bpo-34373: Fix time.mktime() on AIX

[vstinner]

• Fix time.mktime() on AIX: fix code checking for mktime() failure
  for years before 1970.
• mktime(): rename variable 'buf' to 'tm'.
• _PyTime_localtime(): avoid abs() which is limited to int type,
  use "localtime" rather than "ctime" in the error message.
• _PyTime_localtime(): EINVAL constant is now always available.
• _PyTime_localtime(): initialize errno to 0 just in case if
  localtime_r() doesn't set errno on error.

https://bugs.python.org/issue34373

[vstinner]

cc @aixtools@ncoghlan

[native-api]

The purpose of this chunk is unclear.
Probably need to comment how exactly AIX is faulty like in the others.

[aixtools]

The issue AIX mktime() cannot handle on it's own - is that mktime() does not accept a tm structure when the value for tm_year is < 70.

The new version has an _AIX specific block before the call to mktime() (to add years to tm_year until it is >= 70 and remember this 'delta' in number of days per 4 years including leap-day.

Once adjusted (only when tm_year is < 70) the code is "normal". If mktime() fails, a failure is returned.

If mktime() was successful - and _AIX and delta_days != 0 then the structure adjusted by mktime() needs to be adjusted further (back in time):

• the original year is restored
• the tm_wday is adjusted according to delta_days
• the number of seconds (tt) is adjusted according to delta_days.

Hope this helps.

[vstinner]

@native-api: I added more comments. Is it better now?

@aixtools: Would you mind to review again my change please? (review the comments)

[vstinner]

This change also fix this compiler warning:

./Modules/timemodule.c:999:13: warning: 'buf.tm_year' is used uninitialized in this function [-Wuninitialized] 

https://buildbot.python.org/all/#/builders/10/builds/2358

[native-api]

Wow. So my word can actually influence something here.
Seeing the "awaiting core review" label, I thought it would be treated as a passing remark at best.
I'll put more thought into it from now on now that I see that others actually risk relying on it!

[vstinner]

Don't worry. If anything goes wrong, it will be all my fault and I can handle it :-)

[EGuesnet]

Hi,
I have an error during regression tests with Python3.8.1 on AIX 6.1, and it seems related to this PR.
It occurs only on 64 bit. Test passes on 32 bit.

====================================================================== ERROR: test_add_file_after_2107 (test.test_zipfile.StoredTestsWithSourceFile) ---------------------------------------------------------------------- Traceback (most recent call last): File "/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/test/test_zipfile.py", line 606, in test_add_file_after_2107 self.assertRaises(struct.error, zipfp.write, TESTFN) File "/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/unittest/case.py", line 816, in assertRaises return context.handle('assertRaises', args, kwargs) File "/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/unittest/case.py", line 202, in handle callable_obj(*args, **kwargs) File "/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/zipfile.py", line 1739, in write zinfo = ZipInfo.from_file(filename, arcname, File "/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/zipfile.py", line 523, in from_file mtime = time.localtime(st.st_mtime) OverflowError: localtime argument out of range

Is the code 32 bit only, or maybe the test was not updated?

---

I can reproduce the behavior as follow:

$ python3.8_32 Python 3.8.1 (default, Jan 27 2020, 11:34:59) [GCC 8.3.0] on aix Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: timestamp out of range for platform time_t $ python3.8_64 Python 3.8.1 (default, Jan 27 2020, 11:30:15) [GCC 8.3.0] on aix Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: localtime argument out of range $ python3.7_32 Python 3.7.4 (default, Jan 15 2020, 15:50:53) [GCC 8.3.0] on aix6 Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: timestamp out of range for platform time_t $ python3.7_64 Python 3.7.4 (default, Jan 15 2020, 15:46:22) [GCC 8.3.0] on aix6 Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) time.struct_time(tm_year=2107, tm_mon=1, tm_mday=27, tm_hour=10, tm_min=7, tm_sec=32, tm_wday=3, tm_yday=27, tm_isdst=0)

Before the PR, localtime after 2038 worked on 64 bit.

[aixtools]

Been awhile since I worked on this. From memory, the (initial) patch was only meant to affect the 32-bit build - as 64-bit did pass, iirc. Although, I do not recall ever encountering a failure with test_zipfile.

…

On 27/01/2020 15:08, EGuesnet wrote: Hi, I have an error during regression tests with Python3.8.1 on AIX 6.1, and it seems related to this PR. It occurs only on 64 bit. Test passes on 32 bit. ====================================================================== ERROR: test_add_file_after_2107 (test.test_zipfile.StoredTestsWithSourceFile) ---------------------------------------------------------------------- Traceback (most recent call last): File"/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/test/test_zipfile.py", line 606,in test_add_file_after_2107 self.assertRaises(struct.error, zipfp.write, TESTFN) File"/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/unittest/case.py", line 816,in assertRaises return context.handle('assertRaises', args, kwargs) File"/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/unittest/case.py", line 202,in handle callable_obj(*args,**kwargs) File"/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/zipfile.py", line 1739,in write zinfo = ZipInfo.from_file(filename, arcname, File"/opt/freeware/src/packages/BUILD/Python-3.8.1/64bit/Lib/zipfile.py", line 523,in from_file mtime = time.localtime(st.st_mtime) OverflowError: localtime argument out of range Is the code 32 bit only, or maybe the test was not updated? ------------------------------------------------------------------------ I can reproduce the behavior as follow: $ python3.8_32 Python 3.8.1 (default, Jan 27 2020, 11:34:59) [GCC 8.3.0] on aix Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: timestamp out of range for platform time_t $ python3.8_64 Python 3.8.1 (default, Jan 27 2020, 11:30:15) [GCC 8.3.0] on aix Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: localtime argument out of range $ python3.7_32 Python 3.7.4 (default, Jan 15 2020, 15:50:53) [GCC 8.3.0] on aix6 Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) Traceback (most recent call last): File "<stdin>", line 1, in <module> OverflowError: timestamp out of range for platform time_t $ python3.7_64 Python 3.7.4 (default, Jan 15 2020, 15:46:22) [GCC 8.3.0] on aix6 Type "help", "copyright", "credits" or "license" for more information. >>> import time >>> time.localtime(4325562452) time.struct_time(tm_year=2107, tm_mon=1, tm_mday=27, tm_hour=10, tm_min=7, tm_sec=32, tm_wday=3, tm_yday=27, tm_isdst=0) Before the PR, localtime after 2038 worked on 64 bit. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#12726?email_source=notifications&email_token=ACSZR5MN2RWVMC3IL2ENNNDQ73THJA5CNFSM4HEIKDVKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJ7TQGI#issuecomment-578762777>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACSZR5N3LUXRAUMYDXLKRGLQ73THJANCNFSM4HEIKDVA>.

[vstinner]

I have an error during regression tests with Python3.8.1 on AIX 6.1

This PR is closed. Please open a new issue at https://bugs.python.org/