khamidou / lptrace Goto Github PK

View Code? Open in Web Editor NEW

699.0 699.0 43.0 27 KB

Trace any Python program, anywhere!

Home Page: http://khamidou.com/lptrace/

License: GNU General Public License v3.0

Python 100.00%

lptrace's People

Contributors

Stargazers

Watchers

lptrace's Issues

Funky Zero length field name error.

Traceback (most recent call last):
  File "lptrace", line 116, in <module>
    main()
  File "lptrace", line 112, in main
    strace(pid)
  File "lptrace", line 50, in strace
    runfile(pid, trace_code)
  File "lptrace", line 39, in runfile
    cmd = 'execfile(\\"{}\\")'.format(name)
ValueError: zero length field name in forma

I went through the instructions in your README and this error was returned. Any thoughts?

Publish on PyPI

As easy integration with production environments is a concern, this would be much easier to integrate into workflows if it was on PyPI and could be installed with pip. Even better would be to have pip install lptrace register it such that it can be called from the command line without providing an explicit path to the script (as pytest and many others do).

octal literal in os.chmod?

os.chmod(tmp.name, 0666)
os.chmod(fifo_name, 0777)

So 0666 === 438 and 0777 == 511 is these two lines intentional or a typo?

Python version should be explicit

With more and more distributions switching to python3 by default and python2 being clearly deprecated putting a little shebang

#!/bin/env python2

sounds like a reasonable thing to do. Sure it's little trouble to write an alias to python2 lptrace.py but it shouldn't be necessary.

Python3 mode

I haven't gotten the script to work with Python3 yet. Things I did so far:

print-statements (obviously)
octal literals with 0o666 instead of 0666
cmd = 'exec(open(\\"{}\\", "r").read())'.format(name) because Python3 doesn't have execfile
tmp.write(script.encode()) due to bytes/string conversion

I'm guessing that I'm either screwing up with bytes/strings/file reads or the Python3 interface differs in a way that doesn't raise an exception explaining the problem. Anybody with an insight here?

detaching results in SIGABRT

I tried running this trivial Python program with lptrace:

$ cat loop_forever.py
#! /usr/bin/env python

import os
print os.getpid()

while True:
    pass

I ran it:

$ ./loop_forever.py 
25116

then I ran lptrace in another terminal.

$ sudo python lptrace -p 25116

I had to use Ctrl-C twice to get it to quit.

$ sudo python lptrace -p 25116
^CReceived Ctrl-C, quitting
^CReceived Ctrl-C, quitting

After this, loop_forever.py exited with signal 6 (SIGABRT):

$ ./loop_forever.py 
25116
Fatal Python error: This thread state must be current when releasing
Aborted (core dumped)

This may be the same bug as #11, but I wanted to demonstrate it is easy to reproduce on a trivial program, not just Guake.

I am on Debian 9.3 (Stretch), running Python 2.7.13, GDB version 7.12-6, and the latest Github version of lptrace.

$ lsb_release -a
No LSB modules are available.
Distributor ID:	Debian
Description:	Debian GNU/Linux 9.3 (stretch)
Release:	9.3
Codename:	stretch
$ python --version
Python 2.7.13
$ gdb --version | head -n 1
GNU gdb (Debian 7.12-6) 7.12.0.20161007-git
$ git rev-parse HEAD
537053c5fd6c3d84d0fb73d675fe926d470c82d5

Full backtrace of dumped core file is below:

Core was generated by `python ./loop_forever.py'.
Program terminated with signal SIGABRT, Aborted.
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
51	../sysdeps/unix/sysv/linux/raise.c: No such file or directory.
set = {__val = {0, 140451948041416, 140726353628400, 4294967301, 140451948040560, 140726353628384, 93833867754558, 252833149, 4294967295, 1, 140451930217944, 140451947630592, 93833869879752, 93833869879750, 140451946566352, 8}}
pid = <optimized out>
tid = <optimized out>
sig = 6
  Id   Target Id         Frame 
* 1    Thread 0x7fbd84723700 (LWP 31965) __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51

Thread 1 (Thread 0x7fbd84723700 (LWP 31965)):
#0  __GI_raise (sig=sig@entry=6) at ../sysdeps/unix/sysv/linux/raise.c:51
        set = {__val = {0, 140451948041416, 140726353628400, 4294967301, 140451948040560, 140726353628384, 93833867754558, 252833149, 4294967295, 1, 140451930217944, 140451947630592, 93833869879752, 93833869879750, 140451946566352, 8}}
        pid = <optimized out>
        tid = <optimized out>
#1  0x00007fbd836b73fa in __GI_abort () at abort.c:89
        save_stage = 2
        act = {__sigaction_handler = {sa_handler = 0x0, sa_sigaction = 0x0}, sa_mask = {__val = {0, 93833891493728, 0, 1, 140451943638619, 93833891511584, 93833868582196, 93833891493728, 0, 0, 2880, 0, 140451933972304, 140451933967648, 140451933972304, 140451947624192}}, sa_flags = -2089902068, sa_restorer = 0x7fbd83a1d6e0 <stderr>}
        sigs = {__val = {32, 0 <repeats 15 times>}}
#2  0x00005557665fb8b2 in Py_FatalError (msg=<optimized out>) at ../Python/pythonrun.c:1700
No locals.
#3  0x000055576664aff8 in PyGILState_Release (oldstate=PyGILState_LOCKED) at ../Python/pystate.c:635
        tcur = <optimized out>
#4  0x00007ffd68518f0f in ?? ()
No symbol table info available.
#5  0x0000000000000300 in ?? ()
No symbol table info available.
#6  0xcc00000000000330 in ?? ()
No symbol table info available.
#7  0x000000000000000c in ?? ()
No symbol table info available.
#8  0x4fab7054cad2ce00 in ?? ()
No symbol table info available.
#9  0x0000000000000002 in ?? ()
No symbol table info available.
#10 0x00007fbd8465e5d0 in ?? ()
No symbol table info available.
#11 0x00007fbd845fbf80 in ?? ()
No symbol table info available.
#12 0x0000000000000000 in ?? ()
No symbol table info available.

Check if gdb exists

Right now if you launch lptrace without having gdb on your system, it will just end up waiting doing nothing. I guess it would be better if an error was returned.

seems not work well.

in one terminal run the script:
from time import sleep
for i in range(10000):
print "in loop"
sleep(2)

in another terminal run the:
sudo lptrace -p PID_OF_ABOVE_SCRIPT

no output in the second terminal, could you help to take a look why? Thanks in advance.

Detaching from Guake terminal crashes Guake process.

I attached to Guake and then pressed Ctrl-C. Guake is closed.

Expected result: guake continue to work

(guake version - 0.8.8, lptrace - 1.0)
https://dl.dropboxusercontent.com/1/view/k5kard6wqahrskb/Apps/Shutter/%C3%90%C2%92%C3%91%C2%8B%C3%90%C2%B4%C3%90%C2%B5%C3%90%C2%BB%C3%90%C2%B5%C3%90%C2%BD%C3%90%C2%B8%C3%90%C2%B5_003.png

lptrace hangs with no error if gdb is missing

I figured this out by inspecting the err result of p.communicate() on line 86. It would be nice if that were printed :)

Lptrace just hang

When I use lptrace, i just simply run python lptrace -p 21640. But this command just hang forever.
When i add some print log in lptrace like this
`
signal.signal(signal.SIGINT, sigint_handler)

print("bbbbbbbbbb")
with open(fifo_name) as fd:
    print("ssssssssssss")
    while True:
        print("aaaaaaaa")
        data = fd.read()
        print(111111111111)
        if data != '':
            print data

print("mmmmmmmm")

`
it will only print "bbbbbbb", but not any other logs. Did i miss anything

Failed to run on macOS

First of all, thanks for providing this little cool script to help debugging Python programs.

I've been using it successfully on ubuntu server, but when I tried this on my Mac, sudo python lptrace -p 96365 hanged there and the target program was failed by segmentation fault:

$ python -m SimpleHTTPServer 10000
Serving HTTP on 0.0.0.0 port 10000 ...
[1]    96365 segmentation fault  python -m SimpleHTTPServer 10000

I was using Python 2.7.12 installed by homebrew. My system is OS X 10.11.6.

khamidou / lptrace Goto Github PK

lptrace's People

Contributors

Stargazers

Watchers

Forkers

lptrace's Issues

Recommend Projects

Recommend Topics

Recommend Org