Re: [ewg] OFA Management maintainership
Hi, I'm finishing my work for Voltaire those days and wish to transfer my role as OFA management packages maintainer to Alex Netes ale...@voltaire.com which I know many years as a great experienced engineer and very good and positive person. So starting from today his trees should be considered as master development trees: git://git.openfabrics.org/~alexnetes/libibumad git://git.openfabrics.org/~alexnetes/opensm git://git.openfabrics.org/~alexnetes/libibmad git://git.openfabrics.org/~alexnetes/infiniband-diags git://git.openfabrics.org/~alexnetes/ibsim It is also likely that in a near feature maintainerships of libibumad and infiniband-diags will be taken by Ira Weiny wei...@llnl.gov. I would like to wish to Alex and Ira a lot of success with their roles. Also I would like to thank a whole community for good working time. I still be reachable by my email address sashakv...@gmail.com, so feel free to contact me in case of any question. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFA Management maintainership
On 14:17 Wed 09 Feb , Hal Rosenstock wrote: It is also likely that in a near feature maintainerships of libibumad and infiniband-diags will be taken by Ira Weiny wei...@llnl.gov. Do you mean libibmad rather than libibumad ? Yes, sure - libibmad. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [PATCH] OpenSM Release Notes update
OpenSM Release Notes update. Signed-off-by: Sasha Khapyorsky sas...@voltaire.com --- On 10:35 Tue 14 Sep , Vladimir Sokolovsky wrote: Please update release notes and other relevant documentation for the packages that you own. URL: git://git.openfabrics.org/~tziporet/docs.git Branch: ofed_1_5 opensm_release_notes.txt | 81 - 1 files changed, 79 insertions(+), 2 deletions(-) diff --git a/opensm_release_notes.txt b/opensm_release_notes.txt index 9b5de67..bb7c589 100644 --- a/opensm_release_notes.txt +++ b/opensm_release_notes.txt @@ -3,14 +3,14 @@ Version: OpenSM 3.3.x Repo:git://git.openfabrics.org/~sashak/management.git -Date:Dec 2009 +Date:Sep 2010 1 Overview -- This document describes the contents of the OpenSM 3.3 release. OpenSM is an InfiniBand compliant Subnet Manager and Administration, and runs on top of OpenIB. The OpenSM version for this release -is opensm-3.3.5. +is opensm-3.3.7. This document includes the following sections: 1 This Overview section (describing new features and software @@ -127,6 +127,13 @@ f02f40e opensm: implement 'connect_roots' option in fat-tree routing 748d41e opensm SA DB dump/restore: added option to dump SA DB on every sweep b03a95e complib/cl_fleximap: add cl_fmap_match() function b7a8a87 opensm/include/iba/ib_types.h: adding Congestion Control definitions +fa356f8 opensm: Add support for optimized SLtoVLMappingTable programming +8aaae91 Dimension port order file support +7662eec opensm: Add option to specify prefix to syslog messages +2382cf3 opensm: Add update_desc command to opensm console +7cbe193 opensm: toggle sweeping V3 +6f61d8f opensm/osmeventplugin: added new events to monitor SM +84cf603 opensm/main.c: force stdout to be line-buffered 1.3 Library API Changes @@ -247,6 +254,11 @@ fa90512 opensm/osm_vendor_*_sa: fix incompatibility with QLogic SM 55f9772 opensm: Return single PathRecord for SubnAdmGet with DGID/SGID wild carded 5ec0b5f opensm: compress IPV6 SNM groups to use a single MLID +26e7e83 opensm/osm_lid_mgr: fix couple of duplicate LIDs bugs +fedc419 opensm: Multicast root switch calculation +6772fdb opensm: Fix sl2vl configuration +dd3470f complib/cl_timer.c: fixing cl_timer calculation +ae1bcdd opensm/osm_lid_mgr.c: Allow switch lids to be non LMC aligned 4.2 Other Bug Fixes @@ -507,6 +519,70 @@ c6b4d4a opensm/osm_vendor_ibumad.c: Add transaction ID to osm_vendor_send 520af84 opensm/osm_sa_path_record.c: don't set dgid pointer for local subnet 4a878fb opensm/osm_mcast_mgr.c: fix osm_mcast_mgr_compute_max_hops for managed switch +7c48590 opensm/osm_log.c: add OSM_LOG_SYS to default flags +89f7cb6 opensm/osm_lid_mgr: use 'first_time_master_sweep' flag +0cb7fab opensm: conversion to osm_get_port_by_lid() +9d14fc0 opensm/osm_lid_mgr.c: fix memory leak +c364aa1 opensm/opensm.init.in: fix install warning on SLES11 +1010c9c opensm/osm_sa_path_record.c: livelock in pr_rcv_get_path_parms +4b2cd5e opensm/vendor: fix portguids array size +52bf5b2 opensm/osm_subnet.c: fixing some options to not hot-swappable +8900da0 opensm/osm_subnet.{c,h}: passing options to the event plugins +051c57f Delete port only after GID OUT trap was sent +d4ebf7e opensm/complib/cl_passivelock.h: remove unneeded casting +8fdb17c opensm/complib/cl_types.h: convert cl_status_t to int +fd7fb1e opensm/osm_mcast_mgr.c: preserve root switch calculation functionality +fcb0f3a opensm/osm_mcast_mgr.c: code simplifications +444f559 opensm/osm_mcast_mgr.c: fix bug in MC root switch calculation +041ebcb opensm/osm_mcast_mgr.c: remove redundant casting +3717f53 opensm/osm_sa_pkey_record.c: optimize port selection logic +48352be opensm/osm_mcast_mgr.c: fix memory leak +f3cf83f opensm/complib/cl_ptr_vector.c: fix bug/compiler warning +27c8ebd opensm/osm_subnet.h: remove redundant function definition +f296938 opensm/osm_vl_arb_rcv.c: fix double mutex release bug +00bc48e opensm/osm_port_info_rcv.c: fix compilation warning +8823800 opensm/osm_sa.{c,h}: osm_sa_db_file_dump() return values +f4581f3 opensm/osm_qos.c: Fix typo in OSM_LOG message +e3c790a opensm/osm_update_node_desc(): minor prototype improvement +3cc68cb opensm/osm_vl_arb_rcv.c: Dump table after validating block number +7dbb96e opensm SA DB: dump only if modified +fa2106d opensm/osm_sa_infrominfo.c: fixes and simplifications in lid range check +051a1dd opensm/osm_qos.c: split switch external and end ports setup +a6c0189 opensm/osm_qos.c: merge SL2VL mapping capability check +3fe8efe opensm/osm_slvl_map_rcv.c: verify port number values received from + network +88c372c opensm/osm_slvl_map_rcv.c: fix mutex double release bug +d282093 opensm/osm_slvl_map_rcv.c: fix port parsing on BE machine +8e9dbd3 osm_sa_path_record.c: use PR DGID by reference +7c9d375 osm_sa_path_record.c: separate mutlicast processing code +cb2d18e opensm/osm_sa_path_record.c: MGID must be specified explicitly +bd3932b opensm
Re: [ewg] [ANNOUNCE] management tarballs release
On 13:17 Thu 08 Jul , Hal Rosenstock wrote: d3586e7a17bca99fd384a943f00e259e libibumad-1.3.5.tar.gz 754d93f567393d3b9987a65326f40917 libibmad-1.3.5.tar.gz 5c94d6ee49e9c51c801f6634823b5ad5 opensm-3.3.6.tar.gz ba28f6b5323e6067ca019a999eeaf907 infiniband-diags-1.5.6.tar.gz Shouldn't these versions be labeled/tagged in your management git tree ? Would you do that ? I did, but forgot to push a tags to openfabrics free. Fixed now. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] libibumad.so version bump in OFED 1.5.2
On 12:58 Wed 07 Jul , Todd Rimmer wrote: Is there a way to accomplish this feature without an ABI change? Or should ABI altering features be only included in Major releases (such as OFED 1.6?) I urged about ABI change on the list when this patch was posted. And I don't think that we can change the version back. Keep in mind, that ABI changes mean that 3rd parties must rebuild and re-release their applications. While this isn't so bad for OFED itself and IB vendors, it's a major issue for other 3rd party apps. Introducing a ABI change in a maintenance release just makes things that much more confusing. One alternative would be for OFED 1.5.2 to include both the .2 and .3 versions of the library. We discussed such option - to have a package compat-libs or so, but as far as I can remember no one wanted to take responsibility to maintain this. Another alternative would be to design the structures so there is room for expansion or pointers to opaque datatypes for fields only used by the library. Currently it would mean API redesign and more version changes. Sasha Todd Rimmer Chief Architect QLogic Network Systems Group Voice: 610-233-4852 Fax: 610-233-4777 todd.rim...@qlogic.com www.QLogic.com -Original Message- From: Sasha Khapyorsky [mailto:sashakv...@gmail.com] On Behalf Of Sasha Khapyorsky Sent: Wednesday, July 07, 2010 9:34 AM To: Todd Rimmer Cc: e...@openfabrics.org Subject: Re: libibumad.so version bump in OFED 1.5.2 Hi Todd, On 06:48 Tue 06 Jul , Todd Rimmer wrote: I recently discovered that 3rd party applications built for OFED 1.5 will not run on OFED 1.5.2rc2. The issue I found was that libibumad had changed from version 2 to version 3. Can you provide insight as to why this change was made? In reviewing the source code and change logs there do not seem to be any changes which would impact the application API nor application ABI. I diffed the code and the main change seems to be: diff -r --exclude CVS ./src/src/umad.c-orig ./src/src/umad.c 161a162,166 if (sys_read_string(port_dir, SYS_PORT_LINK_LAYER, port-link_layer, UMAD_CA_NAME_LEN) 0) /* assume IB by default */ sprintf(port-link_layer, IB); The above change might imply an update to an rpm dependency on ib- kernel versions, but should not result in a new library version. If the library version remained at 2, then applications built for OFED 1.5 would also run on OFED 1.5.2rc2 without recompile. This is not an only change. The whole commit is: commit 99d0901f3702e22c178352524af18ec567781d51 Author: Yevgeny Kliteynik klit...@dev.mellanox.co.il Date: Tue Mar 9 12:39:40 2010 +0200 libibumad: added link_layer for RoCEE support and updated umad version Added link_layer field to umad_port_t. The field is implemented as char[]. If the relevant file doesn't exist in sysfs, the link layer is IB. Otherwise, the content of link_layer file is used. The libibumad version is promoted. Signed-off-by: Yevgeny Kliteynik klit...@dev.mellanox.co.il Signed-off-by: Sasha Khapyorsky sas...@voltaire.com diff --git a/libibumad/include/infiniband/umad.h b/libibumad/include/infiniband/umad.h index 1f82183..f9da204 100644 --- a/libibumad/include/infiniband/umad.h +++ b/libibumad/include/infiniband/umad.h @@ -116,6 +116,7 @@ typedef struct ib_user_mad { #define SYS_PORT_RATE rate #define SYS_PORT_GUID port_guid #define SYS_PORT_GID gids/0 +#define SYS_PORT_LINK_LAYERlink_layer typedef struct umad_port { char ca_name[UMAD_CA_NAME_LEN]; @@ -132,6 +133,7 @@ typedef struct umad_port { uint64_t port_guid; unsigned pkeys_size; uint16_t *pkeys; + char link_layer[UMAD_CA_NAME_LEN]; } umad_port_t; typedef struct umad_ca { diff --git a/libibumad/libibumad.ver b/libibumad/libibumad.ver index 57cddbd..d36fbbc 100644 --- a/libibumad/libibumad.ver +++ b/libibumad/libibumad.ver @@ -6,4 +6,4 @@ # API_REV - advance on any added API # RUNNING_REV - advance any change to the vendor files # AGE - number of backward versions the API still supports -LIBVERSION=2:1:0 +LIBVERSION=2:2:0 diff --git a/libibumad/src/umad.c b/libibumad/src/umad.c index 277ae6b..d16e750 100644 --- a/libibumad/src/umad.c +++ b/libibumad/src/umad.c @@ -159,6 +159,11 @@ static int get_port(char *ca_name, char *dir, int portnum, umad_port_t * port) if (sys_read_uint(port_dir, SYS_PORT_CAPMASK, port-capmask) 0) goto clean; + if (sys_read_string(port_dir, SYS_PORT_LINK_LAYER, + port-link_layer, UMAD_CA_NAME_LEN) 0) + /* assume IB by default */ + sprintf(port-link_layer, IB); + port-capmask = htonl(port-capmask); if (sys_read_gid(port_dir
Re: [ewg] libibumad.so version bump in OFED 1.5.2
Hi Todd, On 06:48 Tue 06 Jul , Todd Rimmer wrote: I recently discovered that 3rd party applications built for OFED 1.5 will not run on OFED 1.5.2rc2. The issue I found was that libibumad had changed from version 2 to version 3. Can you provide insight as to why this change was made? In reviewing the source code and change logs there do not seem to be any changes which would impact the application API nor application ABI. I diffed the code and the main change seems to be: diff -r --exclude CVS ./src/src/umad.c-orig ./src/src/umad.c 161a162,166 if (sys_read_string(port_dir, SYS_PORT_LINK_LAYER, port-link_layer, UMAD_CA_NAME_LEN) 0) /* assume IB by default */ sprintf(port-link_layer, IB); The above change might imply an update to an rpm dependency on ib-kernel versions, but should not result in a new library version. If the library version remained at 2, then applications built for OFED 1.5 would also run on OFED 1.5.2rc2 without recompile. This is not an only change. The whole commit is: commit 99d0901f3702e22c178352524af18ec567781d51 Author: Yevgeny Kliteynik klit...@dev.mellanox.co.il Date: Tue Mar 9 12:39:40 2010 +0200 libibumad: added link_layer for RoCEE support and updated umad version Added link_layer field to umad_port_t. The field is implemented as char[]. If the relevant file doesn't exist in sysfs, the link layer is IB. Otherwise, the content of link_layer file is used. The libibumad version is promoted. Signed-off-by: Yevgeny Kliteynik klit...@dev.mellanox.co.il Signed-off-by: Sasha Khapyorsky sas...@voltaire.com diff --git a/libibumad/include/infiniband/umad.h b/libibumad/include/infiniband/umad.h index 1f82183..f9da204 100644 --- a/libibumad/include/infiniband/umad.h +++ b/libibumad/include/infiniband/umad.h @@ -116,6 +116,7 @@ typedef struct ib_user_mad { #define SYS_PORT_RATE rate #define SYS_PORT_GUID port_guid #define SYS_PORT_GID gids/0 +#define SYS_PORT_LINK_LAYERlink_layer typedef struct umad_port { char ca_name[UMAD_CA_NAME_LEN]; @@ -132,6 +133,7 @@ typedef struct umad_port { uint64_t port_guid; unsigned pkeys_size; uint16_t *pkeys; + char link_layer[UMAD_CA_NAME_LEN]; } umad_port_t; typedef struct umad_ca { diff --git a/libibumad/libibumad.ver b/libibumad/libibumad.ver index 57cddbd..d36fbbc 100644 --- a/libibumad/libibumad.ver +++ b/libibumad/libibumad.ver @@ -6,4 +6,4 @@ # API_REV - advance on any added API # RUNNING_REV - advance any change to the vendor files # AGE - number of backward versions the API still supports -LIBVERSION=2:1:0 +LIBVERSION=2:2:0 diff --git a/libibumad/src/umad.c b/libibumad/src/umad.c index 277ae6b..d16e750 100644 --- a/libibumad/src/umad.c +++ b/libibumad/src/umad.c @@ -159,6 +159,11 @@ static int get_port(char *ca_name, char *dir, int portnum, umad_port_t * port) if (sys_read_uint(port_dir, SYS_PORT_CAPMASK, port-capmask) 0) goto clean; + if (sys_read_string(port_dir, SYS_PORT_LINK_LAYER, + port-link_layer, UMAD_CA_NAME_LEN) 0) + /* assume IB by default */ + sprintf(port-link_layer, IB); + port-capmask = htonl(port-capmask); if (sys_read_gid(port_dir, SYS_PORT_GID, gid) 0) Altering 'struct umad_port' effectively breaks ABI. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Allowing ib dignostics to be run without being logged in as root.
