Steve Dower added the comment:
Well, the good news is I can repro it on a PGO build in WSL. So now I just need
to go learn all about gcc's PGO build, debug info, and how gdb renders it!
--
___
Python tracker
<https://bugs.python.org/is
Steve Dower added the comment:
I confirmed earlier that removing the new code from builtins_id fixes this
issue, which I suspect means that PGO is deciding to make different
optimizations and produce different output.
Other functions in the same stack also do not show their arguments, so I
Steve Dower added the comment:
New changeset 51394b8c3d42e6e6d368251ff6f0612495724fc0 by Steve Dower (Paul
Monson) in branch 'master':
bpo-36511: Ensure error code propagates out of batch files (GH-13529)
https://github.com/python/cpython/commit/51394b8c3d42e6e6d368251ff6f061
New submission from Steve Dower :
I'd like to be able to build libpython and have it already have the install
path stripped out so I can use rpath instead. This libpython is going to be
distributed with other applications.
Right now, I'm running `install_name_tool -
Steve Dower added the comment:
Okay, fixing the regex isn't an option, as most of the tests (just not the one
copy-pasted above) rely on verifying the parameter value.
I'll figure out how to skip the test on PGO build instead.
--
Change by Steve Dower :
--
keywords: +patch
pull_requests: +13466
stage: -> patch review
___
Python tracker
<https://bugs.python.org/issue37023>
___
___
Py
Steve Dower added the comment:
Since hooks can't be removed, it probably looks like a leak when some are added
for the tests.
I'll update all the tests to run in new processes.
--
___
Python tracker
<https://bugs.python.o
Steve Dower added the comment:
New changeset 6de4574c6393b9cf8d7dfb0dc6ce53ee5b9ea841 by Steve Dower in branch
'master':
bpo-37023: Skip test_gdb under PGO (GH-13555)
https://github.com/python/cpython/commit/6de4574c6393b9cf8d7dfb0dc6ce53
Steve Dower added the comment:
Test is now skipped if PGO was used.
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Steve Dower added the comment:
New changeset 9ddc416e9f6635376312c3615193f19480ac772a by Steve Dower in branch
'master':
bpo-36842: Fix reference leak in tests by running out-of-proc (GH-13556)
https://github.com/python/cpython/commit/9ddc416e9f6635376312c3615193f1
Change by Steve Dower :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Steve Dower added the comment:
Great, thanks Zackery! (And thanks for the ping - I don't notice GitHub
notifications.)
Had a few comments about error handling, but it looks great.
--
___
Python tracker
<https://bugs.python.org/is
Steve Dower added the comment:
I don't think we can change the assumption that sys.prefix is in sys.path at
this point, though technically it's not required (certainly not in 3.7).
Maybe we could create a new marker file that means "never treat this directory
as a na
Steve Dower added the comment:
New changeset 8f96c9f8ed2a4795e34b333411451e24f28f74d2 by Steve Dower (Zackery
Spytz) in branch 'master':
bpo-37007: Implement socket.if_nametoindex(), if_indextoname() and
if_nameindex() on Windows (GH-13522)
https://github.com/python/cpyt
Steve Dower added the comment:
Honestly, no idea. Because it always has been :)
If you can find when it was added, we may be able to answer that.
--
___
Python tracker
<https://bugs.python.org/issue37
Change by Steve Dower :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Steve Dower added the comment:
New changeset c8d5bf6c3fa09b43f6a5ee779d493d251dbcc53c by Steve Dower (Zackery
Spytz) in branch 'master':
bpo-12639: msilib.Directory.start_component() fails if *keyfile* is not None
(GH-13688)
https://github.com/python/cpyt
Steve Dower added the comment:
Thanks, Zackery!
The backports are running CI now and should merge if everything passes.
--
stage: patch review -> backport needed
___
Python tracker
<https://bugs.python.org/issu
Steve Dower added the comment:
Looks like the tests need to be customized for 2.7 and 3.6:
Traceback (most recent call last):
File "D:\a\1\s\lib\test\test_msilib.py", line 64, in
test_directory_start_component_keyfile
self.addCleanup(db.Close)
AttributeError: '_msi.Datab
Change by Steve Dower :
--
resolution: -> fixed
stage: backport needed -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Steve Dower added the comment:
New changeset bfc1f605609218b9734d3cf3eab3531a2f4624e1 by Steve Dower (Miss
Islington (bot)) in branch '2.7':
[2.7] bpo-12639: msilib.Directory.start_component() fails if *keyfile* is not
None (GH-13688)
https://github.com/python/cpyt
Steve Dower added the comment:
I agree with Paul. Consistency with open() is the most important aspect here,
and when we investigated changing that to UTF-8 it was decided to be too
disruptive.
That said, I'm not opposed to adding encoding/errors to basicConfig if someone
wants to do
Steve Dower added the comment:
I have no idea. I don't ever use msilib, but there are a few people who do that
don't have any way to get bugs fixed, so I'll help merge.
If it's just the test leak, probably it's in the test. I don't recall the
change having any
Change by Steve Dower :
--
pull_requests: +13698
pull_request: https://github.com/python/cpython/pull/13812
___
Python tracker
<https://bugs.python.org/issue36
Change by Steve Dower :
--
pull_requests: +13701
pull_request: https://github.com/python/cpython/pull/13815
___
Python tracker
<https://bugs.python.org/issue36
Steve Dower added the comment:
New changeset f61599b050c621386a3fc6bc480359e2d3bb93de by Steve Dower in branch
'2.7':
bpo-36742: Corrects fix to handle decomposition in usernames (GH-13812)
https://github.com/python/cpython/commit/f61599b050c621386a3fc6bc480359
New submission from Steve Dower :
Mostly a note to self to fix the tag. Right now the x64 build gets an extra
suffix, which is incorrect and causes installer builds to fail.
--
assignee: steve.dower
components: Windows
messages: 344638
nosy: paul.moore, steve.dower, tim.golden
Steve Dower added the comment:
The reason I'd prefer sys.platform in most cases is because it's a compile-time
constant, based on the one that includes/excludes APIs near completely, and
most of our tests ought to be switching on these.
I personally don't see any need to sw
Steve Dower added the comment:
> How does select() know the FD_SETSIZE?
It doesn't have to. The actual number of descriptors is passed in, and the API
presumably assumes that FD_SETSIZE is bigger than that (or else the caller
would have crashed writing to invalid memory).
I don&#
Steve Dower added the comment:
There are certainly people relying on sys.prefix being in sys.path (for .pth
files and embedding), but probably not so many that we couldn't remove it in a
major release. It is a bit strange in a venv, where there's typically nothing
in the r
Steve Dower added the comment:
> So it's really FD_SETSIZE that matters, and it's a compile-time constant.
The only way this constant can matter is if the array has to end with an empty
slot, which doesn't appear to be the case. The pre-compiled function can't tell
ho
Change by Steve Dower :
--
keywords: +patch
pull_requests: +13742
stage: needs patch -> patch review
pull_request: https://github.com/python/cpython/pull/13866
___
Python tracker
<https://bugs.python.org/issu
Steve Dower added the comment:
New changeset e0c0c7e8c9f8153a54b92e43aa3d09e69a9fd0c0 by Steve Dower in branch
'master':
bpo-37156: Fix libssl DLL tag in MSI sources (GH-13866)
https://github.com/python/cpython/commit/e0c0c7e8c9f8153a54b92e43aa3d09
Change by Steve Dower :
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python.or
Steve Dower added the comment:
Changing our policy here (from "no policy" to "here's a recommendation")
probably deserves a python-dev discussion first.
--
___
Python tracker
<https:
New submission from Steve Dower :
New changeset e7e5039d6940e41839dcef0433262ff363408dad by Steve Dower (Paul
Monson) in branch 'master':
bpo-37181: Fix test_regrtest failures on Windows arm64 (GH-13872)
https://github.com/python/cpython/commit/e7e5039d6940e41839dcef0433262f
Steve Dower added the comment:
Guessing Victor may have touched something in this area.
Was it exported before? Or did we just have a macro for it? Maybe the macro was
moved to an internal header by mistake?
--
nosy: +vstinner
___
Python tracker
Steve Dower added the comment:
This is due to how the pip.exe launcher is generated, which is part of pip
itself (actually distlib, I think, but someone over there can redirect).
Please report this at https://github.com/pypa/pip
--
resolution: -> third party
stage: -> re
Steve Dower added the comment:
I'm with Petr here. ensurepip/venv should only go as far as we support, which
is the version of the wheels included in our bundle. We should continue having
the stdlib pretend that PyPI doesn't exist.
External network access would justify the longer
Steve Dower added the comment:
This is intended behavior.
When a variable has an assignment anywhere in a function, it becomes a local.
Once a local variable exists, it will shadow the non-local variable.
So by having the "n =", you are marking "n" as a local variab
Steve Dower added the comment:
New changeset 04856c2193eb72d72c46b57fa08095235d732a73 by Steve Dower (Paul
Monson) in branch 'master':
bpo-37238: Enable building for Windows using Visual Studio 2019 (GH-13988)
https://github.com/python/cpython/commit/04856c2193eb72d72c46b57fa08095
Steve Dower added the comment:
Not a whole lot of process, really. It's done now :)
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.pyth
Steve Dower added the comment:
Why is this different from the nuget package? Should we just add more things
into that package?
--
___
Python tracker
<https://bugs.python.org/issue37
Steve Dower added the comment:
"tp_print" has been marked as reserved for all of Python 3. To me, that means
deprecated and do not use. (But perhaps that ought to have been properly
defined in the docs?)
Cython should not be using this field directly. If all they're doing is
Steve Dower added the comment:
> Jeroen Demeyer proposed PR 14009 to fix the source compatibility:
>
> #define tp_print tp_vectorcall
This is a terrible idea, FWIW. Please don't do this.
--
nosy: +ncoghlan
___
Python t
Steve Dower added the comment:
New changeset daf62627518ad97ce66a48c49496aa0573cf0731 by Steve Dower (Paul
Monson) in branch 'master':
bpo-37201: fix test_distutils failures for Windows ARM64 (GH-13902)
https://github.com/python/cpython/commit/daf62627518ad97ce66a48c49496aa
Steve Dower added the comment:
In fact, any reason you can't use "--preset-nuget --include-..."? to add in
whatever else you need? They are designed to compose like this deliberately.
--
___
Python tracker
<https://bugs.pyt
Steve Dower added the comment:
New changeset ff6bb0aa95259413f359d42410526ff0b4dccfb7 by Steve Dower (Paul
Monson) in branch 'master':
bpo-37236: pragma optimize off for _Py_c_quot on Windows arm64 (GH-13983)
https://github.com/python/cpython/commit/ff6bb0aa95259413f359d42410526f
Steve Dower added the comment:
> Any solution that we apply in Cython will require users to regenerate their
> .c sources with a new Cython version in order to make it compile in Py3.8.
> The main decision point in this ticket is: should they need to or not?
Considering Cython
Steve Dower added the comment:
I think Vinay gets to make the call, but I also agree that some sort of error
handling by default makes sense for logging. Especially now that it's (about to
be) easier to override the default.
FWIW, I'd leave the default arguments as None and just
Steve Dower added the comment:
Windows does not have fork, and so the fork functionality is not built into it
(PyOS_AfterFork() is empty).
You will need to report this to mod_wsgi to be able to build without fork
support.
--
resolution: -> third party
stage: -> resolved
Steve Dower added the comment:
You should have a set of log files in %TEMP%. Can you zip those up and attach
them here?
Pip may be failing to install for some reason, and while this ought to be
reported, it is not supposed to cause the entire installation to fail
Steve Dower added the comment:
Isn't the point that device_encoding(FD) gets the encoding of the specified
file? In this case stdout?
It seems odd that chcp doesn't actually update the console code page here, as
that is its entire purpose. Perhaps TextIOWrapper is actually g
Steve Dower added the comment:
Hmm... interesting.
Thanks for the heads-up, I think I know what this may be (there's a conditional
planning stage in the bootstrapper that is probably being skipped on Modify).
I'll take a look when I get a chance.
--
assignee: ->
Steve Dower added the comment:
New changeset 00f6493084c385033fe5574314223217d9a26672 by Steve Dower (Paul
Monson) in branch 'master':
bpo-37288: Fix Windows build when --no-tkinter is specified (GH-14096)
https://github.com/python/cpython/commit/00f6493084c385033fe55743142232
Steve Dower added the comment:
Has anyone tried building with the newer SQLite and confirmed that the tests
and any scenarios they use still work?
We don't currently have a SQLite expert to do these kinds of things.
--
___
Python tracker
&
Steve Dower added the comment:
Thanks, Paul!
--
assignee: -> steve.dower
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.python
Steve Dower added the comment:
The canonical source of versions used on Windows is in PCbuild/python.props
I'll pull the 1.1.1c sources into cpython-source-deps and run a build. If all
goes smoothly, we can consider it, but I don't have a huge amount of time for
CPython this week
Change by Steve Dower :
--
pull_requests: +14003
pull_request: https://github.com/python/cpython/pull/14161
___
Python tracker
<https://bugs.python.org/issue34
Steve Dower added the comment:
I haven't fully tested this, but a suitable test using ctypes might look like:
py = ctypes.PyDLL("", handle=sys.dllhandle)
missing = {name for name in EXPECTED_NAMES if not hasattr(py, name)}
# assert &#x
Steve Dower added the comment:
New changeset d8e3a8af775d683c606f3618d04f2be4e97ac3c0 by Steve Dower in branch
'2.7':
bpo-34631: Updated OpenSSL to 1.0.2s in Windows installer (GH-14161)
https://github.com/python/cpython/commit/d8e3a8af775d683c606f3618d04f2b
Change by Steve Dower :
--
pull_requests: +14005
pull_request: https://github.com/python/cpython/pull/14163
___
Python tracker
<https://bugs.python.org/issue34
Steve Dower added the comment:
The tests seem to pass fine for 1.1.1c against master, so I'll merge that and
see if the backport is also okay.
--
___
Python tracker
<https://bugs.python.org/is
Steve Dower added the comment:
New changeset a268edd6a411480281222b1fdb0f78053434d17f by Steve Dower in branch
'master':
bpo-34631: Updated OpenSSL to 1.1.1c in Windows installer (GH-14163)
https://github.com/python/cpython/commit/a268edd6a411480281222b1fdb0f78
Change by Steve Dower :
--
pull_requests: +14007
pull_request: https://github.com/python/cpython/pull/14165
___
Python tracker
<https://bugs.python.org/issue34
Steve Dower added the comment:
New changeset c28c1358245b9fe42e9559c34eae01befce73a1f by Steve Dower (Miss
Islington (bot)) in branch '3.8':
bpo-34631: Updated OpenSSL to 1.1.1c in Windows installer (GH-14163)
https://github.com/python/cpython/commit/c28c1358245b9fe42e9559c34eae01
Steve Dower added the comment:
I just pushed sources with tag sqlite-3.28.0.0, so now anybody should be able
to submit a CPython PR for it.
--
___
Python tracker
<https://bugs.python.org/issue35
Steve Dower added the comment:
Ned - the 3.7 backport seems to be okay (PR 14165). Do we want it?
--
___
Python tracker
<https://bugs.python.org/issue34
Steve Dower added the comment:
Ah, that's what ctypes.pythonapi is :) I looked at PyDLL first and figured it
out from there.
Should we add a regression test to avoid this happening in the future?
--
___
Python tracker
<https://bugs.py
Steve Dower added the comment:
New changeset 14bac0088271d0a5c428e3468ef94fe7c73e93f7 by Steve Dower in branch
'3.7':
bpo-34631: Updated OpenSSL to 1.1.1c in Windows installer (GH-14163)
https://github.com/python/cpython/commit/14bac0088271d0a5c428e3468ef94f
Steve Dower added the comment:
I'm still seeing this, maybe 1 in 20 builds, so it's semi-random. A new
deadlock, maybe?
--
nosy: +steve.dower
___
Python tracker
<https://bugs.python.o
Steve Dower added the comment:
> If we reintroduce it, why not put it back at its previous place, to provide
> ABI compatibility?
I agree. If compatibility matters here, then we should add new elements at the
end.
> First of all, we longer care about ABI compatibility of PyTypeOb
Steve Dower added the comment:
> Don't confuse *ABI* and *API* compatibility. For *API* compatibility, it
> doesn't matter where we put tp_print.
Don't worry, I'm not. (Though I did blur the lines for the sake of a
tongue-in-cheek reply to your comment. Probably
Steve Dower added the comment:
> Personally, I don't think that your "read of it" in #msg345990 is relevant to
> the discussion of the PR (I don't mean this in a bad way).
It's relevant to the *existence* of the PR, because when you follow through the
rest o
Change by Steve Dower :
--
pull_requests: +14057
pull_request: https://github.com/python/cpython/pull/14219
___
Python tracker
<https://bugs.python.org/issue37
Steve Dower added the comment:
New changeset 43b97ae15c8cbb8c383314bb75c867d0f2f34376 by Steve Dower in branch
'3.7':
bpo-37156: Fix libssl DLL tag in MSI sources (GH-14219)
https://github.com/python/cpython/commit/43b97ae15c8cbb8c383314bb75c867
Steve Dower added the comment:
If we should put in the workaround for 3.8, how does that make it okay to
remove in 3.9? It's been deprecated since 3.0 already, so if we can't remove it
after eight releases, I don't see how it's consistent to remove it after a
ninth. Onc
Change by Steve Dower :
--
keywords: +patch
pull_requests: +14075
stage: -> patch review
pull_request: https://github.com/python/cpython/pull/14240
___
Python tracker
<https://bugs.python.org/issu
Change by Steve Dower :
--
assignee: -> steve.dower
type: -> compile error
versions: +Python 3.8
___
Python tracker
<https://bugs.python.org/i
Steve Dower added the comment:
> Can you give a least an argument of why that's better than putting back
> tp_print at the end?
Minimal change, that's why. If there are compatibility issues with the current
change that we need to fix, then we should make the least inc
Steve Dower added the comment:
New changeset 12f1c726d8328e5e096c35c36901f6d19bc942d3 by Steve Dower in branch
'master':
bpo-37333: Ensure IncludeTkinter has a value (GH-14240)
https://github.com/python/cpython/commit/12f1c726d8328e5e096c35c36901f6
Steve Dower added the comment:
Backport will automerge when it's done.
--
resolution: -> fixed
stage: patch review -> resolved
status: open -> closed
___
Python tracker
<https://bugs.pyth
Steve Dower added the comment:
New changeset f355069a3337711642c3403429afb9faef93f512 by Steve Dower (Paul
Monson) in branch 'master':
bpo-36511: Add buildbot scripts and fix tests for Windows ARM32 buildbot
(GH-13454)
https://github.com/python/cpyt
Steve Dower added the comment:
> It will break ABI compatibility in any case where vectorcall is used, while
> my proposal of putting tp_print at the end does not.
ABI compatibility with what? We haven't released vectorcall yet, so it has no
compatibili
Steve Dower added the comment:
> a class compiled with 3.7 could subclass a class compiled with 3.8 which uses
> vectorcall.
This isn't a supported scenario though. The only way to mix extension modules
compiled against different Python versions is to use the limited ABI, whic
Steve Dower added the comment:
> I was trying to tell you that already in msg345969 but you insisted in
> msg345988 that we should support that anyway.
I made a general note that "extremely unlikely" isn't unlikely enough to be an
argument.
You'll notice I didn&
Steve Dower added the comment:
> Note that the PR for the other similar issue #37221 has been accepted for
> merging.
I've already explained why that issue is different (it breaks a compatibility
guarantee; this one changes something that was explicitly deprecated).
Petr'
Steve Dower added the comment:
New changeset a1952122a3a20272e4db9e3c6205009a91c73ccf by Steve Dower (Paul
Monson) in branch 'master':
bpo-36511: Improve ARM32 buildbot scripts (GH-14251)
https://github.com/python/cpython/commit/a1952122a3a20272e4db9e3c620500
New submission from Steve Dower :
The step of generating a MinGW import library for Python releases is very
rarely used, forces us to take extra build dependencies, and is better handled
by the end-users who need it.
We will drop the libpython38.a file from the main distribution, and update
Steve Dower added the comment:
Realising that this may have an impact on
distutils.cygwincompiler.Mingw32CCompiler, but I can't tell from the source
code how/whether it would pick up the pre-packaged libpython38.a. Does anyone
know?
--
components: +Distutils
nosy: +ds
Steve Dower added the comment:
Turns out that using distutils/setuptools with --compiler=mingw32 is just
doomed to fail when it hits get_msvcr(), which has never been updated for
Python 3.5 onwards.
So I think it's probably safe to assume that this scenario isn't a real thing
Change by Steve Dower :
--
keywords: +patch
pull_requests: +14100
stage: needs patch -> patch review
pull_request: https://github.com/python/cpython/pull/14276
___
Python tracker
<https://bugs.python.org/issu
Steve Dower added the comment:
> While Cython may be able to respond within the beta period, scientific
> projects that depend on it may lag longer as rebuilds and releases will need
> to happen on PyPI, conda, and conda-forge. Release versioning is critical in
> the science
Change by Steve Dower :
--
pull_requests: +14102
pull_request: https://github.com/python/cpython/pull/14279
___
Python tracker
<https://bugs.python.org/issue36
Steve Dower added the comment:
New changeset 43615485f1222cd9a94421a96cb7110bb52e1b8b by Steve Dower in branch
'3.8':
bpo-36511: Improve ARM32 buildbot scripts (GH-14251)
https://github.com/python/cpython/commit/43615485f1222cd9a94421a96cb711
Steve Dower added the comment:
Yes, I noticed that they changed with OpenSSL 1.1.1a, but as far as I can tell
we didn't do anything to cause that - our build definition is still the same,
and uses "perl Configure VC-WIN64A-masm && nmake".
Is this causing issues? Or is
Steve Dower added the comment:
FWIW, I agree totally with Petr's analysis.
Jeroen's analysis ignores some points, has technical inaccuracies (nobody can
use a static type compiled for 3.7 with 3.8), and is based on precedents that
do not apply to this situation. (All of which I&#x
Steve Dower added the comment:
One thing to note is that if we sign this file, it'll have to bypass the text
substitution step completely to avoid modifying line endings or encoding. So
there could be code changes in venv too.
This would be a great contribution from a PowerShell expert
Steve Dower added the comment:
> I'm curious what you mean here...
msg345409
--
___
Python tracker
<https://bugs.python.org/issue37250>
___
___
Py
2401 - 2500 of 5794 matches
Mail list logo
