Mirror
/
cpython
mirror of https://github.com/python/cpython
4
1
Fork 0
Code Issues Releases Wiki Activity
cpython/Lib/test/output/test_profile

88 lines
5.2 KiB
Plaintext
Raw Normal View History

Hopefully fix the profiler right. Add a test suite that checks that it deals correctly with some anomalous cases; according to this test suite I've fixed it right. The anomalous cases had to do with 'exception' events: these aren't generated when they would be most helpful, and the profiler has to work hard to recover the right information. The problems occur when C code (such as hasattr(), which is used as the example here) calls back into Python code and clears an exception raised by that Python code. Consider this example: def foo(): hasattr(obj, "bar") Where obj is an instance from a class like this: class C: def __getattr__(self, name): raise AttributeError The profiler sees the following sequence of events: call (foo) call (__getattr__) exception (in __getattr__) return (from foo) Previously, the profiler would assume the return event returned from __getattr__. An if statement checking for this condition and raising an exception was commented out... This version does the right thing.
2001-10-03 21:58:24 -03:00
test_profile
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
128 function calls (108 primitive calls) in 1.000 CPU seconds
Hopefully fix the profiler right. Add a test suite that checks that it deals correctly with some anomalous cases; according to this test suite I've fixed it right. The anomalous cases had to do with 'exception' events: these aren't generated when they would be most helpful, and the profiler has to work hard to recover the right information. The problems occur when C code (such as hasattr(), which is used as the example here) calls back into Python code and clears an exception raised by that Python code. Consider this example: def foo(): hasattr(obj, "bar") Where obj is an instance from a class like this: class C: def __getattr__(self, name): raise AttributeError The profiler sees the following sequence of events: call (foo) call (__getattr__) exception (in __getattr__) return (from foo) Previously, the profiler would assume the return event returned from __getattr__. An if statement checking for this condition and raising an exception was commented out... This version does the right thing.
2001-10-03 21:58:24 -03:00
Ordered by: standard name
ncalls tottime percall cumtime percall filename:lineno(function)
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
4 0.000 0.000 0.000 0.000 :0(append)
4 0.000 0.000 0.000 0.000 :0(exc_info)
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
1 0.000 0.000 1.000 1.000 :0(exec)
Using repr() generates entries that the current stats package can't collate, so setting it back to the function name
2004-07-12 20:38:02 -03:00
12 0.000 0.000 0.012 0.001 :0(hasattr)
8 0.000 0.000 0.000 0.000 :0(range)
1 0.000 0.000 0.000 0.000 :0(setprofile)
Merge ast-branch to head This change implements a new bytecode compiler, based on a transformation of the parse tree to an abstract syntax defined in Parser/Python.asdl. The compiler implementation is not complete, but it is in stable enough shape to run the entire test suite excepting two disabled tests.
2005-10-20 16:59:25 -03:00
1 0.000 0.000 1.000 1.000 <string>:1(<module>)
Hopefully fix the profiler right. Add a test suite that checks that it deals correctly with some anomalous cases; according to this test suite I've fixed it right. The anomalous cases had to do with 'exception' events: these aren't generated when they would be most helpful, and the profiler has to work hard to recover the right information. The problems occur when C code (such as hasattr(), which is used as the example here) calls back into Python code and clears an exception raised by that Python code. Consider this example: def foo(): hasattr(obj, "bar") Where obj is an instance from a class like this: class C: def __getattr__(self, name): raise AttributeError The profiler sees the following sequence of events: call (foo) call (__getattr__) exception (in __getattr__) return (from foo) Previously, the profiler would assume the return event returned from __getattr__. An if statement checking for this condition and raising an exception was commented out... This version does the right thing.
2001-10-03 21:58:24 -03:00
0 0.000 0.000 profile:0(profiler)
1 0.000 0.000 1.000 1.000 profile:0(testfunc())
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
8 0.064 0.008 0.080 0.010 test_profile.py:103(subhelper)
28 0.028 0.001 0.028 0.001 test_profile.py:115(__getattr__)
1 0.270 0.270 1.000 1.000 test_profile.py:30(testfunc)
23/3 0.150 0.007 0.170 0.057 test_profile.py:40(factorial)
20 0.020 0.001 0.020 0.001 test_profile.py:53(mul)
2 0.040 0.020 0.600 0.300 test_profile.py:60(helper)
4 0.116 0.029 0.120 0.030 test_profile.py:78(helper1)
2 0.000 0.000 0.140 0.070 test_profile.py:89(helper2_indirect)
8 0.312 0.039 0.400 0.050 test_profile.py:93(helper2)
Ordered by: standard name
Function called...
:0(append) ->
:0(exc_info) ->
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
:0(exec) -> <string>:1(<module>)(1) 1.000
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
:0(hasattr) -> test_profile.py:115(__getattr__)(12) 0.028
:0(range) ->
:0(setprofile) ->
<string>:1(<module>) -> test_profile.py:30(testfunc)(1) 1.000
profile:0(profiler) -> profile:0(testfunc())(1) 1.000
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
profile:0(testfunc()) -> :0(exec)(1) 1.000
:0(setprofile)(1) 0.000
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
test_profile.py:103(subhelper) -> :0(range)(8) 0.000
test_profile.py:115(__getattr__)(16) 0.028
test_profile.py:115(__getattr__) ->
test_profile.py:30(testfunc) -> test_profile.py:40(factorial)(1) 0.170
test_profile.py:60(helper)(2) 0.600
test_profile.py:40(factorial) -> test_profile.py:40(factorial)(20) 0.170
test_profile.py:53(mul)(20) 0.020
test_profile.py:53(mul) ->
test_profile.py:60(helper) -> test_profile.py:78(helper1)(4) 0.120
test_profile.py:89(helper2_indirect)(2) 0.140
test_profile.py:93(helper2)(6) 0.400
test_profile.py:78(helper1) -> :0(append)(4) 0.000
:0(exc_info)(4) 0.000
:0(hasattr)(4) 0.012
test_profile.py:89(helper2_indirect) -> test_profile.py:40(factorial)(2) 0.170
test_profile.py:93(helper2)(2) 0.400
test_profile.py:93(helper2) -> :0(hasattr)(8) 0.012
test_profile.py:103(subhelper)(8) 0.080
Ordered by: standard name
Function was called by...
:0(append) <- test_profile.py:78(helper1)(4) 0.120
:0(exc_info) <- test_profile.py:78(helper1)(4) 0.120
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
:0(exec) <- profile:0(testfunc())(1) 1.000
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
:0(hasattr) <- test_profile.py:78(helper1)(4) 0.120
test_profile.py:93(helper2)(8) 0.400
:0(range) <- test_profile.py:103(subhelper)(8) 0.080
:0(setprofile) <- profile:0(testfunc())(1) 1.000
Patch #1550800: make exec a function.
2006-09-06 03:51:57 -03:00
<string>:1(<module>) <- :0(exec)(1) 1.000
Added the cProfile module. Based on lsprof (patch #1212837) by Brett Rosen and Ted Czotter. With further editing by Michael Hudson and myself. History in svn repo: http://codespeak.net/svn/user/arigo/hack/misc/lsprof * Module/_lsprof.c is the internal C module, Lib/cProfile.py a wrapper. * pstats.py updated to display cProfile's caller/callee timings if available. * setup.py and NEWS updated. * documentation updates in the profiler section: - explain the differences between the three profilers that we have now - profile and cProfile can use a unified documentation, like (c)Pickle - mention that hotshot is "for specialized usage" now - removed references to the "old profiler" that no longer exists * test updates: - extended test_profile to cover delicate cases like recursion - added tests for the caller/callee displays - added test_cProfile, performing the same tests for cProfile * TO-DO: - cProfile gives a nicer name to built-in, particularly built-in methods, which could be backported to profile. - not tested on Windows recently!
2006-02-08 08:53:56 -04:00
profile:0(profiler) <-
profile:0(testfunc()) <- profile:0(profiler)(1) 0.000
test_profile.py:103(subhelper) <- test_profile.py:93(helper2)(8) 0.400
test_profile.py:115(__getattr__) <- :0(hasattr)(12) 0.012
test_profile.py:103(subhelper)(16) 0.080
test_profile.py:30(testfunc) <- <string>:1(<module>)(1) 1.000
test_profile.py:40(factorial) <- test_profile.py:30(testfunc)(1) 1.000
test_profile.py:40(factorial)(20) 0.170
test_profile.py:89(helper2_indirect)(2) 0.140
test_profile.py:53(mul) <- test_profile.py:40(factorial)(20) 0.170
test_profile.py:60(helper) <- test_profile.py:30(testfunc)(2) 1.000
test_profile.py:78(helper1) <- test_profile.py:60(helper)(4) 0.600
test_profile.py:89(helper2_indirect) <- test_profile.py:60(helper)(2) 0.600
test_profile.py:93(helper2) <- test_profile.py:60(helper)(6) 0.600
test_profile.py:89(helper2_indirect)(2) 0.140
Hopefully fix the profiler right. Add a test suite that checks that it deals correctly with some anomalous cases; according to this test suite I've fixed it right. The anomalous cases had to do with 'exception' events: these aren't generated when they would be most helpful, and the profiler has to work hard to recover the right information. The problems occur when C code (such as hasattr(), which is used as the example here) calls back into Python code and clears an exception raised by that Python code. Consider this example: def foo(): hasattr(obj, "bar") Where obj is an instance from a class like this: class C: def __getattr__(self, name): raise AttributeError The profiler sees the following sequence of events: call (foo) call (__getattr__) exception (in __getattr__) return (from foo) Previously, the profiler would assume the return event returned from __getattr__. An if statement checking for this condition and raising an exception was commented out... This version does the right thing.
2001-10-03 21:58:24 -03:00