1 - Take into account the compile time setting of the default hash size
2 (reported by Dash Four)
5 - Manpage fix to make it clear how ipset works on setlist type
6 of sets (John Brendler, bugzilla id 640)
9 - Checking null entries when listing/saving hash types of sets
10 deleted because it's unnecessary and can mask possible errors.
13 - Manpage fixes and corrections (Jan Engelhardt)
16 - New protocol is introduced to handle aligment issues properly
17 (bug reported by Georg Chini)
18 - Binding support is removed
21 - Correct format specifiers and change %i to %d (Jan Engelhardt)
24 - New kernel-userspace protocol release
25 - Bigendian and 64/32bit fixes (Stefan Gula, bugzilla id 593)
26 - tests/runtests.sh changed to support old bash shells
29 - On parisc architecture cast increases required aligment (bugzilla
31 - Respect LDFLAGS settings at compile time (Peter Volkov).
34 - In order to disable the extra warning flags, NO_EXTRA_WARN_FLAGS
35 variable added to userspace Makefile
38 - Some compiler warning options are too aggressive and
42 - Premature checking prevents to add valid elements to hash
43 types, fixed (bug reported by JC Janos).
44 - Local variable shadows another variable, fixed (reported
46 - More compiler warning options added and warnings fixed.
49 - Include file <limits.h> was missing from userspace set type
50 modules, reported by Krzysztof Oledzki and Sven Wegener.
53 - Only kernel part changes, see kernel/ChangeLog
56 - macipmap type reported misleading deprecated separator
57 tokens and printed the old one at listing set elements
58 (bug reported by Krzysztof Oledzki)
59 - Warn only once about deprecated separator tokens in
63 - Added KBUILD_OUTPUT support (Sven Wegener)
64 - Fix memory leak in ipset_iptreemap (Sven Wegener)
65 - Fix multiple compiler warnings (Sven Wegener)
66 - ipportiphash, ipportnethash and setlist types added
67 - binding marked as deprecated functionality
68 - element separator token changed to ',' in anticipating
69 IPv6 addresses, old separator tokens are still supported
70 - unnecessary includes removed
71 - ipset does not try to resolve IP addresses when listing
72 the content of sets (default changed)
74 - ChangeLog forked for kernel part
77 - Fix to compile ipset with 2.4.26.x tree statically (bug reported by
81 - compatibility for the 2.6.x kernel tree improved and compiler warnings
82 fixed (Jan Engelhardt)
83 - compatibility fixes for the 2.4.36.x kernel tree added
86 - including limits.h for UINT_MAX is required with glibc-2.8 (pud)
87 - needless cast from and to void pointers cleanups in iptreemap (Sven Wegener)
88 - Initial ipset release with kernel modules included.
91 - segfault on --unbind :all: :all: fixed (reported by bugzilla,
92 report and patch sent by Tom Eastep)
93 - User input parameters are sanitized everywhere
94 - Initial testsuite added and 'test' target to the Makefile
95 added: few bugs discovered and fixed
96 - typo in macipmap type prevented to use max size set of this type
97 - *map types are made sure to allow and use max size of sets
100 - jiffies rollover bug in iptree type fixed (reported by Lukasz Nierycho
102 - endiannes bug in iptree type fixed (spotted by Jan Engelhardt)
103 - iptreemap type added (submitted by Sven Wegener)
104 - 2.6.22/23 compatibility fixes (Jeremy Jacque)
105 - typo fixes in ipset (Neville D)
106 - separator changed to ':' from '%' (old one still supported) in ipset
109 - use correct type (socklen_t) for getsockopt (H. Nakano)
110 - incorrect return codes fixed (Tomasz Lemiech, Alexey Bortnikov)
111 - kernel header dependency removed (asm/bitops.h)
112 - ipset now tries to load in the ip_set kernel module if the protocol
116 - 'ipset -N' did not generate proper return code
117 - 'limit' module parameter added to the kernel modules of the
118 iphash, ipporthash, nethash and iptree type of sets so that
119 the maximal number of elements can now be limited
120 - zero valued entries (port 0 or IP address 0.0.0.0) were
121 detected as members of the hash/tree kind of sets
122 (reported by Andrew Kraslavsky)
123 - list and save operations used the external identifier
124 of the sets for the bindings instead of the internal one
125 (reported by Amin Azez)
128 - Nasty off-by-one bug fixed in iptree type of sets
129 (bug reported by Pablo Sole)
132 All patches were submitted by Jones Desougi
133 - missing or confusing error message fixes for ipporthash
134 - minor correction in debugging in nethash
135 - copy-paste bug in kernel set types at memory allocation
137 - unified memory allocations in ipset
140 - memory allocation in iptree is changed to GFP_ATOMIC because
141 we hold a lock (bug reported by Radek Hladik)
142 - compatibility fix: __nocast is not defined in all 2.6 branches
143 (problem reported by Ming-Ching Tiew)
144 - manpage corrections
147 - garbage collector of iptree type of sets is fixed: flushing
148 sets/removing kernel module could corrupt the timer
149 - new ipporthash type added
150 - manpage fixes and corrections
153 - half-fixed memory allocation bug in iphash and nethash finally
154 completely fixed (bug reported by Nikolai Malykh)
155 - restrictions to enter zero-valued entries into all non-hash type sets
157 - Too strict check on the set size of ipmap type was corrected
160 - memory allocation bug in iphash and nethash in connection with the SET
161 target was fixed (bug reported by Nikolai Malykh)
162 - lockhelp.h was removed from the 2.6.13 kernel tree, ip_set.c is
163 updated accordingly (Cardoso Didier, Samir Bellabes)
164 - manpage is updated to clearly state the command order in restore mode
167 - Jiffies rollover bug in ip_set_iptree reported and fixed by Rob Nielsen
168 - Compiler warning in the non-SMP case fixed (Marcus Sundberg)
169 - slab cache names shrunk in order to be compatible with 2.4.* (Marcus
173 - Magic number in ip_set_nethash.h was mistyped (bug reported by Rob
175 - ipset can now test IP addresses in nethash type of sets (i.e. addresses
176 in netblocks added to the set)
179 - Locking bug in ip_set_nethash.c (Clifford Wolf and Rob Carlson)
180 - Makefile contained an unnecessary variable in IPSET_LIB_DIR (Clifford
182 - Safety checkings of restore in ipset was incomplete (Robin H. Johnson)
183 - More careful resizing by avoiding locking completely
184 - stdin stored internally in a temporary file, so we can feed 'ipset -R'
186 - iptree maptype added
189 - Lock debugging used with debugless lock definiton (Piotr Chytla and
191 - Bindings were not properly filled out at listing (kernel)
192 - When listing sets from kernel, id was not added to the set structure
194 - nethash maptype added
195 - ipset manpage corrections (macipmap)
198 - Missing -fPIC in Makefile (Robert Iakobashvili)
199 - Cut'n'paste bug at saving macipmap types (Vincent Bernat).
200 - Bug in printing/saving SET targets reported and fixed by Michal
204 - Chaining of sets are changed: child sets replaced by bindings
205 - Kernel-userspace communication reorganized to minimize the number
207 - Save and restore functionality implemented
208 - iphash type reworked: clashing resolved by double-hashing and by
209 dynamically growing the set
213 - Rewritten to support child pools
214 - portmap, iphash pool support added
215 - too much other mods here and there to list...