On 12:44 Tue 01 Jun , Yevgeny Kliteynik wrote: $ cat /etc/udev/rules.d/80-ib-umad.rules KERNEL==umad*, NAME=infiniband/%k, MODE=0666 It is not the same. Your propose to expose /dev/infiniband/umad device access to all world, which is obviously even more dangerous than SUIDing diagnostic programs. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Allowing ib dignostics to be run without being logged in as root.
Hi Woody, On 13:51 Tue 25 May , Woodruff, Robert J wrote: Some people were asking me if it would be possible to allow some of the IB diagnostic tools to be run without requiring being logged in as root. Would there be any problem in changing the installation to set their permissions to setuid root to allow this, i.e., chmod +s /usr/sbin/ibnetdiscover chmod +s /usr/sbin/ibaddr chmod +s /usr/sbin/smpquery chmod +s /usr/sbin/perfquery As many others I would also suggest to not make it (at least in default installation). However you can try to run diagnostic tools as non-root user by doing follow: 1. create some dedicated group, let's say 'umad'. 2. add dedicated users to be a members of this group 3. chown root:umad /dev/infiniband/umad* 4. chmod 0660 /dev/infiniband/umad* 5. update ib related udev rules file to match above This is how device access is granted to users normally. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ (listed in http://www.openfabrics.org/downloads/management/latest.txt) md5sum: d3586e7a17bca99fd384a943f00e259e libibumad-1.3.5.tar.gz 754d93f567393d3b9987a65326f40917 libibmad-1.3.5.tar.gz 5c94d6ee49e9c51c801f6634823b5ad5 opensm-3.3.6.tar.gz ba28f6b5323e6067ca019a999eeaf907 infiniband-diags-1.5.6.tar.gz All component versions are from recent master branch. Full list of changes is below. Sasha Al Chu (10): add libibnetdisc caching to libibnetdiscover support --cache and --load-cache options in ibnetdiscover support --load-cache in iblinkinfo and ibqueryerrors fix libibnetdisc corner case when user tries to re-cache a loaded cache support --diff in ibnetdiscover support --diffcheck in ibnetdiscover support lid and nodedesc diffchecks in ibnetdiscover fix libibnetdisc portguid hashing corner case fix libibnetdisc cache error path memleak check for duplicate port guids in libibnetdisc cache Arputham Benjamin (1): opensm: Add option to specify prefix to syslog messages Dale Purdy (1): Dimension port order file support Dave Olson (1): opensm/opensm.init.in: fix install warning on SLES11 Eli Dorfman (Voltaire) (3): opensm: Add update_desc command to opensm console Delete port only after GID OUT trap was sent opensm: Always set enabled speed and width Hal Rosenstock (13): infiniband-diags/ibqueryerrors.c: Fix output format of counters opensm: Add support for optimized SLtoVLMappingTable programming opensm/osm_vl_arb_rcv.c: Dump table after validating block number libibmad: Add support for optional PortRcvErrorDetails counter infiniband-diags/perfquery: Add support for optional PortRcvErrorDetails counter infiniband-diags/ibportstate.8: Add new supported ops libibmad: Add support for PortSamplesControl counter infiniband-diags/perfquery: Add support for PortSamplesControl counter infiniband-diags/ibqueryerrors: Add support for optional PortRcvErrorDetails counter opensm/osm_qos.c: Fix typo in OSM_LOG message infiniband-diags/ibqueryerrors.c: Fix typo in counter name opensm/osm_mcast_mgr.c: Cosmetic changes infiniband-diags/vendstat: Indicate general info for IS4 as well as IS3 Ira Weiny (5): libibnetdisc: fix typo in man page for ibnd_show_progress infiniband-diags: Ignore PortInfo data on down port. libibnetdisc: Convert to a multi-smp algorithm libibnetdisc: add ibnd_set_max_smps_on_wire call ibnetdisc: Separate calls to umad and mad layer to avoid race condition on response MAD's Line Holen (4): opensm/osm_vl_arb_rcv.c: fix double mutex release bug opensm/osm_pkey.h: Fix typo in comment opensm/osm_sa_path_record.c: livelock in pr_rcv_get_path_parms opensm/osm_sa_path_record.c: Lower max number of hops allowed Mike Heinz (4): Perfquery can be too noisy. management: adding mad_dump_fields to libibmad RE: [PATCH] management: adding mad_dump_fields to libibmad RE: [PATCH] management: adding mad_dump_fields to libibmad Ralph Campbell (1): infiniband-diags/ibportstate: allow changes to CA portinfo parameters Sasha Khapyorsky (54): opensm/osm_mcast_mgr.c: strip log-only variable opensm/osm_sa_path_record.c: MGID must be specified explicitly osm_sa_path_record.c: separate mutlicast processing code osm_sa_path_record.c: use PR DGID by reference osm_sa_path_record.c: cleanups opensm/osm_slvl_map_rcv.c: fix port parsing on BE machine opensm/osm_slvl_map_rcv.c: fix mutex double release bug opensm/osm_slvl_map_rcv.c: verify port number values received from network opensm/osm_qos.c: merge SL2VL mapping capability check opensm/osm_qos.c: split switch external and end ports setup opensm/osm_sa_informinfo.c: minor simplifications and fixes opensm/osm_sa_infrominfo.c: fixes and simplifications in lid range check infiniband-diags: -V, --version: exit with zero status opensm/osm_update_node_desc(): minor prototype improvement infiniband-diags: indentation fixes ibqueryerrors.c: more readable details dump function infiniband-diags/iblinkinfo,ibqueryerrors: code refactoring opensm/osm_vl_arb_rcv.c: indentation improvement opensm/osm_mcast_mgr.c: fix memory leak opensm/osm_sa_pkey_record.c: optimize port selection logic opensm/osm_mcast_mgr.c: remove redundant casting opensm/osm_mcast_mgr.c: fix some indentation errors opensm/osm_mcast_mgr.c: fix bug in MC root switch calculation opensm/osm_mcast_mgr.c: code simplifications opensm/osm_mcast_mgr.c: preserve root switch calculation functionality libibnetdisc: code cleanups libibnetdisc: restore
Re: [ewg] When will we have the new management package for OFED 1.5.2
Hi Tziporet, On 11:23 Tue 11 May , Tziporet Koren wrote: As we agreed OFED 1.5.2 will include a new management package I wish to start testing of this package When will you have the new tarball ready? I would suggest to release two versions of management - first in 7-10 days from now for initial testing and final one somewhere at beginning of June. This will enter 1.5.2. Does it seem good? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] [PATCH] infiniband-diags/scripts/ibcheckerrs.in: emulate all ports if necessary.
On 09:18 Thu 06 May , Ira Weiny wrote: Upon thinking about this a bit more, and seeing Mikes patch. I think that the patch which Mike sent some time ago is a better fix. This will work fine for ibcheckerrs. However ibcheckerrors will run AllPortSelect and then go on to query all the ports individually. The patch below will cause a double read for each port which will kill ibcheckerrors performance on a large cluster. Sasha, what is the status of Mikes patch? Applied now. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] infiniband-diags 1.5.5 doesn't build?
Hi Albert, On 13:14 Wed 24 Mar , Albert Strasheim wrote: I'm trying to build infiniband-diags 1.5.5 from http://www.openfabrics.org/downloads/management/infiniband-diags-1.5.5.tar.gz on Fedora Core 12, but the compilation is failing: ibnetdiscover.o: In function `out_switch_port': /home/alberts/rpmbuild/BUILD/infiniband-diags-1.5.5/src/ibnetdiscover.c:301: undefined reference to `IBDEBUG' ibnetdiscover.o: In function `router_iter_func': /home/alberts/rpmbuild/BUILD/infiniband-diags-1.5.5/src/ibnetdiscover.c:420: undefined reference to `IBDEBUG' ibnetdiscover.o: In function `ca_iter_func': /home/alberts/rpmbuild/BUILD/infiniband-diags-1.5.5/src/ibnetdiscover.c:401: undefined reference to `IBDEBUG' ibnetdiscover.o: In function `switch_iter_func': /home/alberts/rpmbuild/BUILD/infiniband-diags-1.5.5/src/ibnetdiscover.c:380: undefined reference to `IBDEBUG' You may want to look at http://www.openfabrics.org/downloads/management/README for build instructions: In order to build infiniband-diags you will need libibumad and libibmad installed (also there in management/). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Is this the right place for questions on OFED?
On 14:48 Tue 12 Jan , Ross Smith wrote: However, it looks like I don't need OFED, and I've today found the opensm packages on the openfabrics.org website, and have downloaded: libibumad-1.3.4.tar.gz libibmad-1.3.4.tar.gz opensm-3.3.5.tar.gz infiniband-diags-1.5.5.tar.gz I followed the instructions in the README file to install them in order. However while everything has compiled, and appears to be installed ok, I still cannot run opensm: # opensm opensm: error while loading shared libraries: libosmvendor.so.3: cannot open shared object file: No such file or directory Does anybody have any ideas as to what I might have done wrong? By default the libraries are installed in /usr/local/lib. Assuming so you likely need to add this to your ld.so.conf paths and rerun ldconfig. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 1281c3497086f3348b8ede7c3c5dcacf libibumad-1.3.4.tar.gz 02d0c65079c067e0e933c4822dad8b28 libibmad-1.3.4.tar.gz ad3f5181f9e8c91de73c0165212f11ea opensm-3.3.4.tar.gz 7a784405b8c3e4766c543658da5da7f4 infiniband-diags-1.5.4.tar.gz All component versions are from recent master branch. Full list of changes is below. Sasha Al Chu (8): remove unused headers in libibnetdiscover remove ibnd_update_node consistently type lids and lmcs libibnetdisc fix potential segfault in ibnd_node_t destroy path split out scan specific data from ibnd_node_t [attempt #2] remove unnecessary casting in libibnetdiscover use nodes/switches lists instead of nodesdist array in libibnetdiscover splice out chassis scan specific from general scan in libibnetdiscover Doron Shoham (1): Add different print macros Eli Dorfman (Voltaire) (1): infiniband-diags/saquery: Report SA MAD Class specific status. Hal Rosenstock (23): opensm/main.c: Use strtoul rather than strtol for parsing transaction timeout opensm/osm_mcast_tbl: Fix size of port mask table array opensm/osm_mcast_tbl.c: In osm_mcast_tbl_clear_mlid, use memset to clear port mask entry opensm/osm_switch.h: Remove dead osm_switch_get_physp_ptr routine infiniband-diags/ibqueryerrors: Add support for PortXmitDiscardDetails opensm/ib_types.h: Fix some commentary typos opensm/osm_base.h: Update IBA reference to use 1.2.1 opensm: Only clear SMP beyond end of PortInfo attribute infiniband-diags/ibsendtrap.c: Add support for SysImageGuid changed trap (145) opensm/osm_link_mgr.c: Fix IBA reference for PortInfo attribute infiniband-diags/saquery.c: Change lids and port numbers to decimal libibmad: Add support for new PortInfo:McastPkeyTrapSuppressionEnabled field opensm: Cosmetic changes opensm/osm_ucast_ftree.c: Cosmetic commentary changes opensm/libvendor/osm_vendor_ibumad.c: Cosmetic change to OSM_LOG message opensm/osm_vendor_api.h: Fix commentary typo opensm/complib/cl_event.c: Cosmetic changes opensm/osm_vl15intf.c: Cosmetic commentary change opensm/osm_sm_mad_ctrl.c: Fix qp0_mads_accounting opensm/osm_trap_rcv.c: Eliminate heavy sweep on receipt of trap 145 opensm/osm_vendor_ibumad.c: Change LID format to decimal in log message opensm/osm_sm_mad_ctrl.c: Make some minor simplifications opensm/osm_sm_mad_ctrl.c: Add missing call to return mad to mad pool Ira Weiny (9): infiniband-diags/ibnetdisc.h: Clarify ports member of ibnd_node_t with more descriptive comment OpenSM: Fix unused variable compiler warning. infiniband-diags/ibqueryerrors: Reformat the output of the xmtdiscard details. infiniband-diags/libibnetdisc: clean up PortInfo queries infiniband-diags/libibnetdisc: Remove nearly duplicate get_node_info function infiniband-diags/ibqueryerrors: Fix remote node name printing infiniband-diags/ibqueryerrors: Change realloc of suppressed fields to a static array infiniband-diags/ibqueryerrors: Properly exit and clean up resources when node info query fails libibmad: libibumad: Print warnings and errors to stderr not stdout Keshetti Mahesh (1): opensm/: Removed empty comments Line Holen (8): opensm/osm_state_mgr.c Added more info to some error messages opensm/osm_ucast_ftree.c Changed log level opensm/osm_helper.c Log format changes opensm/osm_vl15intf.c Log change opensm/osm_sm_mad_ctrl.c Changes to some error messages opensm/osm_vendor_ibumad.c: Move error info into single message opensm/osm_vendor_ibumad.c: Move error info into single message opensm: Added print_desc to various log messages Michael Reed (1): opensm/main.c: opensm cannot be killed while asking for port guid Sasha Khapyorsky (39): opensm/osm_sa_path_record.c: flow simplification opensm/osm_trap_rcv.c: some simplification opensm/osm_trap_rcv.c: use source_lid and port_num for logging opensm/osm_trap_rcv.c: some consolidation libibnetdisc: remove man page of removed API opensm/partition: keep multicast group pointer opensm/osm_mcast_mgr.c: fix return value on alloc_mfts() failures infiniband-diags/ibqueryerrors: minor improvements opensm/main.c: kill unneeded braces opensm: cleanup intn_t uses complib/fleximap: make compar callback to return int opensm/osm_mgrp_new(): add subnet db insertion opensm/osm_console.c: make const functions opensm/osm_perfmgr_db.c: remove casting opensm: remove braces around value in return statements osmtest/osmt_mtl_regular_qp.c: cleaning uintn_t use complib: replace intn_t types by C99 intptr_t
[ewg] [ANNOUNCE] management tarballs release
opensm: Add configurable retries for transactions opensm/osm_trap_rcv.c: Minor reorganization of trap_rcv_process_request Ira Weiny (45): Create a new library libibnetdisc Convert iblinkinfo.pl to C and use new ibnetdisc library. clean up iblinkinfo conversion Convert ibnetdiscover to use new ibnetdisc library libibnetdisc: change ibnd_discover_fabric to receive ibmad_port Add mad_field_name function Convert ibqueryerrors.pl to C and use new ibnetdisc library. ibqueryerrors: switch_guid_str is required for the string resolve function. Fix reversal of DRSLID and DRDLID in MAD_FIELDS enum. Add combined routing support to libibnetdisc Modify '-S' option of iblinkinfo and ibqueryerrors to do a limited scan of the Fix 2 formatting diff's from old ibqueryerrors. Clean up printing of switch heading when printing down links only. Move _with_node_name_map rpm spec option from infiniband-diags to opensm. ibnetdiscover: only report max hops discovered when requested iblinkinfo, ibqueryerrors: prevent core when switch is not found iblinkinfo: remove unused file pointer. iblinkinfo: convert to ibdiag_process_opts ibqueryerrors: fix custom opts string iblinkinfo: optomize -D operation ibqueryerrors: optomize -D operation ibprintca.pl: fix corner case: print switch information for CA's found at the end of the cache file. iblinkinfo, ibqueryerrors: clean up ibdiag_process_opts call Fix error return code from mad_send_via Add ib_resolve_gid_via and make portid an out only parameter of ib_resolve_portid_str_via libibmad: Ensure the proper length is passed into umad_recv on each call opensm: remove extra 0x from debug message. infiniband-diags/libibnetdisc: fix DBGFLAGS variable libibmad: clear packet buffer correctly before formating and sending libibnetdisc: fix potential memory leak of port object infiniband-diags/libibnetdisc: remove all IBPANIC's and clean up error handling libibmad: make accessors function for timeout values used in libibmad libibmad: make accessors function for retry values used in libibmad libibnetdisc: make all fields of ibnd_node_t public libibnetdisc: make all fields of ibnd_port_t public libibnetdisc: make all fields of ibnd_fabric_t public opensm/complib: account for nsec overflow in timeout values infiniband-diags/libibnetdisc: add missing '\n' to error message libibnetdisc: add retract_dpath function infiniband-diags/src/ibqueryerrors.c: fix bug when attempting a sub-fabric scan infiniband-diags/src/ibqueryerrors.c: Remove --all option and replace it with --switch, --ca, --router opensm/opensm/osm_subnet.c: adjust buffer to ensure a '\n' is printed infiniband-diags/libibnetdisc/ibnetdisc.c: fix bug in single node processing. infiniband-diags/src/ibqueryerrors: Add clear errors and counters options infiniband-diags/ibqueryerrors: Use remap'ed node name in clear port error message Jason Gunthorpe (1): Clarify the syntax of the hop_weights_file Joachim Fenkes (1): libibmad: Handle MAD redirection Keshetti Mahesh (3): infiniband-diags/: Cosmetic changes, mostly typos 'ibcheckportwidth' : Exit if LWS is 1X infiniband-diags/libibnetdisc: Removed trivial comments line.ho...@sun.com (7): opensm/osm_ucast_ftree.c Increase the size of the hop table osm_sa_patch_record.c Remove unneeded break statement osm_dump.c dump port if lft is set up osm_ucast_ftree.c Count number of hops instead of calculating it opensm: Add new Sun vendor ID opensm/osm_ucast_ftree.c Fix assert comparing number of CAs to CN ports osm_ucast_ftree.c missing reset of ca_ports Nicolas Morey-Chaisemartin (9): management: Fixed custom_release in SPEC files OpenSM: include/vendor/osm_vendor.h - Replaced #elif with no condition by #else infiniband-diags: Added libibnetdiscover to .spec file opensm/osm_ucast_ftree.c: Enhanced Fat-Tree algorithm opensm/osm_ucast_ftree.c: Fixed issue with reverse_hops introduced by commit 1d7dd18b531c1d6370f80cc7303493d6f3e3e777. opensm/osm_ucast_ftree.cd: Added support for same level links osm_ucast_ftree.c: Removed reverse_hop parameters from fabric_route_upgoing_by_going_down osm_ucast_ftree.c: Cleaned up many comments osm_ucast_ftree.c: Applied osm_indent Sasha Khapyorsky (117): ibnetdiscover: fix types to avoid portability castings infiniband-diags: cosmetic fixes infiniband-diags: remove not used ibnetdiscover.h ibdiag_common: remove duplicated ibdebug variable Merge branch 'pq/ibn4' libibumad: keep port capmask as 32-bit variable ibstat.c: use htohl() for 32-bit capmask conversion opensm/osm_sa_node_record.c: simplify flows opensm
[ewg] Re: [OPENSM] update functions to match .h prototypes
On 15:52 Mon 28 Sep , Stan C. Smith wrote: Hello, The following patches address inconsistencies between header file function prototypes and .c function definitions; missing 'const' attribute. Attached is a Linux EOL patch file in case a mailer hacks/reformats the text. Signed-off-by: Stan Smith (stan.sm...@intel.com) Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [OPENSM] use standard varargs syntax in macro OSM_LOG()
On 16:50 Mon 28 Sep , Stan C. Smith wrote: Signed-off-by: Stan Smith (stan.sm...@intel.com) Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [OPENSM] osm_helper.c use explicit value for struct init
On 09:44 Tue 29 Sep , Stan C. Smith wrote: Signed-off-by Stan Smith (stan.sm...@intel.com) Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [OPENSM] update osm_db_files.c osm_pack.c functions to match .h prototypes
On 09:05 Fri 25 Sep , Stan C. Smith wrote: Hello, The following patches address inconsistancies between header file function prototypes and .c function definitions; missing 'const' attribute. Attched is a Linux EOL patch file in case a mailer hacks/reformats the text. Signed-off-by: Stan Smith (stan.sm...@intel.com) Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [OPENSM] cast to remove warnings about signed vs. unsigned comparisons
On 13:30 Wed 30 Sep , Stan C. Smith wrote: Use (unsigned) cast to remove compiler warnings for signed component in comparison (for loops) . In a couple of cases use unsigned instead of int for the variable declaration. Signed-off-by: Stan Smith stan.sm...@intel.com Applied. Thanks. Basically I prefer to avoid castings when possible - much better is to use proper types. So I'm adding the patch below, hope that it is fine for you. In general to simplify a detection of such cases we can consider to use -Wsign-compare gcc flag in linux environment. Sasha commit 3b30f3b1311b44f3e2042ea2c4e180ffa8291532 Author: Sasha Khapyorsky sas...@voltaire.com Date: Thu Oct 1 17:50:04 2009 +0200 opensm/osm_mesh.c: remove some castings Instead of casting for preventing different sign comparison warnings use unsigned types for affected variables. Signed-off-by: Sasha Khapyorsky sas...@voltaire.com diff --git a/opensm/opensm/osm_mesh.c b/opensm/opensm/osm_mesh.c index 8235e55..e5c53d9 100644 --- a/opensm/opensm/osm_mesh.c +++ b/opensm/opensm/osm_mesh.c @@ -58,7 +58,7 @@ static const struct mesh_info { int dimension; /* dimension of the torus */ int size[MAX_DIMENSION];/* size of the torus */ - int degree; /* degree of polynomial */ + unsigned int degree;/* degree of polynomial */ int poly[MAX_DEGREE+1]; /* polynomial */ } mesh_info[] = { {0, {0}, 0, {0}, }, @@ -263,7 +263,7 @@ static char *poly_print(int n, int *coeff) * * return a nonzero value if polynomials differ else 0 */ -static int poly_diff(int n, const int *p, switch_t *s) +static int poly_diff(unsigned int n, const int *p, switch_t *s) { if (s-node-num_links != n) return 1; @@ -591,13 +591,13 @@ static int get_switch_metric(lash_t *p_lash, int sw) { osm_log_t *p_log = p_lash-p_osm-log; int ret = -1; - int i, j, change; + unsigned int i, j, change; int sw1, sw2, sw3; switch_t *s = p_lash-switches[sw]; switch_t *s1, *s2, *s3; int **m; mesh_node_t *node = s-node; - int num_links = node-num_links; + unsigned int num_links = node-num_links; OSM_LOG_ENTER(p_log); @@ -622,7 +622,7 @@ static int get_switch_metric(lash_t *p_lash, int sw) s2 = p_lash-switches[sw2]; if (s2-node-temp == LARGE) continue; - for (j = 0; (unsigned)j s2-node-num_links; j++) { + for (j = 0; j s2-node-num_links; j++) { sw3 = s2-node-links[j]-switch_id; s3 = p_lash-switches[sw3]; @@ -925,8 +925,8 @@ static void make_geometry(lash_t *p_lash, int sw) int num_switches = p_lash-num_switches; int sw1, sw2; switch_t *s, *s1, *s2, *seed; - int i, j, k, l, n, m; - int change; + unsigned int i, j, k, l, n, m; + unsigned int change; OSM_LOG_ENTER(p_log); @@ -956,7 +956,7 @@ static void make_geometry(lash_t *p_lash, int sw) /* * ignore chain fragments */ - if ((unsigned)n seed-node-num_links n = 2) + if (n seed-node-num_links n = 2) continue; /* @@ -1068,11 +1068,11 @@ static void make_geometry(lash_t *p_lash, int sw) * find switch (other than s1) that neighbors i and j * have in common */ - for (k = 0; (unsigned)k s1-node-num_links; k++) { + for (k = 0; k s1-node-num_links; k++) { if (s1-node-links[k]-switch_id == sw1) continue; - for (l = 0; (unsigned)l s2-node-num_links; l++) { + for (l = 0; l s2-node-num_links; l++) { if (s2-node-links[l]-switch_id == sw1) continue; @@ -1204,11 +1204,11 @@ static int reorder_node_links(lash_t *p_lash, mesh_t *mesh, int sw) static int make_coord(lash_t *p_lash, mesh_t *mesh, int seed) { osm_log_t *p_log = p_lash-p_osm-log; - int i, j, k; + unsigned int i, j, k; int sw; switch_t *s, *s1; - int change; - int dimension = mesh-dimension; + unsigned int
[ewg] Re: [OPENSM] cast to remove compiler warning.
On 13:34 Wed 30 Sep , Stan C. Smith wrote: MS compiler squawks about uint32_t -- const uint16_t conversion with possible loss of data. Cast to eliminate compiler warning. Signed-off-by: stan smith stan.sm...@intel.com Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [PATCH] opensm: use mgrp pointer in port mcm_info
On 08:45 Mon 14 Sep , Hal Rosenstock wrote: Does this mean consolidate_ipv6_snm_req does not work now ? No, it doesn't. As you may remember 'consolidate_ipv6_snm_req' workaround does nothing with MGIDs to MLID mapping, but instead enforces all IPv6 SNM matching requests to join a single multicast group (MGID). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] EWG/OFED meeting minutes for Aug 24
Hi Tziporet, On 18:35 Wed 26 Aug , Tziporet Koren wrote: Not clear who is the owner of ibsim - Sasha is that you? Not an owner, but maintainer - yes, it is mine. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] OFA server fs is full
Hi, I'm not able to push changes to OFA server because / fs is full. Is it really needed to build things there? Why to not build on local machine(s) and then to upload tarballs? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFA server fs is full
On 18:36 Wed 22 Jul , Vladimir Sokolovsky wrote: I cleaned up some build directories. There are about 5GB free space now. Now there is: Filesystem 1K-blocks Used Available Use% Mounted on /dev/sda1151873632 139556036 4602784 97% / We will have a next overflow just in few days. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: Can you point me to management RN or change-log
On 09:48 Mon 20 Jul , Tziporet Koren wrote: I wish to include main management package changes in OFED 1.5 RN Sure, it is in management tree - opensm/doc/opensm_release_notes-3.3.txt. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [PATCH v4] libibmad: Handle MAD redirection
On 16:20 Tue 07 Jul , Joachim Fenkes wrote: Previously, libibmad reacted to GSI MAD responses with a redirect status by throwing an error. IBM eHCA adapters use redirection, so most infiniband_diags tools didn't work against eHCA. Fix: Modify mad_rpc() so that it resends the request to the redirection target if a redirect GS response is received. This is repeated until no redirect response is received, allowing for multiple levels of indirection. The dport argument is updated with the redirection target, so subsequent MADs will not go through the redirection process again but reach the target directly. Tested using perfquery between ehca, mlx4 and mthca in all possible combinations. Signed-off-by: Joachim Fenkes fen...@de.ibm.com Applied. Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: Is management package is ready for OFED 1.5 alpha?
On 06:35 Thu 16 Jul , Hal Rosenstock wrote: I have couple of changes on the queue, then we will release. Are you going to handle your patch backlog before releasing ? Yes. And also note that we will have couple of management releases during OFED 1.5 cycle. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED Nightly OFED-1.5-20090714-0600 build break in infiniband-diags
On 10:32 Tue 14 Jul , Jon Mason wrote: Can you regress the infiniband-diags RPM in the nightlies to the version found in 07/02 until Sasha has it working again? Why should we regress things when we are even in pre-alpha phase yet? We need to fix problems instead. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED Nightly OFED-1.5-20090714-0600 build break in infiniband-diags
On 11:26 Tue 14 Jul , Jon Mason wrote: If you want assistance with test compiling on certain architectures, I'll be happy to help. Your reports are very helpful. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED Nightly OFED-1.5-20090714-0600 build break in infiniband-diags
On 22:59 Tue 14 Jul , Tziporet Koren wrote: was this issue resolved? Yes. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: Is management package is ready for OFED 1.5 alpha?
Hi Tziporet, On 14:34 Mon 13 Jul , Tziporet Koren wrote: What is the status of management package? That is fine. But I'm not on feature freeze point. What is the release schedule you plan? I have couple of changes on the queue, then we will release. We wish to build OFED 1.5 alpha soon Ok. Let's use our regular practice here. OFED 1.5 can include latest management packages releases or daily builds. Once things will be stable we can switch major version number (3.3.x - 3.4.x, etc). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED-1.5-20090713-0600 infiniband-diags build break
On 21:18 Mon 13 Jul , Tziporet Koren wrote: Please take care I've sent the patch now. Have you resolved the management daily build issue? Hope so. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED-1.5-20090706-0600 build break
On 14:39 Thu 09 Jul , Tziporet Koren wrote: Sasha - can you create a new package with the fix? Yes, the problem was with management daily build, I'm fixing this now. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] RE: OFED libraries - on the web at last
Hi Tziporet, On 13:03 Sun 14 Jun , Tziporet Koren wrote: I send this to Sasha who is the management maintainer Only he can decide if to bring libibcommon back to life I think the status of libibcommon as non-public library was discussed many times (as well as its removal). Personally I don't see any good reason to get it back. At least not in the management tree - management doesn't use it. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
-diags/ibsendtrap.c: Set producer type according to node type opensm/osm_helper.c: Convert remaining helper routines for GID printing format opensm: Some cosmetic formatting changes infiniband-diags/man/vendstat.8: Indicate IS4 config group config not persistent across IS4 reboot opensm/osm_perfmgr.c: Eliminate duplicated error number opensm/man/opensm.8.in: Add mention of backing documentation for QoS policy file and performance manager opensm/include/opensm/osm_pkey.h: Fix commentary typo opensm/osm_pkey_mgr.c: Fix pkey endian in log message opensm/doc/performance-manager-HOWTO.txt: Indicate (previously implied) master state libibmad: Add decode support for SwitchInfo OptimizedSLtoVLMappingProgramming Ira Weiny (14): libibmad: Clean up new interface infiniband-diags: Convert ibaddr to new ibmad interface infiniband-diags: Convert ibportstate to new ibmad interface infiniband-diags: Convert ibroute to new ibmad interface infiniband-diags: Convert ibsendtrap to new ibmad interface infiniband-diags: Convert ibtracert to new ibmad interface infiniband-diags: Convert ibsysstat to new ibmad interface infiniband-diags: Convert mcm_rereg_test to new ibmad interface infiniband-diags: Convert perfquery, saquery, sminfo, smpquery, and vendstat to new ibmad interface infiniband-diags: convert ibnetdiscover to new ibmad interface Fix further bugs around console closure and clean up code. Fix ibidsverify.pl to use the correct cache file change missed LID conversion functions from hex to uint libibmad: bump library interface version Itai Baz (3): ib_types.h: Adding BKEY violation trap (259) libibmad/serv.c: Fixed respond function to return proper result code libibmad: added support for handling of BM (Baseboard management) MADs - FIXED without rmpp Julia Volynsky (1): Added send trap for trap 129 (local link integrity) line.ho...@sun.com (3): opensm/osm_link_mgr.c initialize SMSL opensm/osm_link_mgr.c Remove __osm_ prefix opensm/osm_link_mgr.c: indentation fixes Nicolas Morey Chaisemartin (9): opensm: Added io_guid_file and max_reverse_hops options opensm/osm_ucast_ftree.c: Added possible reverse hops for Ftree algorithm. Added documentation for io_guid_file and max_reverse_hop feature opensm/osm_ucast_ftree.c: Removed useless initialisation on switch indexes opensm/osm_switch.h : Fixed wrong comment about return value of osm_switch_set_hops opensm/console: Fixed osm_console poll to handle POLLHUP opensm/osm_console_io.h: Modify osm_console_exit so only the connection is killed, not the socket Fixed cio_close use when ENABLE_OSM_CONSOLE_SOCKET is not set opensm/osm_ucat_ftree.c Enhance min hops counters usage Or Gerlitz (7): generic libibmad perf query/reset api libimad implementation of PortXmtDataSL and PortRcvDataSL perfquery PortXmtDataSL/PortRcvDataSL support infiniband-diags: update configure.in check for libibmad API fix offset used for parsing of XmtDataSL RcvDataSL infiniband-diags/perfquery: add srcport param ib-diag/vendstat: counter-group-info config-counter-group vendor mads Sasha Khapyorsky (38): opensm/main.c: cosmetic opensm: indentation fixes opensm/osm_console.c: kill warning: defined but not used opensm/osm_ucast_mgr.c: code simplifications opensm/osm_lid_mgr: use single array for used_lids opensm: initialize all switch ports opensm: remove unneeded anymore physp initializations opensm: PortInfo requests for discovered switches opensm: remove casting of ib_smp_get_payload_ptr() osmtest: remove useless prototypes opensm/osm_console_io.c: remove 'osm_' prefix from static function names opensm: fix build warning with --disable-console-socket libibmad: cleanup deprecated function use opensm/console: move cio_open() function opensm/osm_console_io.c: move cio_close() function libibmad/rpc: fix class registration bug libibmad: per port timeout and retires setup libibmad: add mad_rpc_class_agent() call libibmad: deprecate old API calls libibmad/rpc: fix _do_madrpc() parameter value opensm/osm_ucast_ftree: indentation fixes opensm: some init functions simplification opensm/osm_sa_link_record.c: improve get_base_lid() opensm: kill __osm_ prefixes in static functions opensm/osm_sw_info_rcv.c: consolidate flows infiniband-diags/vendstat.c: code moving opensm: fix indentations with osm_indent. opensm: clean OSM_CDECL macro infiniband-diags/ibsendtrap: code consolidation opensm/osm_helper.c: return then log is inactive opensm/osm_helper.c: consolidate dr path printing code opensm/osm_helper.c: use single buffer in osm_dump_dr_smp
Re: [ewg] Moving to new server
Hi Jeff, On 16:24 Fri 27 Mar , Jeff Becker wrote: Just thought this would be a good chance to clean up. Thanks for your work. I copied my trees and basically it works. Few things are missing: 1. ~username symbolic links under /pub/scm. So things like: git clone git://69.55.231.68/~sashak/management , don't work yet. 2. Mail server (guess postfix) is not configured and mailx is not installed - both can be useful (and was used on current server) for git commit notification. 3. No gitweb yet. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 97b2609f5eaaf4320b39f44a50500b70 libibumad-1.3.1.tar.gz e60b1c787d7cd2768967ca4766238210 libibmad-1.3.1.tar.gz 8c8c153f21d9f6cee51fc3d501c54fe7 opensm-3.3.1.tar.gz 6b6c87ed01291a2a3322b0ff696c5a11 infiniband-diags-1.5.1.tar.gz All component versions are from recent master branch. Full change log is below. Sasha Arlin Davis (3): libibmad: add os dependent definitions. libibmad: remove c99 definitions within the ib_mad_f structure libibmad: minor changes to source to allow portability to WinOF. David McMillen (1): infiniband-diags/src/ibnetdiscover.c missing LID information on --ports Eli Dorfman (1): opensm/osm_inform.c report IB traps to plugin Eli Dorfman (Voltaire) (10): opensm/osm_subnet.c Fix memory leak for QOS string parameters. libibmad add PortXmitWait and CounterSelect2 to fields. opensm: Add new partition keyword for all hca, switches and routers docs update documenatation about new partition keywords infiniband-diags support PortXmitWait get and set opensm/osm_log.c save log_max_size in subnet opt in MB opensm/osm_subnet.c support subnet configuration rescan and update libibmad/src/dump.c fix dump functions for big endian machines opensm/osm_subnet.c enable log_max_size opt update opensm/osm_subnet.c fix parse functions for big endian machines Hal Rosenstock (23): opensm/libvendor/osm_vendor_sa_api.h: Fix commentary typo opensm/osm_inform.c: Eliminate compile warning opensm/osm_perfmgr_db.h: Remove unused typedef opensm/osm_perfmgr.c: In osm_perfmgr_init, eliminate memory leak on error libibmad/(mad.h fields.c): Add support for PerfMgt ClassPortInfo opensm/include/iba/ib_types.h: Add xmit_wait for PortCounters opensm/PerfMgr: Mainly cosmetic changes opensm/osm_node.h: Fix osm_node_get_num_physp description opensm/PerfMgr: Primarily fix enhanced switch port 0 perf manager operation opensm/doc/perf-manager-arch.txt: Fix some commentary typos opensm/PerfMgr: Add copyrights libibmad: lid print format changed to unsigned libibumad/umad.c: Change lid print format to unsigned infiniband-diags/perfquery: Change option name for extended counters opensm/osm_inform.c: Fix sense of zero GID compare in __match_inf_rec management/libibmad.txt: Remove madrpc_lock/unlock opensm/man/opensm.8.in: Indicate ROUTER_EXP obsoleted opensm/osm_console.c: Improve perfmgr print_counters error message infiniband-diags/smpdump.c: Fix usage examples infiniband-diags/smpdump.c: Release umad resources on exit opensm/console: Enhance perfmgr print_counters for better nodenames libibmad/fields.c: Dump LIDs as unsigned decimal infiniband-diags/saquery.c: Convert more LID prints to unsigned decimal Ira Weiny (3): opensm/opensm/osm_console.c: move reporting of plugins to status command. OpenSM: update osmeventplugin example for the new TRAP event. libibmad: Use enum types for function parameters Mike Heinz (1): opensm/osm_vendor_*_sa: fix incompatibility with QLogic SM Nicolas Morey Chaisemartin (4): Corrected incoherency in __osm_ftree_fabric_route_to_non_cns comments opensm/osm_ucast_ftree.c: Fixed bug on index port incrementation opensm/osm_ucast_ftree.c Fixed bad init value for down port index opensm/osm_console.c : Added dump_portguid function to console to generate a list of port guids matching one or more regexps Ralph Campbell (2): libibumad: get_ca() can call release_ca() with uninitialized data opensm: fix structure definition for trap 257-258 Robert Pearson (10): mesh analysis - skeleton mesh analysis - mesh_t data structure mesh analysis - node and link structures mesh analysis - matrix/polynomial routines mesh analysis - local geometry mesh analysis - mesh info table mesh analysis - induce global geometry mesh analysis - reorder links mesh analysis - lash preparation mesh analysis - integrate into lash core Sasha Khapyorsky (111): opensm: remove some unused variables and funcs opensm/osm_ucast_mgr: indentation fix infiniband-diags/saquery: indentation fixes infiniband-diabs/saquery: unify SA queries processors infiniband-diags/saquery: separate queries and commands infiniband-diags/saquery: PortInfoRecord query infinabd-diags: convert type uint - unsigned int opensm: remove unused header osm_pkey_mgr.h opensm/osm_sm.c: fix MC group creation in race condition opensm/osm_sa_mcmember_record: improve __cleanup_mgrp() opensm/multicast: remove some unused parameters. opensm/osm_subnet: consolidate some duplicated code
[ewg] Re: What is the branch that should be used for management packagefor OFED 1.4.1?
Hi Yevgeny, On 11:12 Mon 16 Feb , Yevgeny Kliteynik wrote: Below is the list of the OpenSM commits that should go into OFED 1.4.1: 1. Commit aa25fcba61e75ac94af2b950c200cdde433f110c Don't clear sw-need_update if port 0 is active. In a single switch subnet, after a fast switch reset, OpenSM doesn't configure LFT of the switch and the routing is lost. Bug #1507. 2. Commit 595f2e30026f9ec85674e6d70661d26de6f77a6b Update LFTs when entering master state Bug #1469. 3. Commit b44c398e7363e9c9605e9ee1d5893f872ef41feb Invalidate routing cache when entering master state Bug #1508 4. Commit 22da81f8cd432195f6c78cfe5f508f04b081aef7 Fix full topology dump in ftree routing. Topology dump was missing all the leaf switches Bug #1509 5. Commit c07d245651035cc7ba3b89ac0e81d54bff829f4d Don't complain about invalid QoS options when its default values are used. Bug #1451. 6. Commit 72a2fa241dffacc2eea3c2b0c131f8bcb4792162 Dump SA MAD before sending it, not after. Dumping of the SA MAD was done after the MAD was sent and returned to the pool, resulting in garbage in the OpenSM log. Bug #1510. Anything I missed? Don't know. From the list only bug#1507 was considered as major (which is strange, the case should be pretty rear and hardly reproducible), the rest have normal severity. As far as I remember dot release policy is to include only critical bug fixes. Isn't it? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: What is the branch that should be used for management package for OFED 1.4.1?
Hi Tziporet, On 17:51 Sun 08 Feb , Tziporet Koren wrote: As you know we are going for OFED 1.4.1 release. What should be the branch to use for this release. This branch should start from the sources in 1.4 and include only critical bug fixes. opensm-3.2 Also - what is the list of bugs that you think should be included in the management package in 1.4.1 If there are no major complains about OpenSM in OFED-1.4 then we probably could not care. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [PATCH] libibumad: get_ca() can call release_ca() with uninitialized data
On 11:53 Mon 02 Feb , Ralph Campbell wrote: In umad.c, get_ca(), the error path when the call to scandir() fails calls release_ca(ca) when the structure has not yet been initialized. This can cause free() to be called on random memory addresses. The fix is to initialize ca-numports and ca-ports earlier in get_ca(). Signed-off-by: Ralph Campbell ralph.campb...@qlogic.com Applied, Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 0cbae1312313c46b9a7d6abae11d9e13 opensm-3.2.5.tar.gz 50b8b23a800e6a703fd8eddd274f94f0 infiniband-diags-1.4.4.tar.gz All component versions are from recent master branch. Full change log is below. Sasha Eli Dorfman (1): opensm/osm_trap_rcv.c disable the port with the least hop count Hal Rosenstock (4): opensm/osm_port_info_rcv.c: Remove SM from sm_guid_tbl when IsSM is not opensm.8.in: Update email address opensm/osm_prefix_route.h: prefix and guid are in network rather than host endian order opensm/osm_lid_mgr.c: Commentary fix Nicolas Morey Chaisemartin (1): OpenSM: Fixed GUID check against cn_guid_file usinf Ftree. Sasha Khapyorsky (23): opensm/osm_sa_sminfo_record: remove unused variable opensm/osm_subnet.c: fix high_limit sign when printing opensm/osm_subnet: don't reassign zeroed config params opensm: fix QoS config bug opensm/osm_trap_rcv.c: separate port disabling code opensm: disable switch ports only opensm: remove function names in OSM_LOG() string opensm/osm_sa_link_record: prevent potential endless recursion opensm/osm_sw_info_rcv: eliminate osm_node_get_any_physp_ptr() use opensm: remove osm_node_get_any_dr_part_ptr() function opensm: remove osm_node_get_any_physp_ptr() function opensm: free lft_buf right after use opensm: rename switch lft_buf to new_lft opensm: fix possible crash when disabling babbling port opensm/ftree: save lft_buf memory allocations opensm/man/opensm.8: add some missing stuff infiniband-diags/grouping: add 10G IP router devid opensm/osm_sa_path_record.c: minor indentation fix opensm: more 3.2.4 related things for RN opensm: fix race in main OpenSM flow. opensm: add RN to distributed docs list opensm: Release Notes update for next version (3.2.5) management: update package versions Yevgeny Kliteynik (3): opensm/osm_lid_mgr.c: cosmetics in log message opensm/osm_state_mgr.c: bug fix in unicast cache opensm: release notes for OSM 3.2.4 ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] ibsim-0.5 tarball release
