delta/python-packages/psutil.git/psutil/tests, branch malloc-info github.com: giampaolo/psutil.git merge from master 2021-05-27T17:49:19+00:00 Giampaolo Rodola g.rodola@gmail.com 2021-05-27T17:49:19+00:00 0682d554f85d9182e5e7f03a504fb768fa88d6d5 Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
Add tolerance to test_linux.TestSystemVirtualMemory.test_total (#1935) 2021-04-15T15:16:26+00:00 Miro Hrončok miro@hroncok.cz 2021-04-15T15:16:26+00:00 ea5b2df02e39816c1f6626dd9d93ae7ede422687 We see this test as very flaky without tolerance in Fedora and CentOS 
We see this test as very flaky without tolerance in Fedora and CentOS
 Speed up cpu_frequncy() on Linux systems (#1851) (#1852) 2021-01-07T11:04:51+00:00 marxin marxin.liska@gmail.com 2021-01-07T11:04:51+00:00 6e494bd4024c3d25769688b53e95942e529392d6 The change is about using /proc/cpuinfo when available. It provides cached values for frequencies and one can fill up minimum and maximum frequency from /sys/devices/system/cpu/cpufreq/policy/* sub-system (which is fast). Fixes #1851. 
The change is about using /proc/cpuinfo when available. It provides
cached values for frequencies and one can fill up minimum and maximum
frequency from /sys/devices/system/cpu/cpufreq/policy/* sub-system
(which is fast).

Fixes #1851.
 Rename cpu_count_physical() to cpu_count_cores() 2020-12-21T00:20:18+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-21T00:20:18+00:00 a4c0a0eb0d2a872ab7a45e47fcf37ef1fde5b012 This has always been cause of confusion, e.g. see: https://github.com/giampaolo/psutil/pull/1727#issuecomment-698934643 Removed the reference to "physical" from dostrings, functions and test. I still left it in the doc though, as it's more explanatory. Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
This has always been cause of confusion, e.g. see:
https://github.com/giampaolo/psutil/pull/1727#issuecomment-698934643

Removed the reference to "physical" from dostrings, functions and test.
I still left it in the doc though, as it's more explanatory.

Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
update INSTALL.rst + fix CI failures 2020-12-19T11:58:27+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-19T11:58:27+00:00 a12dda352d8309048400b337521a780b8884e8cd Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
fix around macos failure on CI 2020-12-19T00:45:00+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-19T00:45:00+00:00 b760f4848db4db49de57918660f6a5059666b720 Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
[macOS, UNIX] prefer _SC_PAGESIZE over (partially) deprecated getpagesize() (#1891) 2020-12-17T17:59:13+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-17T17:59:13+00:00 6bdde37049ba9fae31ccf83620468f77a6b9f396 Add a reusable `psutil_getpagesize()` utility common to all UNIXes. Related to #1885 (`getpagesize()` is deprecated on recent macOS, POSIX.1-2001 and possibly other UNIXes). The problem emerged on macOS but `getpagesize()` is also used in FreeBSD, NetBSD, OpenBSD and AIX, so it makes sense to do this in one place only, similarly to Windows which also provide a `psutil_getpagesize()` utility. Follow cPython's `mmapmodule.c` and `resourcemodule.c` lead and rely on `sysconf(_SC_PAGESIZE)` instead, but leave `getpagesize()` in place as last resort/attempt for systems where it's not deprecated and/or they still legitimately rely on it. Also provide a python wrapper so we can test the return value of this C function against Python's stdlib modules. Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Add a reusable `psutil_getpagesize()` utility common to all UNIXes.

Related to #1885 (`getpagesize()` is deprecated on recent macOS, POSIX.1-2001 and possibly other UNIXes).

The problem emerged on macOS but `getpagesize()` is also used in FreeBSD, NetBSD, OpenBSD and AIX, so it makes sense to do this in one place only, similarly to Windows which also provide a `psutil_getpagesize()` utility.

Follow cPython's `mmapmodule.c` and `resourcemodule.c` lead and rely on `sysconf(_SC_PAGESIZE)` instead, but leave `getpagesize()` in place as last resort/attempt for systems where it's not deprecated and/or they still legitimately rely on it.

Also provide a python wrapper so we can test the return value of this C function against Python's stdlib modules.

Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
 workaround for environ() on CI + MACOS 2020-12-14T21:10:03+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-14T21:10:03+00:00 cc624f17789a1df4d03f632a943b8442081781d7 Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
[Windows] #1877: turn OpenProcess -> ERROR_SUCCESS into AD or NSP (#1887) 2020-12-14T19:27:47+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-12-14T19:27:47+00:00 b6699b41e2c4873deac532e925561686bdb827d5 Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com> 
Signed-off-by: Giampaolo Rodola <g.rodola@gmail.com>
Remove Travis and Cirrus, use GH also for FreeBSD (#1880) 2020-11-15T20:54:42+00:00 Giampaolo Rodola g.rodola@gmail.com 2020-11-15T20:54:42+00:00 e58b0fdaeaedd73a0ca19ad23a63874708d86b91