[pypy-commit] [Git][pypy/pypy][branch/py3.7] 36 commits: fix 0a6fc92dba19, add missing import

Wed Jul 22 12:37:54 EDT 2020

Matti Picus pushed to branch branch/py3.7 at PyPy / pypy

Commits:
0a10e74f by Matti Picus at 2020-07-14T10:51:17+03:00
fix 0a6fc92dba19, add missing import

- - - - -
13582da4 by Matti Picus at 2020-07-14T14:22:57+03:00
typo (thanks nulano)

- - - - -
b802d0ab by Matti Picus at 2020-07-16T15:03:19+03:00
backport test (issue 3269)

- - - - -
10620d7c by Armin Rigo at 2020-07-17T13:10:06+02:00
Older versions of CPython 2.7.x (like 2.7.6) don't support this new syntax
when called in a function involving nested scopes.  These fixes are enough
for me to translate pypy on some strange machine, but I agree that we're
not really aiming to support CPython 2.7.6 any more---so, no test for that

- - - - -
3241c482 by Armin Rigo at 2020-07-19T08:52:49+02:00
Issue #3246

Got it once on aarch64.  Cannot reproduce, of course, so I'm still in the dark.
Trying to add a couple of extra asserts

- - - - -
518c25a2 by Matti Picus at 2020-07-20T10:01:09+03:00
fix building wstr and usc4 buffers when sizeof(Py_UNICODE) == 2 (windows)

--HG--
branch : win-unicode

- - - - -
d1019841 by Armin Rigo at 2020-07-20T13:26:54+02:00
A branch for issue #3250: W_LongObject causes missed opportunity to use
IntegerListStrategy

--HG--
branch : list-with-longs

- - - - -
51f78ec0 by Armin Rigo at 2020-07-20T15:30:19+02:00
Accept W_LongObject inside integer lists as long as they are within range

--HG--
branch : list-with-longs

- - - - -
e8f1908f by Armin Rigo at 2020-07-20T15:40:33+02:00
same with dicts

--HG--
branch : list-with-longs

- - - - -
fca4ffd2 by Armin Rigo at 2020-07-20T15:52:59+02:00
same with sets

--HG--
branch : list-with-longs

- - - - -
19b4ae2d by Armin Rigo at 2020-07-20T15:56:01+02:00
oops

--HG--
branch : list-with-longs

- - - - -
ddc501a5 by Armin Rigo at 2020-07-20T15:57:27+02:00
same with specialized tuples

--HG--
branch : list-with-longs

- - - - -
3fc843c3 by Armin Rigo at 2020-07-20T16:04:06+02:00
same with versioned types

--HG--
branch : list-with-longs

- - - - -
5bff8fc9 by Armin Rigo at 2020-07-20T16:11:35+02:00
another case for rbigint.fits_int()

--HG--
branch : list-with-longs

- - - - -
e31fcfd2 by Armin Rigo at 2020-07-20T16:26:27+02:00
fix this test for the BadUtf8 exception we get now

--HG--
branch : py3.6

- - - - -
7290a0c0 by Matti Picus at 2020-07-20T18:20:56+03:00
close branch to be merged

--HG--
branch : win-unicode

- - - - -
da92ba59 by Matti Picus at 2020-07-20T18:22:03+03:00
merge win-unicode which fixes PyUnicode for sizeof(wchar_t) ==2

--HG--
branch : py3.6

- - - - -
a5fb54a0 by Armin Rigo at 2020-07-20T19:31:04+02:00
Don't use space.int_w() on plain ints, because this is a generic method that
contain many cases including recursion.  A recursion test fails at the moment,
probably because of that.

--HG--
branch : list-with-longs

- - - - -
20b96819 by Armin Rigo at 2020-07-20T20:44:21+02:00
document branch

--HG--
branch : list-with-longs

- - - - -
e23154e5 by Armin Rigo at 2020-07-20T22:26:52+02:00
merge branch list-with-longs

--HG--
branch : py3.6

- - - - -
191cacc2 by Armin Rigo at 2020-07-21T16:18:36+02:00
#3220 in-progress

Support the RPython-level pypysig_set_wakeup_fd() optionally using send()
instead of write()

- - - - -
15b82b07 by Armin Rigo at 2020-07-21T16:46:05+02:00
Revert the direction in which I was taking this, and instead just check
in the C code if the fd appears to be a socket on Windows

- - - - -
5f142e6c by Armin Rigo at 2020-07-21T16:54:06+02:00
add a win64 comment here

- - - - -
c04de469 by Armin Rigo at 2020-07-21T17:17:27+02:00
hg merge default

--HG--
branch : py3.6

- - - - -
059a3593 by Armin Rigo at 2020-07-21T17:19:08+02:00
For now, don't do any checking in the app-level signal.set_wakeup_fd() on
Windows

--HG--
branch : py3.6

- - - - -
8e1d16dd by Armin Rigo at 2020-07-21T17:33:30+02:00
missing include

- - - - -
5d0b3e77 by Armin Rigo at 2020-07-21T17:34:08+02:00
hg merge default

--HG--
branch : py3.6

- - - - -
a13a3bdb by Armin Rigo at 2020-07-21T17:47:39+02:00
fix

- - - - -
9be4ab8c by Armin Rigo at 2020-07-21T17:47:56+02:00
hg merge default

--HG--
branch : py3.6

- - - - -
c510b896 by Ronan Lamy at 2020-07-21T18:07:56+01:00
Fix skipping of backend-specific jit tests to not crash test collection on the other platforms

- - - - -
51a0030d by Ronan Lamy at 2020-07-21T21:22:42+01:00
Add different way of skipping for when pytest is called on a single file (sadness)

- - - - -
bc611d51 by Matti Picus at 2020-07-22T12:39:27+03:00
disable WinConsoleIO

--HG--
branch : py3.6

- - - - -
2119d0af by Armin Rigo at 2020-07-22T15:30:37+02:00
Issue #3188: need FORBID_TEMP_BOXES on aarch64, like on arm32

Otherwise there is no guarantee that calling make_sure_var_in_reg()
twice, with [const, box] respectively, will work: it will allocate a
temporary register and load it with the value of 'const', but then
the second call might spill this same register.

- - - - -
cb6af5dc by Armin Rigo at 2020-07-22T15:34:32+02:00
Goes with the previous check-in.  See a243e4e0b21c

- - - - -
22ecb4a9 by Armin Rigo at 2020-07-22T15:36:04+02:00
hg merge default

--HG--
branch : py3.6

- - - - -
3f5213d0 by Matti Picus at 2020-07-22T19:37:11+03:00
merge py3.6 into branch

--HG--
branch : py3.7

- - - - -

30 changed files:

- pypy/doc/whatsnew-pypy3-head.rst
- pypy/module/_io/interp_win32consoleio.py
- pypy/module/_io/test/test_win32consoleio.py
- pypy/module/cpyext/test/test_unicodeobject.py
- pypy/module/cpyext/unicodeobject.py
- pypy/module/signal/interp_signal.py
- pypy/objspace/std/dictmultiobject.py
- pypy/objspace/std/listobject.py
- pypy/objspace/std/longobject.py
- pypy/objspace/std/setobject.py
- pypy/objspace/std/specialisedtupleobject.py
- pypy/objspace/std/test/test_dictmultiobject.py
- pypy/objspace/std/test/test_liststrategies.py
- pypy/objspace/std/test/test_setstrategies.py
- pypy/objspace/std/test/test_specialisedtupleobject.py
- pypy/objspace/std/test/test_stdobjspace.py
- pypy/objspace/std/test/test_versionedtype.py
- pypy/objspace/std/typeobject.py
- rpython/annotator/specialize.py
- rpython/jit/backend/aarch64/regalloc.py
- rpython/jit/backend/aarch64/test/conftest.py
- rpython/jit/backend/arm/test/conftest.py
- rpython/jit/backend/llsupport/asmmemmgr.py
- rpython/jit/backend/llsupport/descr.py
- rpython/jit/backend/ppc/test/conftest.py
- rpython/jit/backend/zarch/test/conftest.py
- rpython/jit/metainterp/warmspot.py
- rpython/rlib/entrypoint.py
- rpython/rlib/jit.py
- rpython/rlib/rbigint.py

View it on GitLab: https://foss.heptapod.net/pypy/pypy/-/compare/17c1c86ca1184662ead002f5f75a5a69f29d0051...3f5213d07933ed5dfa8a246933410fe1b1b0beeb

-- 
View it on GitLab: https://foss.heptapod.net/pypy/pypy/-/compare/17c1c86ca1184662ead002f5f75a5a69f29d0051...3f5213d07933ed5dfa8a246933410fe1b1b0beeb
You're receiving this email because of your account on foss.heptapod.net.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.python.org/pipermail/pypy-commit/attachments/20200722/d01834c2/attachment-0001.html>