Hi, There is a new release of the ibsim tarball available in: http://www.openfabrics.org/downloads/management/ md5sum: d5383979b5728bb24b9fedc59dd2dc32 ibsim-0.5.tar.gz The version is from recent master branch. Full change log is below. Sasha Al Chu (3): ibsim: parse sim cmds via full name fix error message typo ibsim: add ReLink command Hal Rosenstock (15): ibsim/sim_net.c: Fix some typos ibsim/ibsim.c: Fix usage display ibsim/README: Clarify point of attachment/SIM_HOST use ibsim/README: Clarify point of attachment/SIM_HOST use ibsim/sim.h: Fix NodeDescription size ibsim/sim_cmd.c: Cosmetic changes to help ibsim/sim_mad.c: Cosmetic change to some debug messages ibsim/ibsim.c: Cosmetic change to some debug messages ibsim/README: Cosmetic commentary changes ibsim/TODO: Eliminate inet sockets from todo per previous README change ibsim: Remove some unused routines ibsim/sim.h: Allow max ports based on 36 port switches ibsim/sim_mad.c: Fix some typos ibsim/sim_mad.c: Cosmetic changes ibsim: Add support for vendor ID and system image GUID Ira Weiny (1): ibsim: Add per attribute drop error. Sasha Khapyorsky (50): defs.mk: append CFLAGS, LDFLAGS umad2sim: prevent recursive read()s ibsim.spec.in: export CFLAGS and LDFLAGS ibsim.spec.in: export CFLAGS and LDFLAGS with make install ibsim: prevent print buffer overflow ibsim: match client by nodeid and nodedesc ibsim: drop duplicated parameters from link_ports() tests/mcast_storm.c: test program example tests/mcast_storm.c: indentation fixes dist.sh: add tests directory Remove trailing whitespaces umad2sim/sim_client: fix debug prints format strings ibsim: fix do_cmd() prototype ibsim: remove out file from disconnect_client() parameter list ibsim: drop port capmask IsSM bit when SM client is disconnected ibsim: disconnect client when connection failed ibsim: reply timeout if peer client is unreachable umad2sim/sim_client: force port's capmask IsSM bit set option Update License: field in spec files tests/mcast_storm: fix length value passed to umad_send() tests/mcast_storm: check fopen() return status tests/mcast_storm: ib_gid_t to ibmad_gid_t rename ibsim: rename defport to default_port ibsim: fix nodeid size usages ibsim: don't fail on unknown options README: inet socket rediness support for short RMPP transactions umad2sim: fix snprintf() usage ibsim: fix snprintf() usage Merge branch 'master' of ssh://git.openfabrics.org/~sashak/scm/ibsim ibsim/sim_cmd: consolidate flows umad2sim: cosmetic update Voltaire copyright dates Encode agent id in request transaction id. ibsim: fix compilation warning ibsim: drop mad when attr method failed ibsim: make some stuff static tests/mcast_storm: fix uninitialized subnet prefix tests/mcast_storm: consolidate parameters tests/mcast_storm: various improvements tests/mcast_storm: add command line processor tests/mcast_storm: Add -G, -M and -I options tests/mcast_storm: add yet trivial single join test tests/mcast_storm: fix return value of make_gids_list() tests/mcast_storm: fix MGID command line value parsing tests/mcast_storm: use as default guid of self port tests/mcast_storm: 'joins' test sends a lot of requests tests/mcast_storm: make 'leave' test tests/mcast_storm: various improvements ibsim: version 0.5 Vincent Ficet (1): ibsim: handle EAGAIN error ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 89a49b57015524bc3f6ca8667b640b2d libibumad-1.2.3.tar.gz bf172da0e70dc4ce6cc625fde8707d00 libibmad-1.2.3.tar.gz 93e14f69ce5004bfdef1009f84a53eb7 opensm-3.2.4.tar.gz 32665d7fb2fe2bf734118b8530d4 infiniband-diags-1.4.3.tar.gz All component versions are from recent master branch. Full change log is below. Sasha Al Chu (4): opensm: fix manpage typos fix documentation typos opensm: verify config inputs when config file is rescanned fix qos config parsing bugs Albert Chu (1): support dump_conf console command Doron Shoham (3): install QoS_management_in_OpenSM.txt change log_max_size to MB export osm_log_max in MB Eli Dorfman (3): opensm/osm_sa_path_record.c print port guids in error message opensm/osm_mcast_tbl.c wrong max mcast lid cause the sm to set invalid MFT block. opensm/osm_sa_mcmember_record.c print multicast lid in error message Hal Rosenstock (4): OpenSM/osm_subnet.c: Fix log_max_size conversion to MB libibumad: Add UMAD_MAX_DEVICES define infiniband-diags/ibstat.c: Use UMAD_MAX_DEVICES define opensm/osm_trap_rcv.c: Fix typo Ira Weiny (3): opensm/opensm/osm_state_mgr.c: Add check for valid physical port before using pointer. Fix max parameter passed to umad_get_cas_names opensm: Add check for previous versions of plugins. Or Gerlitz (1): opensm: fix iser service-id used for SL assignment Sasha Khapyorsky (29): opensm/osm_ucast_lash: fix extra memory allocations opensm/osm_ucast_lash: simplify get_phys_connection() prototype opensm/scripts: unify scripts' config opens/osm_inform.c: cosmetic changes opensm/opens.spec: add -D option for logrotate file install command opensm: remove update_master_sm_base_lid field in PortInfo madw context libibmad/src/mad.c: indentation fix libibmad/dump: print more PortInfo:CapabilityMask bits opensm: support more PortInfo:CapabilityMask bits opensm: osm_send_trap144() function opensm: send trap144 to master SM when priority is raised opensm: notify master SM with trap 144 opensm: hide function name with OSM_LOG_MSG_BOX() macro opensm: rename sm signal opensm: sweep on SIGCONT opensm/include/opensm/osm_switch.h: minor simplifications opensm/osm_switch.c: minor: shorter flow opensm/osm_ucast_cache.[ch]: indentation fixes make.dist: don't use ${date}git suffix for release opensm/osm_ucase_cache: simplify cached links allocation code opensm/osm_subnet.c: consolidate logging code opensm/osm_subnet.c: use strdup() function opensm/osm_subnet.c: consolidate qos parameters verification code opensm/osm_subnet.c: move osm_subn_rescan_conf_files() function opensm/osm_sa_mcmember_record: return a real port JoinState on update opensm/osm_sa_mcmember_record: simplify query code infiniband-diags/ibstat.c: remove casting opensm/osm_trap_rcv.c: kill some empty lines management: update versions Tim Meier (1): opensm: osm_opensm.c added a method to remove plugins Yevgeny Kliteynik (16): opensm/scripts/opensm.conf: remove obsolete config file opensm/opensm/Makefile.am: allow 'make dist' from non-source directory opensm: replace switch's fwd_tbl with simple LFT opensm: replace switch's fwd_tbl with simple LFT - remove obsolete files opensm/osm_ucast_ftree.c: some simplification in LFT handling opensm: free lft_buf if it matches switch's lft opensm/osm_ucast_cache: fixing coredump opensm/osm_sa.c: adding missing include opensm/osm_pkey.c: cosmetics in some log message opensm/ib_types.h: rename IB_MC_REC_STATE_SEND_ONLY_MEMBER opensm/osm_multicast.c: bug with joining/leaving mcast group opensm/Makefile.am: install QoS_management_in_OpenSM.txt osmtest/osmt_multicast.c: some refinements to the multicast flow opensm/osm_lid_mgr.c: ignore and overwrite guid2lid (windows) opensm/osm_sa_mcmember_record.c: bad return state when leaving mcast opensm/osmtest: fixing some comments in mcast flow of osmtest ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [PATCH ofed-1.4] ofabuild: proper detection of ibsim version
Detect ibsim version properly as it is defined in the tree rather than using hard coded 0.4 value. Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED] --- The patch is against ofed_1_4 branch. make.srpms |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/make.srpms b/make.srpms index 515384a..d3b92b7 100755 --- a/make.srpms +++ b/make.srpms @@ -66,7 +66,7 @@ case $NAME in NAME=dapl ;; ibsim) -VERSION=0.4 +VERSION=`sed -ne '/#define IBSIM_VERSION/s/^.*\\(.*\)\.*$/\1/p' ibsim/ibsim.c` ;; qlvnictools) VERSION=0.0.1 -- 1.6.0.3.517.g759a ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] RE: OFED meeting agenda for today (Oct 22)
Hi Woody, On 09:23 Wed 22 Oct , Woodruff, Robert J wrote: Attached are the 0.3 version of the BOF slides that we can review today. On slide 11. ibnetdiscover library idea was only briefly reviewed in OFED-1.4 period. It didn't enter OFED-1.4. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] ***SPAM*** [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 5ce057a956435f818621802849fe4d51 libibcommon-1.1.2.tar.gz be74f2916d1901a789b3b558d77b2a8a libibumad-1.2.2.tar.gz f78425228d34ff80440a7669027732ad libibmad-1.2.2.tar.gz 7fbb303e8cbf3f0b8418a0f565ddd7ce opensm-3.2.3.tar.gz 12a32e384dfa92f6a8df8fbaba8bf3b3 infiniband-diags-1.4.2.tar.gz All component versions are from recent master branch. Full change log is below. Sasha Al Chu (21): opensm: reroute console option opensm: rename ucast_file and ucast_dump_file to lfts_file opensm: fix instruction typo in config file opensm: fix segfault corner case when osm_console_init fails opensm/console: close console socket on cleanup path opensm: fix comment typo opensm: remove old comment opensm: fix routing algorithm description add clarification in perfquery manpage use 0xff vs. 255 consistently in perfquery add more detail to error message on perfquery workaround add --loop_ports option to perfquery infiniband-diags/perfquery: support ehanced port 0 with --loop-ports infiniband-diags/perfquery: fix comments infiniband-diags/ibclearerrors: specify -a in call to perfquery tweak notes about port 255 in perfquery manpage infiniband-diags/perfquery: error out if AllPortSelect is not supported infinibanf-diags/perfquery: remove single port CA AllPortSelect workaround infiniband-diags/perfquery: if --loop_ports is specified always loop through all ports if desired perfquery code cleanup infiniband-diags/perfquery: if -a is specified loop through ports if required and aggregate output Albert Chu (2): opensm/opensm.init: do not specify run levels on SuSE opensm: routing chaining Doug Ledford (3): opensm/*/Makefile.am: remove install-exec-hook opensm: remove -rpath from LDFLAGS Trivial: update usage info for ibaddr.c in infiniband-diags-1.4.1 Hal Rosenstock (23): opensm/osm_lin_fwd_tbl.h: Cosmetic changes opensm/osm_rand_fwd_tbl.h: Cosmetic changes opensm/osm_console_io.c: Eliminate compile warnings opensm/osm_qos_parser.y: Eliminate bison warning libibmad/ChangeLog: Fix typo libibumad/umad.c: Cosmetic change opensm/osm_inform.c: Fix compile warning complib/cl_event_wheel.c: Fix some printf typos opensm/osm_lid_mgr.c: Convert lid range prints to decimal opensm/osm_lid_mgr.c: Cosmetic format change opensm/include/opensm/osm_subnet.h: Update some comments opensm/osm_helper.c: Change lids_per_port to decimal OpenSM: More conversion to (unsiged) decimal lid format osmtest/osmtest.8: Fix log_file option description in man page osm_(helper trap_rcv).c: Change output format of notice type to unsigned decimal OpenSM: More man and doc changes for opensm.conf OpenSM: Display port number in decimal in log messages OpenSM: Display port number in decimal in log messages infiniband-diags/ibsysstat.c: Fix a couple of latent bugs OpenSM release notes: Clarify OpenSM compatibility due to change in default SM/SA keys OpenSM release notes: Indicate InfiniScale-IV support infiniband-diags/saquery: Support MFTRecord infiniband-diags/saquery.8: Update saquery man page Ira Weiny (9): infiniband-diags/src/saquery.c: convert GID prints to use inet_ntop infiniband-diags/src/ibaddr.c: convert GID prints to use inet_ntop OpenSM: convert GID prints to use inet_ntop opensm: Add ib_trap_str function opensm: Add a Node Description check on light sweep. Fix some missing node name map substitutions opensm: move vendor specific compilation flags to config.h ibnetdiscover.c: continue processing other ports even if smpquery fails on one port opensm: Add osm_config.h file Jack Morgenstein (2): libibmad: eliminate compiler warnings on x86_64 infiniband-diags: eliminate compiler warnings Keshetti Mahesh (3): complib: trivial change in description of cl_list_insert_tail function opensm/osm_ucast_lash.c: remove unused variables opensm/osm_ucast_lash: find_port_from_lid() function is not used Sasha Khapyorsky (106): libibmad: Bump a library version opensm/osm_sa_lft_record: validate LFT block number opensm/sa: remove local *get_port_by_guid() wrappers opensm/osm_sa_lft_record: pass block parameter in host byte order opensm: speedup and improve ipv6 snm handling opensm: improve port_prof_ignore handling infiniband-diags/saquery: remove unused variable infiniband-diags/src/ibaddr.c: remove unused variables opensm: minor: move GID print buffer definitions under nearest condition opensm/osm_log: reverse log level check flow opensm/osm_log: osl_log
Re: [ewg] Dependencies in OpenSM libraries
On 08:53 Mon 29 Sep , John Russo wrote: The concern is that use of basic utilities and diagnostic tools should not require opensm to be installed. Not opensm, opensm-libs. As far as I remember the concern was about some specific third parties (non-OFA) using non-public libraries. Given the previous discussion it would have seemed that only opensm was going to use the non-public libraries. We are unclear on what the permitted scope of use of these libraries are. And the discussion was about stable and non-stable library APIs, not about permissions for others to use or not to use the libraries. I think everything was on the list. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Dependencies in OpenSM libraries
Hi John, On 14:12 Mon 22 Sep , John Russo wrote: A while back we discussed various dependencies on opensm libraries which were going to become non-public Was that intended to occur in 1.4 or later versions of OFED? In 1.4 I noticed that ibutils and infiniband-diags still have a dependency on libopensm.so and other opensm-libs supplied files. Is this a mistake or something planned to be resolved in a later release of OFED (1.5, etc). I don't think it is mistake - what is the issue with infiniband-diags using opensm-libs? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] perfquery error
Hi Hal, On 12:37 Mon 21 Jul , Hal Rosenstock wrote: or directory) [EMAIL PROTECTED] ~]# perfquery ibpanic: [3926] madrpc_init: can't open UMAD port ((null):0): (No such file or directory) [EMAIL PROTECTED] ~]# This sounds like an additional aspect of the libibumad API issue introduced when some functionality related to this was eliminated. The previous fix related to OpenSM appears to not be the complete solution to these mixed configurations. I think you are referring changes which were done in OpenSM ibumad vendor layer, specifically commits 5cf395cb107ef76091d554cd4b42456dc53b38a2 and 36cfd8e6967d8a7aea74b9a180fac0275ef549dd and had nothing to do with libibumad API. Regarding to an issue itself. Steve, does something like this: diff --git a/libibumad/src/umad.c b/libibumad/src/umad.c index dcc7275..9010307 100644 --- a/libibumad/src/umad.c +++ b/libibumad/src/umad.c @@ -499,6 +499,19 @@ umad_done(void) return 0; } +static unsigned is_ib_type(char *ca_name) +{ + char dir_name[256]; + unsigned type; + + snprintf(dir_name, sizeof(dir_name), %s/%s, SYS_INFINIBAND, ca_name); + + if (sys_read_uint(dir_name, SYS_NODE_TYPE, type) 0) + return 0; + + return type = 1 type = 3 ? 1 : 0; +} + int umad_get_cas_names(char cas[][UMAD_CA_NAME_LEN], int max) { @@ -512,7 +525,7 @@ umad_get_cas_names(char cas[][UMAD_CA_NAME_LEN], int max) for (i = 0; i n; i++) { if (strcmp(namelist[i]-d_name, .) strcmp(namelist[i]-d_name, ..)) { - if (j max) + if (j max is_ib_type(namelist[i]-d_name)) strncpy(cas[j++], namelist[i]-d_name, UMAD_CA_NAME_LEN); } helps? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
Hi Hal, On 05:42 Thu 10 Jul , Hal Rosenstock wrote: Also, what about backward compatibility with apps built with the previous version of these libraries ? What sort of compatibility problem can we expect? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] List of libraries in OFED
