyright 2016 Sasha Goldshtein
......@@ -24,7 +24,7 @@ examples = """examples:
./ucalls 6712 -S # trace only syscall counts
./ucalls -l ruby 1344 -T 10 # trace top 10 Ruby method calls
./ucalls -l ruby 1344 -L # trace Ruby calls including latency
./ucalls -l ruby 1344 -LS # trace Ruby calls and syscalls with latency
./ucalls -l php 443 -LS # trace PHP calls and syscalls with latency
./ucalls -l python 2020 -mL # trace Python calls including latency in ms
"""
parser = argparse.ArgumentParser(
......@@ -34,7 +34,8 @@ parser = argparse.ArgumentParser(
parser.add_argument("pid", type=int, help="process id to attach to")
parser.add_argument("interval", type=int, nargs='?',
help="print every specified number of seconds")
parser.add_argument("-l", "--language", choices=["java", "python", "ruby"],
parser.add_argument("-l", "--language",
choices=["java", "python", "ruby", "php"],
help="language to trace (if none, trace syscalls only)")
parser.add_argument("-T", "--top", type=int,
help="number of most frequent/slow calls to print")
......@@ -49,8 +50,8 @@ parser.add_argument("-m", "--milliseconds", action="store_true",
args = parser.parse_args()
# We assume that the entry and return probes have the same arguments. This is
# the case for Java, Python, and Ruby. If there's a language where it's not the
# case, we will need to build a custom correlator from entry to exit.
# the case for Java, Python, Ruby, and PHP. If there's a language where it's
# not the case, we will need to build a custom correlator from entry to exit.
if args.language == "java":
# TODO for JVM entries, we actually have the real length of the class
# and method strings in arg3 and arg5 respectively, so we can insert
......@@ -70,6 +71,11 @@ elif args.language == "ruby":
return_probe = "method__return"
read_class = "bpf_usdt_readarg(1, ctx, &clazz);"
read_method = "bpf_usdt_readarg(2, ctx, &method);"
elif args.language == "php":
entry_probe = "function__entry"
return_probe = "function__return"
read_class = "bpf_usdt_readarg(4, ctx, &clazz);"
read_method = "bpf_usdt_readarg(1, ctx, &method);"
elif not args.language:
if not args.syscalls:
print("Nothing to do; use -S to trace syscalls.")
......
......@@ -2,7 +2,7 @@ Demonstrations of ucalls.
ucalls summarizes method calls in various high-level languages, including Java,
Python, Ruby, and Linux system calls. It displays statistics on the most
Python, Ruby, PHP, and Linux system calls. It displays statistics on the most
frequently called methods, as well as the latency (duration) of these methods.
Through the syscalls support, ucalls can provide basic information on a
......@@ -60,7 +60,7 @@ METHOD # CALLS
USAGE message:
# ./ucalls.py -h
usage: ucalls.py [-h] [-l {java,python,ruby}] [-T TOP] [-L] [-S] [-v] [-m]
usage: ucalls.py [-h] [-l {java,python,ruby,php}] [-T TOP] [-L] [-S] [-v] [-m]
pid [interval]
Summarize method calls in high-level languages.
......@@ -71,7 +71,7 @@ positional arguments:
optional arguments:
-h, --help show this help message and exit
-l {java,python,ruby}, --language {java,python,ruby}
-l {java,python,ruby,php}, --language {java,python,ruby,php}
language to trace (if none, trace syscalls only)
-T TOP, --top TOP number of most frequent/slow calls to print
-L, --latency record method latency from enter to exit (except
......@@ -88,5 +88,5 @@ examples:
./ucalls 6712 -S # trace only syscall counts
./ucalls -l ruby 1344 -T 10 # trace top 10 Ruby method calls
./ucalls -l ruby 1344 -L # trace Ruby calls including latency
./ucalls -l ruby 1344 -LS # trace Ruby calls and syscalls with latency
./ucalls -l php 443 -LS # trace PHP calls and syscalls with latency
./ucalls -l python 2020 -mL # trace Python calls including latency in ms
......@@ -4,7 +4,7 @@
# uflow Trace method execution flow in high-level languages.
# For Linux, uses BCC, eBPF.
#
# USAGE: uflow [-C CLASS] [-M METHOD] [-v] {java,python,ruby} pid
# USAGE: uflow [-C CLASS] [-M METHOD] [-v] {java,python,ruby,php} pid
#
# Copyright 2016 Sasha Goldshtein
# Licensed under the Apache License, Version 2.0 (the "License")
......@@ -27,7 +27,7 @@ parser = argparse.ArgumentParser(
description="Trace method execution flow in high-level languages.",
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog=examples)
parser.add_argument("language", choices=["java", "python", "ruby"],
parser.add_argument("language", choices=["java", "python", "ruby", "php"],
help="language to trace")
parser.add_argument("pid", type=int, help="process id to attach to")
parser.add_argument("-M", "--method",
......@@ -140,6 +140,13 @@ elif args.language == "ruby":
enable_probe("cmethod__return", "ruby_creturn",
"bpf_usdt_readarg(1, ctx, &clazz);",
"bpf_usdt_readarg(2, ctx, &method);", is_return=True)
elif args.language == "php":
enable_probe("function__entry", "php_entry",
"bpf_usdt_readarg(4, ctx, &clazz);",
"bpf_usdt_readarg(1, ctx, &method);", is_return=False)
enable_probe("function__return", "php_return",
"bpf_usdt_readarg(4, ctx, &clazz);",
"bpf_usdt_readarg(1, ctx, &method);", is_return=True)
if args.verbose:
print(usdt.get_text())
......
......@@ -4,8 +4,8 @@ Demonstrations of uflow.
uflow traces method entry and exit events and prints a visual flow graph that
shows how methods are entered and exited, similar to a tracing debugger with
breakpoints. This can be useful for understanding program flow in high-level
languages such as Java, Python, and Ruby, which provide USDT probes for method
invocations.
languages such as Java, Python, Ruby, and PHP, which provide USDT probes for
method invocations.
For example, trace all Ruby method calls in a specific process:
......@@ -88,12 +88,13 @@ thread running on the same CPU.
USAGE message:
# ./uflow -h
usage: uflow.py [-h] [-M METHOD] [-C CLAZZ] [-v] {java,python,ruby} pid
usage: uflow.py [-h] [-M METHOD] [-C CLAZZ] [-v] {java,python,ruby,php} pid
Trace method execution flow in high-level languages.
positional arguments:
{java,python,ruby} language to trace
{java,python,ruby,php}
language to trace
pid process id to attach to
optional arguments:
......
......@@ -5,7 +5,7 @@
# method calls, class loads, garbage collections, and more.
# For Linux, uses BCC, eBPF.
#
# USAGE: ustat [-l {java,python,ruby,node}] [-C]
# USAGE: ustat [-l {java,python,ruby,node,php}] [-C]
# [-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d]
# [interval [count]]
#
......@@ -132,7 +132,7 @@ class Tool(object):
formatter_class=argparse.RawDescriptionHelpFormatter,
epilog=examples)
parser.add_argument("-l", "--language",
choices=["java", "python", "ruby", "node"],
choices=["java", "python", "ruby", "node", "php"],
help="language to trace (default: all languages)")
parser.add_argument("-C", "--noclear", action="store_true",
help="don't clear the screen")
......@@ -158,6 +158,11 @@ class Tool(object):
"function__entry": Category.METHOD,
"gc__start": Category.GC
}),
"php": Probe("php", ["php"], {
"function__entry": Category.METHOD,
"compile__file__entry": Category.CLOAD,
"exception__thrown": Category.EXCP
}),
"ruby": Probe("ruby", ["ruby", "irb"], {
"method__entry": Category.METHOD,
"cmethod__entry": Category.METHOD,
......
......@@ -4,7 +4,7 @@ Demonstrations of ustat.
ustat is a "top"-like tool for monitoring events in high-level languages. It
prints statistics about garbage collections, method calls, object allocations,
and various other events for every process that it recognizes with a Java,
Python, Ruby, or Node runtime.
Python, Ruby, Node, or PHP runtime.
For example:
......@@ -48,7 +48,7 @@ PID CMDLINE METHOD/s GC/s OBJNEW/s CLOAD/s EXC/s THR/s
USAGE message:
# ./ustat.py -h
usage: ustat.py [-h] [-l {java,python,ruby,node}] [-C]
usage: ustat.py [-h] [-l {java,python,ruby,node,php}] [-C]
[-S {cload,excp,gc,method,objnew,thread}] [-r MAXROWS] [-d]
[interval] [count]
......@@ -60,7 +60,7 @@ positional arguments:
optional arguments:
-h, --help show this help message and exit
-l {java,python,ruby,node}, --language {java,python,ruby,node}
-l {java,python,ruby,node,php}, --language {java,python,ruby,node,php}
language to trace (default: all languages)
-C, --noclear don't clear the screen
-S {cload,excp,gc,method,objnew,thread}, --sort {cload,excp,gc,method,objnew,thread}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment