Openembedded Core Discussions
 help / color / mirror / Atom feed
From: Ross Burton <ross.burton@intel.com>
To: openembedded-core@lists.openembedded.org
Subject: [PATCH 4/4] python3: fix multiprocessing
Date: Mon, 10 Sep 2018 14:31:23 +0100	[thread overview]
Message-ID: <20180910133123.22372-4-ross.burton@intel.com> (raw)
In-Reply-To: <20180910133123.22372-1-ross.burton@intel.com>

glibc 2.28 slightly changed the behaviour of sigaddset() which broke
multiprocessing.  Backport a patch from Python 3.6 to solve this.

Signed-off-by: Ross Burton <ross.burton@intel.com>
---
 meta/recipes-devtools/python/python3/signal.patch | 56 +++++++++++++++++++++++
 meta/recipes-devtools/python/python3_3.5.5.bb     |  1 +
 2 files changed, 57 insertions(+)
 create mode 100644 meta/recipes-devtools/python/python3/signal.patch

diff --git a/meta/recipes-devtools/python/python3/signal.patch b/meta/recipes-devtools/python/python3/signal.patch
new file mode 100644
index 00000000000..534a097771c
--- /dev/null
+++ b/meta/recipes-devtools/python/python3/signal.patch
@@ -0,0 +1,56 @@
+Upstream-Status: Backport
+Signed-off-by: Ross Burton <ross.burton@intel.com>
+
+From 4315389df3c4e8c1f94a18ab11a4b234762132b1 Mon Sep 17 00:00:00 2001
+From: Antoine Pitrou <pitrou@free.fr>
+Date: Mon, 23 Apr 2018 22:22:49 +0200
+Subject: [PATCH] [3.6] bpo-33329: Fix multiprocessing regression on newer
+ glibcs (GH-6575) (GH-6582)
+
+Starting with glibc 2.27.9000-xxx, sigaddset() can return EINVAL for some
+reserved signal numbers between 1 and NSIG.  The `range(1, NSIG)` idiom
+is commonly used to select all signals for blocking with `pthread_sigmask`.
+So we ignore the sigaddset() return value until we expose sigfillset()
+to provide a better idiom.
+(cherry picked from commit 25038ecfb665bef641abf8cb61afff7505b0e008)
+---
+ .../next/Library/2018-04-23-13-21-39.bpo-33329.lQ-Eod.rst  |  1 +
+ Modules/signalmodule.c                                     | 14 ++++++++------
+ 2 files changed, 9 insertions(+), 6 deletions(-)
+ create mode 100644 Misc/NEWS.d/next/Library/2018-04-23-13-21-39.bpo-33329.lQ-Eod.rst
+
+diff --git a/Modules/signalmodule.c b/Modules/signalmodule.c
+index e0d06b434d..138e74e8a9 100644
+--- a/Modules/signalmodule.c
++++ b/Modules/signalmodule.c
+@@ -744,7 +744,6 @@ iterable_to_sigset(PyObject *iterable, sigset_t *mask)
+     int result = -1;
+     PyObject *iterator, *item;
+     long signum;
+-    int err;
+ 
+     sigemptyset(mask);
+ 
+@@ -766,11 +765,14 @@ iterable_to_sigset(PyObject *iterable, sigset_t *mask)
+         Py_DECREF(item);
+         if (signum == -1 && PyErr_Occurred())
+             goto error;
+-        if (0 < signum && signum < NSIG)
+-            err = sigaddset(mask, (int)signum);
+-        else
+-            err = 1;
+-        if (err) {
++        if (0 < signum && signum < NSIG) {
++            /* bpo-33329: ignore sigaddset() return value as it can fail
++             * for some reserved signals, but we want the `range(1, NSIG)`
++             * idiom to allow selecting all valid signals.
++             */
++            (void) sigaddset(mask, (int)signum);
++        }
++        else {
+             PyErr_Format(PyExc_ValueError,
+                          "signal number %ld out of range", signum);
+             goto error;
+-- 
+2.11.0
+
diff --git a/meta/recipes-devtools/python/python3_3.5.5.bb b/meta/recipes-devtools/python/python3_3.5.5.bb
index 02decf470c8..15ec7d7fda6 100644
--- a/meta/recipes-devtools/python/python3_3.5.5.bb
+++ b/meta/recipes-devtools/python/python3_3.5.5.bb
@@ -41,6 +41,7 @@ SRC_URI += "\
             file://0002-Makefile-add-target-to-split-profile-generation.patch \
             file://float-endian.patch \
             file://ftplib.patch \
+            file://signal.patch \
            "
 SRC_URI[md5sum] = "f3763edf9824d5d3a15f5f646083b6e0"
 SRC_URI[sha256sum] = "063d2c3b0402d6191b90731e0f735c64830e7522348aeb7ed382a83165d45009"
-- 
2.11.0



  parent reply	other threads:[~2018-09-10 13:31 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-09-10 13:31 [PATCH 1/4] python3: fix ftplib with TLS 1.3 Ross Burton
2018-09-10 13:31 ` [PATCH 2/4] python3: remove redundant patch Ross Burton
2018-09-10 16:12   ` Khem Raj
2018-09-10 16:16     ` Burton, Ross
2018-09-10 16:23       ` Burton, Ross
2018-09-10 17:35         ` Khem Raj
2018-09-10 13:31 ` [PATCH 3/4] python3: remove patch that inadvertently disables xattrs Ross Burton
2018-09-10 13:31 ` Ross Burton [this message]
2018-09-10 14:02 ` ✗ patchtest: failure for "python3: fix ftplib with TLS 1..." and 3 more Patchwork

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180910133123.22372-4-ross.burton@intel.com \
    --to=ross.burton@intel.com \
    --cc=openembedded-core@lists.openembedded.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox