1 Summary of changes in iSCSI-SCST between versions 1.0.1 and 1.0.2
2 -----------------------------------------------------------------
4 - Support bidirectional transfers added
7 Summary of changes in iSCSI-SCST between versions 1.0.0 and 1.0.1
8 -----------------------------------------------------------------
10 - Support for variable lenght CDB added
12 - Sending and receiving padding bytes as well as sesnse data
13 reimplemented to make it more reliable
15 - Support for SCSI Async. Event Notifications added
17 - Sessions and connections reinstatement reimplemented to be conforming
20 - iscsi-scst-adm fixed
22 - Removed artificial limitation of sg_tablesize. It isn't needed,
23 because the regular iSCSI negotiated MaxRecvDataSegmentLength-related
24 facilities do all the job pretty well.
26 - Merge with IET r170-211:
28 * Update MaxConnections documentation
30 * Remove isns config description from man page
32 * Check return values of chdir(), ftruncate() and write(), because
33 recent versions of the glibc insist on the return value being
34 checked by introducing __attribute__((warn_unused_result)) to these
37 * Fix snprintf use in isns.c
39 * Take \0-termination into account when passing strings to
40 isns_tlv_set() to solve incompatibility with MS iSNS 3.0 as IQN
41 length is multiple of 4
43 * Fix list corruption if SCST target registration fails
45 * Register the target port actually used instead of the default iSCSI port.
47 * Remove unused connection->pad
49 * Refactor cmnd_execute()
51 * Introduce min*/max*() and more list_* macros
53 * Multipart PDU support for discovery sessions
55 * Add support for Reject PDUs and make use of them in appropriate places.
57 * Ignore SIGPIPE in iscsi-scstd
59 * Allow removal of kernel target objects if the user space object does not exist
61 * Use netlink_socket_release() to clean up netlink sockets correctly
62 (taking care of namespaces)
64 * More scalable proc seq_file implementation
66 * Segfault in ietd buffer overrun in dump_line
68 * Add iscsi-scst-adm support to display list of CHAP accounts and their data
70 * Fixed compilation on glibc 2.8
72 * Fix CHAP account handling bugs
74 * Fix bugs in multi PDU handling code
76 - Support for CentOS 5.2 added (Bart Van Assche)
78 - Fixed NOP-Out with payload handling
80 - Several minor iSCSI RFC violations fixed
82 - Update for kernels up to 2.6.29
84 - Zero-copy feature, introduced by put_page_callback patch made
85 kernel compile time configurable via
86 CONFIG_TCP_ZERO_COPY_TRANSFER_COMPLETION_NOTIFICATION option. If you
87 upgrade from earlier versions, see the corresponding "[CAUTION]"
88 e-mail in scst-devel mailing list about possible compatibility
91 - A lot of cleanups and minor fixes
93 Summary of changes in iSCSI-SCST since it was IET
94 -------------------------------------------------
96 - Commands processing flow changed to work with SCST. Significant
97 cleanup was done. Modules and config files were renamed to allow to
98 run with IET on the same host.
100 - In IET the iSCSI negotiation isn't fully iSCSI RFC confirmed: it doesn't
101 support ranges of values and, more important, violates RFC in case
102 when in the IET config file some value for some parameter is set and
103 a remote initiator doesn't initiate the negotiation for this
104 parameter or declare its value. According to RFC, in this case IET
105 shall use the RFC-specified default value, but it uses config file
106 specified one instead. Looks like the implementation confuses IET
107 config file default values and iSCSI RFC ones. The default values
108 handling was fixed. But support for ranges remains unfixed (see ToDo
111 - All shutdown/restart problems (resource leaks, hangups, etc.), especially
112 under load, were fixed.
114 - Full duplex network IO implemented.
116 - Threading model reimplemented.
118 - Digests calculation made multi-threaded.
120 - TX data digest errors handling iSCSI RFC violation was fixed. Another one
121 (SNACK handling) remains unfixed.
123 - Ability to send data with sense in a single response added.
125 - Ability to handle initiators misbehavior a bit improved: several BUG()'s
126 were relaced by the proper handling. Hovewer, there is a plenty of work in
127 this area left, IET is known to have a lot of weaknesses here.
129 - Task management made more robust. Particularly, possible data
130 corruption scenarios were fixed.
132 - Implemented SN based iSCSI IO flow control
134 - Response data send timeout implemented: now if initiator doesn't
135 accept data for too long (30 sec), target closes connection
137 - A lot of other bugfixes and code cleanups.