Your message dated Tue, 28 Feb 2017 06:59:00 +0000
with message-id <e5b58f7c-0a0a-79d7-5200-82a33a393...@thykier.net>
and subject line Re: Bug#856228: unblock: libnids/1.23-2.1
has caused the Debian Bug report #856228,
regarding unblock: libnids/1.23-2.1
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 ow...@bugs.debian.org
immediately.)


-- 
856228: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=856228
Debian Bug Tracking System
Contact ow...@bugs.debian.org with problems
--- Begin Message ---
Package: release.debian.org
Severity: normal
User: release.debian....@packages.debian.org
Usertags: unblock

Hi,

Please unblock package libnids

This fixes 2 RC bugs in the package:
- #851060 where the package doesn't work on armhf. The package contains
numerous violations of the strict-aliasing rule so I added
-fno-strict-aliasing as an easy workaround which does fix the bug on armhf.

- #855602 where the package assumes the old gnu89 inline semantics but
this was never caught because the undefined references only occur when
building another package which links against libnids. The bug only
happens after being rebuilt and on mips64el where it's already broken.
two separate patches are needed to fix this - one concerning 'after' and
'before' which is already applied upstream, and another concerning some
i386 only functions.

Thanks,
James

unblock libnids/1.23-2.1

-- System Information:
Debian Release: 9.0
  APT prefers unstable-debug
  APT policy:(500, 'unstable')
Architecture: amd64 (x86_64)
Foreign Architectures: i386

Kernel: Linux 4.9.0-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)
diff -Nru libnids-1.23/debian/changelog libnids-1.23/debian/changelog
--- libnids-1.23/debian/changelog       2010-07-21 20:23:34.000000000 +0100
+++ libnids-1.23/debian/changelog       2017-02-26 16:25:37.000000000 +0000
@@ -1,3 +1,13 @@
+libnids (1.23-2.1) unstable; urgency=medium
+
+  * Non-maintainer upload.
+  * Fix assembly of TCP streams on armhf by adding -fno-strict-aliasing.
+    (Closes: #851060)
+  * Fix use of "inline" with GCC >= 5 which causes undefined references in
+    applications linked against libnids. (Closes: #855602)
+
+ -- James Cowgill <jcowg...@debian.org>  Sun, 26 Feb 2017 16:25:37 +0000
+
 libnids (1.23-2) unstable; urgency=high
 
   * Update my email address (closes: #574042).
diff -Nru libnids-1.23/debian/patches/01_before-after.patch 
libnids-1.23/debian/patches/01_before-after.patch
--- libnids-1.23/debian/patches/01_before-after.patch   1970-01-01 
01:00:00.000000000 +0100
+++ libnids-1.23/debian/patches/01_before-after.patch   2017-02-26 
16:25:37.000000000 +0000
@@ -0,0 +1,52 @@
+Description: fix before and after declarations
+ Fix declarations of before and after functions so that they just happen in 
the header file to fix undefined references in libnids.so.
+Origin: upstream, 
http://downloads.sourceforge.net/project/libnids/libnids/1.24/libnids-1.24.tar.gz
+Bug-Debian: https://bugs.debian.org/855602
+Applied-Upstream: 1.24
+Last-Update: 2015-12-06
+---
+This patch header follows DEP-3: http://dep.debian.net/deps/dep3/
+--- a/src/util.c
++++ b/src/util.c
+@@ -29,18 +29,6 @@ test_malloc(int x)
+   return ret;
+ }
+ 
+-inline int
+-before(u_int seq1, u_int seq2)
+-{
+-  return ((int)(seq1 - seq2) < 0);
+-}
+-
+-inline int
+-after(u_int seq1, u_int seq2)
+-{
+-  return ((int)(seq2 - seq1) < 0);
+-}
+-
+ void
+ register_callback(struct proc_node **procs, void (*x))
+ {
+--- a/src/util.h
++++ b/src/util.h
+@@ -23,8 +23,18 @@ struct lurker_node {
+ 
+ void nids_no_mem(char *);
+ char *test_malloc(int);
+-inline int before(u_int seq1, u_int seq2);
+-inline int after(u_int seq1, u_int seq2);
++
++static inline int
++before(u_int seq1, u_int seq2)
++{
++  return ((int)(seq1 - seq2) < 0);
++}
++
++static inline int
++after(u_int seq1, u_int seq2)
++{
++  return ((int)(seq2 - seq1) < 0);
++}
+ void register_callback(struct proc_node **procs, void (*x));
+ void unregister_callback(struct proc_node **procs, void (*x));
+ 
diff -Nru libnids-1.23/debian/patches/02_inline.patch 
libnids-1.23/debian/patches/02_inline.patch
--- libnids-1.23/debian/patches/02_inline.patch 1970-01-01 01:00:00.000000000 
+0100
+++ libnids-1.23/debian/patches/02_inline.patch 2017-02-25 17:50:03.000000000 
+0000
@@ -0,0 +1,45 @@
+Description: Fix more undefined references when using GCC-5.
+ Avoids making the functions ip_fast_csum, ip_compute_csum, my_tcp_check and
+ my_udp_check inline. See https://github.com/aol/moloch/issues/440 as well.
+Author: Robert Scheck <rob...@fedoraproject.org>
+Origin: vendor, 
http://pkgs.fedoraproject.org/cgit/rpms/libnids.git/commit/?id=ecafb692f20e0acad555f66c3cc1646997a82dae
+Bug-Debian: https://bugs.debian.org/855602
+---
+This patch header follows DEP-3: https://dep.debian.net/deps/dep3/
+
+--- a/src/checksum.c
++++ b/src/checksum.c
+@@ -120,7 +120,7 @@ csum_partial(const u_char * buff, int le
+   By Jorge Cwik <jo...@laser.satlink.net>, adapted for linux by Arnt
+   Gulbrandsen.
+ */
+-inline u_short ip_fast_csum(u_char * iph, u_int ihl)
++u_short ip_fast_csum(u_char * iph, u_int ihl)
+ {
+   u_int sum;
+   if (dontchksum(((struct ip*)iph)->ip_src.s_addr))
+@@ -191,13 +191,13 @@ csum_tcpudp_magic(u_int saddr, u_int dad
+   this routine is used for miscellaneous IP-like checksums, mainly in
+   icmp.c
+ */
+-inline u_short
++u_short
+ ip_compute_csum(u_char * buff, int len)
+ {
+   return (csum_fold(csum_partial(buff, len, 0)));
+ }
+ 
+-inline u_short
++u_short
+ my_tcp_check(struct tcphdr *th, int len, u_int saddr, u_int daddr)
+ {
+   if (dontchksum(saddr))
+@@ -205,7 +205,7 @@ my_tcp_check(struct tcphdr *th, int len,
+   return csum_tcpudp_magic(saddr, daddr, len, IPPROTO_TCP,
+                          csum_partial((char *)th, len, 0));
+ }
+-inline u_short
++u_short
+ my_udp_check(void *u, int len, u_int saddr, u_int daddr)
+ {
+   if (dontchksum(saddr))
diff -Nru libnids-1.23/debian/patches/series libnids-1.23/debian/patches/series
--- libnids-1.23/debian/patches/series  2010-07-21 01:13:10.000000000 +0100
+++ libnids-1.23/debian/patches/series  2017-02-25 17:31:41.000000000 +0000
@@ -1 +1,3 @@
 debian-changes-1.23-2
+01_before-after.patch
+02_inline.patch
diff -Nru libnids-1.23/debian/rules libnids-1.23/debian/rules
--- libnids-1.23/debian/rules   2010-07-19 19:10:46.000000000 +0100
+++ libnids-1.23/debian/rules   2017-02-25 17:01:45.000000000 +0000
@@ -16,7 +16,7 @@
 DEB_BUILD_GNU_TYPE  ?= $(shell dpkg-architecture -qDEB_BUILD_GNU_TYPE)
 
 
-CFLAGS = -Wall -g
+CFLAGS = -Wall -g -fno-strict-aliasing
 
 ifneq (,$(findstring noopt,$(DEB_BUILD_OPTIONS)))
        CFLAGS += -O0

Attachment: signature.asc
Description: OpenPGP digital signature


--- End Message ---
--- Begin Message ---
James Cowgill:
> Package: release.debian.org
> Severity: normal
> User: release.debian....@packages.debian.org
> Usertags: unblock
> 
> Hi,
> 
> Please unblock package libnids
> 
> This fixes 2 RC bugs in the package:
> - #851060 where the package doesn't work on armhf. The package contains
> numerous violations of the strict-aliasing rule so I added
> -fno-strict-aliasing as an easy workaround which does fix the bug on armhf.
> 
> - #855602 where the package assumes the old gnu89 inline semantics but
> this was never caught because the undefined references only occur when
> building another package which links against libnids. The bug only
> happens after being rebuilt and on mips64el where it's already broken.
> two separate patches are needed to fix this - one concerning 'after' and
> 'before' which is already applied upstream, and another concerning some
> i386 only functions.
> 
> Thanks,
> James
> 
> unblock libnids/1.23-2.1
> 
> [...]

Unblocked, thanks.

~Niels

--- End Message ---

Reply via email to