Hi Oren, On 13:31 Wed 09 Jul , Oren Kladnitsky wrote: No. Ibutils use opensm libvendor for mad sending interface. Right, it is not used directly. But it is used in LDFLAGS (due to libosmvendor - libibumad - libibcommon dependency): config/osm.m4: OSM_LDFLAGS=$OSM_LDFLAGS -lopensm -losmvendor -losmcomp -libumad -libcommon ibis/config/osm.m4: OSM_LDFLAGS=$OSM_LDFLAGS -lopensm -losmvendor -losmcomp -libumad -libcommon ibmgtsim/config/osm.m4: OSM_LDFLAGS=$OSM_LDFLAGS -lopensm -losmvendor -losmcomp -libumad -libcommon Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
On 11:57 Wed 09 Jul , Hal Rosenstock wrote: 891c907cf7fb56191c1cd4224608ef63 libibmad-1.2.1.tar.gz There have been API additions and changes in the libraries since last release. Although the tarball versions have changed, have the ones in the library map files changed ? Don't they need to be updated to reflect the API additions/changes ? Yes, I missed thiat (thought it was done with the API addition patch). I will update the library version, this will take effect in the next daily build. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
On 13:22 Wed 09 Jul , Hal Rosenstock wrote: Thanks. Is it just libibmad or was it also libibumad as well ? No API related changes were in libibumad. Is it just the *.ver file or also the version in the *.map file ? It is just *.ver file, *.map is symbol versioning - there were additions, no changes in existing functions API. this will take effect in the next daily build. The tar balls just issued won't be reissued ? There were some commits after this already. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
On 13:53 Wed 09 Jul , Hal Rosenstock wrote: It is just *.ver file, *.map is symbol versioning - there were additions, no changes in existing functions API. Isn't that needed too when APIs are added ? I don't think. Also, some existing functions had minor parameter changes (like const being added). Yes, but it does not touch ABI and doesn't affect a linker. this will take effect in the next daily build. The tar balls just issued won't be reissued ? There were some commits after this already. Shouldn't there be a branch for this if that is an issue ? And then to do libibmad-1.2.1.1? Looks like an overkill for this particular case - 1.2.* is in development period and the issue is pretty minor - we will fix it in today's daily build and in the next tarball release. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] List of libraries in OFED
Hi Tziporet, On 18:01 Sun 06 Jul , Tziporet Koren wrote: This is the list of libraries in OFED We wish to define which are the libraries that are public and their API is being preserved from release to release. I agree with this concept and suggest to keep only public and others lists. Actually to have only public list (libraries with stable API) should be enough. There are libraries I already classified, and there are some I am not sure about Each owner - please reply regarding the libraries that are under unknown Also - if I missed any library please add it to the list Thanks, Tziporet Public: === * libibcommon I wanted to remove (merge with libibumad and libibmad) libibcommon yet in OFED 1.3 (we discussed this on the list). Not sure I will be able to do it now (in OFED 1.4 timeframe), but at least let's move it from public list. * libdat2 * libdat * libibcm * libibverbs * libibmad * libibumad * libsdp * librdmacm Private: * libcxgb3 * libehca * libipathverbs * libmlx4 * libmthca * libnes Unknown: * libibdmcom * libdaplcma * libdaplofa * libibdm * libibis * libibmscli * libopensm * libosmcomp * libosmvendor * libosmvendor_openib 'libosmvendor_openib' is not a library, just symbolic link. * libumad2sim Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] IB management daily build tarballs
Hi, I'm starting to run daily builds for IB management packages (libibcommon, libibumad, libibmad, opensm, infiniband-diags). It will be placed at: http://www.openfabrics.org/downloads/management/daily First build tarballs are available already. Some details: 1. Files format are: package_name-base_version-MMDD-git_abbreviation.tar.gz Example: opensm-3.2.1-20080707-9d027ca.tar.gz 2. I think to run it via crontab (on my local machine, not on the OFA server). 3. New tarballs will be generated only for packages where actual changes have been made. Any feedback will be appreciated. Enjoy! Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Compatibility in OFED
On 01:04 Tue 03 Jun , Sasha Khapyorsky wrote: Separately, we should discuss how me manage version changes - introducing a version change in the middle of the RCs seems a bit late in the process. I agree that number of changes (at all) should be minimized in RC period. BTW the change you are complaining about was done Jun 13 2007. I think it was even before OFED-1.3 started. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] OFED meeting sumary on June 02, 08
On 15:25 Tue 03 Jun , Tziporet Koren wrote: 4. Backward compatibility of libraries: Problem: Qlogic found late in OFED 1.3 release process that one of opensm libraries had changed the binary API, and this brake some of the customers' application. Decisions: 1. Need to preserve binary API of libraries and not break backward compatibility (even if it requires extra efforts) It could be impossible for some libraries. 2. In case binary API must be changed, then both versions will be included in OFED to prevent broken applications. Any ideas about how it should be done (technically)? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Compatibility in OFED
On 09:55 Mon 02 Jun , John Russo wrote: In OFED 1.3rc2 we noticed that libosmcomp's version had changed from 1 to 2. Unfortunately this has caused forward compatibility problems for existing applications which were compiled against OFED 1.2.5.1. It would be preferred if when such upgrades occur, both the .1 and .2 version of the library are provided such that existing applications do not need to be recompiled nor reinstalled. What is the problem to recompile? Anyway you can keep old version if you like. Alternatively limiting the necessity for such library version changes (never break old interfaces) would be preferred. It is hard to promise for 100%. libosmcomp is not a great candidate for stable API (it is even not documented). I'm sure that changes were done for a reason and obviously it was published/discussed on the list. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: OpenSM 3.1.11 for OFED 1.3.1
On 04:55 Thu 15 May , Hal Rosenstock wrote: I haven't seen any email yet to Vlad requesting it to be pulled into OFED 1.3 so I assume that is pending. I noticed that this was pulled without request. Also, the release notes would need some minor updating now (renumbering and perhaps mention of additional bug fixes). I think Hal was asking about OpenSM version bumping and tarball release for 1.3.x. If this is not done, then how is this identified as different from 3.1.10 ? 1) by OFED version, 2) by git version suffix (but not sure it appears when OpenSM is compiled from SRPM). We're back to which OFED did the OpenSM come from and I thought the agreed direction was to get away from this as much as possible with tarballs. Correct, and this is how this will work with next major release. Unlike this OFED-1.3.1 is *bug fix* release, which should mean that only limited number of critical fixes should enter there. We probably will generate new tarball at end of 1.3.x cycle (since there are couple of changes already), but I will try to limit number of such releases by only one - don't see a dig value to put version bump commit on every single change. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: OpenSM 3.1.11 for OFED 1.3.1
Hi Ira, On 13:28 Thu 15 May , Ira Weiny wrote: I just sent a patch with a HOWTO for the PerfMgr as well. That could be applied to 1.3 as well.[*] Hmm, is it really needed? PerfMgr is disabled in OFED build, so when one needs this it will be necessary to pull source tree anyway. Right, it could be extracted from SRPM, but I guess latest and greatest will be more useful then. Although it could be useful as promotion. Other opinions? [*] Sasha, sorry for all the last minute patches... :-( No, this is not a problem :) Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Compiling OFED 1.3 on Gentoo
Hi Olga, On 17:18 Mon 12 May , Olga Shern wrote: We are trying to compile OFED 1.3 on Gentoo and see the following error, snip... But if I install source RPM file and then running 'rpmbuild -ba libibcommon.spec' then I can build RPM, so only rpmbuild --rebuild command causing to problems. Basically Gentoo doesn't use RPM as package manager, but builds packages from sources using portage/emerge stuff. So it is nice that it builds somehow. As another workaround rpm2targz probably can be used too. Of course better would be to have native *.ebuild files. Have someone succeeded to build OFED 1.3 on Gentoo? I'm using Gentoo as my workstation, don't do RPMs there however. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: OpenSM 3.1.11 for OFED 1.3.1
Hi Hal, On 04:24 Wed 14 May , Hal Rosenstock wrote: There was just a change has been made for OpenSM on ofed_1_3 branch. I also think there was a doc change to the OpenSM release notes there as well post 3.1.10 (which was what was also made part of OFED 1.3). Also, perhaps Ira's client rereg patch for this version as well ? I don't think - nobody asked for this change in OFED, and I assume nobody tested it there, and 1.3.1-rc1 is already out. Will there be another OpenSM release (say 3.1.11) available for OFED 1.3.1 ? Not sure, but we will see. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Re: OpenSM 3.1.11 for OFED 1.3.1
On 17:19 Wed 14 May , Tziporet Koren wrote: There is the seg-fault bug that was fixed by Yevgeny that we need. This single commit was applied to 1.3 branch. I think Hal was asking about OpenSM version bumping and tarball release for 1.3.x. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
On 14:45 Thu 03 Apr , Scott M. Ferris wrote: I get compile errors for opensm-3.2.1 because osm_console_io.h is missing. Does the make dist target need to be updated to put that file in the tarball? Sure, it should be. I will re upload fixed tarball. If you're going to respin the package for that, could you also do a quick test of opensm with no IB cable attached to the HCA? Unfortunately I cannot do it now - don't have any equipment available. I found that opensm 3.2.0 would spin and hog a CPU when there was no cable attached. It's a pathological case, but sometimes happens in my lab. Thanks for reporting (although it would be better to have this report right after 3.2.0). I will look at this after Sonoma. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ANNOUNCE] management tarballs release
On 21:35 Thu 03 Apr , Sasha Khapyorsky wrote: Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: b398ef1246a392338053c8e382b3e6ee libibcommon-1.1.0.tar.gz abce72fbb91530a97493eba7a28a0de6 libibumad-1.2.0.tar.gz fe7a6b80b28e56cf74ffbe09c8819c71 libibmad-1.2.0.tar.gz b0695f75cda10051c8846fd22b77491a opensm-3.2.1.tar.gz OpenSM tarball was replaced by: 997d10f81896a0d70e0f21f0e78eca92 opensm-3.2.1.tar.gz (due to compilation issue). Sorry about inconsistency. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [ANNOUNCE] management tarballs release
On 22:18 Thu 03 Apr , Bernd Schubert wrote: In file included from main.c:61: ../include/opensm/osm_opensm.h:56:35: error: opensm/osm_console_io.h: No such file or directory Same here, Should be fixed now. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] OFED March 10 teleconfrance meeting summary
On 16:28 Tue 11 Mar , Tziporet Koren wrote: OFED meeting summary: = 1. Decide on fix release criteria and process (see below) 2. OFED 1.4: - Target date: Sep 2008 - Will start with kernel 2.6.25 once it released - Next meeting will be devoted to define the features set. In addition to this I would suggest to switch to tarball based distribution for userspace components with OFED 1.4 (as was discussed many times - maintainers will release tarballs, no ofed_1_4/package any more, etc). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: Please make sure your document is updated for OFED 1.3
On 09:57 Tue 26 Feb , Yevgeny Kliteynik wrote: The version is inside the opensm_release_notes.txt file: @@ -1,4 +1,4 @@ -OpenSM Release Notes 3.1.9 +OpenSM Release Notes 3.1.10 Of course... Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [PATCH] opensm: rename OpenSM startup script to opensmd
On 08:05 Mon 25 Feb , Hal Rosenstock wrote: On Sun, 2008-02-24 at 14:15 +, Sasha Khapyorsky wrote: Rename OpenSM startup script /etc/init.d/opensm - /etc/init.d/opensmd. Does any OFED README or build or install script need updating due to the name change ? Not that I know. In OFED-1.2 script was named with 'd'. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: Please make sure your document is updated for OFED 1.3
On 00:02 Tue 26 Feb , Yevgeny Kliteynik wrote: opensm_release_notes.txt - OpenSM version should be 3.1.10. If you prefer, I can provide a patch. The content is same. Under OFED docs it is w/out version prefix... Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] SLES10 X86_64 build problem found with 1.3rc4
Hi, On 07:22 Wed 20 Feb , Todd Rimmer wrote: Hello, I was testing some aspects of 1.3rc4 build and discovered that I could not install it on SLES10 x86_64 with the --build32 option for 32 bit libraries, it worked fine if I used the 64 bit libraries only. Here is the relevant information from the install/build failure. Seems like opensm srpm is using an incorrect library path. I can see earlier in the build where it built the 32 bit libosmcomp library which should have the missing symbols. Error Message from install.pl when installing all and using --build32 command line option: Build opensm-libs RPM Running rpmbuild --rebuild --define '_topdir /var/tmp/OFED_topdir' --target x86_64 --define '_prefix /usr' --define '_exec_prefix /usr' --define '_sysconfdir /etc' --define '_usr /usr' /root/todd/OPENIB_INSTALL/stage/QLogicIB-IFS.0root/QLogic-OFED-1.3-rc4/S RPMS/opensm-3.1.8-0.1.ofed20080207.src.rpm Running LDFLAGS=' -m32 -g -O2 -L/usr/lib' CFLAGS=' -m32 -g -O2' CPPFLAGS=' -m32 -g -O2' CXXFLAGS=' -m32 -g -O2' FFLAGS=' -m32 -g -O2' LDLIBS=' -m32 -g -O2 -L/usr/lib' rpmbuild --rebuild --define '_topdir /var/tmp/OFED_topdir' --target i586 --define '_prefix /usr' --define '_exec_prefix /usr' --define '_sysconfdir /etc' --define '_usr /usr' --define '_lib lib' --define '__arch_install_post %{nil}' /root/todd/OPENIB_INSTALL/stage/QLogicIB-IFS.0root/QLogic-OFED-1.3-rc4/S RPMS/opensm-3.1.8-0.1.ofed20080207.src.rpm Failed to build opensm RPM See /tmp/OFED.22149.logs/opensm.rpmbuild32bit.log I have attached the referenced log Currently I cannot reproduce it with RHEL. Will try to find SLES10... FYI, how do I go about getting a bugzilla account so I can enter problems like this directly into bugzilla? https://bugs.openfabrics.org/createaccount.cgi Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [PATCH] ofed/docs: remove ofed 1.2 related bugs from OpenSM Release Notes
Remove list of bugs which was fixed in OFED-1.2 period from current OpenSM Release Notes. Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED] --- Tziporet, sorry but there is already fix for this. Sasha opensm_release_notes.txt | 142 -- 1 files changed, 0 insertions(+), 142 deletions(-) diff --git a/opensm_release_notes.txt b/opensm_release_notes.txt index 8c6f3e4..40a253f 100644 --- a/opensm_release_notes.txt +++ b/opensm_release_notes.txt @@ -278,148 +278,6 @@ or visible bugs were also fixed. * Protect SMInfo response against port moving issue. -* osm_sminfo_rcv.c: Add SMInfo self query check. OpenSM can query - itself for SMInfo occasionally due to port moving during subnet - discovery process. Don't create remote SM entry in this case to - prevent deadlocks. - -* osm_ucast_updn.c: Two similar bugs in up/down routing fixed. - 8-bit integers were used as indexes when scanning subnet, which - in one case caused OpenSM to crash when ranking path is longer - than 256 switches, and in the other case, caused OpenSM to go into - an infinite loop when fabric has more than 256 roots. - -* osm_sm_state_mgr.c: In __osm_sm_state_mgr_send_master_sm_info_req, - handle master GUID port not found properly - -* osm_sa_multipath_record.c: In __osm_mpr_rcv_get_path_parms, return - IB_NOT_FOUND rather than IB_ERROR when can't route to LID from switch - -* osm_sa_path_record.c: In __osm_pr_rcv_get_path_parms, return IB_NOT_FOUND - rather than IB_ERROR when can't route to LID from switch - -* osm_vendor_ibumad.c: In osm_vendor_set_sm, set issmfd to - -1 on open error - -* osm_vendor_ibumad: Termination crash fix - When OpenSM is terminated umad_receiver thread still running even after - the structures are destroyed and freed, this causes to random (but easily - reproducible) crashes. The reason is that osm_vendor_delete() does not - care about thread termination. This patch adds the receiver thread - cancellation (by using pthread_cancel() and pthread_join()) and cares to - keep have all mutexes unlocked upon termination. There is also minor - termination code consolidation - osm_vendor_port_close() function. - -* osm_port_profile.h: Fix reinsertion issue in osm_port_prof_set_ignored_port - -* osm_matrix.h: Fix segfault with up/down and root nodes file - -* osm_sa_path_record.c: In osm_pr_rcv_process, fix endian of hop_limit - -* osm_vendor_ibumad.c: Close umad port in osm_vendor_delete - -* osm_sa_(multipath path)_record.c: Fix MultiPathRecord/PathRecord issues - with using MTU/rate/PktLife explicitly ignoring selectors - - OpenSM just uses the resulting path MTU/rate/pkt-life and fail the - query even though the selector might be allowing for selecting an - appropriate value. - - After this fix, the following results are obtained for a case of - path allowing maximal 2K MTU. - -In standard mode: - -MTU greater than ... 256 (0x01) - equal to ... 2K -MTU less than .. 256 (0x41) - NO PATHS -MTU equal to ... 256 (0x81) - equal to ... 256 -MTU largest possible 256 (0xc1) - equal to ... 2K -MTU greater than ... 512 (0x02) - equal to ... 2K -MTU less than .. 512 (0x42) - equal to ... 256 -MTU equal to ... 512 (0x82) - equal to ... 512 -MTU largest possible 512 (0xc2) - equal to ... 2K -MTU greater than ... 1K (0x03) - equal to ... 2K -MTU less than .. 1K (0x43) - equal to ... 512 -MTU equal to ... 1K (0x83) - equal to ... 1K -MTU largest possible 1K (0xc3) - equal to ... 2K -MTU greater than ... 2K (0x04) - NO PATHS -MTU less than .. 2K (0x44) - equal to ... 1K -MTU equal to ... 2K (0x84) - equal to ... 2K -MTU largest possible 2K (0xc4) - equal to ... 2K -MTU greater than ... 4K (0x05) - NO PATHS -MTU less than .. 4K (0x45) - equal to ... 2K -MTU equal to ... 4K (0x85) - NO PATHS -MTU largest possible 4K (0xc5) - equal to ... 2K - - -With enable_quirks (when one of the ends is a Tavor device): - -MTU greater than ... 256 (0x01) - equal to ... 1K -MTU less than .. 256 (0x41) - NO PATHS -MTU equal to ... 256 (0x81) - equal to ... 256 -MTU largest possible 256 (0xc1) - equal to ... 2K -MTU greater than ... 512 (0x02) - equal to ... 1K -MTU less than .. 512 (0x42) - equal to ... 256 -MTU equal to ... 512 (0x82) - equal to ... 512 -MTU largest possible 512 (0xc2) - equal to ... 2K -MTU greater than ... 1K (0x03) - NO PATHS -MTU less than .. 1K (0x43) - equal to ... 512 -MTU equal to ... 1K (0x83) - equal to ... 1K -MTU largest possible 1K
Re: [ewg] Trying to install ibutils (from OFED-1.3)
On 05:24 Sun 17 Feb , Hal Rosenstock wrote: I think --with-osm option is used only to specify where OpenSM is located. Couldn't this default to the default place where OpenSM is installed ? I think it is when no any '--with(out)-osm' is specified. And why dos it state gen1? This is to select the vendor layer. Just looked in ibutils/config/osm.m4 - it tries to find OpenSM in standard plces and if it is not found decides that it should be under IBGold installation and go to 'gen1' option. When failed to detect it ibutils thinks it is in IBGold somewhere. Should it default to OpenIB (gen2 or whatever it's called in ibutils) rather than gen1 ? It does so by default (when no --without-osm option is not specified). I think you need to install at least opensm-libs and opensm-devel packages. Also, libibumad. Right, but it is in opensm-libs dependencies list. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] Trying to install ibutils (from OFED-1.3)
On 09:13 Sun 17 Feb , Hal Rosenstock wrote: At one point, single stepping OpenSM state used to be supported. Maybe still in the ibutils side (I know it's been ripped out of OpenSM). It would not pass compilation then. For my best knowledge 'single step' was used by osmsh, not ibutils. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [PATCH] ofed_scripts/install.pl: add dist_req_build for opensm-libs
Add dist_req_build for opensm-libs package - this should be same as for opensm, because its build is similar. Without this ofed_1_3 build where only opensm-libs was selected (sometimes implicitly) fails in build time on machine where bison is not installed. Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED] --- install.pl |2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/install.pl b/install.pl index e9ea645..eed0dc5 100755 --- a/install.pl +++ b/install.pl @@ -683,7 +683,7 @@ my %packages_info = ( 'opensm-libs' = { name = opensm-libs, parent = opensm, selected = 0, installed = 0, rpm_exist = 0, rpm_exist32 = 0, -available = 1, mode = user, dist_req_build = [], +available = 1, mode = user, dist_req_build = [bison, flex], dist_req_inst = [], ofa_req_build = [libibumad-devel], ofa_req_inst = [libibumad], install32 = 1, exception = 0 }, -- 1.5.4.rc2.60.gb2e62 ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Sasha Khapyorsky (1): management: bump versions Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
Re: [ewg] [ANNOUNCE] management tarballs release
On 13:57 Tue 12 Feb , Tziporet Koren wrote: Is this the last planed release of OSM for OFED 1.3? It would be ideal (if there will not be new bugs or so). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 1c9764865b4c4f03529494f4272a7daf libibcommon-1.0.8.tar.gz 879688a264e982600628dfa576c84cea libibumad-1.1.7.tar.gz de275d0aba15b587f00ac1e30ba8b0e8 libibmad-1.1.6.tar.gz 17678a2eaeb0cae9273b091da00cc7dc infiniband-diags-1.3.6.tar.gz 92385653112cd915ec31542eb2d94714 opensm-3.1.9.tar.gz 0f43afda110cfc285ef89637fcbf6ed7 opensm-3.2.0.tar.gz opensm-3.1.9 is recent OFED version. opensm-3.2.0 is recent master. The rest is the same. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Ira K. Weiny (2): Move opensm.8 man page in prep for making config file changes. Update man page for configurable partition and prefix-routes file Ira Weiny (1): Add node name map, partition config, and QOS policy config files to the FILES section of man page. Sasha Khapyorsky (1): opensm: scripts/opensmd - fix opensm path. Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [GIT PULL] ~sashak/management.git
On 18:08 Mon 28 Jan , Sasha Khapyorsky wrote: Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Plus those, last minute ones: Hal Rosenstock (3): opensm/osm_subnet.c: Better clarity in opensm.opts file for perfmgr_redir opensm/osm_perfmgr.c: Fix duplicated error code opensm/osm_perfmgr.c: If redirection requested but disabled, don't rerequest Sasha Khapyorsky (3): opensm/perfmgr: remove unused label - fix compilation warning opensm/perfmgr: wakeup perfmgr discover only after NO_PENDING_TRANSACTION was signaled opensm/perfmgr: break perfmgr discovery if osm_exit_flag is on Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Ira K. Weiny (1): infiniband-diags/scripts/iblinkinfo.pl: Fix switch to switch output with new Ira Weiny (11): infiniband-diags/scripts/IBswcountlimits.pm: Fix comment infiniband-diags/scripts/IBswcountlimits.pm: clean up long lines infiniband-diags/scripts/iblinkinfo.pl: clean up output format Add -C ca_name and -P ca_port to ibprint[ca|rt|switch].pl Update perl script man pages for -C and -P options infiniband-diags/scripts/ibprintswitch.pl: fix regex when searching for switch by name infiniband-diags/scripts/ibprintrt.pl: fix regex when searching for router by name infiniband-diags/scripts/ibprintca.pl: fix regex when searching for node by name opensm/opensm/osm_subnet.c: add a comment of valid force_link_speed values opensm/scripts/redhat-opensm.init: fix starting opensm when using daemon mode opensm/opensm/osm_subnet.c: update valid force_link_speed values to v1.2.1 of the spec Sasha Khapyorsky (11): opensm/opensm.spec.in: add Obsoletes rule infiniband-diags/infiniband-diags.spec.in: add Obsoletes rule opensm/lash: fix possible segfault in osm_get_lash_sl() infiniband-diags/saquery: usage field to struct query_cmd infiniband-diags/saquery: code consolidation opensm/scripts: fix sldd.sh paths opensm/opensm.spec: install scripts/sldd.sh management/libib*: remove trailing whitespaces opensm: remove trailing whitespaces infiniband-diags: remove trailing whitespaces management: clean whitespaces Tim Meier (2): opensm: diags add DR path support to some utils opensm: update man pages for diags DR support Timothy A. Meier (1): opensm: osm_subnet.c log and print the path of the cached option file Yevgeny Kliteynik (2): opensm/scripts: removing trailing blanks opensm/scripts: fixing MAXSMPS values to the right default Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: Ramifications of outstanding ibutils patches relative to management
Hi Hal, On 05:25 Sat 26 Jan , Hal Rosenstock wrote: AFAIK there currently appear to be 2 outstanding patches against ibutils which have not yet been incorporated for OFED 1.3. Are you about pkey index related patches? Assuming so... What are the ramifications of this in terms of OFED 1.3 ? Will ibutils work with the portions of management that it uses ? Is there something that wouldn't ? I think it works now (for unclear reasons). Anyway I'm planning to apply pkey_ix renaming patch next week - it will break ibutils build then. Hope to have any feedback before this. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] the OFA server is out of free space
Hi, Recently I tried to git-push and got 'No space left on device' error: $ git push Counting objects: 28, done. Compressing objects: 100% (19/19), done. Writing objects: 100% (19/19), 3.45 KiB, done. Total 19 (delta 16), reused 0 (delta 0) error: file write error (No space left on device) fatal: unable to write sha1 file ... Fast look shows: $ df / Filesystem 1K-blocks Used Available Use% Mounted on /dev/sda1151873632 14415874476 100% / $ du -s /tmp 3030328 /tmp/ (and most big things there are from 2007) $ time du -s /home 106945588 /home real36m43.422s ( - yes, this is true) user0m7.660s sys 0m49.910s Guys! What about to clean things up? Thanks! Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Sasha Khapyorsky (4): infiniband-diags/configure.in: complib doesn't have opensm dependencies anymore opensm/perfmgr: use pkey at index 0 libibumad: increase the version of the library libibumad/man: umad_get_pkey man page Yevgeny Kliteynik (2): opensm/osm_ucast_ftree.c: fixing coredump in fat-tree routing opensm/osm_ucast_ftree.c: cosmetics in log messages Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Ira Weiny (1): Fix perfmgr enable in spec file. Sasha Khapyorsky (19): management/*/Makefile.am: fix ChangeLog generation rule management/gen_ver.sh script opensm: update OpenSM version according to the tree state opensm/Makefile.am: update configure.in version on 'make dist' management/*/Makefile.am: update configure.in version on 'make dist' opensm/opensm/Makefile.am: fix version update rule osmtest: remove unused error.h file infiniband-diags: use common build version infiniband-diags: remove not needed anymore __BUILD_VERSION_TAG__ infiniband-diags: update version according to the tree state opensm/vendor: filter out non-IB devices in osm_vendor_get_all_port_attr() opensm/osm_sa_slvl_record: remove unused variable opensm/osm_sa_slvl_record: fix overflow crash libibumad: umad_get_pkey() function opensm/vendor: use valid pkey index value for gsi mads infiniband-diags/saquery: attribute names support infiniband-diags/saquery: support for SL2VLTableRecord attribute infiniband-diags/saquery: support ports with LinkRecord query infiniband-diags/saquery: support for VLArb and PKey Table Records Yevgeny Kliteynik (2): opensm/osm_qos_policy.c: trivial fix in passing wrong pointer opensm/osm_mcast_mgr.c: fixing a seg. fault in processing mcast groups Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [GIT PULL] ~sashak/management.git
Vlad, hi again, On 01:00 Tue 15 Jan , Sasha Khapyorsky wrote: Please pull recent ofed_1_3 branch of ~sashak/management.git. I found that your ofed_1_3/management tree already consists most of our recent changes (which is fine IMO :)). Does it mean that you run git-pull automatically and I don't need to send [GIT PULL] request to you? Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Sasha Khapyorsky (5): opensm: Revert opensm/osm_pkey_mgr.c: setting only outbound partition enforcement on switch opensm: mcast mgr improvements infiniband-diags/dump_lfts.sh: adopt DR Path parser infiniband-diags/dump_lfts.sh: fix switch DR Path parser management: bump versions Yevgeny Kliteynik (2): opensm/osm_pkey_mgr.c: setting only outbound partition enforcement on switch opensm: osm_state_mgr.c - stop idle queue processing if heavy sweep requested Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: df15dde16ee5b28c25affa2dedaa3d0f libibcommon-1.0.7.tar.gz 124f9f5c3e0c28afb86bb8bbe4a60ecd libibumad-1.1.6.tar.gz 047a6d9f834d7012185e5a28f0f6141f libibmad-1.1.5.tar.gz 53675686246f0cf92853358483f2d0f3 opensm-3.1.8.tar.gz 892dab82b783a4ff2063a438aa93418f infiniband-diags-1.3.5.tar.gz Complete list of changes since last release is below: Sasha Al Chu (6): osm cache file extra info patch OpenSM: Fix error return corner case OpenSM: Fix comment typo. OpenSM: osm routing engine type OpenSM: Fix incorrect reporting of routing engine used OpenSM: Fix incorrect identification of routing engine used Albert L. Chu (2): support minhop as a routing engine input change 'min-hop' to 'minhop' for consistency with routing engine input Erez Strauss (1): ibnetdiscover - ports report Hal Rosenstock (3): OpenSM/libvendor/osm_vendor_ibumad.c: Make error code in osm_log message unique libibmad/dump.c: Use bit mask approach to decoding LinkWidth/Speed Enabled/Supported opensm/libvendor/osm_vendor_ibumad_sa.c: In __osmv_sa_mad_rcv_cb, handle attribute offset of 0 Ira Weiny (5): opensm/include/opensm/osm_event_plugin.h: fix comment opensm/opensm/osm_event_plugin.c: clean up version check error message a bit opensm/opensm/osm_event_plugin.c: remove duplicate header include opensm/opensm/osm_perfmgr_db.c: fix clearing previous count when out of band opensm: Add perfmgr print_counters node to the console to print individual values Rolf Manderscheid (1): opensm: allow multiple scopes in a partition Sasha Khapyorsky (50): opensm: move vendor specific header files to include/vendor opensm: remove unused flag opensm: move IBA constants from osm_sa_mcmember_record.h to ib_types.h opensm: move OpenSM constants from osm_sa_mcmember-record.h to osm_base.h opensm: cosmetic fixes opensm: make osm_pkey_get_tables static opensm: remove testability_mode option libibumad: fix memory leak infiniband-diags/saquery: add get_any_records() function infinibad-diags/saquery: move lid resolving functions infiniband-diags/saquery: LinkRecord query support infiniband-diags/saquery: allow empty src and/or dst with --src-to-dst option infiniband-diags/man: add -x option to saquery man page opensm: fix lmc_mask bit order in osm_sa_link_record.c opensm: don't break name_map using when routing_engine was not found. opensm: remove unused osm_port_lid_category_t enum opensm: minor cleaning opensm/lash: fix wrong allocation size opensm/lash: cosmetic opensm/osm_ucast_updn.c: indentation fixes opensm/libvendor: indentation fixes infiniband-diags/ibcheckerrors: for CAs query only single ports opensm/config/osmvsel.m4: update LDADD variable, not LDFLAGS infiniband-diags/ibcheckerrors: fix port errors count infiniband-diags/scripts: fix perfquery usage opensm: don't zero base LID when invalid value is received opensm: remove old style code formatters opensm/Makefile: remove opensm_CXXFLAGS opensm: recover only for base LID values = 0xc000 opensm/osm_port_info_rcv: node instead of port as parameter for osm_pi_rcv_process_set() opensm/osm_node_new: move p_node-print_desc setup libibmad: initialize sm portid in ib_resolve_smlid() libibcommon: fix overflow in debug/log prints opensm: rename __osm_epi_plugin_t to osm_event_plugin_t opensm/osm_event_plugin.h: add names to structures opensm: remove useless osm_node_get_remote_type() opensm: indentation fixes manangement: kill __WORDSIZE macro checks complib: make __cl_thread_wrapper() static opensm: make some functions static opensm/osm_helper: make some functions static opensm: make some functions statics opensm: some micro-optimizations opensm/updn: report fallback properly opensm/updn: rename __osm_subn_calc_up_down_min_hop_table() opensm: Revert opensm/osm_pkey_mgr.c: setting only outbound partition enforcement on switch opensm: mcast mgr improvements infiniband-diags/dump_lfts.sh: adopt DR Path parser infiniband-diags/dump_lfts.sh: fix switch DR Path parser management: bump versions Yevgeny Kliteynik (10): opensm: Remove unnecessary ntoh and hton conversions in LinkRecord processing opensm: adding missing comparison by to_lid/from_lid in LinkRecord processing opensm: Fixing broken logic in 'process world' part of LinkRecord processing opensm: printing to stderr note about error in QoS policy file opensm: trivial change of log message opensm: fixing coredump in QoS policy pkey validation
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Al Chu (4): OpenSM: Fix comment typo. OpenSM: osm routing engine type OpenSM: Fix incorrect reporting of routing engine used OpenSM: Fix incorrect identification of routing engine used Erez Strauss (1): ibnetdiscover - ports report Ira Weiny (1): opensm: Add perfmgr print_counters node to the console to print individual values Sasha Khapyorsky (14): libibmad: initialize sm portid in ib_resolve_smlid() libibcommon: fix overflow in debug/log prints opensm: rename __osm_epi_plugin_t to osm_event_plugin_t opensm/osm_event_plugin.h: add names to structures opensm: remove useless osm_node_get_remote_type() opensm: indentation fixes manangement: kill __WORDSIZE macro checks complib: make __cl_thread_wrapper() static opensm: make some functions static opensm/osm_helper: make some functions static opensm: make some functions statics opensm: some micro-optimizations opensm/updn: report fallback properly opensm/updn: rename __osm_subn_calc_up_down_min_hop_table() Yevgeny Kliteynik (1): opensm/osm_pkey_mgr.c: trivial fix in log message Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Sasha Khapyorsky (3): opensm: recover only for base LID values = 0xc000 opensm/osm_port_info_rcv: node instead of port as parameter for osm_pi_rcv_process_set() opensm/osm_node_new: move p_node-print_desc setup Thanks. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Al Chu (1): OpenSM: Fix error return corner case Hal Rosenstock (1): opensm/libvendor/osm_vendor_ibumad_sa.c: In __osmv_sa_mad_rcv_cb, handle attribute offset of 0 Sasha Khapyorsky (9): opensm/osm_ucast_updn.c: indentation fixes opensm/libvendor: indentation fixes infiniband-diags/ibcheckerrors: for CAs query only single ports opensm/config/osmvsel.m4: update LDADD variable, not LDFLAGS infiniband-diags/ibcheckerrors: fix port errors count infiniband-diags/scripts: fix perfquery usage opensm: don't zero base LID when invalid value is received opensm: remove old style code formatters opensm/Makefile: remove opensm_CXXFLAGS Yevgeny Kliteynik (3): opensm: trivial change of log message opensm: fixing coredump in QoS policy pkey validation opensm: QoS policy - fixing pkey range implementation Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [PATCH] infiniband-diags/ibcheckerrors: for CAs query only single ports
Hi Hal, On 10:41 Tue 11 Dec , Hal Rosenstock wrote: I'm not close that 'all ports' simulation in perfquery is great thing. In a sense, it's no different than what the agent itself might be doing; albeit over a larger time span. perfquery is low level tool and it should be able to indicate in clear way that 'all ports' option is not supported by port instead of hiding this behind simulation. Maybe 'all ports' simulation should optional... Guess you did a 180 turn on this. Last I recall on the list you wanted this functionality. Just wanted it fully functional... Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [PATCH] infiniband-diags/ibcheckerrors: for CAs query only single ports
On 10:30 Wed 12 Dec , Hal Rosenstock wrote: Hi Sasha, On Wed, 2007-12-12 at 17:48 +, Sasha Khapyorsky wrote: Hi Hal, On 10:41 Tue 11 Dec , Hal Rosenstock wrote: I'm not close that 'all ports' simulation in perfquery is great thing. In a sense, it's no different than what the agent itself might be doing; albeit over a larger time span. perfquery is low level tool and it should be able to indicate in clear way that 'all ports' option is not supported by port instead of hiding this behind simulation. Maybe 'all ports' simulation should optional... Guess you did a 180 turn on this. Last I recall on the list you wanted this functionality. Just wanted it fully functional... And now ? And now too. But think it should be off by default, a new command line option could be added to turn it on. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [PATCH] 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. Signed-off-by: Sasha Khapyorsky [EMAIL PROTECTED] --- infiniband-diags/scripts/ibcheckerrors.in | 32 ++-- 1 files changed, 12 insertions(+), 20 deletions(-) diff --git a/infiniband-diags/scripts/ibcheckerrors.in b/infiniband-diags/scripts/ibcheckerrors.in index cac2475..5cfabc6 100644 --- a/infiniband-diags/scripts/ibcheckerrors.in +++ b/infiniband-diags/scripts/ibcheckerrors.in @@ -79,15 +79,15 @@ echo $text | awk ' BEGIN { ne=0 } -function check_node(lid) +function check_node(lid, port) { nodechecked=1 if (system('$IBPATH'/ibchecknode '$ca_info' '$gflags' '$verbose' lid)) { ne++ - badnode=1 + print \n# ntype : nodeguid 0x nodeguid failed return } - if (system('$IBPATH'/ibcheckerrs '$ca_info' '$gflags' '$verbose' '$brief' lid 255)) + if (system('$IBPATH'/ibcheckerrs '$ca_info' '$gflags' '$verbose' '$brief' lid port)) nodeerr=1; } @@ -105,30 +105,22 @@ function check_node(lid) lid = substr($0, index($0, port 0 lid ) + 11) lid = substr(lid, 1, index(lid, ) - 1) - check_node(lid) + check_node(lid, 255) } /^\[/ { nports++ port = $1 - if (!nodechecked) { - lid = substr($0, index($0, lid ) + 5) - lid = substr(lid, 1, index(lid, ) - 1) - check_node(lid) - } - if (badnode) { - print \n# ntype : nodeguid 0x nodeguid failed - next - } sub(\\(.*\\), , port) gsub([\\[\\]], , port) - if (nodeerr) - if (system('$IBPATH'/ibcheckerrs '$ca_info' '$gflags' '$verbose' '$brief' lid port)) { - if (!'$v' oldlid != lid) { - print # Checked ntype : nodeguid 0x nodeguid with failure - oldlid = lid - } + if (ntype != Switch) { + lid = substr($0, index($0, lid ) + 5) + lid = substr(lid, 1, index(lid, ) - 1) + check_node(lid, port) + if (nodeerr) pcnterr++; - } + } else if (nodeerr + system('$IBPATH'/ibcheckerrs '$ca_info' '$gflags' '$verbose' '$brief' lid port)) + pcnterr++; } /^ib/ {print $0; next} -- 1.5.3.rc2.29.gc4640f ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ofa-general] [PATCH] infiniband-diags/ibcheckerrors: for CAs query only single ports
On 16:46 Tue 11 Dec , Sasha Khapyorsky wrote: On 07:25 Tue 11 Dec , Hal Rosenstock wrote: On Tue, 2007-12-11 at 15:27 +, Sasha Khapyorsky wrote: On 06:57 Tue 11 Dec , Hal Rosenstock wrote: On Tue, 2007-12-11 at 13:46 +, Sasha Khapyorsky wrote: For CAs query performance counters only for single ports by lid and port number, and not whole node with 'all ports' option. Should the description also reference the bug # ? I will add. Will a similar thing be done to the other diag scripts which have this same issue (but haven't been reported yet) ? It is reasonable. I will try to check other scripts too. Would it be better to fix this in the underlying tool used (perfquery) and in that way address it for all the diag scripts ? I think perfquery could/should be improved as well, but it is not the same issue. Why not ? If perfquery paved over the lack of support for all ports, then all the scripts would be fine as is, right ? Another aspect of this. I'm not close that 'all ports' simulation in perfquery is great thing. perfquery is low level tool and it should be able to indicate in clear way that 'all ports' option is not supported by port instead of hiding this behind simulation. Maybe 'all ports' simulation should optional... I'm not sure yet. Also I think that when perfquery targets CA port just by LID and when port number is not specified 'all ports' should not be default, but instead port number of this LID. Such behavior seems to be more native for me. Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Al Chu (1): osm cache file extra info patch Albert L. Chu (2): support minhop as a routing engine input change 'min-hop' to 'minhop' for consistency with routing engine input Hal Rosenstock (2): OpenSM/libvendor/osm_vendor_ibumad.c: Make error code in osm_log message unique libibmad/dump.c: Use bit mask approach to decoding LinkWidth/Speed Enabled/Supported Ira Weiny (4): opensm/include/opensm/osm_event_plugin.h: fix comment opensm/opensm/osm_event_plugin.c: clean up version check error message a bit opensm/opensm/osm_event_plugin.c: remove duplicate header include opensm/opensm/osm_perfmgr_db.c: fix clearing previous count when out of band Rolf Manderscheid (1): opensm: allow multiple scopes in a partition Sasha Khapyorsky (19): opensm: move vendor specific header files to include/vendor opensm: remove unused flag opensm: move IBA constants from osm_sa_mcmember_record.h to ib_types.h opensm: move OpenSM constants from osm_sa_mcmember-record.h to osm_base.h opensm: cosmetic fixes opensm: make osm_pkey_get_tables static opensm: remove testability_mode option libibumad: fix memory leak infiniband-diags/saquery: add get_any_records() function infinibad-diags/saquery: move lid resolving functions infiniband-diags/saquery: LinkRecord query support infiniband-diags/saquery: allow empty src and/or dst with --src-to-dst option infiniband-diags/man: add -x option to saquery man page opensm: fix lmc_mask bit order in osm_sa_link_record.c opensm: don't break name_map using when routing_engine was not found. opensm: remove unused osm_port_lid_category_t enum opensm: minor cleaning opensm/lash: fix wrong allocation size opensm/lash: cosmetic Yevgeny Kliteynik (4): opensm: Remove unnecessary ntoh and hton conversions in LinkRecord processing opensm: adding missing comparison by to_lid/from_lid in LinkRecord processing opensm: Fixing broken logic in 'process world' part of LinkRecord processing opensm: printing to stderr note about error in QoS policy file Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [ANNOUNCE] management tarballs release
Hi, There is a new release of the management (OpenSM and infiniband diagnostics) tarballs available in: http://www.openfabrics.org/downloads/management/ md5sum: 25b9491f90c7e851f5bafd556bcac5f6 libibcommon-1.0.6.tar.gz cec79c655914b83d13457d9a4096ef2b libibumad-1.1.5.tar.gz 169ddc65af6d1cfe8a92988ce7a60627 libibmad-1.1.4.tar.gz f8abde643dab46ce9b1dfec2fa644ab5 infiniband-diags-1.3.4.tar.gz 30f2260c4fe2ad0737d644e737b0d3bf opensm-3.1.7.tar.gz (libibcommon was not changed since last release). Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg
[ewg] Re: [GIT PULL] ~sashak/management.git
Hi Vlad, Please pull recent ofed_1_3 branch of ~sashak/management.git. The changes are: Hal Rosenstock (1): libibmad/dump.c: Support link speed and width vendor extensions Ira Weiny (7): Break out a CONF_DIR variable in configure to base other config files on. Add option to change the default opensm dir under sysconfdir as the config dir Add the default in the with-node-name-map help string opensm/configure.in: remove unecessary comment Add --with-qos-policy-conf to configure Add --with-partitions-conf to configure opensm/config/osmvsel.m4: Convert help strings to AC_HELP_STRING Rolf Manderscheid (2): opensm: support multiple routers in a subnet PATCH diags: add saquery option to get path records by GIDs Sasha Khapyorsky (30): opensm: fix PortInfo update issues. opensm: eliminate some unneeded PortInfo Set requests opensm/osm_link_mgr: use return status when PortInfo is updated opensm: update default config names in man pages and doc opensm/scripts/opensm.conf: fix log file name in comment opensm: merge redhat-opensm.conf and opensm.conf opensm/main.c: trivial cosmetic improvements opensm/man: break long lines, remove utf-8 characters opensm/doc: remove utf-8 characters from text files infiniband-diags: remove utf-8 characters from sources and man pages opensm/man/opensm.8: remove utf-8 characters libibumad: fix NULL pointer referencing opensm/libvendor: remove not used umad_ca field opensm/libvendor: fix umad_port leak opensm/vendor: handle guid = 0 with osm_vendor_bind() opensm/libvendor: make statics libvendor: osm_vendor_get_all_port_attr() rework opensm/main: merge gen2 code in get_port_guid() opensm/osmtest: adopt to osm_vendor_get_all_port_attr() changes opensm: no interactive games in daemon mode infiniband-diags/ibtracert: print lids in decimal form opensm/osmtest: update port chooser opensm/libvendor: remove __osmv_get_lid_and_sm_lid_by_port_guid() opensm/vendor: be sure that node has IB type opensm/configure: add OPENSM_CONFIG_DIR variable opensm: use OPENSM_CONFIG_DIR for prefix-routes file definition libibumad: version 1.1.5 libbmad: version 1.1.4 opensm: version 3.1.7 infiniband-diags: version 1.3.4 Yevgeny Kliteynik (10): osm: fixing memory leak in QoS policy osm: fixing memory leak in node description osm: remove unused 'is_ipoib' parameter when adding mcast group osm: use OSM_DEFAULT_PARTITION_CONFIG_FILE instead of hard-coded string osm: moving call to QoS policy file parser to osm_qos_setup() osm: cosmetics - removing obsolete comment osm: improving error reporting function of the QoS parser osm: broken logic when scanning subnet for PIR request osm: Using new format of QoS parser error function osm: QoS - adding simplified syntax for policy definition Thanks, Sasha ___ ewg mailing list ewg@lists.openfabrics.org http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg