Consider this multi-threaded program:
import threading
class SomeThread(threading.Thread):
def run(self):
a = 1
print a
def main():
print 'hola'
someThread = SomeThread()
someThread.start()
if __name__ == '__main__':
main()
When I debug this program with pdb, at the prompt I first set a break point at each of the two print statements. Then I continue. pdb breaks at print 'hola'
. I continue again and see the effect of the print in the other thread, but pdb doesn't break.
The help commands don't list anything to switch thread contexts like gdb... so... is it just not possible in one thread context to set a breakpoint that will trip in another context?
This works for me:
import threading
import pdb
class SomeThread(threading.Thread):
def run(self):
a = 1
print a
pdb.set_trace()
def main():
print 'hola'
pdb.set_trace()
someThread = SomeThread()
someThread.start()
if __name__ == '__main__':
main()
Which gives me:
C:\Code>python b.py
hola
> c:\code\b.py(13)main()
-> someThread = SomeThread()
(Pdb) l
8 pdb.set_trace()
9
10 def main():
11 print 'hola'
12 pdb.set_trace()
13 -> someThread = SomeThread()
14 someThread.start()
15
16 if __name__ == '__main__':
17 main()
[EOF]
(Pdb) c
1
--Return--
> c:\code\b.py(8)run()->None
-> pdb.set_trace()
(Pdb) l
3
4 class SomeThread(threading.Thread):
5 def run(self):
6 a = 1
7 print a
8 -> pdb.set_trace()
9
10 def main():
11 print 'hola'
12 pdb.set_trace()
13 someThread = SomeThread()
(Pdb)
This is under Windows 7 and with Python 2.7.2. What OS & Python version are you using?
after you hit your first breakpoint, I assume you are stepping with (n) next line
when you get to this line
someThread.start()
make sure you are using (s) step-into and not (n) next.
pdb commands