2015-12-12  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibsendtrap.c: Eliminate unused parameter from send_trap
	src/ibsendtrap.c: In function ?send_trap?:
	src/ibsendtrap.c:124: warning: unused parameter ?name?

	ibportstate.c: Fix unsigned comparison warnings
	src/ibportstate.c:450: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:454: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:458: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:462: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:483: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:500: warning: comparison of unsigned expression < 0 is always false
	src/ibportstate.c:504: warning: comparison of unsigned expression < 0 is always false

	Remove unused pisize parameter from dump_portinfo in ibdiag_common
	src/ibdiag_common.c: In function ?dump_portinfo?:
	src/ibdiag_common.c:856: warning: unused parameter ?pisize?

2015-10-13  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibqueryerrors.c: Removed unused passed parameters to print_port_config and query_and_dump
	src/ibqueryerrors.c: In function ?print_port_config?:
	src/ibqueryerrors.c:176: warning: unused parameter ?node_name?
	src/ibqueryerrors.c: In function ?query_and_dump?:
	src/ibqueryerrors.c:367: warning: unused parameter ?node?

	ibcacheedit.c: Eliminate unused parameter passed to update_switchportguids
	src/ibcacheedit.c: In function ?update_switchportguids?:
	src/ibcacheedit.c:168: warning: unused parameter ?guid?

2015-10-13  Ira Weiny  <ira.weiny@intel.com>

	update README/configure.ac for 1.6.6

2015-10-13  Dan Ben Yosef  <danby@mellanox.com>

	libibnetdisc: Avoid pushing same pointer to the hash table
	This patch avoids pushing the same pointer to the hash table that causes
	endless loop during dumping.

2015-10-13  Hal Rosenstock  <hal@dev.mellanox.co.il>

	perfquery.c: Fix smp_query_via return value checks
	smp_query_via returns pointer so < 0 comparison is wrong:
	src/perfquery.c: In function ?is_rsfec_mode_active?:
	src/perfquery.c:481: warning: ordered comparison of pointer with integer zero
	src/perfquery.c: In function ?main?:
	src/perfquery.c:919: warning: ordered comparison of pointer with integer zero
	src/perfquery.c:928: warning: ordered comparison of pointer with integer zero

	Also fix the smp_query_via call in is_port_info_extended_supported in
	ibdiag_common.c

	Reported-by: David Binderman <dcb314@hotmail.com>

2015-10-13  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibdiag_common.c: Move static to beginning of get_build_version declaration
	Eliminate compiler warning:
	src/ibdiag_common.c:85: warning: ?static? is not at beginning of declaration

2015-09-20  Ira Weiny  <iweiny@gmail.com>

	infiniband-diags: rdma-ndd: remove udev logging when not supported
	udev_set_log_* functions are deprecated on newer versions of libudev
	Conditionally compile out the use of these functions.

2015-09-20  Dan Ben Yosef  <danby@mellanox.com>

	perfquery -T (print Extended Speed Counters) times out on nodes supporting RS_FEC capability.
	Before sending ExtendedSpeedCounters PerfMgt GMP, perfquery sends
	PortInfoExtended SMP to get CapMask and FECModeActive fields.
	Upon reception of PortInfoExtended SMP, fec_mode_active field is
	erroneously decoded to 16 bit buffer instead of 32 bit buffer.
	As a result, memory corruption occurs and portid.dlid field is set to 0,
	so that ExtendedSpeedCounters GMP is sent to wrong destination.

2015-09-20  Vladimir Koushnir  <vladimirk@mellanox.com>

	saquery.c: Fix saquery -D option
	-D and -list saquery options are operational only
	when -N option is explicitly invoked (default option)

	This patch allows using -D or --list options when no other option is
	invoked.

2015-08-02  Ana Guerrero López  <ana@ekaia.org>

	rdma-ndd: fix compiler warnings.
	This patch fixes -Wformat-security warnings, that in Debian (and Ubuntu)
	are enabled by default:

	src/rdma-ndd.c: In function 'update_node_desc':
	src/rdma-ndd.c:149:3: error: format not a string literal and no format arguments [-Werror=format-security]
	   fprintf(f, new_nd);
	   ^
	src/rdma-ndd.c: In function 'udev_log_fn':
	src/rdma-ndd.c:247:2: error: format not a string literal and no format arguments [-Werror=format-security]
	  syslog(LOG_ERR, msg);

2015-08-02  Hal Rosenstock  <hal@dev.mellanox.co.il>

	iblinkinfo.c: Close additional file descriptor in advance
	Additional file descriptor for SMP MADs should be closed before running
	ibnd_discover_fabric() to avoid parallel usage of two SMP file descriptors

2015-07-14  Vladimir Koushnir  <vladimirk@mellanox.com>

	vendstat: mad_rpc_close_port not called in corner cases
	Call mad_rpc_close_port in all corner cases.
	This is needed so fds are not leaked with ibsim.

2015-07-14  Michal Schmidt  <mschmidt@redhat.com>

	build-sys: avoid overlinking to libudev
	Not all built binaries need to link to udev, only rdma-ndd.
	Use pkg-config to detect udev in configure.ac.

	rdma-ndd: never use udev_get_sys_path()
	The function was deprecated and later dropped by udev entirely
	because it was pointless.
	Quoting kernel's Documentation/sysfs-rules.txt:
	  sysfs is always at /sys
	  Parsing /proc/mounts is a waste of time. Other mount points are
	  a system configuration bug you should not try to solve. [...]

2015-07-03  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibdiag_sa.c: In sa_get_handle, handle umad_open_port and umad_register failures
	Treat umad_open_port and umad_register errors as failures for sa_get_handle

2015-07-03  Vladimir Koushnir  <vladimirk@mellanox.com>

	ibqueryerrors: improve code related to DR option
	The call to ibnd_find_port_dr() is removed
	as couple of lines after that the port is overridden by result of
	ibnd_find_port_guid()

2015-07-03  Vladimir Koushnir  <vladimirk@mellanox.com>

	ibqueryerrors: Close global file descriptor before running ibnd_discover_fabric
	Global file descriptor for SMPs and GMPs should be closed before running
	ibnd_discover_fabric() to avoid parallel usage of two SMP file descriptors


	With fix ups to the goto labels

2015-07-03  Vladimir Koushnir  <vladimirk@mellanox.com>

	ibqueryerrors: Fix crash when no SM is running
	Handle sa_get_handle error in path_record_query

	ibqueryerrors: code improvement
	goto is performed after IBEXIT call

	dump_fts: Open global file descriptor after calling ibnd_discover_fabric
	Global file descriptor for SMPs should be opened only after running
	ibnd_discover_fabric() to avoid parallel usage of two SMP file descriptors

2015-07-03  Vladimir Koushnir  <vladimirk@mellanox.com>

	query_smp.c: Avoid busy looping in process_one_recv
	Today, process_one_recv calls umad_recv with zero timeout.
	That means umad_recv will use non-blocking read() without polling,
	In many cases, this causes busy loop.

	In the fix, umad_recv is invoked with -1 (infinite timeout) and the read()
	will occur only when poll exits. Also, no longer need to check for
	EWOULDBLOCK.

2015-07-03  Vladimir Koushnir  <vladimirk@mellanox.com>

	Remove redundant umad file descriptor from libibnetdisc
	Today, two umad SMP file descriptors are used in libibnetdisc.
	One of them is needed only for retrieving LID of the local port
	for combined routing.

	The patch removes the need for 2 files descriptors by retrieving
	LID of the local port in advance.

2015-04-22  Vladimir Koushnir  <vladimirk@mellanox.com>

	ibqueryerrors: Resource leak in path_record_query
	sa_free_handle function is not called at the end of
	path_record_query function.
	The patch fixes file descriptor leak when path_record_query
	function is called.

2015-04-12  Hal Rosenstock  <hal@dev.mellanox.co.il>

	perfquery.c: Change format of capability mask in IBWARN for consistency

2015-03-16  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags/rdma-ndd: add --pidfile option
	rdma-ndd start up script works better with a pidfile.  This was only partially
	implemented in the start up script and not at all in the daemon.

	Reported-by: Dan Ben Yosef <danby@mellanox.com>

2015-03-11  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags/rdma-ndd: Fix issues with install
	On Debian in particular it was reported that the binary was attempting to run
	from /sbin when it was installed to /usr/sbin.  This was not seen on RHEL as
	RHEL defines /sbin to be a softlink to /usr/sbin.

	In addition fix the /var path by using the localstatedir macro.

	Reported-by: Dan Ben Yosef <danby@mellanox.com>

2015-02-11  Ira Weiny  <ira.weiny@intel.com>

	update README for 1.6.5

2015-02-07  Dan Ben Yosef  <danby@mellanox.com>

	ibtracert.c: Remove checking the port 0 state for base switch port 0
	The port 0 state check needed to be only for HCA/Routers and
	EnhancedPort0 switches.
	For BasePort0 switches, the PortState field, in PortInfo arrtibute, for port0 is
	undefined.

	Add new configure flag to enable/disable the rdma-ndd build
	Add new flag --enable-rdma-ndd (default=yes)

2015-02-07  Hal Rosenstock  <hal@dev.mellanox.co.il>

	Align infiniband-diags MLNX device IDs with OpenSM
	Same device IDs as in opensm/osm_subnet.c:is_mlnx_ext_port_info_supported

	Switch-IB is device ID 0xcb20 rather than 0xcb84

2015-01-28  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: update for udev changes.
	RHEL7 brings changes to the udev library.

	The udev library no longer exports udev_get_sys_path function.  Check for this and remove
	this call if it is not supported.

	No longer require the libudev package for installation.  The library requirement is
	automatically built in during the build.

2015-01-28  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags/test-code: remove unused variables
	Unused variables were found while building on RHEL7 with gcc 4.8.2

2015-01-19  Hal Rosenstock  <hal@dev.mellanox.co.il>

	vendstat.c: Add additional vendor IDs to ext_fw_info_device
	Add additional supported device IDs to ext_fw_info_device.
	These IDs are ConnectX-4, ConnectX-IB, and Switch-IB.
	This affects is_ext_fw_info_supported which determine
	which Mellanox specific vendor MADs are issued to a device.

	ibnetdisc.c: Add additional supported device IDs to is_mlnx_ext_port_info_supported
	Similar to previous ibdiag_common.c patch, this adds additional device IDs
	to be supported to the ibnetdiscover internal is_mlnx_ext_port_info routine.
	These IDs are ConnectX-4, ConnectX-IB, and Switch-IB.

2015-01-19  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibdiag_common.c: Add more supported device IDs in is_mlnx_ext_port_info_supported
	Add support for some ConnectX-IB, ConnectX-4, and Switch-IB in
	is_mlnx_ext_port_info_supported.

	This only affects tool(s) which invoke this routine
	(is_mlnx_ext_port_info_supported) which is
	currently just ibportstate.

	is_mlnx_ext_port_info_supported determines whether or not the MLNX
	vendor extended port info SMP is sent which is used to determine FDR10
	speed.

	Not having the device in the list would only result in FDR10 being
	misinterpreted as QDR.

2015-01-10  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: add rdma-ndd daemon
	rdma-ndd is a system daemon which watches the procfs hostname file as well as
	monitoring for InfiniBand libudev device changes.

	Upon detecting changes it will update the Node Descriptions of the RDMA devices
	in the system.

	This deamon is intended to work with kernels which support polling of the
	procfs hostname.  If your kernel does not support this feature the daemon
	will set the Node Descriptions to the hostname at start up only.

	ibdiag.conf can be used to specify a node description format with the following
	format specifier support.

	%h == current hostname
	%d == device name

	Regardless of previous settings, all NDs are written at startup and
	individually when a new device is detected.  Otherwise the ND is compared to
	the existing one and not written if it has not changed.

2015-01-10  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibccconfig.c: Account for enhanced switch port 0 in SwitchPortCongestionSetting handling
	Enhanced switch port 0 can support CC (indicated in CC ClassPortInfo:CapabilityMask.EnhancedPort0CC).

	portnum 0 is a valid port number and offsets should be calculated based on port 0
	rather than port 1.

	Also, fix some error messages to indicate switch port congestion setting
	rather then switch congestion setting failure.

2015-01-10  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibccquery.c: Account for enhanced switch port 0 in SwitchPortCongestionSetting handling
	Enhanced switch port 0 can support CC (indicated in CC ClassPortInfo:CapabilityMask.EnhancedPort0CC).

	portnum 0 is a valid port number and offsets should be calculated based on port 0
	rather than port 1.

	Also, fix some error messages to indicate switch port congestion setting
	rather then switch congestion setting failure.

2014-11-18  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: update smpquery man page

2014-11-18  Dan Ben Yosef  <danby@dev.mellanox.co.il>

	New MAD SM:PortInfoExtended and changing PM:PortExtendedSpeedsCounters
	This change needed for EDR support.
	This change includes:
	1. Adding new MAD option to smpquery: PortInfoExtended (PIE).
	2. Changing perfquery to support PortExtendedSpeedsCounters with
	RSFEC mode active.

2014-11-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	saquery.c: Return proper status from query_sa_cpi on bad result status
	When result status was other than IB_SA_MAD_STATUS_SUCCESS,
	0 was being returned rather than EIO as intended.

2014-11-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	saquery.c: Fix handling of cpi and ClassPortInfo options
	Need to check if query_type is CLASS_PORT_INFO
	in order to handle saquery [cpi | ClassPortInfo].
	Otherwise, it just dumps ClassPortInfo attribute
	as all 0s.

	Note that saquery -c works fine because it
	set command to SAQUERY_CMD_CLASS_PORT_INFO.

2014-10-18  Hal Rosenstock  <hal@mellanox.com>

	ibping.c: Fix support of OUIs other than IB_OPENIB_OUI

2014-06-30  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibportstate.c: Cosmetic changes to MLNX Extended PortInfo
	for consistency with smpquery mepi output:

	smpquery mepi 24 1
	StateChangeEnable:...............0x00
	LinkSpeedSupported:..............0x01
	LinkSpeedEnabled:................0x01
	LinkSpeedActive:.................0x00

	and to prepare for IBTA extended PortInfo attribute

2014-04-20  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: If rst2man is available fix missing source dir in configure

2014-04-20  Dan Ben Yosef  <danby@mellanox.com>

	Fix missing source dir in configure.ac
	This fixes running configure outside the infiniband-diags dir

2014-03-04  Hal Rosenstock  <hal@dev.mellanox.co.il>

	saquery.c: In query_sa_cpi, make error message consistent
	with similar message in get_any_records

	saquery.c: In print_node_record, change LID print format
	"Standard" for LID printing is unsigned rather than signed decimal

2014-01-26  Ira Weiny  <ira.weiny@intel.com>

	update README for 1.6.4

	Revert "libibnetdisc/ibnetdisc.c: improve log information to stdout"
	This reverts commit a3f2fcd468b459e0f3b1f5b7c35a7db181d22ee9.

2014-01-25  Ira Weiny  <ira.weiny@intel.com>

	update README for 1.6.3

	ibtracert.8.in: change default pre-built man page

2014-01-25  Al Chu  <chu11@llnl.gov>

	infiniband-diags: Check node-name-map strings when querying by name

2014-01-25  Dan Ben Yosef  <danby@mellanox.com>

	libibnetdisc/ibnetdisc.c: improve log information to stdout
	Write to stdout the route-path when smp returns
	with status error.

	libibnetdisc/ibnetdisc.c: fix insert of invalid lid 0xFFFF into lid_port hash_table

2013-12-20  Dan Ben Yosef  <danby@mellanox.com>

	ibtracert.c: Add 'force' flag to man page
	Add 'force' route option to man page.

2013-12-19  Dan Ben Yosef  <danby@mellanox.com>

	infiniband-diags/vendstat: Fix GeneralInfo SW version
	Fix the offset for GeneralInfo SWInfo for IS4

2013-10-21  Dan Ben Yosef  <danby@mellanox.com>

	ibportstate: on/off option is missing in help

2013-10-11  Dan Ben Yosef  <danby@mellanox.com>

	iblinkinfo: Remove "-R" option from man page
	remove "-R" option from iblinkinfo man

	ibaddr: Add missing -t option to manpage
	Add missing 'timeout' option

2013-09-22  Hal Rosenstock  <hal@dev.mellanox.co.il>

	ibdiag_sa.c: Output attribute ID in hex rather than decimal
	so that it's easier to correlate to IBA spec

	For example:
	saquery PKTR 20/1/0 --smkey 1
	ibwarn: sa_query: umad_send failed: attr 51: Bad file description

	0x33 would be easier than 51 as no conversion is required to
	look it up in IBA spec.

2013-08-20  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: fail configure if glib2 is not found

2013-08-19  Bart Van Assche  <bvanassche@acm.org>

	Fix errors triggered by autoregen.sh
	Avoid that running autoregen.sh with recent versions of autoconf and
	automake triggers the following errors:

	libibnetdisc/Makefile.am:35: compiling `testleaks.c' with per-target flags requires `AM_PROG_CC_C_O' in `configure.in'
	Makefile.am: installing `./INSTALL'
	Makefile.am: required file `./NEWS' not found

2013-08-19  Bart Van Assche  <bvanassche@acm.org>

	get_lid_from_name(): Remove an unused variable
	This patch avoids that the following compiler warning is triggered:

	src/saquery.c:869:18: warning: variable 'p_ni' set but not used [-Wunused-but-set-variable]

2013-08-18  Ira Weiny  <ira.weiny@intel.com>

	Update README and version to 1.6.2

2013-08-12  Hal Rosenstock  <hal@dev.mellanox.co.il>

	perfquery.c: Fix dumping of PortSamplesControl
	was stopping after CounterSelect10 field as insufficient buffer size for
	dumping this attribute.

2013-07-19  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: Eliminate unneeded clean_nodedesc calls
	in ibroute and dump_fts
	clean_nodedesc is called if needed within remap_node_name

2013-07-19  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: convert IBERROR to IBEXIT
	NOTE: IBERROR has always exited; this simply clarifies the macro and function behavior

	infiniband-diags: umad_init needs to be called before sa_get_handle

2013-07-05  Hal Rosenstock  <hal@dev.mellanox.co.il>

	saquery: Add SMInfoRecord support
	This patch is the combination of 2 patches:
	     infiniband_diags/src/saquery.c: Add SMInfoRecord support
	     infiniband_diags/doc/rst/saquery.8.in.rst: Add support to the manual

2013-07-05  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: saquery: Dump SwitchInfoRecord based on IsSetMulticastFDBTopSupported SA capability
	Only include MulticastFDBTop when SA
	ClassPortInfo:CapabilityMask2.IsSetMulticastFDBTopSupported is set

	In order to do this, SA CPI is queried prior to other queries and passed to all
	dump functions via query_params when required.

	Original patch supplied by Hal Rosenstock.

	Changes from V1:
		use cl_ntoh* instead of ntoh*

2013-06-23  Albert Chu  <chu11@llnl.gov>

	ibnetdiscover: Output vlcap in full output format
	The additional vlcap information is useful when simulating a fabric
	in ibsim and fabric VL information is needed.

	ibnetdiscover: Add missing -f option to manpage

2013-06-22  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibstat fix strncpy coverity check and -l bug
	Changes since V1:
		Fix port parameter usage.

2013-06-12  Dan Ben Yosef  <danby@dev.mellanox.co.il>

	saquery.c: resource leak
	Overwriting "node_name" in call "node_name = remap_node_name(node_name_map,
	nr->node_info.node_guid, (char*)nr->node_desc.description)" leaks the
	storage that "node_name" points to.

	ibtracert.c: Uninitialized variable
	Reading from uninitialized variable

	ibsysstat.c: fix resource leak
	Memory leak.

2013-06-12  Dan Ben Yosef  <danby@dev.mellanox.co.il>

	ibtracert.c: fix resource leak
	Call return without freeing the buffer.

	Changes from original patch:
		Removed return after IBERROR; IBERROR performs an exit.

2013-06-12  Hal Rosenstock  <hal@dev.mellanox.co.il>

	libibnetdisc/Makefile.am: Add missing makefile dependency
	Found-by: Doron Tsur <doront@mellanox.com>

2013-05-29  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: Fix gen_ver.sh for use with new Autoconf file

2013-05-22  Hal Rosenstock  <hal@dev.mellanox.co.il>

	saquery: Add SwitchInfoRecord support
	This patch is the combination of the 2 patches supplied by Husam Kahalah
	        [v2] infiniband-diags/saquery.c: switchinfo support added
	        switchinfo support added manual

	In addition, this version changes the field names to match those in Vol1 v1.2.1
	and Errata.

2013-05-22  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: remove doc/man/ibclearcounters.8.in
	ibclearcounters is part of the compatibility package and the man page remains in
	the man directory.

	infiniband-diags: add .gitignore

	infiniband-diags: updates for new autoconf tools

2013-03-28  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags/dump_fts: fix block calculation for LFT's
	This fix was found by Dan Ben Yosef in ibroute.  ibroute and dump_fts should
	share this code but for now, fix the bug in dump_fts.

2013-03-28  Dan Ben Yosef  <danby@mellanox.com>

	infiniband-diags/ibroute.c: fix pull extra block for LFT
	when we pull LFT we pull extra block, this may cause to failure.

2013-03-26  Dan Ben Yosef  <danby@mellanox.com>

	infiniband-diags/ibportstate.c: Add ON and OFF options.
	Add new Options :
	                1.ON: Changed the port state to enable only in case
	                      the current state is disable.
	                2.OFF: Change the port state to disable.

2013-03-26  Ira Weiny  <ira.weiny@intel.com>

	infiniband-diags: update emails in default pages

	infiniband-diags: update emails

	infiniband-diags: fix rst2man when using upstream docutils
	The upstream docutils package does not install "rst2man".  Instead the executable
	is "rst2man.py".  Check for both options before giving up.

	infiniband-diags: add skip-sl to default ibqueryerrors.8.in

2013-03-18  Dan Ben Yosef  <danby@mellanox.com>

	infiniband-diags/ibqueryerrors.c: obtain destination sl before perf query
	1) if use -G or -D option : we obtain sl before doing perf query.
	2) if no destination is given : we obtain sl for every pair source-destination.
	3) if no destination is given and use --skip-sl option : we don't obtain
	sl to all nodes in the fabric,in this case sl=0 for all node pairs.

2013-02-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix rpm requires for glib2

	infiniband-diags: check_lft_balanc: support either DR or Lid based dump lfts output

	infiniband-diags: deprecate dump_[m|l]fts.sh scripts

2013-02-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add dump_fts tool
	dump_fts adds a faster version of the functionality of dump_[l|m]fts.sh.  This
	code is based off of the ibroute code and simply uses libibnetdisc to scan the
	fabric instead of using ibnetdiscover and letting ibroute requery all that data
	over again.  This improves things in 3 ways.

	1) performance improves by nearly 2 orders of magnitude.
	2) this version greatly reduces the mads required and thus reduces the
	   impact on the fabric.
	3) Everything is queried with DR paths which ensures if the routing tables are
	   bad on the cluster the query will still complete and give you the
	   information you were looking for.  (To be fair dump_lft.sh has the DR option
	   but it is currently buggy.)

	Example runs on the ~1400 nodes of the Hyperion test cluster show:

	13:45:46 > time ./dump_lfts.sh > /dev/null

	real    4m58.175s
	user    0m6.407s
	sys     0m17.983s

	13:53:12 > time ./dump_fts > /dev/null
	dump tables: linear forwarding table get failed

	real    0m8.121s
	user    0m3.032s
	sys     0m3.342s

	Changes from V1:
		Add status and query information to error messages.
		Add man page files which were missed in the first patch

	Changes from V2:
		Additional status and address information on error messages.
		clean up error handling when fabric scan fails.

2013-02-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: libibnetdisc add find node by lid
	NOTE: this change adds a glib requirement to the package.

	Changes since V1:
		Use GINT_TO_POINTER rather than allocating keys

	Changes since V2:
		Use internal object everywhere rather than just hacked into
		discover_fabric
		Generate lid2port hash when reading cached fabrics

2013-02-15  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add change to pregenerated saquery.8.in

2013-02-15  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/saquery: Add support for ServiceID in PathRecord querying

2013-02-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix saquery rst/man page

2013-02-13  Hal Rosenstock  <hal@mellanox.com>

	infiniband-diags/saquery.c: Display NodeRecord LID in decimal
	rather than hex

2013-02-07  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add comment for path_portid

	infiniband-diags: ibroute: add MAD status to query failure message

2013-01-22  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibroute fix resolve failure message

2013-01-05  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibportstate.c: Fix PortInfo sets with extended speeds
	Attribute Modifier bit 31 should be on (set to 1) in this case

2012-11-27  Albert Chu  <chu11@llnl.gov>

	infiniband-diags: Support node-name-map in ibroute

2012-11-26  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add make check target for shell scripts

2012-11-26  Garrett Cooper  <yanegomi@gmail.com>

	Remove bashisms from infiniband-diags

2012-11-05  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Fix ibsysstat man page

2012-11-01  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibsysstat.c: Fix non default OUI operation

2012-11-01  Dan Ben Yosef  <danby@mellanox.com>

	infiniband-diags/src/saquery.c: New option to saquery InformInfoRecord (IIR)
	Add the option to display the InformInfoRecord by SubscriberGid

	Applied with additional changes:

	Cleaner error message:
	-          fprintf(stderr, "Failed to find InformInfoRecord SubscriberGID for \"%s\"\n",str);
	+          fprintf(stderr, "Invalid SubscriberGID specified: \"%s\"\n",str);

	Added parens around assignment:
	-           if(rc = parse_iir_subscriber_gid(argv[0], &ir) < 1)
	+           if((rc = parse_iir_subscriber_gid(argv[0], &ir)) < 1)

2012-10-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Fix SA error reporting for common MAD errors

2012-09-21  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix bug with loop ports where SMP queries are needed

	infiniband-diags: move sa query code into it's own common file
	Also remove bind from handle name

2012-09-21  Albert Chu  <chu11@llnl.gov>

	infiniband-diags/src/perfquery.c: Fix all_ports corner case
	if AllPortSelect is not supported and --all_ports is specified, it should
	only be emulate AllPortSelect if no port is specified or the all ports
	port number (255) is specified.

	infiniband-diags: Allow user to specify multiple ports
	Allow user to select multiple ports via comma or range input.  May
	be particularly useful for gathering aggregate performance counters
	for groups of ports, such as all the uplinks or downlinks from a
	switch.

2012-08-15  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add default <man>.in files

2012-08-03  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: combine common alias option processing

2012-07-30  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: update README and configure.in for 1.6.1 release

2012-07-24  Albert Chu  <chu11@llnl.gov>

	infiniband-diags: Fix command line parsing bug in ibccconfig, ibccquery, saquery, and smpquery.
	Ensure user's inputted operation directly matches an operation name or
	operation alias.  Do not base it on the length of the input, which can
	lead to conflicts with operation selection.

2012-07-13  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/vendstat.c: Commentary change
	Add device names to go along with device IDs

2012-07-11  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/perfquery.c: Fix compile warning
	src/perfquery.c: In function main:
	src/perfquery.c:662: warning: integer constant is too large for long type

2012-07-11  Daniel Klein  <danielk@dev.mellanox.co.il>

	Subject: [PATCHv2] infiniband-diags/vendstat: updated devices supporting extended fw info
	Changed devices ID to an array of device ID ranges.

	Version 2:
	Added range for ConnectX devices: 0x1003 - 0x1011

2012-07-11  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: Support MLNX ExtendedPortInfo for ConnectIB device

2012-07-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	iblinkinfo: fix no port 0 info segmentation fault
	Fix iblinkinfo segfault when switch port 0 not responding to PortInfo
	If switch port 0 doesn't respond to PortInfo, the speed and width of switch
	ports is unknown.


	With additional brackets.

2012-07-08  Ira Weiny  <weiny2@llnl.gov>

	Merge branch 'master' of git://git.openfabrics.org/~iraweiny/infiniband-diags

2012-07-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	mcm_rereg_test.c: Fix a couple of output messages
	In rereg_send, it's not a leave but query
	Fix endian of guid in rereg_send_all

	ibnetdiscover: fix not showing FDR10 speeds
	Fix: ibnetdiscover shows FDR10 speeds as QDR due to bug in setting
	default parameters.

2012-07-08  Ira Weiny  <weiny2@llnl.gov>

	Merge branch 'from-hal'

2012-07-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	mcm_rereg_test.c: Fix a couple of output messages
	In rereg_send, it's not a leave but query
	Fix endian of guid in rereg_send_all

	ibnetdiscover: fix not showing FDR10 speeds
	Fix: ibnetdiscover shows FDR10 speeds as QDR due to bug in setting
	default parameters.

	ibnetdiscover: fix not showing FDR10 speeds
	Fix: ibnetdiscover shows FDR10 speeds as QDR due to bug in setting
	default parameters.

2012-06-29  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibnetdiscover: fix no port 0 info segmentation fault
	Fix ibnetdiscover egmentation fault when it reaches a switch
	that does not respond to a PortInfo Get for port 0.

	Note:
	When ibnetdiscover can't read the port capabilites field from switch port 0,
	'???' is printed for link speed and width.

2012-06-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: saquery avoid a potential seg fault in libibmad
	The common function sa_query should pass the correct data size to mad_build_pkt
	for it's internal memcpy.

2012-06-01  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: update README for 1.6.0 release

	infiniband-diags: udpate smp documentation in infiniband-diags page

	infiniband-diags: Add SMP/QP0 usage explanation to general man page

	infiniband-diags: saquery clarify default smkey option value

2012-06-01  Jim Foraker  <foraker1@llnl.gov>

	infiniband-diags: Add SA smkey option to config file
	Defines a default for the smkey used in SA requests

	ib-diags/saquery: Fix smkey handling
	smkey is already defined as a global inside saquery.c, so remove
	broken support for passing it around as a function parameter

	infiniband-diags: Add m_key option to config file

	infiniband-diags/ibportstate.c: Support changing MKey, lease, and protect bits

	infiniband-diags: Allow specification of an mkey to use on the command line

	infiniband-diags/ibportstate.c: Display MKey, lease, and protect bits
	By default, display mkey lease and protection bits.
	With a flag, also display mkey itself

	infiniband-diags: install config file mode 400
	Necessary so that we may add credentials (ie, mkey) to it

	infiniband-diags: Obfuscate sensitive output by default
	Cause security credentials to not be displayed by default. Instead,
	the "-K" flag is supported, which must be passed to the command
	to display valid results.

2012-05-30  Jim Foraker  <foraker1@llnl.gov>

	infiniband-diags: clean up exclude arguments to ibdiag_process_opts()
	Alphabetize for standardization and greater clarity

2012-05-29  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibportstate.c: Remove an extra space from warning message

2012-05-25  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibportstate.c: Add newly added options to usage
	espeed and fdr10

2012-05-16  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix autogen with new rst documentation

2012-05-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: move man/check_lft_balance,ibcacheedit,ibstatus,ibidsverify to rst documentation

	infiniband-diags: move man/saquery,vendstat,ibsysstat,ibfindnodesusing to rst documentation

	infiniband-diags: move man/perfquery,sminfo,smpdump,smpquery to rst documentation

	infiniband-diags: move man/ibportstate,ibstat,ibroute,ibtracert to rst documentation

	infiniband-diags: move man/ibping,ibnodes,ibhosts,ibswitches,ibrouters to rst documentation

	infiniband-diags: move man/ibcc[config|query] to rst documentation

	infiniband-diags: move man/dump_[ml]fts to rst documentation

2012-05-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: convert documentation from nroff to rst
	    Having documentation in ReStructured Text has the following benefits

	    1) The addition of common functionality (options, config files, etc.) can be
	       documented once and then included in documents for multiple tools.

	       1a) Documentation source is more straight forward and writers can
	           concentrate on the features which are unique to each tool.

	    2) Documentation can be generated into multiple formats (man, html) not just
	       man pages.

	    3) Documentation for any individual tool is complete within that page.  (There
	       is no referencing of other documents and guessing which "common" options
	       apply.)

	    Included as part of this patch is a complete audit of the options for these
	    tools.

	    Included in this initial patch are 4 tool which covered most of the options
	    and the main infiniband-diags man page.

2012-05-09  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: check_lft_balance.pl remove unused options

2012-05-07  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibidsverify exit status 1 on errors found

	infiniband-diags: fix ibidsverify portguid validation.
	PortGUID may be equal to the NodeGUID of it's own node for 1 port.  Duplicate
	PortGUID's on that node will be detected in insert_portguid

	infiniband-diags: move ibidsverify man page to main rpm
	It was incorrectly put in the compat rpm.

	infiniband-diags: move deprecated scripts to "compat" rpm
	commit 7cdad904b8d846cf3958a779561ede8bc59b5d05 missed ibclearerrors and
	ibclearcounters.

	infiniband-diags: update ib[hosts|switches|routers] to standard options

	infiniband-diags: smpdump fix option 'string'

	infiniband-diags: remove -u common option

2012-05-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: remove SMI class registration where possible

2012-05-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: remove mad.h from ibstat
	ibstat does not require libibmad directly.

	The change to ibdiag_common was required because of the conflict between the
	define in ib_types.h and enum in mad.h which presented itself after the fix in
	ibstat.

2012-05-02  Albert Chu  <chu11@llnl.gov>

	infiniband-diags ibcconfig: Update documentation examples
	Update examples to clearly show that the CACongestionSetting operation
	should be used multiple times to configure different SLs.

2012-05-01  Albert Chu  <chu11@llnl.gov>

	infiniband-diags/src/ibccquery.c: Fix CACongestionSetting inputs
	Make inputs to CACongestionSetting more like MAD packet, allowing
	multiple SLs to be configured at one time.

2012-05-01  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: move deprecated scripts to "compat" rpm
	In addition remove them from the make install target unless autoconf'ed on.

	Changes from V1:
		Properly Rebased against current master
		Reorganize tool list in Makefile.am and spec.in files
		Add set_nodedesc.sh

2012-05-01  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: mark duplicate/lower performing scripts as deprecated.
	Inform user of alternatives if available.

	Changes since V1:
		Add set_nodedesc.sh

2012-04-23  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/vendstat.c: Support IS4/SX firmware version info

2012-04-17  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: preserve long_opts global for use in ibdiag_show_usage
		Calling ibdiag_show_usage after ibdiag_process_opts will result in
		using free'd memory.

2012-04-14  Jim Foraker  <foraker1@llnl.gov>

	infiniband-diags: migrate resolve_portid_str in from libibmad
	ib_resolve_portid_str_via() is replaced with resolve_portid_str(),
	which operates identically but does not issue any SMPs.  Data is
	pulled from the umad layer, which necessitates it taking 2 new
	arguments.

	infiniband-diags: Remove dependencies on ib_resolve_smlid/ib_resolve_self
	ib_resolve_smlid() and ib_resolve_self() issue SMP queries to the local
	interface for data that can be acquired thru other means.
	resolve_sm_portid() (originally defined in 3179ae8c and made public here)
	and the new function resolve_self() replace these functions, pulling
	information from the umad layer instead.

	infiniband-diags/mcm_rereg_test.c: Include ib_types.h
	Remove IB_MCM_COMPMASK defines, which should be pulled from
	ib_types.h instead.  This is necessary in order to allow future
	inclusion of ibdiag_common.h.

	infiniband-diags/mcm_rereg_test.c: dead code removal
	Remove old #if 0'd code

	infiniband-diags/ibroute.c: Dead code removal
	Program has already exited in the case where argc == 0

2012-03-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: saquery further error value clean up.

	infiniband-diags: remove unused options from check_lft_balance.pl

	infiniband-diags: remove ib_resolve_smlid_via requirement for sa queries
	ib_resolve_smlid_via, in the MAD library, issues a SubnGet(PortInfo) on the
	local port to resolve the SM LID and SL.  saqueries do not require the use of
	SMP's.

2012-02-18  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Fix saquery error reporting
	The sa_query function can fail in 2 ways
		1) the system calls can fail
		2) the SA returns a failure result

	The first case was not being handled properly.  In addition ensure the program
	returns a proper errno value as an exit code regardless of the failure mode.

2012-02-08  Ira Weiny  <weiny2@llnl.gov>

	Fix ibping server when client stops sending.
	ibping -S treats receive errors as if they were valid incoming MAD packets and tries to send responses to them.

	Original fix provided by Mike Heinz <michael.heinz@qlogic.com>

2012-01-17  Albert Chu  <chu11@llnl.gov>

	infiniband-diags: Support ibccconfig, congestion control setting tool

2012-01-12  Ira Weiny  <weiny2@llnl.gov>

	ibswitches: fix output when switch names have "base" or "enhanced" in them

2012-01-12  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/configure.in: Fix typo

2011-12-05  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibccquery Fix field name

2011-12-05  Albert Chu  <chu11@llnl.gov>

	infiniband-diags/src/ibccquery.c: Fix off by one output bug, ccti_limit is 0 based not 1 based

2011-12-05  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add config file option for Mellanox ExtPortInfo queries
	In addition add a common option to specify an alternate config file.

	infiniband-diags: move is_mlnx_ext_port_info_supported to common code

2011-12-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix printing of extended speeds when port is down

2011-12-01  Hal Rosenstock  <hal@mellanox.com>

	infiniband-diags/perfquery.c: Fix reset for PortExtendedSpeedsCounters
	---
	Changes since v1:
	Patch now against separated/recent infiniband-diags tree

2011-12-01  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/perfquery.c: Fix display of PortExtendedSpeedsCounters
	Add additional buffer space for output
	so no longer chopped after FECCorrectableBlockCtrLane11

2011-12-01  Albert Chu  <chu11@llnl.gov>

	Fix typo in ibccquery manpage

2011-11-07  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: check_lft_balance.pl add -C/-P options

	check_lft_balance.pl: Add extra check when using -e heuristic flag in check_lft_balance.pl

2011-11-04  Albert L.Chu  <chu11@llnl.gov>

	Add -e heuristic flag to check_lft_balance.pl to detect common scenarios where unbalanced routing will occur.

	Update check_lft_balance.pl to work with newer infiniband-diag tools

2011-11-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors: don't exit on reset error
	IBERROR exits the process.  ibqueryerrors should continue despite this error.

2011-10-27  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add a general config file
	This supports the common options for CA, Port, and timeout.

	infiniband-diags: add ibccquery man page to build

	infiniband-diags: clean up build

2011-10-26  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: iblinkinfo flag ports which were not scanned
	When using the -n option iblinkinfo may not scan the entire fabric.  As such
	some ports will not be available to print.  Output that information rather than
	blank data as before.

2011-10-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: saquery: use ib-node-name-map file

	infiniband-diags: iblinkinfo: fix error message

	infiniband-diags: fix installation of error_thresholds file

	Master branch is not moving on to 1.6

2011-10-01  Albert Chu  <chu11@llnl.gov>

	support --switches-only and --cas-only options in iblinkinfo

2011-09-29  Ira Weiny  <weiny2@llnl.gov>

	Update version in prep for next release

2011-09-29  Albert Chu  <achu@hyperion1.llnl.gov>

	Support ibccquery, congestion control query tool.

2011-09-29  Marcus R Epperson  <marcus.r.epperson@gmail.com>

	infiniband-diags/src/ibqueryerrors.c: don't increment bad_ports for port ALL

	infiniband-diags/src/ibqueryerrors.c: fix segv due to CA port 0 reference

2011-09-26  Alex Netes  <alexne@mellanox.com>

	infiniband-diags/scripts/ibcheckport.in: ignore lid/sm lid checks on non zero port switches

2011-09-23  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors add query failure count
	Report query failure count as part of summary

	infiniband-diags: ibnetdisc Wrap header for use in C++ code

2011-09-23  Hefty, Sean  <sean.hefty@intel.com>

	RE: [PATCH v2] ib-diags: Add cast to fix build on windows

2011-09-01  Ira Weiny  <weiny2@llnl.gov>

	Update AUTHORS and make README release notes.
	Increment version

	In addition move detailed text from README to the new infiniband-diags man page.

2011-09-01  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: Fix cap_mask extended link speed supported endian checks
	Similar to previous patch for ibnetdiscover.c

	infiniband-diags/ibqueryerrors.c: Add support for FDR10

2011-09-01  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/iblinkinfo.c: Add support for FDR10
	    Signed-off-by: Hal Rosenstock <hal@mellanox.com>
	    Signed-off-by: Ira Weiny <weiny2@llnl.gov>

	infiniband-diags: Add support for FDR10 to ibnetdiscover
	    Signed-off-by: Hal Rosenstock <hal@mellanox.com>
	    Signed-off-by: Ira Weiny <weiny2@llnl.gov>

	infiniband-diags: Add FDR10 support to libibnetdisc
	    Signed-off-by: Hal Rosenstock <hal@mellanox.com>
	    Signed-off-by: Ira Weiny <weiny2@llnl.gov>

2011-09-01  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibstat.c: Add support for FDR10
	without changing libibumad umad_port_t structure

	infiniband-diags/ibnetdiscover.c: Fix cap_mask endian for extended link speeds

	infiniband-diags/ibportstate: Add support for FDR10

2011-08-31  Albert Chu  <chu11@llnl.gov>

	fix manpage typos

2011-08-24  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors Clean up exit codes
	Update man page for exit status

2011-08-23  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add human readable output for data counters
	conversion function is provided in ibdiag_common

2011-08-22  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: move get_msg to common code
	iblinkinfo and ibqueryerrors should share this.

2011-08-22  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/man/ibportstate.8: Update for fdr

2011-08-17  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibnetdiscover.c: Changes for extended link speeds

	infiniband-diags/libibnetdisc/ibnetdisc.c: Changes for extended link speeds

	infiniband-diags/ibqueryerrors.c: Changes for extended link speeds

	infiniband-diags/iblinkinfo.c: Changes for extended link speeds

	infiniband-diags/ibportstate.c: Changes for extended link speeds

	infiniband-diags/perfquery: Add support for extended link speeds
	PortExtendedSpeedsCounters attribute

	infiniband-diags/saquery: Add extended link speed support for PortInfoRecord queries

	infiniband-diags/smpquery: Add support for extended link speeds

2011-08-06  shefty  <sean.hefty@intel.com>

	ibnetdisc: Fix build for windows
	Fix a couple of compile errors that occur building on windows.

2011-08-06  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: make library dependancies explicit

2011-08-03  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: update check for latest dump function in libibmad

	Merge branch 'master' of git://git.openfabrics.org/~iraweiny/infiniband-diags

2011-08-03  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/smpquery: Add Mellanox extended portinfo query

	infiniband-diags/libibnetdisc/ibnetdisc.c: Remove compile warning
	src/ibnetdisc.c: In function 'ibnd_find_port_dr':
	src/ibnetdisc.c:676: warning: 'rc' may be used uninitialized in this function

2011-08-03  Albert L.Chu  <chu11@llnl.gov>

	perfquery: support many optional performance counters, including congestion control counters

2011-08-03  Ira Weiny  <weiny2@llnl.gov>

	Merge branch 'master' of git://git.openfabrics.org/~iraweiny/infiniband-diags

2011-08-03  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/saquery.c: PortInfoRecord now has an options field
	rather than reserved field

	This is for coordination with the latest ib_types.h

2011-08-03  Ira Weiny  <weiny2@llnl.gov>

	Merge branch 'conv-iblinkinfo-portguid'

2011-07-21  Albert L.Chu  <chu11@llnl.gov>

	perfquery: support many optional performance counters, including congestion control counters

2011-07-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors fix null threshold_str
	When printing "(null)" was being printed if there was no threshold file

2011-07-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: convert ib[linkinfo|queryerrors] to use PortGUIDs
	NodeGUID's are no longer output nor accepted as input.

	Furthermore expand search to cover all node types:

	iblinkinfo CA and router nodes are output in a slightly different manner due to
	each port containing a unique PortGUID.  "-l" output is unaffected.

	ibqueryerrors correctly reports PortGUID for each port printed.

2011-07-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: libibnetdisc Allow a DR Path partial fabric query starting at a CA
	The old algorithm would attempt to extend a DR path through the CA and fail.
	The new algorithm retracts the path properly.

	This does not work when combined routing is used with the CA as a starting
	point.  :-(

2011-07-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: libibnetdisc add port interface
	Add ability to find port's based on PortGUID and DRPath

2011-07-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: move core saquery functionality to ibdiag_common
	While this is sub-optimal it is the easiest way to get saquery support to the
	other diags.

	infiniband-diags: saquery; remove "result" global
	make struct sa_query_result a paramenter of sa_query.

2011-07-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: change perl install to vendorlib

	infiniband-diags: configure.in; check for more recent function in libibmad

	infiniband-diags: Add sysconfdir to spec for config files
	commit cbe10407a14e57c8febe6e039ae617737f0b6e0f added a configuration file
	directory in sysconfdir.  I missed this in the spec file.

2011-07-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: Initial changes for extended speed support in libibmad
	Need larger buffer for mad_dump_portinfo due to additional extended
	speed fields

	Signed-off-by: Hal Rosenstock

2011-07-06  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Remove reference to OpenSM from COPYING

2011-06-25  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: add package requires to spec file.

	infiniband-diags: test for libosmcomp with a function that is used in the code

2011-06-25  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibnetdiscover.c: Fix typo in help

2011-06-16  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibqueryerrors.c: Fix typo in help

2011-06-10  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/libibnetdisc/ibnetdisc_cache.c: Fix a couple of memory leaks

	infiniband-diags/libibnetdisc/ibnetdisc.c: Fix possible overrun of path
	when path->cnt is 63 in add_port_to_dpath

	infiniband-diags/ibportstate.c: Fix sizeof
	It shouldn't be sizeof the pointer but this doesn't affect execution
	since this parameter is not used by mad_dump_portstates

2011-05-28  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: perfquery; Support PortCountersExtendedNoIETF
	Changes since V1:
	fix typo
	   if (cap_mask & IB_PM_EXT_WIDTH_NOIETF_SUP) {
	to
		if (cap_mask & IB_PM_EXT_WIDTH_SUPPORTED) {

	infiniband-diags: perfquery; print cap_mask for information
	Also fix cap_mask endianess in error print

2011-05-28  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/saquery.c: Fix SA_ERR_UNKNOWN define

2011-05-26  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/configure.in: Support override for IBPATH
	Basing IBPATH on sbindir is not always what is desired.
	Allow for override of IBPATH, when this is not the case.

2011-05-24  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/libibnetdisc/ibnetdisc.c: Better error message
	include node GUID and port number

2011-05-02  Eli Dorfman  <elido@dev.mellanox.co.il>

	infiniband-diags/vendstat: Fixed general info query
	Fixed general info query

	There was no output when the query succeeded because the if statement was
	incomplete

2011-05-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors; add --counters option
	Display the data counters (optionally extended data counters) for all ports in the fabric.

	Chages in V2:
	Only print Unicast/Multicast fields when IsExtendedWidthSupported is indicated.

2011-05-02  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: ibqueryerrors --data; show/clear PortCountersExtended when supported
	Chages in V2:
	Only print/clear Unicast/Multicast fields when IsExtendedWidthSupported is
	indicated.

	infiniband-diags: ibqueryerrors; use #defines for CapabilityMask

	infiniband-diags: ibqueryerrors; add ibcheckerrors like summary
	ibcheckerrors reports a summary of nodes checked, node errors, port checked, and
	port errors.  Track and print same as a summary.

	infiniband-diags: ibqueryerrors; add error thresholds

2011-04-28  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/perfquery.c: Use ib_types.h defines for CapabilityMask
	rather than hard coded values now that these have been added to ib_types.h

2011-04-27  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/ibnetdiscover.c: Eliminate blank line when sysimgguid is 0

	infiniband-diags/libibnetdisc/ibnetdisc.c: Handle calloc failure
	in recv_node_info

2011-04-27  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: fix "dist" target and subsequent rpm building

2011-04-26  Marcus Epperson  <mrepper@sandia.gov>

	infiniband-diags/ibnetdiscover: Make --ports option use remap_node_name()

2011-04-26  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: improve backwards/standalone compatibility
	specifically allow for umad_port_t to not have link_layer
	Furthermore "mad.h" should not have libibnetdisc OS specific defines.

	infiniband-diags: iblinkinfo; remove non-supported -c option from man page

2011-04-13  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: saquery, clean up error code/return types

	infiniband-diags: saquery, properly free mad in get_and_dump_any_records

2011-04-11  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags: Make sure buffers are set to zero before calling xxx_query_via
	where xxx is smp and pma. This is a cosmetic change based on
	IBA 1.2.1 vol 1 p.725 line 9:

	When constructing a response MAD that contains all or part of
	the corresponding request MAD, it is acceptable to include the
	contents of reserved fields in the request MAD in the response
	MAD without regard to their content. That is, such fields need not
	be set to zero in the response MAD.

2011-04-08  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags.spec.in: Remove duplication of ibdiscover.pl

2011-03-28  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: saquery; clean up name/lid processing
	For NodeRecord querys in particular if a name was not found all NodeRecords
	were printed.  Correct behavior is to report "node not found".

2011-03-28  Hal Rosenstock  <hal@mellanox.com>

	infiniband-diags/saquery: Add GUIDInfoRecord support

2011-03-25  Jason Gunthorpe  <jgunthorpe@obsidianresearch.com>

	Fix ibdatacounters
	Commit fb3b1e753d527faa08f863a55f7f8392fb01f2ed switched the call
	to ibdatacounts with ibcheckerrs which breaks the point of the script,
	put it back.

	Fix missing $IBPATH in ibdatacounts
	Otherwise it blows up if smpquery is not in the search path.

2011-02-17  Tamir Ronen  <tamirr@voltaire.com>

	add support for ibnetdiscover grouping for Mellanox switches
	    Group Mellanox switches to the same chassis using SYSTEM_GUID
	    Use the node description format:
	    'MF0;<system name>:<system type>/<system slot name>[:board type]/U<node index>'
	    to identify system type, slot type Line/Spine, slotnum and anafanum.
	    Map Line ports 1-18 as external ports 1-18.

2011-02-15  Ira Weiny  <weiny2@llnl.gov>

	Merge branch 'master' of git://git.openfabrics.org/~alexnetes/infiniband-diags

2011-02-15  Albert Chu  <chu11@llnl.gov>

	support --filterdownports in iblinkinfo

	Add lid and node description diff options for --diffcheck in iblinkinfo

	support --diffcheck in iblinkinfo

	add --diff support to iblinkinfo

2011-02-15  Ira Weiny  <weiny2@llnl.gov>

	ibnetdisc.c: Fix partial scan of the fabric.
	   When max_hops was specified it did not result in a correct limited scan of
	   the fabric.
	   In addition the -S and -D options of iblinkinfo/ibqueryerrors should by
	   default only scan 1 hop from the node specified to allow for faster
	   processing.

2011-02-14  Jason Gunthorpe  <jgunthorpe@obsidianresearch.com>

	The shell scripts use bashism, so use bash directly.
	Otherwise it blows up on distributions that don't use bash as /bin/sh,
	like Debian and Ubuntu. One could probably just remove the bashisms, but
	that is much harder.

	An example error is:
	ibswitches: 5: Syntax error: "(" unexpected

2011-02-14  Alex Netes  <alexne@voltaire.com>

	infiniband-diags: package versions update
	infiniband-diags package versions update.

	infiniband-diags: update shared library versions
	Update shared library version in accordance with recent changes.

2011-02-03  Alex Netes  <alexne@voltaire.com>

	Makefile: ChangeLog and version generation script path fix
	Move ChangeLog and version generation script into top source directory and fix
	path in Makefile accordingly.

2011-01-02  Ira Weiny  <weiny2@llnl.gov>

	Further timeout paramater verification (Was: [PATCH] infiniband-diags: Verify timeout value specified to diagnostics)
	On Mon, 20 Dec 2010 13:57:36 -0800
	Jay Fenlason <fenlason@redhat.com> wrote:

	> Sorry if you get this more than once, I seem to have outgoing mail misconfiguration here.
	>
	> On Wed, Dec 15, 2010 at 11:47:09AM -0800, Ira Weiny wrote:
	> >
	> > Verify timeout value specified to diagnostics
	> >
	> >
	> > Signed-off-by: Ira Weiny <weiny2@hype137.llnl.gov>
	> > ---
	> >  infiniband-diags/src/ibdiag_common.c |   10 +++++++---
	> >  1 files changed, 7 insertions(+), 3 deletions(-)
	> >
	> > diff --git a/infiniband-diags/src/ibdiag_common.c b/infiniband-diags/src/ibdia
	> +g_common.c
	> > index 99861f1..8ccf2fc 100644
	> > --- a/infiniband-diags/src/ibdiag_common.c
	> > +++ b/infiniband-diags/src/ibdiag_common.c
	> > @@ -175,9 +175,13 @@ static int process_opt(int ch, char *optarg)
	> >               ibd_dest_type = IB_DEST_GUID;
	> >               break;
	> >       case 't':
	> > -             val = strtoul(optarg, 0, 0);
	> > -             madrpc_set_timeout(val);
	> > -             ibd_timeout = val;
	> > +             val = (int)strtol(optarg, NULL, 0);
	> > +             if (val > 0) {
	> > +                     madrpc_set_timeout(val);
	> > +                     ibd_timeout = val;
	> > +             } else
	> > +                     IBERROR("Invalid timeout \"%s\".  Timeout requires a "
	> > +                             "positive integer value.", optarg);
	> >               break;
	> >       case 's':
	> >               /* srcport is not required when resolving via IB_DEST_LID */
	> > --
	> > 1.5.4.5
	>
	> This only partially detects invalid timeouts.  For example, timeouts
	> of "123skidoo" or 1234534587347895789457897897894578978912902393 will
	> be accepted but will not work as expected.  To fully test the
	> conversion, you need to do something more like:
	>
	>   long val;
	>   char *endp;
	>
	>   ...
	>   errno = 0;
	>   val = strtol ( optarg, &endp 0 );
	>   if ( errno || ( endp && *endp != '\0' ) || val < 0 || val > INT_MAX ) {
	>         /* invalid timeout detected */
	>   ...
	>
	> All of the tests are required to detect invalid inputs.
	>
	> 		-- JF

	Patch below,
	Ira

	From: Ira Weiny <weiny2@hype137.llnl.gov>
	Date: Wed, 22 Dec 2010 11:40:33 -0800
	Subject: [PATCH] Further timeout paramater verification

	suggested by Jay Fenlason <fenlason@redhat.com>

2010-12-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Verify timeout value specified to diagnostics
	Verify timeout value specified to diagnostics

2010-11-30  Ira Weiny  <weiny2@llnl.gov>

	libibmad/fields.c: Change all PortCounter names to match the Specification
	Change from V1:
	   Add name change to scripts in infiniband-diags

2010-11-30  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/iblinkinfo.c: Limit some queries to switches
	infiniband-diags/iblinkinfo.c: Limit certain queries to switches

	iblinkinfo man page says:
	DESCRIPTION
	        iblinkinfo reports the link info for each port of each switch
	active in
	        the IB fabric.

	OPTIONS
	        -S <guid>
	               Output only the switch specified by <guid> (hex format)

	        -D <direct_route>
	               Output only the switch specified by the direct route path.

	Queries by DR path or GUID should (also) be limited to switches.

2010-11-30  Hal Rosenstock  <hal@dev.mellanox.co.il>

	infiniband-diags/saquery.c: In dump_one_mcmember_record, fix flow label endian
	Already in host endian

2010-11-30  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibtracert: Eliminate direct route (-D) option
	Direct route does not make sense as an ibtracert option so eliminate it.

2010-10-25  Yevgeny Kliteynik  <kliteyn@mellanox.co.il>

	ibstat.c: fix core dump if wrong port phys state is reported

2010-10-25  Jason Gunthorpe  <jgunthorpe@obsidianresearch.com>

	Fix autotools to include the necessary M4 files
	Otherwise running autogen.sh with a new version of autotools and then
	building on a system with an older version tends to explode.
	Unfortunately this is sometimes necessary since the new version is
	required by the package.

	This is how GNU envisions this mess works at least..

2010-09-06  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	inifiband-diags: Support Voltaire switch ISR4200
	Support Voltaire switch (ISR4200) grouping.

	infiniband-diags: Do not exit when unexpected node found
	Show error message but do not exit when unexpected node is found.

2010-09-04  Sasha Khapyorsky  <sashak@voltaire.com>

	management: packages versions update
	Management packages versions update.

	management: update shared library versions
	Update shared library version in accordance with recent changes.

2010-09-01  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibnetdiscover: Fix handling of CA ports in recv_port_info
	When multiple ports on the same CA are connected to the same subnet and
	an ibnetdiscover is initiated from one of them, the discovery continues
	past the other CA port and an error occurs.

	The error is:
	src/query_smp.c:188; umad (DR path slid 0; dlid 0; 0,1,20,2 Attr0x11:0) bad status 110; Connection timed out

	Fix this by saving the initiating port number in the fabric structure
	and using that rather than local_port for the comparison in
	recv_port_info.

2010-09-01  Doron Shoham  <dorons@voltaire.com>

	ibnetdiscover: add '-f' flag to show full information (ports' speed and width).
	add '-f' flag to show full information (ports' speed and witdh).
	mainly to work with ibsim (using links real speed and width).

2010-08-01  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibstat: convert to PRIx64 macros GUID printing
	Convert to using PRIx64 macros for printing GUID values.

	Pointed out by Irena Kruchkovsky.

2010-07-26  Hefty, Sean  <sean.hefty@intel.com>

	libibnetdisc: fix cast in unmarshall16
	Uri Habusha reported a build error on windows as a result of an
	incorrect cast to uint64_t.

2010-06-17  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery.8: Add some missing counters to description
	Also, updated email address

2010-06-14  Ira Weiny  <weiny2@hera2.llnl.gov>

	ibqueryerrors.c: Optimize by querying AllPortSelect first
		If errors are seen with AllPortSelect query individual ports for more
		details.

2010-06-14  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	fix MFT record dump
	fix wrong MLID calculation in MFT record dump.

2010-06-10  Al Chu  <chu11@llnl.gov>

	support diffing lids and nodedesc on remoteports in ibnetdiscover
	Hey Sasha,

	A slight tweak to the patch.  Support diffing lids and node descriptions
	on remote ports (previously it diffed only "local" lids and node
	descriptions).  Also add appropriate manpage notes.

	Al

	On Tue, 2010-04-20 at 15:30 -0700, Al Chu wrote:
	> Hey Sasha,
	>
	> This patch supports diffing node descriptions on remote ports
	> (previously diffing of just the "local" node description was supported).
	>
	> Al
	>
	> email message attachment
	> > -------- Forwarded Message --------
	> > From: Albert Chu <chu11@llnl.gov>
	> > Subject: [PATCH] support diffing nodedesc on remoteports in
	> > ibnetdiscover
	> > Date: Tue, 20 Apr 2010 15:09:59 -0700
	> >
	> > Signed-off-by: Albert Chu <chu11@llnl.gov>
	> > ---
	> >  infiniband-diags/src/ibnetdiscover.c |   11 +++++++++++
	> >  1 files changed, 11 insertions(+), 0 deletions(-)
	> >
	> > diff --git a/infiniband-diags/src/ibnetdiscover.c b/infiniband-diags/src/ibnetdiscover.c
	> > index 57f9625..eeb1b9f 100644
	> > --- a/infiniband-diags/src/ibnetdiscover.c
	> > +++ b/infiniband-diags/src/ibnetdiscover.c
	> > @@ -720,6 +720,17 @@ static void diff_ports(ibnd_node_t * fabric1_node, ibnd_node_t * fabric2_node,
	> >  			fabric2_out++;
	> >  		}
	> >
	> > +		if (data->diff_flags & DIFF_FLAG_PORT_CONNECTION
	> > +		    && data->diff_flags & DIFF_FLAG_NODE_DESCRIPTION
	> > +		    && fabric1_port && fabric2_port
	> > +		    && fabric1_port->remoteport && fabric2_port->remoteport
	> > +		    && memcmp(fabric1_port->remoteport->node->nodedesc,
	> > +			      fabric2_port->remoteport->node->nodedesc,
	> > +			      IB_SMP_DATA_SIZE)) {
	> > +			fabric1_out++;
	> > +			fabric2_out++;
	> > +		}
	> > +
	> >  		if (fabric1_out) {
	> >  			diff_iter_out_header(fabric1_node, data,
	> >  					     out_header_flag);
	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 20 Apr 2010 15:09:59 -0700
	Subject: [PATCH] support diffing lids and nodedesc on remoteports in
		ibnetdiscover
	Message-Id: <1271873984.17987.245.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-06-10  Al Chu  <chu11@llnl.gov>

	support diffing nodedesc on remoteports in ibnetdiscover
	Hey Sasha,

	This patch supports diffing node descriptions on remote ports
	(previously diffing of just the "local" node description was supported).

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 20 Apr 2010 15:09:59 -0700
	Subject: [PATCH] support diffing nodedesc on remoteports in ibnetdiscover
	Message-Id: <1271802596.17987.229.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-06-10  Al Chu  <chu11@llnl.gov>

	add ibcacheedit tool
	Hey Sasha,

	This patch adds a new ibcacheedit tool to infiniband-diags.  As the name
	suggests, it offers users options to edit a stored cache.  This tool is
	primarily necessary to allow system administrators to update caches as
	they make minor modifications to a cluster (e.g. node dies, thus HCA is
	replaced) rather than regularly regenerating a new cache.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 19 Apr 2010 11:16:51 -0700
	Subject: [PATCH] add ibcacheedit tool
	Message-Id: <1271702059.17987.209.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-06-10  Al Chu  <chu11@llnl.gov>

	support libibnetdisc caching no-overwrite flag
	Hey Sasha,

	This patch adds a flag to support a no-overwrite caching flag for
	libibnetdisc.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 19 Apr 2010 11:09:37 -0700
	Subject: [PATCH] support libibnetdisc caching no-overwrite flag
	Message-Id: <1271702059.17987.207.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-06-10  Al Chu  <chu11@llnl.gov>

	make libibnetdisc default behavior to overwrite a cache
	Hey Sasha,

	As discussed in a previous thread, this patch makes the default caching
	behavior in libibnetdisc to overwrite a previous cache.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 19 Apr 2010 10:52:14 -0700
	Subject: [PATCH] make libibnetdisc default behavior to overwrite a cache
	Message-Id: <1271702059.17987.211.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-06-09  Hefty, Sean  <sean.hefty@intel.com>

	ib/mgmt: fix build under windows stack
	Fix warnings in windows build.

	These are typecasts in several places, with a couple additional changes
	needed to libibnetdisc.  Windows uses ISO definitions instead of
	POSIX definitions, so calls like _read, _write, etc. are used on
	Windows instead of read, write.  The result is that internal functions
	are renamed with an ibnd prefix.

	In order to handle ibdebug exported from libibmad, Windows need to mark
	the variable as a 'dllexport' from libibmad, but as 'dllimport' for users.
	Since libibnetdisc wants to 'dllimport' ibdebug from libibmad, while
	'dllexporting its own functions, it requires a separate 'MAD_EXPORT' like
	definition to allow this.

2010-06-02  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibnetdiscover.c: Cosmetic changes
	Eliminate unneeded comment
	Also, cosmetic formatting change

2010-06-02  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/libibnetdisc/chassis.c: In group_nodes, set fabric->chassis
	when grouping by SystemImageGUID

	Without doing this, find_chassisguid will always return without
	finding the already discovered chassis so no grouping occurs
	when it should

2010-06-02  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/libibnetdisc/chassis.c: In group_nodes, pass on all nodes
	rather than just switches for common SystemImageGUID
	Now consistent with original grouping.c from which this was ported

2010-05-31  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: move ibmad_port out of smp_engine

2010-05-25  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/libibnetdisc/ibnetdisc.c: Fix compile warns on 32 bit archs
	src/ibnetdisc.c: In function `query_port_info':
	src/ibnetdisc.c:216: warning: long unsigned int format, different type arg (arg 5)
	src/ibnetdisc.c: In function `recv_node_info':
	src/ibnetdisc.c:311: warning: long unsigned int format, different type arg (arg 5)
	src/ibnetdisc.c:335: warning: long unsigned int format, different type arg (arg 6)

2010-05-25  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/vendstat: allow multiple config space records
	Allow to use -R and -W options (config space record access) multiple
	number of times, so that multiple config space records will be reading
	and/or writing. Using both -R and -W is allowed as well. Example:

	  vendstat -R 0x12345,0xffff -R 0x12346 -W 0x12347,0,0xfff <lid>

	Note that now number of records is effectively limited by 18 (so that it
	fit single MAD), this could be increased with using multiple MADs.

2010-05-25  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/vendstat: add config space access options
	Add Mellanox vendor class config space records access options (-R for
	read and -W for write). Example of usage:

	  vendstat -R 0x12345,0xffff <lid> - will read low 16 bits of config
	                                     space record at address 0x12345

	  vendstat -W 0x12345,0,0xfff <lid> - will zero low 12 bits of config
	                                      space record at address 0x12345

2010-05-25  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/vendstat: code simplifications
	Simplify and consolidate the code by using helper function do_vendor().

2010-05-22  Sasha Khapyorsky  <sashak@voltaire.com>

	management: packages versions update
	Management packages versions update.

	management: update shared library versions
	Update shared library version in accordance with recent changes.

2010-05-22  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	opensm: Always set enabled speed and width
	Initialize default link speed and width to "No State Change" and
	always set in SubnSet(PortInfo)

2010-05-22  Ira Weiny  <weiny2@llnl.gov>

	ibnetdisc: Separate calls to umad and mad layer to avoid race condition on response MAD's
		Specify CA/Port to use which allows parallel scanning to other operations.

2010-05-07  sebastien dugue  <sebastien.dugue@bull.net>

	libibnetdisc: Add grouping for Voltaire's ISR4700 switch
	  The ISR4700 features 3 kind of boards:

	  - sLB-4018 line board with a single 36 port asic
	  - sFB-4700 fabric board with a single 36 port asic
	  - sFB-4700X2 double density fabric board with 2 36 port asics

	  The double density fabric board (sFB-4700X2) features external 12X
	connectors that are only an aggregation of 3 4X ports, therefore
	ext_portnum is set to match the number printed on the faceplate.

2010-05-06  Mike Heinz  <michael.heinz@qlogic.com>

	RE: [PATCH] management: adding mad_dump_fields to libibmad
	Sasha asked that I re-submit the patches for perfquery in a slightly different format. This is the third of 3 patches.

	This patch corrects the AllPortSelect error message that is generated by ibcheckerrors when used against switches that do not support that attribute.

	Signed-off-by: Michael Heinz

	--------- snip -------

2010-05-06  Mike Heinz  <michael.heinz@qlogic.com>

	RE: [PATCH] management: adding mad_dump_fields to libibmad
	Sasha asked that I re-submit the patches for perfquery in a slightly different format. This is the second of 3 patches.

	This patch uses the new mad_dump_fields function to suppress the display of extended attributes when querying switches that do not support them.

	Signed off: Michael Heinz

	---------- snip ----------

2010-04-19  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: fix indentations
	Fix source indentations using opensm/ism_indent script.

2010-04-19  Al Chu  <chu11@llnl.gov>

	check for duplicate port guids in libibnetdisc cache
	Hey Sasha,

	This patch checks for duplicate port guids in a libibnetdisc cache when
	it is loaded and report an error back to the user appropriately.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 14 Apr 2010 15:25:11 -0700
	Subject: [PATCH] check for duplicate port guids in libibnetdisc cache
	Message-Id: <1271285070.17987.138.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-18  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc cache error path memleak
	Hey Sasha,

	This patch fixes a mem-leak through error paths in the libibnetdisc
	cache loading.  If some data had not yet been "copied over" to the
	fabric struct and an error occurred, that memory would be leaked.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 14 Apr 2010 14:11:27 -0700
	Subject: [PATCH] fix libibnetdisc cache error path memleak
	Message-Id: <1271284770.17987.131.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-18  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc portguid hashing corner case
	Hey Sasha,

	This patch fixes a corner case in libibnetdisc that was storing
	portguids w/ a guid of 0.

	This bug was relatively innoucuous for ibnetdiscover b/c ibnetdiscover
	does not output these ports.  However, it became a problem for me in the
	caching library as I attempted to reconstruct a fabric, and multiple
	ports were identifying themselves with identical guids [1].

	Al

	[1] - The fact the caching code assumes duplicate guids can't exist is
	also a bug.  But that's for another patch.  This is a bug by itself.

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 13 Apr 2010 14:01:43 -0700
	Subject: [PATCH] fix libibnetdisc portguid hashing corner case
	Message-Id: <1271264633.17987.112.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-18  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: remove not needed num_smps_outstanding counter
	Now 'num_smps_outstanding' represents number of MADs on a wire, which is
	duplicates by using its qmap's cl_qmap_count() function.

	libibnetdisc: remove not needed process_smp_queue() call
	process_smp_queue() is called initially in issue_smp() and post called
	in each process_one_recv() invocation. So no need in yet another call.

	libibnetdiscover: more outstanding MADs counting fix
	Decrease outstanding MADs counter before calling process_smp_queue()
	and smp receive callback, so that there will be room available for
	sending more MADs. This fixes 'ibnetdiscover -o 1' (max outstanding
	SMPs on a wire = 1) bug.

	libibnetdisc: fix memory leak in case of send_smps() failure
	Free smp object in case of send_smp() failure - it is not on the send
	list already and not tracked anymore.

2010-04-14  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: config struct for ibnd_discover_fabric()
	Instead of using static configuration parameters/flags and to have API
	functions to set up each one (and so add new API when any new flag is
	added) use just single configuration structure which is reenterently
	passed by to ibnd_fabric_discover() function.

	In order to not change API a lot we will keep some "padding" space in
	this configuration structure, so it will be extendable for future
	parameters/flags.

	Remove old config related functions.

2010-04-14  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: don't query CA ports not connected to a fabric
	We can save some amount of MADs by not querying CA/Router ports which
	is not connected to our fabric. When discovery reaches CA or Router
	node it will always get PortInfo for a port which was discovered and
	not others.

2010-04-13  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: don't try to cross discovery over CA
	When discovery is running from CA node it shouldn't try to cross over
	all ports, but only via local one (send over non-local ports will fail
	since CA doesn't route MADs).

	libibnetdisc: fix outstanding SMPs countung
	send_smp() failure blocks whole discovery execution (it never returns).
	Fix this by updating all outstanding SMPs related counters only after
	successful MAD sending.

	libibnetdisc: add \n at end of error messages
	Add '\n' new line character at end of error messages printouts.

	infiniband-diags/ibnetdiscover: cleanups
	Merge some flows. Fix indentation using osm_indent script.

2010-04-13  Al Chu  <chu11@llnl.gov>

	support lid and nodedesc diffchecks in ibnetdiscover
	Hi Sasha,

	This patch adds lid and node description diff options for --diffcheck in
	ibnetdiscover.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 31 Mar 2010 11:18:43 -0700
	Subject: [PATCH] support lid and nodedesc diffchecks in ibnetdiscover
	Message-Id: <1270659341.26381.31.camel@crazyclimber.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-13  Al Chu  <chu11@llnl.gov>

	support --diffcheck in ibnetdiscover
	Hi Sasha,

	This patch adds basic --diffcheck support in ibnetdiscover, allowing
	configuration of the diff checks done in the default --diff option.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 31 Mar 2010 11:13:37 -0700
	Subject: [PATCH] support --diffcheck in ibnetdiscover
	Message-Id: <1270659341.26381.29.camel@crazyclimber.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-13  Al Chu  <chu11@llnl.gov>

	support --diff in ibnetdiscover
	Hi Sasha,

	This patch adds the default --diff support in ibnetdiscover.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 31 Mar 2010 11:10:34 -0700
	Subject: [PATCH] support --diff in ibnetdiscover
	Message-Id: <1270659341.26381.27.camel@crazyclimber.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

	Hey Sasha,

	The following sets of patches implement a --diff and --diffcheck options
	in ibnetdiscover to let users diff an ibnetdiscover state to a previous
	ibnetdiscover state.  The goal of this option is to help system
	administrators isolate/determine changes in the network quickly compared
	to a previous state.  Here's an example:

	vendid=0x8f1
	devid=0x5a30
	sysimgguid=0x8f10400411f57
	switchguid=0x8f10400411f56(8f10400411f56)
	Switch  24 "S-0008f10400411f56"         # "ISR9024D Voltaire" base port 0 lid 11 lmc 0
	< [14]  "H-0002c90200219ef0"[1](2c90200219ef1)          # "wopr0" lid 64 4xDDR
	< [19]  "H-0002c9030000ff7c"[1](2c9030000ff7d)          # "wopr9" lid 48 4xDDR
	> [20]  "H-0002c9030000ff7c"[1](2c9030000ff7d)          # "wopr9" lid 4 4xDDR

	< vendid=0x2c9
	< devid=0x6282
	< sysimgguid=0x2c90200219ef3
	< caguid=0x2c90200219ef0
	< Ca    2 "H-0002c90200219ef0"          # "wopr0"
	< [1](2c90200219ef1)    "S-0008f10400411f56"[14]                # lid 64 lmc 2 "ISR9024D Voltaire" lid 11 4xDDR

	In this particular example, port 14 on the switch (which is connected to
	node 'wopr0') was up before but is now down (and the associated CA is
	noted too).  In addition, 'wopr9' is connected to port 20 instead of
	port 19 on the switch.

	By default --diff checks switches, cas, routers, and port connections.
	The --diffcheck option allows the user to specify which diff options
	they want done, and also adds other diff checks for lids and/or node
	descriptions.  More diff checks could be added later as needed.  For
	example, the following only checks for differences of lids on switches.

	vendid=0x8f1
	devid=0x5a30
	sysimgguid=0x8f10400411f57
	switchguid=0x8f10400411f56(8f10400411f56)
	< Switch        24 "S-0008f10400411f56"         # "ISR9024D Voltaire" base port 0 lid 11 lmc 0
	> Switch        24 "S-0008f10400411f56"         # "ISR9024D Voltaire" base port 0 lid 3 lmc 0
	< [13]  "H-0002c90200219e64"[1](2c90200219e65)          # "wopri" lid 4 4xDDR
	> [13]  "H-0002c90200219e64"[1](2c90200219e65)          # "wopri" lid 1 4xDDR

	Others on the list may wonder how this is different than just using the
	normal 'diff' tool.  The differences I can think of are:

	1) This checks differences in the network, not text.  This is
	particularly important when lids, lmc, etc. are changed.  Otherwise
	there are many differences in a normal diff output that aren't
	necessary.

	2) This provides the appropriate "context" in the diff output, showing
	the appropriate system ids to allow a system administrator to identify
	ports on what switch have changed.  Under normal diff output, you may
	not get that appropriate context of information.  The system
	administrator can of course use options like --context in diff, but the
	goal is to make the diff output clear and concise, not outputting
	unnecessary junk.

	3) As parallelization has been added into ibnetdisocver/libibnetdiscover
	this becomes more critical as output in ibnetdiscover/libibnetdiscover
	can be re-ordered.  So a normal diff suddenly is non-functional.

	There's probably other minor advantages.  Even if minor output tweaks
	happen to ibnetdiscover in the future, this can still work against old
	cache files.

	Al

2010-04-10  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc corner case when user tries to re-cache a loaded cache
	Hey Sasha,

	Noticed this corner case.  If I load a ibnetdiscover cache, then try to
	re-cache it, the cache doesn't store things properly b/c I didn't
	reconstruct one fabric data structure properly.  This fixes that bug.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Fri, 9 Apr 2010 10:12:47 -0700
	Subject: [PATCH] fix libibnetdisc corner case when user tries to re-cache a
		loaded cache
	Message-Id: <1270833600.17987.15.camel@auk31.llnl.gov>
	Mime-Version: 1.0
	Content-Transfer-Encoding: 7bit

2010-04-10  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: no backward NodeInfo queries
	Then switch is reached via port N we don't need to query back via this
	port - source node is discovered already. Finally this saves some amount
	of unnecessary MADs.

	infiniband-diags/ibnetdiscover: report number of used MADs
	When '--max_hops,-m' option is specified report (in comment) a total
	number of used MADs for discovery aside of number of hops.

	libibnetdisc: keep total number of used SMPs
	Keep and track total number of used SMPs in smp engine structure
	(total_smps field) and for statistic purpose return it with fabric
	object (total_mads_used field).

	libibnetdisc: restore hops limited discovery
	Restore hops limited discovery.

	libibnetdisc: restore show_progress functionality
	Restore show_progress functionality. This output is used by scripts like
	'dump_lfts.sh -D', also sometimes useful for debug.

2010-04-10  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: code cleanups
	Make newly introduced local functions to be static. Remove not needed
	braces. Simplify some expressions.

	Fix indentations using osm_indent script.

2010-04-08  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: add ibnd_set_max_smps_on_wire call
	        In addition update ibnetdiscover, iblinkinfo, and ibqueryerrors to use this function.

2010-04-08  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: Convert to a multi-smp algorithm
	v3: change DEFAULT_MAX_SMP_ON_WIRE to 2

		Allow for multiple SMP's to be on the wire at a single time.  This
		algorithm splits the processing of SMP's to a small smp engine which
		may be useful to split out in the future.

2010-03-23  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	infiniband-diags/ibstat.c: print link layer for RoCEE support

	infiniband-diags/ibstatus: add link_layer for RoCEE support
	RoCEE introduces new file in sysfs: "link_layer".
	Assume IB if the file doesn't exist (driver w/o RoCEE support).

2010-03-23  Mike Heinz  <michael.heinz@qlogic.com>

	Perfquery can be too noisy.
	When perfquery is run against fabrics that do not support PortXmitWait, it emits this warning for every port:

	ibwarn: [23225] dump_perfcounters: PortXmitWait not indicated so ignore this counter

	When running ibcheckerrors on a large fabric, this leads to a flood of warnings.

	The proposed patch reduces the warning to a verbose message.

2010-03-23  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Ignore PortInfo data on down port.
	According to C14-24.2.1:
	If PortInfo:PortState == Down then only PortInfo:PortState and
	PortInfo:PortPhysicalState _must_ be valid.  Other fields may be invalid
	depending on the vendor.  Therefore ignore all PortInfo data other than those
	fields when reporting PortInfo on a down port.

2010-02-21  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: fix typo in man page for ibnd_show_progress

2010-02-11  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/vendstat: Indicate general info for IS4 as well as IS3
	Update both tool and man page for this
	Also, updated email address

2010-01-25  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibqueryerrors.c: Fix typo in counter name

2010-01-19  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/iblinkinfo,ibqueryerrors: code refactoring
	Some code flow refactoring and indentation fixes (with osm_indent).

2010-01-19  Al Chu  <chu11@llnl.gov>

	support --load-cache in iblinkinfo and ibqueryerrors
	Hey Sasha,

	Here's an updated patch with the cleanup changes as we discussed.

	Al

	On Sat, 2010-01-16 at 16:28 +0200, Sasha Khapyorsky wrote:
	> On 10:23 Fri 15 Jan     , Al Chu wrote:
	> > Hi Sasha,
	> >
	> > This adds the --load-cache options to iblinkinfo and ibqueryerrors.
	> >
	> > Al

2010-01-16  Sasha Khapyorsky  <sashak@voltaire.com>

	ibqueryerrors.c: more readable details dump function
	Make details query and dump function to be more readable - eliminate
	repeated conditional flows by using function parameters.

2010-01-16  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibqueryerrors: Add support for optional PortRcvErrorDetails counter
	to --details option

2010-01-16  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: indentation fixes
	Indentation fixes using opensm/opensm/osm_indent script.

2010-01-16  Al Chu  <chu11@llnl.gov>

	support --cache and --load-cache options in ibnetdiscover
	Hi Sasha,

	This adds the --cache and --load-cache options to ibnetdiscover.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 9 Dec 2009 15:19:47 -0800
	Subject: [PATCH] support --cache and --load-cache options in ibnetdiscover

2010-01-16  Al Chu  <chu11@llnl.gov>

	add libibnetdisc caching to libibnetdiscover
	Hi Sahsa,

	This patch adds caching functionality to libibnetdisc through the new
	functions ibnd_cache_fabric() and ibnd_load_fabric().

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 9 Dec 2009 15:19:24 -0800
	Subject: [PATCH] add libibnetdisc caching to libibnetdiscover

2010-01-16  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery: Add support for PortSamplesControl counter

	infiniband-diags/ibportstate.8: Add new supported ops
	Added by Ralph Campbell <ralph.campbell@qlogic.com>
	in commit 3af8e05e01a92bb504c8e806c0c81dc5ef246f7f

2010-01-12  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery: Add support for optional PortRcvErrorDetails counter

2010-01-12  Ralph Campbell  <ralph.campbell@qlogic.com>

	infiniband-diags/ibportstate: allow changes to CA portinfo parameters
	This patch adds new commands to ibportstate to support initializing
	the link for CAs connected back-to-back. It also allows more than
	one field to be changed at the same time such as "lid 23 arm" or
	"width 1 speed 3 enable".

2010-01-12  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: -V, --version: exit with zero status
	When -V, --version command line option is used exit with zero status
	after a version showing.

2009-12-29  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibqueryerrors.c: Fix output format of counters
	Should be unsigned decimal rather than plain decimal

2009-12-24  Sasha Khapyorsky  <sashak@voltaire.com>

	management: packages versions update
	Update OpenSM and infiniband-diags packages version (3.3.5 and .1.4.5
	respectively).

2009-12-13  Al Chu  <chu11@llnl.gov>

	correct manpages for tools that are no longer scripts
	Hey Sasha,

	Another set of typos I found.  Nothing fancy.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 10 Dec 2009 13:21:02 -0800
	Subject: [PATCH] correct manpages for tools that are no longer scripts

2009-12-13  Al Chu  <chu11@llnl.gov>

	remove obsoleted option from iblinkinfo manpage
	Hi Sasha,

	Nothing fancy, noticed the manpage still had this obsoleted option
	listed.  I copied the documentation that was put into the ibqueryerrors
	manpage.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	>From e0050396ddd4cbd611abef3ac74fca854c038b12 Mon Sep 17 00:00:00 2001
	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 10 Dec 2009 11:25:22 -0800
	Subject: [PATCH] remove obsoleted option from iblinkinfo manpage

2009-12-13  Sasha Khapyorsky  <sashak@voltaire.com>

	management: package versions bump
	Update management package versions:

		libibumad-1.3.4
		libibmad-1.3.4
		opensm-3.3.4
		infiniband-diags-1.5.4

2009-12-13  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update library versions
	Update library versions (version info strings) in accordance to changes
	were made from recent release.

2009-11-26  Al Chu  <chu11@llnl.gov>

	splice out chassis scan specific from general scan in libibnetdiscover
	Hey Sasha,

	With patch 2/3 the chassis code does not depend on the data structures
	used in the scan, so we can splice out the chassis specific chunks.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 16 Nov 2009 17:20:32 -0800
	Subject: [PATCH] splice out chassis scan specific from general scan

2009-11-26  Al Chu  <chu11@llnl.gov>

	use nodes/switches lists instead of nodesdist array in libibnetdiscover
	Hey Sasha,

	I noticed that that 'nodesdist' array was used for going through only
	switches or all nodes in the fabric.  Spoke to Ira about it, and he
	believes it's b/c in early ibnetdiscover code the 'nodesdist' array was
	the only data structure available for iterating over nodes.  Now we can
	use the 'nodes' list or 'switches' list, which makes it clearer to
	understand what's going on and is probably the "right" data structure to
	now use.  It also makes the chassis code section not dependent on the
	scan specific chunks of libibnetdiscover.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 16 Nov 2009 17:14:56 -0800
	Subject: [PATCH] use nodes/switches lists instead of nodesdist array

2009-11-26  Al Chu  <chu11@llnl.gov>

	remove unnecessary casting in libibnetdiscover
	Hey Sasha,

	I noticed a bunch of unnecessary casting.  The typing is done correctly
	throughout the code, so I assume the casting was done due to legacy
	reasons.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Mon, 16 Nov 2009 17:14:34 -0800
	Subject: [PATCH] remove unnecessary casting in libibnetdiscover

2009-11-26  Doron Shoham  <dorons@voltaire.com>

	Add different print macros
	Add different print macros for warning, verbose and
	debug messages.

2009-11-15  Al Chu  <chu11@llnl.gov>

	split out scan specific data from ibnd_node_t [attempt #2]
	This splits out some scan specific data from ibnd_node_t that doesn't
	need to be in the public struct.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From 75f819d436218be0c9bab7d0a6e763b4054748dd Mon Sep 17 00:00:00 2001
	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 29 Oct 2009 18:59:26 -0700
	Subject: [PATCH] split out scan specific data from ibnd_node_t

2009-11-13  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibqueryerrors: Properly exit and clean up resources when node info query fails

2009-11-13  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibqueryerrors: Change realloc of suppressed fields to a static array
	Sasha,

	2 clean up patches which apply in order.

	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Tue, 3 Nov 2009 13:19:33 -0800
	Subject: [PATCH] infiniband-diags/ibqueryerrors: Change realloc of suppressed fields to a static array

		Realloc size was wrong causing a core when enough errors were
		suppressed.  Reproduced by running:

		ibqueryerrors -c -s RcvSwRelayErrors,LinkDowned,VL15Dropped,XmtWait,SymbolErrors,LinkRecovers,RcvErrors

		Change this to a static array which also removes the need to free
		memory.

2009-11-13  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibqueryerrors: Fix remote node name printing
	"-r" option was reporting the current node name, not the remote node
		name of the port as it should have been.

2009-11-13  Al Chu  <chu11@llnl.gov>

	fix potential segfault in ibnd_node_t destroy path
	Hey Sasha,

	There is a potential for the ports array in ibnd_node_t to be NULL if
	you hit an error during scan and eventually destroy the fabric struct
	before returning to the user.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 5 Nov 2009 15:18:27 -0800
	Subject: [PATCH] fix potential segfault in ibnd_node_t destroy path

2009-11-13  Al Chu  <chu11@llnl.gov>

	consistently type lids and lmcs libibnetdisc
	Hey Sasha,

	The ibnd_port_t lid and lmc are typed uint16_t/uint8_t, but the
	ibnd_node_t smalid and smalmc are typed int/int.  This patch turns the
	ints into uint16_t/uint8_t.

	Naturally, we can do the opposite, but it should be consistent one way
	or the other.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 5 Nov 2009 14:58:11 -0800
	Subject: [PATCH] consistently type lids and lmcs libibnetdisc

2009-11-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/libibnetdisc: Remove nearly duplicate "get_node_info" function
	Fold remaining functionality into "query_port_info"

2009-11-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/libibnetdisc: clean up PortInfo queries
	Sasha,

	This and the subsequent patch clean up the port info query code to be more
	straight forward.

	First off I got rid of the "decode_port_info" function as it does very little
	and combined the smp_query_via call into a "query_port_info"

	The second patch removes the "get_port_info" function as it became nearly
	redundant with query_port_info.

	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Tue, 10 Nov 2009 11:22:03 -0800
	Subject: [PATCH] infiniband-diags/libibnetdisc: clean up PortInfo queries

		remove old "decode_port_info" and replace with new "query_port_info"

2009-11-12  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibqueryerrors: minor improvements
	Remove some not needed initializations, fix indentations with
	osm_indent.

2009-11-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibqueryerrors: Reformat the output of the xmtdiscard details.
	Old:
	Errors for 0x2c9020040fec8 "Infiniscale-IV Mellanox Technologies"
	   GUID 0x2c9020040fec8 port 27: [XmtDiscards == 8]
	   GUID 0x2c9020040fec8 port 27: [PortNeighborMTUDiscards == 8]

	New:
	Errors for 0x2c9020040fec8 "Infiniscale-IV Mellanox Technologies"
	   GUID 0x2c9020040fec8 port 27: [XmtDiscards == 8] [PortNeighborMTUDiscards == 8]

2009-11-12  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Change lids and port numbers to decimal
	rather than hex

2009-11-06  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibsendtrap.c: Add support for SysImageGuid changed trap (145)

2009-11-06  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: remove man page of removed API
	Remove man page for removed ibnd_update_node() API. Also removing its
	entry from library version script.

2009-11-06  Al Chu  <chu11@llnl.gov>

	remove ibnd_update_node
	Hey Sasha,

	After talking to Ira about it, we think it's best remove
	ibnd_update_node.

	A) It's not being used.
	B) It probably not implemented properly
	C) Some of Ira's original plans for its use require more API functions
	to be added, which of course isn't done right now.

	So for now, it seems best to just remove it since it's an additional
	API function that can lead to confusion.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 29 Oct 2009 18:56:32 -0700
	Subject: [PATCH] remove ibnd_update_node

2009-11-06  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibnetdisc.h: Clarify "ports" member of ibnd_node_t with more descriptive comment

2009-11-06  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibqueryerrors: Add support for PortXmitDiscardDetails
	When --details selected, if PortCounters has transmit discards, then
	query PortXmitDiscardDetails. On reset, if --details is selected, then
	PortXmitDiscardDetails will also be reset.

	Unfortunately, PortSamplesControl:OptionMask can't be trusted to determine
	which PortXmitDiscardDetails counters are supported so an option was added
	for this.

2009-11-03  Al Chu  <chu11@llnl.gov>

	remove unused headers in libibnetdiscover
	Hey Sasha,

	Just noticed it.  Left over/cut and pasted from the original
	ibnetdiscover tool code or no longer necessary due to recent changes.
	This patch is independent of the other patches I recently posted.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 3 Nov 2009 10:33:08 -0800
	Subject: [PATCH] remove unused headers

2009-11-02  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	infiniband-diags/saquery: Report SA MAD Class specific status.
	Report SA MAD Class specific status.

	In addition to SM status.

2009-11-02  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump package versions
	Bump IB management packages versions:

		libibumad-1.3.3
		libibmad-1.3.3
		opensm-3.3.3
		infiniband-diags-1.5.3

	Update mailing list name in configure.in files.

2009-11-02  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update libraries version information
	Update libraries version information. Since last management release
	APIs were changed in:

	- libibnetdisc
	- libibmad (backward compatible)
	- libibumad
	- libosmvendor

	And only internal changes in:

	- libosmcomp
	- libopensm

2009-11-02  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update/add Voltaire copyright
	Add or update Voltaire copyright for source files changed by Voltaire
	people in 2008-2009 years (no automatic or cosmetic changes).

2009-10-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perfquery: merge some code
	Merge some obviously repeated code in optional counters processing
	functions.

2009-10-30  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery: Add support for optional PortXmitDiscardDetails counter

2009-10-29  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: some cleanups
	Cleanups: kill not needed braces, castings, consolidate flows.

	libibnetdisc: rename ibnd_scan variable to scan
	Rename 'ibnd_scan' variable to shorter 'scan' to highlight its local
	feature.

2009-10-29  Al Chu  <chu11@llnl.gov>

	remove 'dist' field from ibnd_node_t, which was virtually not used
	Remove the 'dist' field from the ibnd_node_t struct and rearch code
	appropriately.  It ends up this field was only used to pass a value from
	create_node() to add_to_nodedist(), of which create_node() is the only
	function that calls add_to_nodedist().  In other words, it served pretty
	much no purpose.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 28 Oct 2009 16:18:39 -0700
	Subject: [PATCH] remove 'dist' field from ibnd_node_t, which was virtually not used

2009-10-29  Al Chu  <chu11@llnl.gov>

	split out ibnd_fabric_t fields that are only used during a scan
	Split out public parameters from ibnd_fabric_t that are useless b/c they
	are only used during the ibnetdiscover scan.

	Note that this patch has similarities to a previous patch from Ira,
	however it is separate and independent of that patch series.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 27 Oct 2009 16:16:14 -0700
	Subject: [PATCH] split out ibnd_fabric_t fields that are only used during a scan

2009-10-28  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/ibqueryerrors: Use remap'ed node name in clear port error message

2009-10-24  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibqueryerrors.c: Cosmetic changes
	Eliminated redundant "\n" in IBWARNs
	Fixed commentary typo

2009-10-23  Keshetti Mahesh  <keshetti.mahesh@gmail.com>

	infiniband-diags/libibnetdisc: Removed trivial comments
	infiniband-diags/libibnetdisc/: Removed trivial comments.

2009-10-23  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibqueryerrors.c: use common node_name
	Get common node_name string using remap_node_name() and use it over
	reporting functions - in this way simplify the code, avoid multiple
	remap_node searches and fix minor memory leaks.

	infiniband-diags/ibqueryerrors.c: use common portid in query_cap_mask()
	Use common portid in query_cap_mask() function instead of generating
	this again in this function.

	infiniband-diags: minor cosmetic fixes
	Couple of minor and cosmetic fixes in ibqueryerrors.c and iblinkinfo.c:
	- simplify get_max() function
	- kill unneeded braces ( {} and () )
	- indentation fixes (with osm_indent)

2009-10-23  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibqueryerrors: Add clear errors and counters options
		V2 add check for XMT_WAIT support on clear

		Add -k and -K options to clear errors and counters.  If both are
		specified they will both be cleared.

		Update man page

		In addition fix 2 bugs
		fix the printing of Xmt Wait errors
		properly skip the counter select field.

2009-10-22  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/libibnetdisc/ibnetdisc.c: fix bug in single node processing.
		Eli fixed an issue with running ibnetdiscover from a switch but it
		introduced a bug in processing a single switch:

	17:19:42 > ./iblinkinfo -S 0x000b8cffff00490c
	Switch 0x000b8cffff00490c MT47396 Infiniscale-III Mellanox Technologies:
	...
	           8   11[  ] ==( 4X 2.5 Gbps   Down/ Polling)==>             [  ] "" ( )
	           8   12[  ] ==( 4X 5.0 Gbps Active/  LinkUp)==>             [  ] "" ( )
	           8   13[  ] ==( 4X 2.5 Gbps   Down/ Polling)==>             [  ] "" ( )
	...

		The port we "come in on" when discovering the switch is not reported properly.

		This patch reverses Eli's patch and fixes his original bug in a way
		which does not introduce the above issue.

2009-10-13  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Modify dump_path_record
	Similar to commit 27ea3c85500f7eff741c93e662a344742dbbea13
	to fix gcc-4.4.1 warnings

	Fix "dereferencing type-punned pointer will break strict-aliasing rules"
	warning catched by gcc-4.4.1.

2009-09-29  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibqueryerrors: simplify node_type_to_print setup.
	Siplify node_type_to_print setup.

2009-09-29  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibqueryerrors.c: Remove --all option and replace it with --switch, --ca, --router
	By default ibqueryerrors should print errors for all node types.
	Adding the other options allows for the limitation of this output.

	Also change the --switch option to be --node-guid which is really more
	accurate and use "-G" for better compliance with other utilities.  "-S"
	is left in for backward compatibility for the time being.

	Update the man page

2009-09-27  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibqueryerrors.c: fix bug when attempting a sub-fabric scan
	Also ibd_sm_id is never valid in this tool as the "-s" option is used
		for "suppress"

2009-09-23  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibportstate.c: Eliminate uninitialized variable compile warning

2009-09-19  sebastien dugue  <sebastien.dugue@bull.net>

	infiniband-diags/ibnetdiscover: Add separator when printing chassis type
	When grouping is enabled add a '#' separator between the switch guid
	and the chassis type and slot.

2009-09-19  Doron Shoham  <dorons@voltaire.com>

	ibportstate: fixes for width option
	Originally was:

	ibportstate: add width option.
	Similar to the speed option, this option can
	explicitly set the port's LinkWidthEnable value.
	It supports values from 0-15 and 255.

	But since similar patch was submitted before, there is only rebase
	results with couple of typos fixes.

2009-09-19  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibportstate: Support changing of link width
	Also, update man page

2009-09-19  Keshetti Mahesh  <keshetti.mahesh@gmail.com>

	'ibcheckportwidth' : Exit if LWS is 1X
	Fix: Modify ibcheckportwidth to exit if LWS is 1X instead of processing
	next lines.
	Trivial: ibcheckportwidth man page cosmetic change

	infiniband-diags/: Cosmetic changes, mostly typos
	Cosmetic changes

2009-08-31  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibroute: minor flow optimization
	Minor flow optimization - skip startlid value validation when it is
	wildcarded.

2009-08-31  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibroute: Add support for MulticastFDBTop
	Add support for SwitchInfo:MulticastFDBTop
	Added by MgtWG errata #4505-4508 and 4640

	If MulticastFDBTop set to other than 0, only fetch MulticastForwardingTable
	blocks up through MulticastFDBTop rather than MulticastFDBCap

	If MulticastFDBTop set to 0xbfff, this means no entries (per 4640)

2009-08-30  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: fix compilation warning
	Newly introduced retract_dpath() was declared as int but no any value
	was returned, this resulted in this warning:

	src/ibnetdisc.c: In function ‘retract_dpath’:
	src/ibnetdisc.c:186: warning: control reaches end of non-void function

	Fixing this by declaring retract_dpath() as void.

2009-08-30  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: add retract_dpath function
	When using combined routing some switches do not handle Hop Count of 0
		well.  Detect when the drpath count is 0 and return to lid based
		routing in this case.

	infiniband-diags/libibnetdisc: add missing '\n' to error message

2009-08-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perfquery: indentation fixes
	Indentation fixes using osm_indent.

2009-08-30  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery.c: Indicate whether PortXmitWait counter is supported
	Indicate extended v. (normal) port counters in output
	Also, some cosmetic formatting changes and commentary typo fixed

2009-08-30  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	Fix IB network discovery from switch node.

2009-08-23  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: make all fields of ibnd_fabric_t public
	On Sun, 16 Aug 2009 14:41:27 +0300
	Sasha Khapyorsky <sashak@voltaire.com> wrote:

	> On 20:42 Thu 13 Aug     , Ira Weiny wrote:
	> >
	> > @@ -108,8 +107,8 @@ typedef struct ibnd_port {
	> >  /** =========================================================================
	> >   * Chassis
	> >   */
	> > -typedef struct chassis {
	> > -	struct chassis *next;
	> > +typedef struct ibnd_chassis {
	> > +	struct ibnd_chassis *next;
	> >  	uint64_t chassisguid;
	> >  	unsigned char chassisnum;
	> >
	> > @@ -124,11 +123,17 @@ typedef struct chassis {
	> >  	ibnd_node_t *linenode[LINES_MAX_NUM + 1];
	> >  } ibnd_chassis_t;
	> >
	> > +/* HASH table defines */
	> > +#define HASHGUID(guid) ((uint32_t)(((uint32_t)(guid) * 101) ^ ((uint32_t)((guid) >> 32) * 103)))
	>
	> Why should this macro be published (by moving from internal.h to
	> ibnetdisc.h)?
	>
	> As far I can see it is only used in ibnetdisc.c, so actually we can keep
	> it internally and to move to this file.
	>

	You are right, good catch.  I just copied it blindly with HTSZ which must be
	there.

	git am is not working now on the last two patches [4/5 and 5/5] so I am
	sending new versions of them so that they apply cleanly.

	V2 below,
	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 13 Aug 2009 20:08:51 -0700
	Subject: [PATCH] libibnetdisc: make all fields of ibnd_fabric_t public

		In addition clean up the name of the chassis struct

2009-08-16  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Allow pkey and qkey to be hidden

	infiniband-diags/saquery.c: Fix CHECK_AND_SET_VAL macro
	Changed check from > to != since using integer comparison
	and some masks can use full range and hence be negative

2009-08-16  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: make all fields of ibnd_port_t public

2009-08-16  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: fix again typo in dump_node()
	Use ';' as statement termination character (re-fix after wrong rebase).

2009-08-16  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: make all fields of ibnd_node_t public

2009-08-16  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Fix typo in option name

2009-08-14  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniabnd-diags: indentation fixes
	Indentation fixes using opensm/osm_indent.

	infiniband-diags/ibportstat.c: simplify conditions
	Simplify conditions like ((a & N) && (b & N)) as (a & b & N).

	infiniband-diags/ibportstate.c: consolidate flows
	Consolidate flows for better code readability.

2009-08-13  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc: indentation fixes
	Fix indentations using opensm/osm_indent script.

	libibnetdisc/ibnetdisc.c: typo fix
	Fix statement completion typo ',' -> ';'.

2009-08-13  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/libibnetdisc: remove all IBPANIC's and clean up error handling
	This patch applies after:

		libibnetdisc: fix potential memory leak of port object

	Which I sent last week but I don't think has made it up stream.

	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Wed, 12 Aug 2009 16:13:56 -0700
	Subject: [PATCH] infiniband-diags/libibnetdisc: remove all IBPANIC's and clean up error handling

2009-08-13  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: fix potential memory leak of port object
	NOTE: This moves the port allocation below the port array allocation
		failure rather than free the port allocation after port array
		allocation fails.

2009-08-05  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsendtrap.c: code consolidation
	Consolidate some duplicated code. Use more descriptive function names.

	infiniband-diags/ibsendtrap.c: fix array size
	Fix 'traps' array size. Also define it as static const.

2009-08-05  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibsendtrap.c: Add support for link_speed_enabled_change trap

2009-08-04  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibsendtrap.c: Fill in capability mask on trap 144

2009-08-03  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/man: break long lines
	Break long (>= 80 characters) lines for better patch readability.

2009-08-03  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibnetdiscover.8: Add max hops option

2009-08-02  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibnetdiscover: remove not used var
	Remove not used anymore 'timeout' global variable.

2009-08-02  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	Remove timeout_ms parameter to ibnd_discover_fabric, timeout parameter
	should be specified via the ibmad_port.  Remove extraneous use of global
	timeout_ms in library.  Adjust ibnetdiscover, ibqueryerrors, iblinkinfo,
	and test code appropriately for adjustment.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 14:16:14 -0700
	Subject: [PATCH] Remove timeout_ms parameter to ibnd_discover_fabric, timeout parameter should be specified via the ibmad_port.  Remove extraneous use of global timeout_ms in library.  Adjust ibnetdiscover, ibqueryerrors, iblinkinfo, and test code appropriately for adjustment.

2009-08-02  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	Check input parameters to libibnetdisc functions

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 14:15:25 -0700
	Subject: [PATCH] Check input parameters to libibnetdisc functions

2009-08-01  Sasha Khapyorsky  <sashak@voltaire.com>

	bibnetdisc/ibnetdisc.c: fix parameter incompatibility warning
	Fix function parameter pointer type incompatibility compilation warning.

2009-08-01  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	Make api more consistent and make struct ibnd_fabric a struct that
	represents just fabric data by removing the ibmad_port and making it a
	function paramete in appropriate functions.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 14:14:57 -0700
	Subject: [PATCH] Make api more consistent and make struct ibnd_fabric a struct that represents just fabric data by removing the ibmad_port and making it a function paramete in appropriate functions.

2009-08-01  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	Fix potential memleak in ibnd_discover_fabric error path.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 11:58:45 -0700
	Subject: [PATCH] Fix potential memleak in ibnd_discover_fabric error path.

2009-08-01  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	Use IBPANIC consistently in libibnetdisc, in particular, since IBPANIC
	calls exit, there's no use in returning a value after an error.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 11:57:35 -0700
	Subject: [PATCH] Use IBPANIC consistently in libibnetdisc, in particular, since IBPANIC calls exit, there's no use in returning a value after an error.

2009-08-01  Al Chu  <chu11@llnl.gov>

	fix libibnetdisc API consistency and bugs
	In libibnetdisc, do not automatically output messages to stderr/stdout
	for warnings or non-fatal errors. Use IBND_DEBUG instead of IBPANIC or
	IBWARN to allow users of the lib to decide output conditions.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 23 Jul 2009 11:56:53 -0700
	Subject: [PATCH] In libibnetdisc, do not automatically output messages to stderr/stdout for warnings or non-fatal errors. Use IBND_DEBUG instead of IBPANIC or IBWARN to allow users of the lib to decide output conditions.

2009-07-31  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/libibnetdisc: fix DBGFLAGS variable

2009-07-20  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibtracert.c: Fix linearcap test in switch_lookup

2009-07-19  Al Chu  <chu11@llnl.gov>

	libibnetdisc cleanup patches
	Make api more consistent by removing convenience pointer from
	ibnd_node_t and requiring ibnd_fabric_t passed to all functions (in this
	case ibnd_update_node).

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	>From a6f7c6bea7f3549bb0bd24a4304f9dc3aa4b73ac Mon Sep 17 00:00:00 2001
	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 7 Jul 2009 09:57:26 -0700
	Subject: [PATCH] Make api more consistent by removing convenience pointer from ibnd_node_t and requiring ibnd_fabric_t passed to all functions (in this case ibnd_update_node).

2009-07-19  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibroute.c: Fix typo in IBWARN message

	infiniband-diags/ibtracert.c: Fix comparison to SwitchInfo.MulticastFDBCap in switch_mclookup

2009-07-15  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/README: Update saquery dependencies

2009-07-14  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/libibnetdisc: add man pages to EXTRA_DIST
	Add man pages files to EXTRA_DIST list so 'make dist' will generate a
	valid tarball.

2009-07-12  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibroute.c: Fix format and typo in printf

2009-07-07  Sasha Khapyorsky  <sashak@voltaire.com>

	libibnetdisc/Makefile.am: add internal.h to SOURCES list
	Without this 'make dist' (used in daily in release builds) generates
	broken infiniand-diags tarball.

2009-06-26  Ira Weiny  <weiny2@llnl.gov>

	iblinkinfo, ibqueryerrors: clean up ibdiag_process_opts call

2009-06-26  Ira Weiny  <weiny2@hyperion1.llnl.gov>

	ibprintca.pl: fix corner case: print switch information for CA's found at the end of the cache file.

2009-06-26  Ira Weiny  <weiny2@llnl.gov>

	ibqueryerrors: optomize -D operation
	only scan part of the fabric when -D is specified.

2009-06-26  Ira Weiny  <weiny2@hyperion1.llnl.gov>

	iblinkinfo: optomize -D operation
	Remove -f option and combine with -D
	   only scan part of the fabric when -D is specified.
	   add -a option which allows for printing "all" of the switches found when
	      performing a partial scan

2009-06-18  Ira Weiny  <weiny2@llnl.gov>

	ibqueryerrors: fix custom opts string

	iblinkinfo: convert to ibdiag_process_opts

	iblinkinfo: remove unused file pointer.

	iblinkinfo, ibqueryerrors: prevent core when switch is not found
	If the switch is not found print nice error message instead of seg faulting

2009-06-18  Ira Weiny  <weiny2@llnl.gov>

	ibnetdiscover: only report max hops discovered when requested
	Added "-m" flag to report this information if the user wants it.  I also
	changed the text in the message which says "reported max hops discovered".  I
	don't know if we want to change that text to something else but I wanted to
	indicate this number is not constant and may change.  This is true not just if
	you change the algorithm of discovery but also if you run from different
	nodes.

	Thoughts,
	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Wed, 6 May 2009 17:56:23 -0700
	Subject: [PATCH] ibnetdiscover: only report max hops discovered when requested

2009-06-17  Ira Weiny  <weiny2@llnl.gov>

	Move _with_node_name_map rpm spec option from infiniband-diags to opensm.
	The node name map is now part of libosmcomp.

2009-06-13  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	infiniband_diags/ibportstate.c perform reset in a single mad
	perform reset in a single mad

	Perform reset by changing physical state to polling instead of
	disable and enable.
	This ensures that reset a port will not disconnect it (assuming negotiation succeeds).
	Only disable operation is not allowed on HCA ports since it will irreversibly
	disconnect it though this may also happen on a switch ingress port.

2009-06-13  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	infiniband-diags: Do not change port state on SubnSet
	Do not change port state on SubnSet

	When changing physical state do not change logical port state to avoid invalid
	transitions in port state machine.
	From the IB spec When writing PortInfo:PortState, only legal transitions are
	valid. So if PortState is ACTIVE and we try to set it to ACTIVE this will fail.

	This patch will allow reset in a single MAD by changing physical state to polling.

2009-06-13  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibdiag_common.c: Eliminate compile warning on x86_64 archs
	src/ibdiag_common.c: In function pretty_print
	src/ibdiag_common.c:95: warning: field precision should have type int,
	but argument 3 has type long int

2009-06-13  Sean Hefty  <sean.hefty@intel.com>

	ib-mgmt/libibnetdisc: fix typecast warning

2009-05-12  Doron Shoham  <dorons@voltaire.com>

	saquery: fix -c arguement
	set SAQUERY_CMD_CLASS_PORT_INFO instead of CLASS_PORT_INFO

2009-05-12  Ira Weiny  <weiny2@llnl.gov>

	Clean up printing of switch heading when printing "down links" only.
	Another corner case:  If there are no down links on a switch and "-d" is selected then the header for that switch should not be printed.

	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 30 Apr 2009 13:41:38 -0700
	Subject: [PATCH] Clean up printing of switch heading when printing "down links" only.

2009-05-12  Ira Weiny  <weiny2@llnl.gov>

	Fix 2 formatting diff's from old ibqueryerrors.
	2 changes I noted in the output from ibqueryerrors.

	"Link Info:" was not being printed when "-r" was used.

	The "header": Errors for 0x<guid> "<node name>"

	Should only be printed when errors are found.

	The following patch cleans those up.

	Ira

	From: Ira Weiny <weiny2@llnl.gov>
	Date: Tue, 28 Apr 2009 14:39:11 -0700
	Subject: [PATCH] Fix 2 formatting diff's from old ibqueryerrors.

2009-05-06  Ira Weiny  <weiny2@llnl.gov>

	Modify '-S' option of iblinkinfo and ibqueryerrors to do a limited scan of the
	fabric first and then fall back to a full scan which searches for the GUID.

2009-05-06  Ira Weiny  <weiny2@llnl.gov>

	Add combined routing support to libibnetdisc
	Also allow a scan to start at a switch.

	Fix ibnd_discover when the specified ib_portid_t starts LID routed.

2009-05-06  Sean Hefty  <sean.hefty@intel.com>

	ib-mgmt: fixup ibsendtrap for windows
	Fix some typecast issues.

2009-05-06  Sasha Khapyorsky  <sashak@voltaire.com>

	ibstat.c: use htohl() for 32-bit capmask conversion
	capmask field was changed to be 32-bit, so use ntohl() instead of
	ntohll(). Casting is also not needed then.

2009-05-05  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibnetdiscover.c: Cosmetic formatting changes

2009-05-05  Nicolas Morey-Chaisemartin  <nicolas.morey-chaisemartin@ext.bull.net>

	infiniband-diags: Added libibnetdiscover to .spec file

2009-05-01  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Display attribute ID in hex rather than decimal
	for easier correlation to IBA spec

2009-05-01  Nicolas Morey-Chaisemartin  <nicolas.morey-chaisemartin@ext.bull.net>

	management: Fixed custom_release in SPEC files
	Removed a space which make rpmbuild fail when _dist and CUSTOM_RELEASE are set:
	error: line 15: Tag takes single token only: Release: ofed1.4.1 .fc11

	This is due to
	Release: %rel%{?dist}
	and %rel having a trailing whitespace.

2009-05-01  Sasha Khapyorsky  <sashak@voltaire.com>

	Merge branch 'pq/ibn4'
	Conflicts:
		libibmad/src/libibmad.map

2009-05-01  Ira Weiny  <weiny2@llnl.gov>

	ibqueryerrors: switch_guid_str is required for the string resolve function.

2009-04-26  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump package versions
	Bump management packages versions:

		libibumad-1.3.2
		libibmad-1.3.2
		opensm-3.3.2,
		infiniband-diags-1.5.2.

2009-04-25  Sasha Khapyorsky  <sashak@voltaire.com>

	ibdiag_common: remove duplicated ibdebug variable
	ibdebug is defined already in libibmad. Remove duplication in
	ibdiag_common.c.

	ibdiag_common: remove duplicated ibdebug variable
	ibdebug is defined already in libibmad. Remove duplication in
	ibdiag_common.c.

	infiniband-diags: remove not used ibnetdiscover.h
	Remove not needed anymore ibnetdiscover.h header file.

	infiniband-diags: cosmetic fixes
	Cosmetic fixes - mostly removing extra empty lines from code and man
	pages, also formatting.

2009-04-25  Ira Weiny  <weiny2@llnl.gov>

	Convert ibqueryerrors.pl to C and use new ibnetdisc library.

2009-04-25  Ira Weiny  <weiny2@llnl.gov>

	libibnetdisc: change ibnd_discover_fabric to receive ibmad_port
	In order to allow ibmad_port to be opened with additional classes
	libibnetdisc should accept an ibmad_port as a parameter. The library
	will error out if the classes it needs are not opened.

	Convert ibnetdiscover and iblinkinfo to use the new interface to
	libibnetdisc.

2009-04-25  Sasha Khapyorsky  <sashak@voltaire.com>

	ibnetdiscover: fix types to avoid portability castings
	We did this before, but somehow it was lost in libibnetdisc patches.

2009-04-25  Sean Hefty  <sean.hefty@intel.com>

	ib-mgmt/ibn3 branch: libibnetdisc add windows support
	Allow libibnetdisc to build and run on Windows as part of the WinOF
	distribution

	ib-mgmt/ibn3 branch: diags updated for continued windows support

2009-04-25  Ira Weiny  <weiny2@llnl.gov>

	Convert ibnetdiscover to use new ibnetdisc library
	All other functionality is preserved

	clean up iblinkinfo conversion
	Clean up a comment
	Fix potential bug

	Convert iblinkinfo.pl to C and use new ibnetdisc library.
	>From a677ae35fe7a5966f05b5859df8f00e9b18df864 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 3 Apr 2009 15:28:18 -0700
	Subject: [PATCH] Convert iblinkinfo.pl to C and use new ibnetdisc library.

2009-04-25  Ira Weiny  <weiny2@llnl.gov>

	Create a new library libibnetdisc
	>From e1c2c10678b0d1d90f7eb31eb1c1441b5ee43311 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 3 Apr 2009 15:28:08 -0700
	Subject: [PATCH] Create a new library libibnetdisc

	This encompasses the functionality of ibnetdiscover in a C library.  It returns
	a single "ibnd_fabric_t" object which represents the data found during the
	scan.  The NodeInfo, PortInfo, and SwitchInfo are preserved from the queries
	made on the fabric to be used by the calling function as they see fit.

	This greatly benefits some diags like iblinkinfo.pl.  This diag in particular
	was re-written using this library in C and has shown an 85% speed up on a ~1000
	node cluster.

	Previous iblinkinfo.pl
	   real    3m35.876s
	   user    0m13.210s
	   sys     1m1.046s

	New iblinkinfotest
	   real    0m32.869s
	   user    0m0.067s
	   sys     0m0.140s

2009-04-21  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/man/vendstat.8: Indicate IS4 config group config not persistent across IS4 reboot

2009-04-19  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibsendtrap.c: Set producer type according to node type
	rather than assuming CA

2009-04-17  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/man/vendstat.8: Fix PortXmit/RcvDataSL examples

2009-04-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perltidy.sh: option to format a single file
	Adding possibility to format a single file.

2009-04-15  Ira Weiny  <weiny2@llnl.gov>

	Fix ibidsverify.pl to use the correct cache file
	Sasha,

	I found this bug when I was testing the libibnetdisc stuff.

	This applies to the master.

	Ira

	>From 656ad88a1f3ca6bcd7601b03da1b3822e4091156 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 3 Apr 2009 16:00:46 -0700
	Subject: [PATCH] Fix ibidsverify.pl to use the correct cache file

	In addition add the -C and -P options for specifying a different HCA and port

2009-04-15  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/ibsendtrap.c: Local link integrity is an "urgent" trap

2009-04-14  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsendtrap: code consolidation
	Code consolidation to prevent duplications.

2009-04-14  Julia Volynsky  <juliav@voltaire.com>

	Added send trap for trap 129 (local link integrity)
	Added send trap for trap 129 (local link integrity).

2009-04-12  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/vendstat: Update man page and examples for PortXmit/RcvDataSL counter support

2009-03-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/vendstat.c: code moving
	Move subsequent functions code to be above main(), so no prototyping is
	needed.

2009-03-24  Or Gerlitz  <ogerlitz@voltaire.com>

	ib-diag/vendstat: counter-group-info & config-counter-group vendor mads
	Implement counter-group-info and config-counter-group IS4 vendor mads.
	The config-counter-group mad should be used to configure a switch port
	to make counters such as PortXmitDataSL or PortRcvDataSL active.

2009-03-21  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery.8: Fix typo in short option for PortXmitDataSL counters

2009-03-20  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery.c: Label PortXmit/RcvDataSL counters in headings
	Also, some cosmetic formatting changes

2009-03-12  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/perfquery.8: Extended counters are now -x rather than -e

	infiniband-diags/vendstat.c: Add missing mad_rpc_close_port call

	infiniband-diags/perfquery.c: Fix some memory leaks on exit

	infiniband-diags/perfquery.8: Update man page for PortXmit/RcvDataSL

2009-03-12  Sasha Khapyorsky  <sashak@voltaire.com>

	libibmad: add mad_rpc_class_agent() call
	Add new thread safe mad_rpc_class_agent() class which gets port as
	parameter. Deprecate old mad_class_agent() and convert its usage in
	ibsysstat.

2009-03-10  David A. McMillen  <davem@systemfabricworks.com>

	infiniband-diags/ibnetdiscover: Apply --node-name-map remapping to names printed by --ports option

2009-03-09  Or Gerlitz  <ogerlitz@voltaire.com>

	infiniband-diags/perfquery: add srcport param
	Add srcport param for the PortXmtDataSL/PortRcvDataSL calls made to libibmad

2009-03-07  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: convert ibnetdiscover to "new" ibmad interface
	On Fri, 20 Feb 2009 09:23:50 -0800
	Ira Weiny <weiny2@llnl.gov> wrote:

	> On Fri, 20 Feb 2009 08:55:57 -0500
	> Hal Rosenstock <hal.rosenstock@gmail.com> wrote:
	>
	> > On Thu, Feb 19, 2009 at 10:05 PM, Ira Weiny <weiny2@llnl.gov> wrote:
	> > > Here is v2 of the patch series.
	> > >
	> > > I used __attribute__ ((deprecated)) on the functions which should aid others
	> > > in realizing that these functions will go away.  (It sure helped me to convert
	> > > all the diags.
	> > >
	> > > Also I did _not_ convert ibnetdiscover as my new libibnetdisc already uses the
	> > > new interface and I am hoping it will be accepted soon.
	> >
	> > A related issue is whether ibnetdiscover will support both the new
	> > library and the old way until the library is more proven via some
	> > build option. If it is to support both, then converting it should be
	> > done.
	>
	> The conversion is easy.  I will do it for now to remove the build warnings.
	> And now that I think about it more leaving in the old and new code to be
	> chosen via configure is probably not a bad idea.  I don't know what is going
	> to happen once we standardize on the mad library for decoding strings.  There
	> are some incompatibilities there (ie 1x vs 1X and 2.5Gbps vs SDR etc.)
	>
	> I will say, however, that I tested the library extensively and the first
	> version's output was identical to the old version with the sole exception of
	> the order ports were printed in.  :-D  So my confidence is high it will be
	> accepted sooner rather than later.
	>
	> Ira

	Patch below:

	>From ad8cbf227a803d64c02872f74d7d542b815c6092 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 20 Feb 2009 14:43:48 -0800
	Subject: [PATCH] infiniband-diags: convert ibnetdiscover to "new" ibmad interface

2009-03-07  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Convert perfquery, saquery, sminfo, smpquery, and vendstat to "new" ibmad interface
	>From e809dfacb08e6c2237ad2d0f197d1227654dde87 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 19 Feb 2009 18:53:10 -0800
	Subject: [PATCH] infiniband-diags: Convert perfquery, saquery, sminfo, smpquery, and vendstat to "new" ibmad interface

	infiniband-diags: Convert mcm_rereg_test to "new" ibmad interface
	>From 4dcd4839baaa7f3bc31d01d5e695fced36b53533 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 19 Feb 2009 18:24:56 -0800
	Subject: [PATCH] infiniband-diags: Convert mcm_rereg_test to "new" ibmad interface

	infiniband-diags: Convert ibsysstat to "new" ibmad interface
	>From 1c19e419e04a98bcfe10b1c597856f43ea36668a Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 19 Feb 2009 18:14:49 -0800
	Subject: [PATCH] infiniband-diags: Convert ibsysstat to "new" ibmad interface

	infiniband-diags: Convert ibtracert to "new" ibmad interface
	>From 0961e0ce048950e65bb78578538cff38b2c8332d Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 19 Feb 2009 17:58:36 -0800
	Subject: [PATCH] infiniband-diags: Convert ibtracert to "new" ibmad interface

	infiniband-diags: Convert ibsendtrap to "new" ibmad interface
	also make mad_send_via public to do the conversion

	infiniband-diags: Convert ibroute to "new" ibmad interface
	>From 5b66b604de9bc43458ca4d295c5ab14cf2c6df10 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Thu, 19 Feb 2009 17:30:14 -0800
	Subject: [PATCH] infiniband-diags: Convert ibroute to "new" ibmad interface

	infiniband-diags: Convert ibportstate to "new" ibmad interface

	infiniband-diags: Convert ibaddr to "new" ibmad interface

2009-03-06  Hal Rosenstock  <hnrose@comcast.net>

	opensm/infiniband-diags: Changes for C rather than C++ style comments

2009-03-02  Or Gerlitz  <ogerlitz@voltaire.com>

	infiniband-diags: update configure.in check for libibmad API
	update infiniband-diags/configure.in check for libibmad API

2009-03-02  Or Gerlitz  <ogerlitz@voltaire.com>

	perfquery PortXmtDataSL/PortRcvDataSL support
	perfquery PortXmtDataSL/PortRcvDataSL (IBA A13.6.5/6) support


	-- changes from V1: set command line args such for unused shortcuts

2009-03-02  Or Gerlitz  <ogerlitz@voltaire.com>

	generic libibmad perf query/reset api
	Export generic api to issue query and reset on performance counters,
	this will allow for easier integration of more counters (e.g QoS,
	CCA related).

	Port perfquery to the generic libibmad api for performance counters
	query and reset.

2009-02-28  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump all package versions
	Bump all management packages versions - libibumad-1.3.1, libibmad-1.3.1,
	opensm-3.3.1, infiniband-diags-1.5.1.

2009-02-27  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibroute: fix warning
	Fix pointer signedness difference warning and potential WinOF warnings
	(verified with -Wsign-compare).

2009-02-26  Sean Hefty  <sean.hefty@intel.com>

	[ib-diag] saquery: add support for WinOF

	[ib-diags] saquery: set correct pkey table field
	port_num is incorrectly set instead of block_num

	[ib-diag] ibsysstat: add support for WinOF

	[ib-diag] mcm_rereg_test: add support for WinOF

	[ib-diag] ibsendtrap: add support for WinOF

	[ib-diag] ibtracert: add support for WinOF

	[ib-diag] ibroute: add support for WinOF

	[ib-diag] ibnetdiscover: add support for WinOF
	Mainly fixing datatypes to avoid type mismatches.

2009-02-26  Hal Rosenstock  <hnrose@comcast.net>

	infiniband-diags/saquery.c: Convert more LID prints to unsigned decimal

	infiniband-diags/smpdump.c: Release umad resources on exit

	infiniband-diags/smpdump.c: Fix usage examples

2009-02-26  David McMillen  <davem@systemfabricworks.com>

	infiniband-diags/src/ibnetdiscover.c missing LID information on --ports
	In commit 496b83de4c4175ef3f8e4517708e1cbb8ee782f3 you added code to get the
	correct information from the proper port, which fixes problems with wrong
	speeds, etc. but appears to have broken the reporting for the LID in the
	--ports output.
	Without --ports, the values used for a switch are node->smalid
	and node->smalmc but with --ports the value is taken from port->lid.  Also,
	when getting the "other side" of a link, a quick look at the code seems like
	it just uses port->lid as well.

	I did a quick hack in get_node() to replace port->lid and port->lmc with the
	values that were saved in node->smalid and node->smalmc and it seems to
	solve the problem.  I just worry that I have not really thought through all
	of the impacts of this.  Also, if my basic idea is correct, it seems like
	other coding choices would be to either leave out lid/lmc decode
	from decode_port_info() and do the decode in the two places where it is
	needed, or to add a boolean argument to decode_port_info() to enable/disable
	lid/lmc decode.

	Dave

	------- quick hack as a patch if you want it ------

	    Fix problem where LID 0 is showing on switch ports

2009-02-18  Sean Hefty  <sean.hefty@intel.com>

	smpquery: add support for WinOF
	Allow smpquery to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

2009-02-18  Sean Hefty  <sean.hefty@intel.com>

	ibping: add support for WinOF
	Allow ibping to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	For portability, use complib to obtain time stamps.

2009-02-18  Sean Hefty  <sean.hefty@intel.com>

	smpdump: add support for WinOF
	Allow smpdump to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	ibstat: add support for WinOF
	Allow ibstat to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	ibportstate: add support for WinOF
	Allow ibportstate to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	perfquery: add support for WinOF
	Allow perfquery to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	ibaddr: add support for WinOF
	Allow ibaddr to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	vendstat: add support for WinOF
	Allow vendstat to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF repository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

	sminfo: add support for WinOF
	Allow sminfo to build and run on both Linux and Windows.  Window
	build files are maintained in the WinOF respository.  These changes
	allow dropping the infiniband-diags into the WinOF build environment.

2009-02-14  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniabnd-diags/common: wrap debug macros with do {} while (0)
	Wrap debug macros which use 'if () {}' with 'do { .. } while (0)' to
	prevent potential 'else' statement mismatching. Also use portable
	__VA_ARGS__ macro.

2009-02-11  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: some code consolidation
	Consolidate repeated code using helper functions
	get_and_dump_any_records() and get_and_dump_all_records().

	infiniband-diags/saquery: fix types and some cleanup
	Fix types - mostly ib_net*_t -> uint*_t conversion. Use host byte order
	SA attributes from mad.h (instead of ib_types.h). Fix functions
	prototypes and return value types. Remove osm* stubs. Remove unused
	'offset' argument in get_any_records() and get_all_gecords() functions.

2009-02-11  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: remove osm vendor layer
	Replace OSM Vendor layer by libibumad and libibmad (rpc) calls.

	This patch is done following "minimum changes" rule to demonstrate osm
	vendor replacement. Many subsequent improvements and simplification can
	be done. All current saquery functionality is preserved.

2009-02-07  Hal Rosenstock  <halr@obsidianresearch.com>

	infiniband-diags/perfquery: Change option name for extended counters
	Sasha,

	Per the RFC, this patch changes the option name for extended counters to
	to not cover up common errors option. This changes it from -e/--extended
	to -x/--extended so -e/--errors can be used to get error information as
	is common with the IB diags.

	-- Hal

	>From 73fc1c902aa712dd2682e2c439098aa1d21e18a0 Mon Sep 17 00:00:00 2001
	From: Hal Rosenstock <hal.rosenstock@gmail.com>
	Date: Thu, 5 Feb 2009 17:57:11 -0500
	Subject: [PATCH] infiniband-diags/perfquery: Change option name for extended counters

	to not cover up common errors option
	Now -x --extended

2009-02-05  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/common: use enum MAD_DEST as ibd_dest_type type
	Use introduced 'enum MAD_DEST' as type of ibd_dest_type variable.

2009-01-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: merge PathRecord query functions
	In order to prevent code duplications and let to user to use all
	introduced query parameters merge all functions querying PathRecord -
	use single query_path_records().

	Backward compatibility is preserved (actually is is setting NumbPath
	when --src-to-dst or --sgid-to-dgid options are used).

2009-01-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: more params for Path and MCMember Records
	This adds many parameters for PathRecord and MCMemberRecord queries
	(such as mtu, rate, etc.). And slso extended MCMemberRecord dumper.
	Backward compatibility with existing options (-m, -g, --src-to-dst,
	--sgid-to-dgid) is preserved.

	infiniband-diags/saquery: adding query params
	This adds query params structure - currently it has only slid, dlid,
	sgid and dgid and used by PathRecord query. Will be extended by other
	parameters and can be used in other queries.
	Also using option allows wildcards now ( 'saquery --src-to-dst 2:' ).

	infiniband-diags/saquery: CHECK_AND_SET_VAL() macro
	CHECK_AND_SET_VAL() macro is SA query encoding helper - prevents many
	code duplications.

	infiniband-diags/saquery: cosmetic
	Indentation fixes, not needed braces removing, shorter names.

2009-01-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: fix encoding of SA queries
	Various queries encoding fixes:

	- PortInfoRecord: PortNum has length 8 bit and should not be converted to
	  network byte order.
	- NodeRecord: initialize lid value to prevent garbage in a query.
	- PkeyTableRecord: BlockNumber has 16 bit length and should be encoded
	  in network byte order.

2009-01-28  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/smpdump: fix SL value encoding
	Fix SL value encoding when using LID routed get requests.

	infiniband-diags/saquery: fix backward compatibility bug
	This fixes long-standing backward compatibility bug when options like
	'-L' and '--src-to-dst' were not processed correctly.

2009-01-27  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsysstat: backward compatibility fixes
	1. For short server responses ( <= 256 ) don't use RMPP flag - when
	communicating with old (non RMPP capable) clients it prevents timeout
	generation.

	2. Drop invalid MADs on a server side - this also will drop timeout
	MAD generated when sending long reply to RMPP not capable client.

	3. Report invalid responses as error on a client side.

2009-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsysstat: use RMPP for client/server communication
	This patch adds support for bigger than (256 - vendor2 data offset) data
	sending by ibsysstat server using RMPP. It fixes bug#1237 - where server
	output was truncated due to MAD size limitation.

	infiniband-diags/saquery: add lid parameter to NodeRecord query
	Add optional LID parameter to NodeRecord SA Query.

2009-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/smpquery: usage improvement
	This makes usage of smpquery operations more user friendly - similar to
	saquery each operation now has a shorter alias, string matching is case
	insensitive and abbreviations are allowed for both operation name and
	alias. And it is how this looks:

	Usage: smpquery [options] <op> <dest dr_path|lid|guid> [op params]

	Supported operations (and aliases, case insensitive):
	  NodeInfo (NI) <addr>
	  NodeDesc (ND) <addr>
	  PortInfo (PI) <addr> [<portnum>]
	  SwitchInfo (SI) <addr>
	  PKeyTable (PKeys) <addr> [<portnum>]
	  SL2VLTable (SL2VL) <addr> [<portnum>]
	  VLArbitration (VLArb) <addr> [<portnum>]
	  GUIDInfo (GI) <addr>

2009-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: remove unneeded includes
	Remove unneeded header files inclusions.

	infiniband-diags: make get_build_version() static
	Now when get_build_version() wrapped by command line options processor
	we can make it static.

	infiniband-diags: remove argv0 global variable
	We don't need this anymore - prog_name is used instead.

2009-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: using common command line option processing
	This converts infiniband-diags tools to use the common command line
	processing framework - it unifies the usages, options and removes a lot
	of duplications. The tools functionality is preserved (should not be a
	problems with backward compatibility), however the usage message is
	changed for many tools and now looks like:

	Usage: ibaddr [options] [<lid|dr_path|guid>]

	Options:
	  --gid_show, -g          show gid address only
	  --lid_show, -l          show lid range only
	  --Lid_show, -L          show lid range (in decimal) only
	  --Ca, -C <ca>           Ca name to use
	  --Port, -P <port>       Ca port number to use
	  --Direct, -D            use Direct address argument
	  --Guid, -G              use GUID address argument
	  --timeout, -t <ms>      timeout in ms
	  --sm_port, -s <lid>     SM port lid
	  --errors, -e            show send and receive errors
	  --verbose, -v           increase verbosity level
	  --debug, -d             raise debug level
	  --usage, -u             usage message
	  --help, -h              help message
	  --version, -V           show version

	Examples:
	  ibaddr 		# local port's address
	  ibaddr 32		# show lid range and gid of lid 32
	  ibaddr -G 0x8f1040023	# same but using guid address
	  ibaddr -l 32		# show lid range only
	  ibaddr -L 32		# show decimal lid range only
	  ibaddr -g 32		# show gid address only

2009-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: command line option processing framework
	The main motivation of this is to unify infiniband-diags command line
	options and tools usage. Also it simplifies programming and can remove
	a lot of duplications. The usage message is also unified over all tools
	and looks like:

	Usage: ibaddr [options] [<lid|dr_path|guid>]

	Options:
	  --gid_show, -g          show gid address only
	  --lid_show, -l          show lid range only
	  --Lid_show, -L          show lid range (in decimal) only
	  --Ca, -C <ca>           Ca name to use
	  --Port, -P <port>       Ca port number to use
	  --Direct, -D            use Direct address argument
	  --Guid, -G              use GUID address argument
	  --timeout, -t <ms>      timeout in ms
	  --sm_port, -s <lid>     SM port lid
	  --errors, -e            show send and receive errors
	  --verbose, -v           increase verbosity level
	  --debug, -d             raise debug level
	  --usage, -u             usage message
	  --help, -h              help message
	  --version, -V           show version

	Examples:
	  ibaddr 		# local port's address
	  ibaddr 32		# show lid range and gid of lid 32
	  ibaddr -G 0x8f1040023	# same but using guid address
	  ibaddr -l 32		# show lid range only
	  ibaddr -L 32		# show decimal lid range only
	  ibaddr -g 32		# show gid address only

	Custom (per tool) option processing is also supported.

2009-01-25  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/dump_mfts.sh: fix -D format parsing
	The fixes are similar to those were done in dump_lfts.sh, plus all
	previous adoption and fixes.

	infiniband-diags/dump_lfts.sh: fix -D format parsing
	DR format was changed slightly - fix format parsing accordingly.
	Also as Ira suggested use '-s' (show) option with ibnetdiscover instead
	of -v (verbose) - it produces less output.

2009-01-23  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibdiag_common: cleanup argv0 prototype
	Cleanup argv0 prototype.

	infiniband-diags/smpdump.c: use common ib definitions
	Use common IB class  definitions from libibmad.

	infiniband-diags: remove duplicated ibdebug prototype
	ibdebug has prototype in ibdiag_common.h so remove duplicated
	occurrences.

	infiniband-diags/ibdiag_common: move get_build_version()
	Move get_build_version() function to source file - so for version change
	only relink will be necessary for all tools rather than full rebuild.

	infiniband-diags/ibdiag_common: cosmetic
	Cosmetic changes - indentation fixes, not needed includes removing, etc.

	infiniband-diags/Makefile.am: use common library
	Use common static library libcommon.a for all tools.

	infiniband-diags/Makefile.am: merge CFLAGS
	Merge per tool CFLAGS - use common AM_CFLAGS definition. saquery doesn't
	need custom CFLAGS anymore since proper OpenSM vendor configuration is
	included via osm_vendor_api.h now.

	infiniband-diags/Makefile.am: kill -rpath
	Kill -rpath usage in LDFLAGS.

2009-01-20  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perfquery.c: fix typo

2009-01-20  Eli Dorfman (Voltaire)  <dorfman.eli@gmail.com>

	infiniband-diags support PortXmitWait get and set
	support PortXmitWait get and set

2009-01-05  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perfquery: indentation fixes
	Use tabs instead of spaces for indentation.

2008-12-25  Sasha Khapyorsky  <sashak@voltaire.com>

	management: remove libibcommon dependencies
	Remove libibcommon dependencies from libibmad and infiniband-diags.
	Actually libibcommon is not used anymore in management tree.

	Macros IBWARN(), IBPANIC(), ALIGN() and functions htonll(), htohll() and
	xdump() are part of libibmad now.

2008-12-25  Sasha Khapyorsky  <sashak@voltaire.com>

	management: move sysfs()_* function to libibumad
	Move sysfs() stuff to libibumad tree. Now libibumad doesn't depend from
	libibcommon anymore - remove this dependency.

2008-12-21  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibstat,smpdump: kill unused includes
	Kill unused headers inclusion.

2008-12-18  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: minor indentation fixes

2008-12-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infinabd-diags: convert type uint -> unsigned int
	Convert uint type to unsignedint in accordance with c99 definitions.

	infiniband-diags/saquery: PortInfoRecord query
	PortInfoRecord query implementation. Alias is "PIR" (case insensitive),
	usage is [lid]/[port], all is optional.

2008-12-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: separate queries and commands
	This new control level 'command' will be used to preserve useful
	backward compatible usage (controlled by command line options) where
	complex SA queries are used. Single queries (controlled by query names)
	will be performed by query id using SAQUERY_CMD_QUERY command.

	Such separation will help us to extend existing functionality and to
	preserve saquery backward compatibility.

	Also rename print_portinfo_records() to print_issm_records() b/c it is
	what this function does (controlled by "command").
	query_portinfo_records() will be implemented later.

2008-12-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diabs/saquery: unify SA queries processors
	Unify single SA queries processors, rename it print_*() -> query_*().

2008-12-13  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: indentation fixes
	Indentation fixes with opensm/osm_indent. Also replace bind_handle local
	variable name by shorter 'h'.

	management: bump all package versions
	Bump all package version - start a new development cycle.

2008-12-06  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update package versions
	Update versions: opensm - 3.2.5, infiniband-diags - 1.4.4

2008-11-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/grouping: add 10G IP router devid
	Add 10G IP router device id for grouping.

2008-11-18  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update versions
	Update versions

2008-11-18  Hal Rosenstock  <halr@obsidianresearch.com>

	infiniband-diags/ibstat.c: Use UMAD_MAX_DEVICES define
	Sasha,

	Please see attached patch.

	-- Hal

	infiniband-diags/ibstat.c: Use UMAD_MAX_DEVICES define

2008-11-16  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibstat.c: remove casting
	Remove umad_get_cas_names() parameter casting.

2008-11-16  Ira Weiny  <weiny2@llnl.gov>

	Fix max parameter passed to umad_get_cas_names
	From a9149f4e38081d206d0be0af2194f4e09f944f21 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 14 Nov 2008 11:36:01 -0800
	Subject: [PATCH] Fix max parameter passed to umad_get_cas_names

2008-10-19  Sasha Khapyorsky  <sashak@voltaire.com>

	management/*/configure.in: bump all versions
	Bump all management versions:

	libibcommon-1.1.2
	libibumad-1.2.2
	libibmad-1.2.2
	opensm-3.2.3
	infiniband-diags-1.4.2

2008-10-19  Sasha Khapyorsky  <sashak@voltaire.com>

	management: update Voltaire copyright
	Update Voltaire copyright message for non-trivially modified files in
	2008.

2008-10-19  Hal Rosenstock  <halr@obsidianresearch.com>

	infiniband-diags/saquery.8: Update saquery man page
	Sasha,

	This patch updates the saquery man page.

	-- Hal

	infiniband-diags/saquery.8: Update saquery man page for additional queries

2008-10-19  Hal Rosenstock  <halr@obsidianresearch.com>

	infiniband-diags/saquery: Support MFTRecord
	Sasha,

	This patch adds support for MFTRecord in saquery.

	-- Hal

	infiniband-diags/saquery: Add support for MFTRecord

2008-10-19  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: add LFT Record handling
	Add SA LFT Record attribute handling.

2008-10-18  Doug Ledford  <dledford@redhat.com>

	Trivial: update usage info for ibaddr.c in infiniband-diags-1.4.1
	Simple patch just to add an option to the usage info that was left out
	but is valid.

2008-10-12  Sasha Khapyorsky  <sashak@voltaire.com>

	management: add Voltaire copyright
	Add Voltaire copyright message for non-trivially modified files.

2008-10-11  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/perfquery: fix constant size warning
	Fix constant size warning (appeared on 32-bit machines).

2008-10-11  Al Chu  <chu11@llnl.gov>

	infiniband-diags/perfquery: if -a is specified loop through ports if required and aggregate output
	And finally the big patch.  If -a is specified and AllPortSelect is not
	supported (and -l isn't specified) loop through all ports and aggregate
	them into one output.

	So the patch is a tad lengthy given the manual packet parsing/counting
	that had to be done.  I'm not aware of any libs/helper funcs in OFED
	that could have made this code shorter.  Please let me know if there are
	some obvious funcs that could make this better.

	Also, my understanding is that in IB, counters don't wrap around.  When
	they get to the max they stay at the max.  So wrap-around checks for all
	of the counters is there as I added things up.  Couldn't find some
	wrapper funcs for this in OFED already.  So hopefully I'm not repeating
	code.

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:39:19 -0700
	Subject: [PATCH] aggregate port counters for single output under --all_ports

2008-10-11  Al Chu  <chu11@llnl.gov>

	perfquery code cleanup
	Hey Sasha,

	This is a code cleanup patch renaming "all" to "all_ports" for
	consistency to other code in perfquery.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:39:13 -0700
	Subject: [PATCH] rename all to all_ports for clarity to optionname

2008-10-11  Al Chu  <chu11@llnl.gov>

	infiniband-diags/perfquery: if --loop_ports is specified always loop through all ports if desired
	> Hey Sasha,
	>
	> As discussed in the other thread, --loop_ports will now iterate through
	> all ports no matter what with this patch.
	>
	> Al
	>
	> _______________________________________________
	> general mailing list
	> general@lists.openfabrics.org
	> http:// lists.openfabrics.org/cgi-bin/mailman/listinfo/general
	>
	> To unsubscribe, please visit http:// openib.org/mailman/listinfo/openib-general
	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:38:48 -0700
	Subject: [PATCH] when --loop_ports is specified always iterate, when --all_ports is specified without AllPortSelect support, iterate through ports

2008-10-11  Al Chu  <chu11@llnl.gov>

	infinibanf-diags/perfquery: remove single port CA AllPortSelect workaround
	Hey Sasha,

	This removes the workaround in perfquery that gets around lack of
	AllPortSelect support on a CA w/ a single port.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:38:43 -0700
	Subject: [PATCH] remove single port CA AllPortSelect workaround

2008-10-11  Al Chu  <chu11@llnl.gov>

	infiniband-diags/perfquery: error out if AllPortSelect is not supported
	Hey Sasha,

	As discussed in the other thread, this patches makes perfquery error out
	if the user requested port 255 and AllPortSelect is not supported.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:38:30 -0700
	Subject: [PATCH] if user inputs port 255 error if AllPortSelect flag not supported

2008-10-11  Al Chu  <chu11@llnl.gov>

	tweak notes about port 255 in perfquery manpage
	Hey Sasha,

	As discussed in the other threads, using -a should be different than
	specifying port 255.  Adjust the manpage appropriately.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:38:25 -0700
	Subject: [PATCH] tweak perfquery note about port 255

2008-10-11  Al Chu  <chu11@llnl.gov>

	infiniband-diags/ibclearerrors: specify -a in call to perfquery
	Hey Sasha,

	As discussed in the other thread, ibclearerrors should now specify -a
	instead of depend on passing port 255 to the tool.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Thu, 9 Oct 2008 16:38:22 -0700
	Subject: [PATCH] specify -a in perfquery call to clear all errors

2008-10-09  Al Chu  <chu11@llnl.gov>

	infiniband-diags/perfquery: fix comments
	Hey Sasha,

	Dumb patch.  I realized the comments weren't in the right place.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 8 Oct 2008 17:28:54 -0700
	Subject: [PATCH] fix perfquery comment

2008-10-09  Al Chu  <chu11@llnl.gov>

	infiniband-diags/perfquery: support ehanced port 0 with --loop-ports
	Hey Sasha,

	Fixes the enhanced port 0 issue Hal referred to in the previous thread.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Wed, 8 Oct 2008 17:28:40 -0700
	Subject: [PATCH] support ehanced port 0 with --loop_ports

2008-10-09  Hal Rosenstock  <halr@obsidianresearch.com>

	infiniband-diags/ibsysstat.c: Fix a couple of latent bugs
	Sasha,

	This patch is based on a code inspection of ibsysstat.c due to the
	buffer overflow observed with more than 2 CPUs. It fixes a couple of
	latent bugs although won't improve that particular issue.

	-- Hal

	infiniband-diags/ibsysstat.c: Fix latent bugs related to build_cpuinfo
	(exceeding MAX_CPUS and not being able to open /proc/cpuinfo)
	Also, issue warning when cpuinfo is truncated due to lack of room

2008-10-08  Al Chu  <chu11@llnl.gov>

	add --loop_ports option to perfquery
	Hey Sasha,

	We have a switch here that does not report the AllPortSelect flag as a
	capability.  It's pretty annoying typing each port on the switch or
	always having to script around this one oddball switch we have.  So I
	added an option --loop_ports for perfquery.  If you want to do something
	to all the ports on the CA/Switch, but AllPortSelect isn't available, it
	loops through all the available ports instead.

	There was already a workaround in the tool for a CA that did not support
	the AllPortSelect flag.  I get the feeling the workaround may have been
	for a specific hardware, so I kept the workaround in there.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 7 Oct 2008 14:05:54 -0700
	Subject: [PATCH] support --loop_ports in perfquery

2008-10-08  Al Chu  <chu11@llnl.gov>

	add more detail to error message on perfquery workaround
	Hey Sasha,

	Didn't know what some error messages really meant.  So added some
	detail.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 7 Oct 2008 09:57:56 -0700
	Subject: [PATCH] clarify error messages

2008-10-08  Al Chu  <chu11@llnl.gov>

	use 0xff vs. 255 consistently in perfquery
	Hey Sasha,

	Nothing fancy.  Was just inconsistent.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 7 Oct 2008 09:20:14 -0700
	Subject: [PATCH] create/use ALL_PORTS macro for clarity

2008-10-08  Al Chu  <chu11@llnl.gov>

	add clarification in perfquery manpage
	Hey Sasha,

	I ran upon a script that was using 255 for the port number.  I had no
	idea what it meant at the time.  Thought a clarification in the
	perfquery manpage would help.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From: Albert Chu <chu11@llnl.gov>
	Date: Tue, 7 Oct 2008 09:15:39 -0700
	Subject: [PATCH] note that port 255 is the same as -a

2008-09-28  Jack Morgenstein  <jackm@dev.mellanox.co.il>

	infiniband-diags: eliminate compiler warnings
	infiniband: eliminate compiler warnings on x86_64

	The printf's below generated warnings of the form:
	warning: format '%lu' expects type 'long unsigned int', but argument 4 has type 'long long unsigned int'
	on 64-bit systems -- due to the influence of the <>ull constants. Casting solves this.

2008-09-13  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibtracert: fix port by direct path resolving
	Then option '-D' is used ports provided to ibtracert in direct path
	format. This option was broken (bug #1136) due to incorrect resolution -
	lack of lid. This addresses bug #1136.

2008-09-12  Sasha Khapyorsky  <sashak@voltaire.com>

	man/ibtracert: remove non-existing -e option
	Remove non-existing '-e' command line option from ibtracert man page.
	This addresses bug #1136.

	infiniband-diags/ibclearcounters: remove unrelated -N option
	Remove unrelated -N command line option. This addresses bug#1132.

	man/ibnetdiscover: cleanup non-existing options
	Cleanup non-existing options from ibnetdiscover man page. This addresses
	bug #1131.

2008-09-10  Ira Weiny  <weiny2@llnl.gov>

	ibnetdiscover.c: continue processing other ports even if smpquery fails on one port

2008-08-19  Ira Weiny  <weiny2@llnl.gov>

	Fix some missing node name map substitutions
	From 1b7ed57320796720a0e8c25f04a2544ac9374aa5 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Mon, 18 Aug 2008 15:42:01 -0700
	Subject: [PATCH] Fix some missing node name map substitutions

	These are 2 cases where the node name map substitution was missed.

2008-07-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/src/ibaddr.c: remove unused variables

2008-07-24  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibaddr.c: convert GID prints to use inet_ntop

2008-07-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: remove unused variable

2008-07-24  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/saquery.c: convert GID prints to use inet_ntop

2008-07-08  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump management versions
	Bump management versions: libibcommon-1.1.1, libibumad-1.2.1,
	libibmad-1.2.1, opensm-3.2.2, infiniband-diags-1.4.1

2008-07-09  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/mcm_rereg_test.c: Cosmetic formatting change
	infiniband-diags/mcm_rereg_test.c: Cosmetic formatting change

2008-07-07  Sasha Khapyorsky  <sashak@voltaire.com>

	management: revert configure.in update configure.in on 'make dist'
	Recently configure.in automatic update was added. This setup there a
	package version in accordance with tree state during 'make dist'. However
	this method has a disadvantage - distributed configure.in template is
	newer than an actual './configure' script, it results to an additional
	(automatic) reconfiguration during build time.

	This patch removes this "smart" configure.in update. Instead we will
	handle versions directly in a daily build script.

2008-07-06  Al Chu  <chu11@llnl.gov>

	Fix regenerate cache corner case.
	Hey Sasha,

	Fixes a corner case I noticed.  Nothing too fancy.

	Al

	--
	Albert Chu
	chu11@llnl.gov
	925-422-5311
	Computer Scientist
	High Performance Systems Division
	Lawrence Livermore National Laboratory

	From d1c5008c9cdc2a1c065b70d772fe11c527b05bcc Mon Sep 17 00:00:00 2001
	From: Albert L. Chu <chu11@llnl.gov>
	Date: Mon, 30 Jun 2008 15:09:49 -0700
	Subject: [PATCH] fix regenerate cache corner case

2008-06-29  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: fix snprintf() usage
	snprintf() can return the value larger than provided buffer size - check
	is needed.

2008-06-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: split print_multicast_group_records() function
	Split logic overflowed function print_multicast_group_records() to two -
	print_multicast_group_records() and print_multicast_member_records().
	Simplify the code as for another print_*_records() functions.

2008-06-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/saquery.c: fix potential core dump
	fix potential core dump and/or incorrect node descriptions from being
	printed if a corresponding node_record is not found for a multicast
	member record.

2008-06-12  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/saquery.c: In print_multicast_group_records, only query NodeRecords when needed
	infiniband-diags/saquery.c: In print_multicast_group_records, only query
	NodeRecords when needed. This is a minor optimization.

2008-06-09  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/saquery.c: In dump_results, use query_result rather than query_svc_rec
	infiniband-diags/saquery.c: In dump_results, use query_result rather
	than query_svc_rec

	Most queries are not for service records.

	If attribute identifier validation is desired, this should be
	implemented differently.

2008-06-08  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibnetdiscover.c: Fix the PortInfo data on the port we discover a switch on.
	I noticed over the last couple days that ibnetdiscover would print the
	incorrect speed on the port I would run ibnetdiscover from.  For example:

	Switch  24 "S-0008f10400411f56"         # "SW1 wopr ISR9024D (MLX4 FW)" base port 0 lid 11 lmc 0
	...
	[13]    "H-0002c90200219e64"[1](2c90200219e65)          # "wopri" lid 32 4xSDR

	                                                                           ^^^
	               (Note from the switch side of things it thinks the speed is SDR.)
	...
	Ca      2 "H-0002c90200219e64"          # "wopri"
	[1](2c90200219e65)      "S-0008f10400411f56"[13]                # lid 32 lmc 0 "SW1 wopr ISR9024D (MLX4 FW)" lid 11 4xDDR

	                                                                                                                     ^^^
	                                                                                                           (but here DDR.)

	It turns out that when you first discover a switch the port object created gets
	the PortInfo of Port "0".

	This patch queries again for the PortInfo of the port we "came in on".

	Ira

	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 5 Jun 2008 19:03:44 -0700
	Subject: [PATCH] infiniband-diags/src/ibnetdiscover.c: Fix the PortInfo data on the port we discover a switch on.

	   Previously the portinfo data for that port object would be the PortInfo of
	   port "0" of the switch.  Specifically this would cause the speed of the link
	   to be printed incorrectly.

2008-06-08  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/man/saquery: add --smkey option to the man page

2008-06-08  Sasha Khapyorsky  <sashak@voltaire.com>

	saquery: --smkey command line option
	This adds possibility to specify SM_Key value with saquery. It should
	work with queries where OSM_DEFAULT_SM_KEY was used.

	If non-numeric string (like 'x') is provided with --smkey option then
	saquery will prompt to get SM_Key value.

2008-06-06  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/ibsendtrap.c: Support CA and port num
	infiniband-diags/ibsendtrap.c: Support CA and port num
	Also, fix usage

2008-06-06  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/mcm_rereg_test.c: Handle error when guid file not found
	infiniband-diags/mcm_rereg_test.c: Handle error when guid file not found

	ERR: cannot open port_guids.list: No such file or directory

2008-06-03  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/saquery.c: Update for change to osm_mad_pool_init API
	infiniband-diags/saquery.c: Update for change to osm_mad_pool_init API

2008-06-01  Hal Rosenstock  <hrosenstock@xsigo.com>

	management: Support separate SA and SM keys as clarified in IBA 1.2.1
	management: Support separate SA and SM keys as clarified in IBA 1.2.1

	v2 is just a rebase to latest tree

2008-06-01  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/scripts/*.pl: prevent some zero exists on errors
	Upon failures break execution and drop error status.

	infiniband-diags/scripts/*.pl: exit 2 on usage errors
	Add non-zero exit status (2) on usage errors for perl scripts.

2008-05-31  Matthias Blankenhaus  <matthias@sgi.com>

	saquery: port selection problem
	The problem appears when -C and/or -P are used.

	I have a patch that fixes the problem:

2008-05-29  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/ibprint[ca|switch|rt].pl: allow printing of multiple matches but print warning to user that multiple matches were found
	I think it is useful to print multiple matches found when searching for
	matches.  Specifically when switches have not been named, ie they all have some
	"Mellanox..." or "Voltaire..." name.

	This prints all matches but also warns the user at the end that it found X
	matches.

	Ira

	From 11b85c9b526b9067aa12eac5d445d8ee43a7d024 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 23 May 2008 16:25:19 -0700
	Subject: [PATCH] infiniband-diags/scripts/ibprint[ca|switch|rt].pl: allow printing of multiple
	 matches but print warning to user that multiple matches were found

2008-05-29  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/ibprint[ca|switch|rt].pl: fix printing by name
	I guess when I added support to search by GUID I must have broken the printing
	by name.  This changes these scripts to use the common convention of "-G" to
	specify that a GUID is to be searched for and fixes the printing when a name is
	specified.

	Ira

	From e9b6766bd6b3661a5bc1e78c9e95784a99a631c0 Mon Sep 17 00:00:00 2001
	From: Ira Weiny <weiny2@llnl.gov>
	Date: Fri, 23 May 2008 16:19:57 -0700
	Subject: [PATCH] infiniband-diags/scripts/ibprint[ca|switch|rt].pl: fix printing by name

2008-05-14  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/iblinkinfo.pl: fix printing of switch name when port 1 is down.
	I found a bug in the printing of the names of switches on iblinkinfo.pl.  The
	name of the switch was being pulled from the first ports "link" structure.  The
	problem is, if the first port is down there was no structure available.  This
	gets the switch name from the first link structure available and prints the
	name correctly.

	Ira

	From 9b69c0ff4c7785be78157ab78e4a4892d64e2fb2 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 1 May 2008 15:46:25 -0700
	Subject: [PATCH] infiniband-diags/scripts/iblinkinfo.pl: fix printing of switch name when port 1
	is down.

2008-05-14  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	infiniband-diags/Makefile.am: fix location of ibdiag_version.h
	Hi Sasha,

	When compiling infiniband-diags not from the source code location,
	compilation fails to find the ibdiag_version.h file - fixing it.

2008-04-03  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump all versions
	Bump all components version in master branch.

2008-03-30  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: pass valid VLArb table size to dump func
	Pass valid VLArb table size to bump function from libibmad.

2008-03-26  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/vendstat.c: Fix port xmit wait handling
	Fix config space accesses for port xmit wait

2008-03-25  Dotan Barak  <dotanb@dev.mellanox.co.il>

	management: Remove extraneous semicolon from several files
	Remove extraneous semicolon from several components in the management.

2008-03-24  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/ibping.c: Remove extraneous semicolon

2008-03-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsendtrap.c: indentation fixes

2008-03-24  Ira Weiny  <weiny2@llnl.gov>

	Add mcm_rereg_test to test-utils option.
	From 60a6563c1f726137a188a024481f25cbab9b140b Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 21 Mar 2008 14:44:31 -0700
	Subject: [PATCH] Add mcm_rereg_test to test-utils option.
	Fix mcm_rereg_test.c to use ibmad_gid_t

2008-03-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibsendtrap.c: add include files
	Add include files - stdlib.h for exit() and string.h for strtoul().

2008-03-24  Ira Weiny  <weiny2@llnl.gov>

	Add an optional test utility 'ibsendtrap'
	From 9294bf0041e88a077918619d5a93be318075cd74 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 21 Mar 2008 14:26:05 -0700
	Subject: [PATCH] Add an optional test utility 'ibsendtrap'
	use --enable-test-utils to build this tool

2008-03-24  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/set_nodedesc.sh: enhance to be able to set names other than hostname and to provide feedback on the names assigned
	From 9553461b42d8dde7418b9eb3c07a8e78b58a32f1 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Mon, 17 Mar 2008 14:46:45 -0700
	Subject: [PATCH] infiniband-diags/scripts/set_nodedesc.sh: enhance to be able to set names other

	than hostname and to provide feedback on the names assigned.

2008-03-20  Sasha Khapyorsky  <sashak@voltaire.com>

	saquery: trivial: remove empty line

2008-03-20  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	infiniband-diags/saquery: print SL in MCast groups
	Printing SL in the MCast groups query.

2008-03-20  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	infiniband-diags/Makefile.am: fix 'make install'
	When the code compiled not from the source dir itself, 'make install'
	fails. Fixing the wrong file location.

	I'd like to see this fix in ofed_1_3 as well, but it's your call.

2008-03-14  Albert L. Chu  <chu11@llnl.gov>

	add router support to check_lft_balance.pl

2008-03-13  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags: Fix install of IBswcountlimits.pm script
	infiniband-diags: Fix install of IBswcountlimits.pm script

2008-03-10  Ira Weiny  <weiny2@llnl.gov>

	Rename ib_gid_t in mad.h to mad_gid_t to prevent name collision with ib_types.h
	Ok, I added that as well, but that only reports a compile error when someone
	uses it.  Without the USE_DEPRECATED_IB_GID_T macro "ib_gid_t" still conflicts
	with ib_types.h.  I really want to fix that problem.  What I like about the
	macro is if someone has a lot of uses in their code this gives them an easy way
	to compile against the new version.

2008-03-04  Albert Chu  <chu11@llnl.gov>

	check_lft_balance script
	Here's the script I mentioned before that I used for the balance checking
	earlier.  Its nothing fancy but probably could be useful to others.

2008-03-04  Ira Weiny  <weiny2@llnl.gov>

	Update documentation for guid format
	From b7581a29d6ad58de01c2dc6c2bce420f300210ab Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Mon, 3 Mar 2008 11:01:00 -0800
	Subject: [PATCH] Update documentation for guid format

	infiniband-diags/scripts/ib[linkinfo][queryerrors].pl: report switch not found

	Fix bug which prevented some GUIDs from being found due to formating issues.

2008-03-01  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/ibprintswitch.pl: fix printing of ports
	When there was only one switch in the system this would fail to print the
	switch ports as designed.  This fixes this and makes the code consistent
	with printca and printrt.

2008-02-29  Ira Weiny  <weiny2@llnl.gov>

	Create script to automate perltidy command

2008-02-20  Jeremy Brown  <jeremy.brown@qlogic.com>

	ibstatus - small script change
	The ibstatus script, in infiniband-diags, doesn't tolerate a missing
	device very well. This small fix makes the output a bit more useful.

	Before: $ ibstatus
	/usr/bin/ibstatus: line 60: cd: /sys/class/infiniband: No such file or directory
	Fatal error:  device '8cpu-mdout.8hoursamber8-8hr-2007-09-06-01:52:23': sys
	files not found
	(/sys/class/infiniband/8cpu-mdout.8hoursamber8-8hr-2007-09-06-01:52:23/ports)

	After: $ ibstatus
	Fatal error:  No devices

	Thanks for considering this!

	Jeremy

2008-02-11  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump versions in master branch

2008-01-31  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: perl scripts formatting
	Perl scripts formatting with:

	  perltidy -pt=2 -sbt=2 -bt=2 -nsfs -b -t -nola -ce -sbl -nbbc

2008-01-31  Timothy A. Meier  <meier3@llnl.gov>

	opensm: diags bug fix in iblinkinfo.pl
	Sasha,

	Sorry, I missed this in my previous patch.

	 From b9bd2d2e5be0121c148fe7087ca7e6cce357a55e Mon Sep 17 00:00:00 2001
	From: Tim Meier <meier3@llnl.gov>
	Date: Wed, 30 Jan 2008 16:00:50 -0800
	Subject: [PATCH] opensm: diags bug fix in iblinkinfo.pl

	Fixes -D bug determining guid from dr

2008-01-31  Tim Meier  <meier3@llnl.gov>

	opensm: diags better error checking for DR option
	The -D option for specifying a direct route for both
	iblinkinfo.pl and ibqueryerrors.pl aborts with an
	error message if the route does not lead to a switch.

2008-01-31  Tim Meier  <meier3@llnl.gov>

	infiniband-diags: formatting some perlscripts
	Some perl scripts formatting with

	  perltidy -pt=2 -sbt=2 -bt=2 -nsfs -b -t -nola -ce -sbl -nbbc

2008-01-27  Hal Rosenstock  <hrosenstock@xsigo.com>

	management: Update License: field in management spec files
	Update License: field to match the exact format given in
	http://fedoraproject.org/wiki/Packaging/LicensingGuidelines
	for a package available under a choice of GPL or BSD license.

	infiniband-diags: Add missing COPYING file
	infiniband-diags: Add missing COPYING (license) file and update spec
	accordingly

2008-01-26  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: remove trailing whitespaces

2008-01-26  Ira K. Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/iblinkinfo.pl: Fix switch to switch output with new
	ibnetdiscover output.

	   ibnetdiscover now prints port guids.  This format change cause the regex to
	   break when parsing switch to switch links.  Fix this by parsing for the
	   remote port guid.  And while we are at it add an option to print the port
	   guids parsed.

2008-01-26  Tim Meier  <meier3@llnl.gov>

	opensm: update man pages for diags DR support
	Added man -D information for ibquerryerrors and iblinkinfo

	opensm: diags add DR path support to some utils
	Added direct route support to iblinkinfo.pl and ibqueryerrors.pl.

2008-01-24  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: code consolidation
	Some code consolidation and cleanup.

2008-01-23  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: usage field to struct query_cmd
	Add usage field to struct query_cmd. When initialized this usage text
	will be shown as output of 'saquery --help' for the specific query type.

2008-01-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/ibprintca.pl: fix regex when searching for node by name
	From 6a94e0a34428a067c125bcb11ed9f306e60c35c9 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:24:37 -0800
	Subject: [PATCH] infiniband-diags/scripts/ibprintca.pl: fix regex when searching for node by
	name

	infiniband-diags/scripts/ibprintrt.pl: fix regex when searching for router by name
	From efcac0f06884c15c9a3abf97dee99953d9c5112f Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:27:40 -0800
	Subject: [PATCH] infiniband-diags/scripts/ibprintrt.pl: fix regex when searching for router by
	name

	infiniband-diags/scripts/ibprintswitch.pl: fix regex when searching for switch by name
	From ac265e5c7130bf9a6b43ae2ed86ae342dbfe6bc0 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:26:50 -0800
	Subject: [PATCH] infiniband-diags/scripts/ibprintswitch.pl: fix regex when searching for switch
	by name

	Update perl script man pages for -C and -P options
	From 967a1c0c87d86ce4d86d3c870692ec6ea494a5ea Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:59:38 -0800
	Subject: [PATCH] Update perl script man pages for -C and -P options

	Add -C <ca_name> and -P <ca_port> to ibprint[ca|rt|switch].pl
	From 5b2317bf383d22e9ccb87d3c1d24735b3e508ab4 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:21:28 -0800
	Subject: [PATCH] Add -C <ca_name> and -P <ca_port> to ibprint[ca|rt|switch].pl

2008-01-20  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/iblinkinfo.pl: clean up output format
	From b29ba8352381f21c178a57d5e0bac0d82bfb64ab Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 16:06:06 -0800
	Subject: [PATCH] infiniband-diags/scripts/iblinkinfo.pl: clean up output format

	   - make "line mode" explicit in the code
	   - Better align output fields

2008-01-19  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/scripts/IBswcountlimits.pm: clean up long lines
	From 76499fd3eef2ec847e9967e10f87eb8ecbbfc97d Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 11:46:27 -0800
	Subject: [PATCH] infiniband-diags/scripts/IBswcountlimits.pm: clean up long lines

	infiniband-diags/scripts/IBswcountlimits.pm: Fix comment
	From 96eb9de7b1918766020e7d9621d79d86949cfc39 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 18 Jan 2008 11:39:01 -0800
	Subject: [PATCH] infiniband-diags/scripts/IBswcountlimits.pm: Fix comment

2008-01-19  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/infiniband-diags.spec.in: add Obsoletes rule
	This addresses bug#854 and adds Obsoletes rpm rule for infiniband-diags
	packages, which should help with upgrades (from OFED-1.2x where it was
	named openib-diags).

2008-01-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/configure.in: complib doesn't have opensm dependencies anymore
	complib doesn't have dependency from opensm libs (libopensm and
	libosmvendor), so cleanup AC_CHECK_LIB flags.

2008-01-10  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: support for VLArb and PKey Table Records
	This adds support for VLArbTableRecord and PKeyTableRecord attributes.

2008-01-10  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: support ports with LinkRecord query
	Ports can be selected now with LinkRecord query. This should be passed
	in commnad line in format [LID]/[port]. Examples:

	  saquery LinkRecord - queries all LinkRecords in a fabric
	  saquery LR 28 - all LinkRecords for lid 28
	  saquery LR 28/3 - all LinkRecords for lid 28 port 3
	  saquery LR 28/3 36/4 - all LinkRecords from lid 28 port 3 to lid 36 port 4
	  saquery LR /3 - all LinkRecords from ports 3 in a fabric

2008-01-10  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: support for SL2VLTableRecord attribute
	This adds support for SL2VLTableRecord attribute, The port numbers could
	be optionally specified in command line together with LID in follow
	format: <LID>/[in-port]/[out-port]. Examples:

	  saquery SL2VLTableRecord 28 - query lid 28
	  saquery SL2VL 28/1/3        - query lid 28, in-port 1, out-port 3
	  saquery SL2VL 28//3         - query lid 28, out-port 3

2008-01-10  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/saquery: attribute names support
	This let to pass requested via command line SA attribute by name.
	Examples:

	  saquery NodeRecord
	  saquery NR <lid>

	Main motivation for this addition is that I cannot find appropriate free
	characters for adding new attributes (specifically PKeyTableRecord and
	SL2VLTableRecord - p, P, s, S are used already).

	This preserves a command line options currently used for same
	purposes.

2008-01-07  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: update version according to the tree state
	Update the version according to the tree state. Similar to OpenSM
	version update.

	infiniband-diags: remove not needed anymore __BUILD_VERSION_TAG__
	Remove not needed anymore __BUILD_VERSION_TAG__ macro.

	infiniband-diags: use common build version
	Use common version (as defined in configure.in) instead of per tool
	__BUILD_VERSION_TAG__.

2008-01-07  Sasha Khapyorsky  <sashak@voltaire.com>

	management/*/Makefile.am: update configure.in version on 'make dist'
	This is similar gen_ver.sh usage in opensm/Makefile.am.

	This updates the package version stored in configure.in in newly created
	by 'make dist' package tarball. The version is generated by gen_ver.sh
	in accordance with the tree state.

2008-01-06  Sasha Khapyorsky  <sashak@voltaire.com>

	management/*/Makefile.am: fix ChangeLog generation rule
	Then individual package is detached from the main source tree, running
	'make dist' in this detached sub-tree will fail, it is because
	../gen_chlog.sh script will not be found. Fix it, make ChangeLog
	generation on dist-hook optional.

2008-01-03  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/man/ibnetdiscover.8: Add newly added ports report option
	infiniband-diags/man/ibnetdiscover.8: Add newly added ports report
	option

2008-01-03  Sasha Khapyorsky  <sashak@voltaire.com>

	management: bump versions
	Bump versions of the management sub projects:
	libibcommon, libibumad, libibmad, opensm, infiniband-diags.

	infiniband-diags/dump_lfts.sh: fix switch DR Path parser
	It is highly possible that ibnetdiscover finds the same switches at
	different paths. In this case the current version of the script will
	dump LFTs multiple times for the same switches. This patch fixes this
	bug - all discovered paths are sorted and unified by switch GUIDs.

	infiniband-diags/dump_lfts.sh: adopt DR Path parser
	Adopt DR Path parser to the new ibnetdiscover format, where DR Path is
	shown as comma separated list.

2007-12-25  Erez Strauss  <erezs@voltaire.com>

	ibnetdiscover - ports report
	The patch introduce additional output mode for ibnetdiscover which is
	focused on the ports, and print one line for each port with the needed
	port information.

2007-12-22  Sasha Khapyorsky  <sashak@voltaire.com>

	manangement: kill __WORDSIZE macro checks
	Kill #if __WORRDSIZE == 64 checks.

2007-12-13  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/scripts: fix perfquery usage
	It similar to ibcheckerrors fix - don't use 'all port' option for CA
	ports querying and reset.

	infiniband-diags/ibcheckerrors: fix port errors count
	This fixes port error counter (pcnterr) update. Also removes unsued
	variables.

2007-12-11  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibcheckerrors: for CAs query only single ports
	For CAs query performance counters only for single ports by lid and port
	number, and not whole node with 'all ports' option.

	This patch targets bug#810
	(https://bugs.openfabrics.org/show_bug.cgi?id=810): diags fail on
	two-port HCAs (2 ports; only 1 supported currently)

2007-12-04  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/man: add -x option to saquery man page
	Add -x option (LinkRecord info) to saquery man page.

	infiniband-diags/saquery: allow empty src and/or dst with --src-to-dst option
	Allow to use empty source and/or destination fields with --src-to-dst
	option, in this case empty field will not be marked in comp_mask.
	Examples: '--src-to-dst :48', '--src-to-dst 4:', '--src-to-dst :'.

	infiniband-diags/saquery: LinkRecord query support
	Add support for SA LinkRecord query.

	infinibad-diags/saquery: move lid resolving functions
	Move lid resolving functions, so it will be useful by query processors.

	infiniband-diags/saquery: add get_any_records() function
	Add get_any_records() function - this gets attribute specific data (id,
	comp_mask, offset, etc.) as parameters.

2007-11-19  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: version 1.3.4
	infiniband-diags version 1.3.4.

2007-11-19  Rolf Manderscheid  <rvm@obsidianresearch.com>

	PATCH diags: add saquery option to get path records by GIDs
	Hi Sasha,

	This patch adds a --sgid-to-dgid option to saquery, useful for
	validating configuration of opensm prefix routes.

	    Rolf

2007-11-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibtracert: print lids in decimal form
	As stated in bug#504 (https://bugs.openfabrics.org/show_bug.cgi?id=504)
	lid output format unification is needed. Print LIDs as decimal in
	ibtracert.

2007-11-14  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: remove utf-8 characters from sources and man pages

2007-11-06  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: version 1.3.3
	Version 1.3.3

2007-11-05  Ira K. Weiny  <weiny2@llnl.gov>

	Change node name map implementation to use qmap in memory storage

2007-11-04  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/grouping.c: For Xsigo grouping, fix seg fault when no peer port
	infiniband-diags/grouping.c: For Xsigo grouping, fix seg fault when no
	peer port

2007-11-04  Ira Weiny  <weiny2@llnl.gov>

	Move nodenamemap out of infiniband-diags into libosmcomp
	From fe2756789ffbc69466eefea3cdffe200a0718561 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 1 Nov 2007 15:00:37 -0700
	Subject: [PATCH] Move nodenamemap out of infiniband-diags into libosmcomp

2007-11-02  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/man/smpquery: fix typo

2007-11-02  Ira Weiny  <weiny2@llnl.gov>

	Change switch map function names to reflect the new functionality of a simple node name map.
	From b4f6bbea815aaa91837d464f882d30405ffe9d98 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Wed, 31 Oct 2007 16:34:44 -0700
	Subject: [PATCH] Change switch map function names to reflect the new functionality of a simple

	node name map.

2007-11-02  Ira Weiny  <weiny2@llnl.gov>

	use lookup_switch_name for all node types.
	From f451c7b352b44747d1eb6f4627109a3f12b757e6 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Wed, 31 Oct 2007 16:22:12 -0700
	Subject: [PATCH] use lookup_switch_name for all node types.

	infiniband-diags/src/smpquery.c : special case situation where nodename len is >= 32
	From 04c67433708cb14ab384a2acfc19755998f7a8b2 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 1 Nov 2007 19:31:30 -0700
	Subject: [PATCH] infiniband-diags/src/smpquery.c : special case situation where nodename len is >= 32

2007-10-27  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/configure.in: fix comment
	From b338078dc970c09513dd1d3023bebff334010c05 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 19 Oct 2007 11:22:40 -0700
	Subject: [PATCH] infiniband-diags/configure.in: fix comment

2007-10-23  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/sminfo: Fix activity count display
	infiniband-diags/sminfo: Fix activity count display

	infiniband-diags/scripts: Eliminate some duplicated messages
	infiniband-diags/scripts: Eliminate some duplicated messages

2007-10-23  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	osm: QoS - adding CPI:CapabilityMask2 and turning on QOS_SUPPORTED bit
	Adding ClassPortInfo:CapabilityMask2 field and turning
	on OSM QoS capability bit (OSM_CAP2_IS_QOS_SUPPORTED).

2007-10-21  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags: Formalize BuildRequires for rpmbuild
	From 33d2c9cca44ce13aa8f35b2228369a33f7a45a70 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Wed, 17 Oct 2007 15:23:55 -0700
	Subject: [PATCH] Formalize BuildRequires for rpmbuild

	   the mock build tool in particular requires specific build requires

2007-10-21  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/scripts: Updated for perfquery support of no AllPortSelect option
	infiniband-diags/scripts: Updated for perfquery support of no
	AllPortSelect option

	Eliminate new ibwarn message added to perfquery to let user know
	AllPortSelect option is not supported by specified PMA

	Tested-by: Greg Kurtzer <gmk@lbl.gov>

2007-10-21  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/perfquery.c: Support PMAs which don't support AllPortSelect option
	infiniband-diags/perfquery.c: Support PMAs which don't support
	AllPortSelect option

	Currently only support single port HCAs are supported in this mode but
	this can be extended for other devices if needed

	Tested-by: Greg Kurtzer <gmk@lbl.gov>

2007-10-14  Sasha Khapyorsky  <sashak@voltaire.com>

	libibcommon,infiniband-diags: move get_build_version() to diags
	Move get_build_version() function (which is ifdefed by
	__BUILD_VERSION_TAG__) to infiniband-diags, where it is only used.

2007-10-13  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/perfquery.c: Fix issues when checking PerfMgt:ClassPortInfo.CapabilityMask
	infiniband-diags/perfquery.c: Fix issues when checking
	PerfMgt:ClassPortInfo.CapabilityMask

	1. bit 9, if we're counting from 0, will have mask of 0x200,
	   not 0x100. mask of 0x100 will be for counter aggregation according
	   to IBA 1.2.

	2. If capmask is 16 bit big-endian word, then we're looking
	   at the wrong byte on x86, we must ntohs(*pc2) first.

	3. Also, change pointer dereference with memcpy,
	e.g.:

	           memcpy (&capmask, pc+2, sizeof(capmask));
	           capmask = ntohs(capmask);

	Those pointer dereferenes are royal pain on ia64 unless you can
	guarantee what pc is always aligned properly.

	Found-by: Max Matveev <makc@sgi.com>

	Compile tested only

2007-09-20  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: update version to 1.3.2
	Update version to 1.3.2.

	management/*/*.spec.in: change Source field to OFA paths
	Change Source field to point OFA site paths

2007-09-19  Sasha Khapyorsky  <sashak@voltaire.com>

	management/*/configure: RELEASE and TARBALL for spec files
	RELEASE and TARBALL variables are used as substitution pattern in
	*.spec.in files, but never defines. As result ./configure script
	produces invalid *.spec files for management sub-projects. This patch
	fixes this.

	Values are defined as "unknown" and "$PACKAGE-$VERSION.tar.gz", but
	could be overwritten with environment variables, like this:

	  RELEASE=ofed_7.8 TARBALL=opensm-7.8.9-customized.tar.gz ./configure

2007-09-19  Sasha Khapyorsky  <sashak@voltaire.com>

	management/*/Makefile.am: generate ChangeLog files when distributed
	Similar to opensm - this will generate up-to-date ChangeLog files for
	all management sub-project on 'make dist'.

	management/*/Makefile.am: clean dist-hook target
	Remove unneeded things from dist-hook target.

	management/*/*.spec.in: don't run autogen.sh
	Don't run autogen.sh script for rpm generation - valid tarballs should
	have already generated ./configure.

	management/*/Makefile.am: fix include paths
	Fix include paths for management libraries and diags builds - since
	it will always linked against installed libs, we will use only installed
	external header files for build.

	libibmad,infiniband-diags: fix include paths
	Use include <infiniband/*.h> (rather than just <*.h>) for external
	management header files from libibcommon and libibumad (also libibmad
	for diags).

	management/*/ChangeLog: trivial spacing fixes

	management/*/Makefile.am: add autogen.sh script to 'make dist'
	Somebody could want to regenerate auto* stuff with distributed tarball.
	So add autogen.sh to archive.

2007-09-18  Hal Rosenstock  <hrosenstock@xsigo.com>

	infiniband-diags/ibnetdiscover: Bump build version
	infiniband-diags/ibnetdiscover: Bump build version

2007-09-18  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags/ibnetdiscover: trivial code simplifications
	Trivial code simplifications - mainly flows consolidation and indentation
	fixes.

2007-09-18  Hal Rosenstock  <hrosenstock@xsigo.com>

	ibnetdiscover: Support Xsigo chassis grouping
	I think this also fixes a bug with grouping of multiple non Voltaire
	chassis as well.

	Note: this patch is against OFED 1.2

2007-09-17  Sasha Khapyorsky  <sashak@voltaire.com>

	opensm and all: fix Source field in spec.in files.

2007-09-15  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: use shared ibdebug var
	Use shared ibdebug variable over diag utils

2007-09-10  Ira Weiny  <weiny2@llnl.gov>

	Fix regexp's for new ibnetdiscover output
	The ibnetdiscover output has changed so this command was failing.  I am not sure when this happened but not matter this should fix it.

	Ira

	From 9aadfb84826a5ea31107624b4b29e90d7c97e55b Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 7 Sep 2007 14:34:09 -0700
	Subject: [PATCH] Fix regexp's for new ibnetdiscover output

2007-09-10  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/ibdiag_common.c: do not print warning of failed default switch map open
	From 59f8772d60a4b061eb2e27ded9abecc9b9e83d5c Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 7 Sep 2007 15:08:20 -0700
	Subject: [PATCH] infiniband-diags/src/ibdiag_common.c: do not print warning of failed default
	switch map open

	   This really clutters up some of the diag scripts output now that more of the
	   tools support the switch map functionality.

2007-09-10  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/smpquery.c: fix compiler warning
	From a20eaa1b0743aa1cc0c11372c2a989911cb5bcde Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Fri, 7 Sep 2007 15:10:51 -0700
	Subject: [PATCH] infiniband-diags/src/smpquery.c: fix compiler warning

2007-09-10  Ira Weiny  <weiny2@llnl.gov>

	Add -C and -P options to perl diags to be able to use alternate CA's and ports
	We have a few nodes which are connected to multiple fabrics.  The perl diags were unable to specify which port or CA to use.  In our case this left us unable to use these tools on one of the subnets attached.  This patch adds that support.

	Ira

	From b2f95d93e1c2a730f554275cf636ccd687d1106e Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 6 Sep 2007 09:37:10 -0700
	Subject: [PATCH] Add -C and -P options to perl diags to be able to use alternate CA's and ports

	infiniband-diags/scripts/IBswcountlimits.pm
	infiniband-diags/scripts/ibfindnodesusing.pl
	infiniband-diags/scripts/iblinkinfo.pl
	infiniband-diags/scripts/ibqueryerrors.pl

2007-08-31  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/Makefile.am: clean up extra directives
	From 8da03fd7526e0879404425609abea7bc2da0d7f0 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 30 Aug 2007 09:22:29 -0700
	Subject: [PATCH] infiniband-diags/Makefile.am: clean up extra directives

2007-08-24  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/src/smpquery.c: Add switch-map support for nodedesc query
	From 5bc16ccf9e476d717c8ad02034086e6cfdae3584 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 23 Aug 2007 14:43:40 -0700
	Subject: [PATCH] infiniband-diags/src/smpquery.c: Add switch-map support for nodedesc query

2007-08-23  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	osm: QoS - adding new PathRecord fields
	Adding QoS fields to PathRecord.

2007-08-15  Ira Weiny  <weiny2@llnl.gov>

	iblinkinfo.pl: Remove annoying double quote around switch name.
	From fc0251b5150afdd5d40d55facdf85c645939b482 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Tue, 14 Aug 2007 18:04:43 -0700
	Subject: [PATCH] iblinkinfo.pl: Remove annoying double quote around switch name.

2007-08-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/set_mthca_nodedesc.sh: change to work for all HCA's
	From 3c78f9f7797a9029b33a1a80476f82929f8a7ad7 Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 9 Aug 2007 18:51:09 -0700
	Subject: [PATCH] Change the name of set_mthca_nodedesc.sh to set_nodedesc.sh because it should

	work for all HCA types now.

2007-08-12  Ira Weiny  <weiny2@llnl.gov>

	infiniband-diags/set_mthca_nodedesc.sh: change to work for all HCA's
	From 3abf9ac10ee90f7272969486704c2989d254a2ac Mon Sep 17 00:00:00 2001
	From: Ira K. Weiny <weiny2@llnl.gov>
	Date: Thu, 9 Aug 2007 16:07:38 -0700
	Subject: [PATCH] Modify set_mthca_nodedesc to set the node_desc for any hca type found in sysfs.

	This has been tested for the MLX4 and mthca cards but not others.

2007-08-08  Bernd Schubert  <bs@q-leap.de>

	ibnetdiscover: man page improvement
	From: Bernd Schubert <bs@q-leap.de>

	> > I'm also not sure about the pathes, IMHO the man page of ibnetdiscover
	> > should give one more example, so
	> >
	> >
	> >
	> >       -D      use directed path address arguments. The path
	> >               is a comma separated list of out ports.
	> >               Examples:
	> >               "0"             # self port
	> >               "0,1,2,1,4"     # out via port 1, then 2, ...
	> >
	> >
	> > "out via port 1, then out via port 2, then out via port 1, ..."
	> >
	> > or
	> >
	> > "out via port 1, then in port 2, then out via port 1, ..."
	> >
	> >
	> > You see what I mean?
	>
	> It's the former. It's the out port on each hop along the path.

	Thanks, do you mind to apply this patch? I know the text already says
	"separated list of out ports", but for those who don't believe it
	like me ;) the comment will make it a bit more convincing.

2007-08-08  davem@systemfabricworks.com  <davem@systemfabricworks.com>

	infiniband-diags/scripts: Fix Bug 239 Error Reporting
	   Fix Bug 239 OpenIB diag scripts don't return error when lacking umad
	   permissions.  Returning the error from the head of a shell pipeline is a
	   problem, so this fix causes the awk scripts to pass error messages through.
	   This will pass all standard error messages.

	   This patch needs [ofa-general] [PATCH] infiniband-diags: Add common flags
	   -P, -C, and -t (posted Tue Jul 31 13:39:27 PDT 2007) applied first.

2007-08-05  davem@systemfabricworks.com  <davem@systemfabricworks.com>

	infiniband-diags: Add common flags -P, -C, and -t
	   Add common flags -P, -C, and -t to infiniband-diags programs and scripts to
	   allow specifying the HCA port number, HCA device name, and query timeout.
	   These diagnostic programs can now be directed to either different fabrics
	   attached to the system, or forced to use different ports should the fabric
	   fail and comparisons are needed.

	   Two of these had conflicting prior use of the flags.  For the ibcheckerrs
	   script, -T is now used to specify the threshold file.  In the saquery
	   program, -p and -c are now used to specify getting the PathRecord info
	   and getting the SA's class port info.

	   Other than the resolution of the three conflicts, all comands behave exactly
	   the same as they did before the change if these common flags are not used.

2007-08-04  Yevgeny Kliteynik  <kliteyn@dev.mellanox.co.il>

	osm: cosmetics - removing trailing blanks
	Hi Sasha.

	This patch removes trailing blanks in bunch of files

2007-07-22  davem@systemfabricworks.com  <davem@systemfabricworks.com>

	infiniband-diags/ibnetdiscover: Fix DDR link speed decode
	   Fix ibnetdiscover DDR link speed decode by moving string from [3] to [2].

	infiniband-diags/scripts: Handle new and old topology file format
	   Fix infiniband-diags scripts to handle changed ibnetdiscover topology file
	   format and remain backward compatible with old file format.

2007-07-19  Sasha Khapyorsky  <sashak@voltaire.com>

	infiniband-diags: remove trailing whitespaces
	Remove trailing whilespaces.

2007-07-10  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags: Updated to version 1.3.1

2007-06-20  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags: Use diag common code ib_error routine

2007-06-18  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/src/ibtracert.c: Cosmetic change to IBWARN message

	infiniband-diags/man/ibaddr.8: Improve description

2007-06-07  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/scripts/ibidsverify.pl: Validate port GUID against node GUIDs Note only one port GUID can duplicate a node GUID

	infiniband-diags/scripts/ibidsverify.pl: Normalize portguid format to be the same as nodeguid format (0x preceeding the guid)

	infiniband-diags/scripts/ibidsverify.pl: Fix parsing of node and port GUIDs for switches

	infiniband-diags/scripts/ibidsverify.pl: Cosmetic formatting changes

2007-06-05  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/ibidsverify.pl: Support port GUID validation
	Note that original topology file format without port GUIDs is also
	supported in which case this validation is omitted.

	infiniband-diags/ibnetdiscover: Add port GUIDs to topology file

	infiniband-diags/man/ibnetdiscover.8: Fix a typo

	infiniband-diags/scripts/ibidsverify.pl: Generalize validate_non_zero_guid for GUID rather than NodeGUID to ready for adding PortGUID verification

	infiniband-diags/src/ibnetdiscover.c: Minor changes to minimize differences for adding port GUID into topology file

	infiniband-diags/src/ibnetdiscover.c: In get_node, set portguid to NodeInfo PortGUID rather than NodeGUID

	infiniband-diags/man/ibnetdiscover.8: Added text description for active link width and link speed

2007-06-04  Hal Rosenstock  <halr@voltaire.com>

	ibnetdiscover: Add link width and speed to topology file output

	infiniband-diags/src/ibnetdiscover.c: Cosmetic formatting change

2007-06-02  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/man/ibnetdiscover.8: Add topology file format section

	infiniband-diags/man/ibnetdiscover.8: Add grouping information

	infiniband-diags/ibnetdiscover: Fix list by nodetype operations

2007-06-01  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/man/ibnetdiscover.8: Cosmetic changes

	infiniband-diags/ibnetdiscover: Add supporti for -R(outer_list) option

	infiniband-diags/Makefile.am: Add ibidsverify

	infiniband-diags/scripts/ibidsverify.pl: Add initial version of ibidsverify.pl script

	infiniband-diags/scripts/IBswcountlimits.pm: Fixed typo and other cosmetic changes

	infiniband-diags/scripts/iblinkinfo.pl: Cosmetic formatting change

	infiniband-diags/scripts/ibqueryerrors.pl: Cosmetic change

	infiniband-diags/man/ibidsverify.8: Add man page for ibidsverify.pl script

2007-05-31  Hal Rosenstock  <halr@voltaire.com>

	management/*.spec: Remove *.spec files from management git repo as committed files. They are generated files.

	management/*.spec: Latest daily versions of .spec files (prior to removing these due to build changes)

	infiniband-diag/man/(ibprintca ibprintswitch ibprintrt).8: Add description of list capability to man pages

	infiniband-diags: Add ibdatacounters

	infiniband-diags/ibdatacounters: Add script and man page for subnet wide data counters

	infiniband-diags/configure.in: Change IBSCRIPTPATH from bindir to sbindir

2007-05-30  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/(Makefile.am configure.in): Add ibrouters and ibprintrt.pl

	infiniband-diags: Add scripts and man pages for display of IB routers (ibrouters.in, ibprintrt.pl, ibrouters.8, ibprintrt.8)

	infiniband-diags/scripts/ibprintca.pl: Fix bug introduced in last commit which caused compilation error

	infiniband-diags/scripts/ibqueryerrors.pl: Add GUID to output line for ports

	infiniband-diags/scripts/(ibcheckerrs ibcheckport ibcheckportstate ibcheckportwidth ibdatacounts).in: Add lid and port into verbose output

	infiniband-diags/scripts/(ibcheckerrs ibcheckport ibdatacounts).in: Changed portnum to port in output

	infiniband-diags: Updated configure.in and Makefile.am for ibdatacounts script and man page

	infiniband-diags/man/ibdatacounts.8: Cosmetic change

	infiniband-diags/ibdatacounts.(in 8): Add script to display only data counters and associated man page
	Also, cosmetic change to ibcheckerrs man page

2007-05-26  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/scripts/IBswcountlimits.pm: Fix node description parsing for switches

	infiniband-diags/scripts/iblinkinfo.pl: Add peer port link width and speed validation

	infiniband-diags/ChangeLog: Add log entry for change to IBswcountlimits.pm for router support

	infiniband-diags/scripts/IBswcountlimits.pm: Add support for routers

2007-05-25  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/ChangeLog: Add log entries for script changes to IBswcountlimits.pm and iblinkinfo.pl

	infiniband-diags/scripts/iblinkinfo.pl: Display remote LID with peer port info

	infiniband-diags/scripts/IBswcountlimits.pm: Add support for rem_lid in get_link_ends subroutine

	infiniband-diags/src/ibportstate.c: Handle peer ports at 1x that should be wider and 2.5 Gbps that should be faster

	infiniband-diags/src/ibportstate.c: Add LinkSpeed/Width related components to output

2007-05-24  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/scripts/ibprintca.pl: Add support for routers

2007-05-23  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/scripts: Add support for IB routers in existing scripts

2007-05-23  Ira K. Weiny  <weiny2@llnl.gov>

	infiniband-diags/infiniband-diags.spec.in: Move commands from bin to sbin

2007-05-22  Hal Rosenstock  <halr@voltaire.com>

	Diags/man: Cosmetic changes

2007-05-21  Hal Rosenstock  <halr@voltaire.com>

	diags/man: Fixed some formatting

2007-05-15  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/Makefile.am: Move diag programs and scripts from bin to sbin

	Merge branch 'master' of git://staging.openfabrics.org/~halr/management
	Merge Doug's RPM packaging changes into master

2007-05-14  Hal Rosenstock  <halr@voltaire.com>

	spec summary changes

2007-05-10  Hal Rosenstock  <halr@voltaire.com>

	infiniband-diags/configure.in: Changes for changing from openib-diags to infiniband-diags

2007-04-12  Hal Rosenstock  <halr@voltaire.com>

	Diags: Renamed openib-diags to infiniband-diags Also, some other changes based on this (openib-diags.spec.in moved to infiniband-diags.spec.in, openib-diags.spec moved to infiniband-diags.spec, and Makefile.am changed for this)
