Your message dated Sat, 22 Jul 2017 13:17:18 +0100
with message-id <[email protected]>
and subject line Closing bugs for 9.1 p-u fixes
has caused the Debian Bug report #868361,
regarding stretch-pu: package socat/1.7.3.1-2+deb9u1
to be marked as done.
This means that you claim that the problem has been dealt with.
If this is not the case it is now your responsibility to reopen the
Bug report if necessary, and/or fix the problem forthwith.
(NB: If you are a system administrator and have no idea what this
message is talking about, this may indicate a serious mail system
misconfiguration somewhere. Please contact [email protected]
immediately.)
--
868361: https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=868361
Debian Bug Tracking System
Contact [email protected] with problems
--- Begin Message ---
Package: release.debian.org
Severity: normal
Tags: stretch
User: [email protected]
Usertags: pu
Hi SRMs,
There's a chance that on some signals (including, but not limited to
SIGSEGV) socat goes to an infinite loop and consume all CPU cycles.
Upstream patched it[1] for 1.7.3.2 release, which is in Sid. Basically
set 'diag_immediate_exit' in the signal handling function to let it
exit. Full debdiff is attached.
Thanks for considering,
Laszlo/GCS
[1]
http://repo.or.cz/socat.git/commitdiff/6b596b8852d8fad2675894e3ceb18a04801eaf23?hp=d34493c18df0a4d0c4fdb5bda74a155ce13e4ccf
diff -Nru socat-1.7.3.1/debian/changelog socat-1.7.3.1/debian/changelog
--- socat-1.7.3.1/debian/changelog 2016-11-24 22:47:30.000000000 +0000
+++ socat-1.7.3.1/debian/changelog 2017-07-14 13:52:03.000000000 +0000
@@ -1,3 +1,10 @@
+socat (1.7.3.1-2+deb9u1) stretch; urgency=medium
+
+ * Backport upstream fix for SIGSEGV and other signals could lead to a
+ 100% CPU loop.
+
+ -- Laszlo Boszormenyi (GCS) <[email protected]> Fri, 14 Jul 2017 13:52:03 +0000
+
socat (1.7.3.1-2) unstable; urgency=low
* Backport upstream fix to build with OpenSSL 1.1.0 (closes: #828550).
diff -Nru socat-1.7.3.1/debian/patches/08-signals_could_lead_CPU_loop.patch socat-1.7.3.1/debian/patches/08-signals_could_lead_CPU_loop.patch
--- socat-1.7.3.1/debian/patches/08-signals_could_lead_CPU_loop.patch 1970-01-01 00:00:00.000000000 +0000
+++ socat-1.7.3.1/debian/patches/08-signals_could_lead_CPU_loop.patch 2017-07-14 13:52:03.000000000 +0000
@@ -0,0 +1,43 @@
+From 6b596b8852d8fad2675894e3ceb18a04801eaf23 Mon Sep 17 00:00:00 2001
+From: Gerhard Rieger <[email protected]>
+Date: Wed, 11 May 2016 20:34:33 +0200
+Subject: [PATCH 1/1] SIGSEGV and other signals could lead to a 100% CPU loop
+
+---
+ CHANGES | 3 +++
+ socat.c | 3 ++-
+ 2 files changed, 5 insertions(+), 1 deletion(-)
+
+diff --git a/CHANGES b/CHANGES
+index ee15bd0..1e1bc5d 100644
+--- a/CHANGES
++++ b/CHANGES
+@@ -1,3 +1,6 @@
++corrections:
++ SIGSEGV and other signals could lead to a 100% CPU loop
++
+ porting:
+ Changes to make socat compile with OpenSSL 1.1.
+ Thanks to Sebastian Andrzej Siewior e.a. from the Debian team for
+diff --git a/socat.c b/socat.c
+index 09039ff..ace006d 100644
+--- a/socat.c
++++ b/socat.c
+@@ -1422,12 +1422,13 @@ void socat_signal(int signum) {
+ diag_in_handler = 1;
+ Notice1("socat_signal(): handling signal %d", signum);
+ switch (signum) {
+- case SIGQUIT:
+ case SIGILL:
+ case SIGABRT:
+ case SIGBUS:
+ case SIGFPE:
+ case SIGSEGV:
++ diag_immediate_exit = 1;
++ case SIGQUIT:
+ case SIGPIPE:
+ diag_set_int('x', 128+signum); /* in case Error exits for us */
+ Error1("exiting on signal %d", signum);
+--
+2.7.4.GIT
+
diff -Nru socat-1.7.3.1/debian/patches/series socat-1.7.3.1/debian/patches/series
--- socat-1.7.3.1/debian/patches/series 2016-11-24 22:47:30.000000000 +0000
+++ socat-1.7.3.1/debian/patches/series 2017-07-14 13:52:03.000000000 +0000
@@ -6,3 +6,4 @@
05-xio-ip.patch
06-socat.1.patch
07-openssl-1.1.patch
+08-signals_could_lead_CPU_loop.patch
--- End Message ---
--- Begin Message ---
Version: 9.1
Hi,
These bugs all relate to updates which were included in today's stretch
point release.
Regards,
Adam
--- End Message ---
