opal_info_support: output component versions #1515

ghost · 2016-04-02T15:00:00Z

Hopefully a better approach than #1514. Aims at fixing #1396.

Tests:

% ompi_info --param btl all
             MCA btl: vader (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: sm (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: self (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: tcp (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: parameter "btl_tcp_if_include" (current value: "",
                      data source: default, level: 1 user/basic, type:
                      string)
                      Comma-delimited list of devices and/or CIDR
                      notation of networks to use for MPI communication
                      (e.g., "eth0,192.168.0.0/16").  Mutually exclusive
                      with btl_tcp_if_exclude.
             MCA btl: parameter "btl_tcp_if_exclude" (current value:
                      "127.0.0.1/8,sppp", data source: default, level: 1
                      user/basic, type: string)
                      Comma-delimited list of devices and/or CIDR
                      notation of networks to NOT use for MPI
                      communication -- all devices not matching these
                      specifications will be used (e.g.,
                      "eth0,192.168.0.0/16").  If set to a non-default
                      value, it is mutually exclusive with
                      btl_tcp_if_include.
             MCA btl: parameter "btl_tcp_progress_thread" (current value:
                      "0", data source: default, level: 1 user/basic,
                      type: int)


% ompi_info --param btl tcp
             MCA btl: tcp (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: parameter "btl_tcp_if_include" (current value: "",
                      data source: default, level: 1 user/basic, type:
                      string)
                      Comma-delimited list of devices and/or CIDR
                      notation of networks to use for MPI communication
                      (e.g., "eth0,192.168.0.0/16").  Mutually exclusive
                      with btl_tcp_if_exclude.
             MCA btl: parameter "btl_tcp_if_exclude" (current value:
                      "127.0.0.1/8,sppp", data source: default, level: 1
                      user/basic, type: string)
                      Comma-delimited list of devices and/or CIDR
                      notation of networks to NOT use for MPI
                      communication -- all devices not matching these
                      specifications will be used (e.g.,
                      "eth0,192.168.0.0/16").  If set to a non-default
                      value, it is mutually exclusive with
                      btl_tcp_if_include.
             MCA btl: parameter "btl_tcp_progress_thread" (current value:
                      "0", data source: default, level: 1 user/basic,
                      type: int)


% ompi_info -a
[...]
 MPI_MAX_DATAREP_STRING: 128
       MCA allocator: basic (MCA v2.1.0, API v2.0.0, Component v3.0.0)
       MCA allocator: bucket (MCA v2.1.0, API v2.0.0, Component v3.0.0)
       MCA backtrace: execinfo (MCA v2.1.0, API v2.0.0, Component v3.0.0)
             MCA btl: vader (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: sm (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: self (MCA v2.1.0, API v3.0.0, Component v3.0.0)
             MCA btl: tcp (MCA v2.1.0, API v3.0.0, Component v3.0.0)
        MCA compress: bzip (MCA v2.1.0, API v2.0.0, Component v3.0.0)
[...]
        MCA sharedfp: sm (MCA v2.1.0, API v2.0.0, Component v3.0.0)
        MCA sharedfp: lockedfile (MCA v2.1.0, API v2.0.0, Component v3.0.0)
        MCA sharedfp: individual (MCA v2.1.0, API v2.0.0, Component v3.0.0)
            MCA topo: treematch (MCA v2.1.0, API v2.2.0, Component v3.0.0)
            MCA topo: basic (MCA v2.1.0, API v2.2.0, Component v3.0.0)
       MCA vprotocol: pessimist (MCA v2.1.0, API v2.0.0, Component v3.0.0)
             MCA mca: parameter "mca_param_files" (current value: "/home/kilo/.openmpi/mca-params.conf:/home/kilo/local/etc/openmpi-mca-params.conf", data source: default, level: 2 user/detail, type: string, deprecated, synonym of: mca_base_param_files)
                      Path for MCA configuration files containing variable values
[...]

@jsquyres / @hjelmn - What are your thoughts, gents?

rhc54 · 2016-04-02T15:54:10Z

FWIW: I like having the dashed lines separating the output - made it a lot easier to read, for me.

ghost · 2016-04-02T15:55:26Z

@rhc54 😄 For me too! Once we nail down a good approach to this bug, I'll be happy to try something similar in another pull request.

ghost · 2016-04-02T15:56:12Z

Mellanox tests (I've no idea what they do 😄) failed:
...
18:42:02 + taskset -c 12,13 timeout -s SIGSEGV 10m /var/lib/jenkins/jobs/gh-ompi-master-pr/workspace/ompi_install1/bin/mpirun -np 2 -bind-to core -mca btl_openib_if_include mlx5_0:1 -x MXM_RDMA_PORTS=mlx5_0:1 -x UCX_NET_DEVICES=mlx5_0:1 -x UCX_TLS=rc,cm -mca pml ob1 -mca btl self,openib /var/lib/jenkins/jobs/gh-ompi-master-pr/workspace/ompi_install1/thread_tests/thread-tests-1.1/message_rate_th 8
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:42:02 Note: Max supported value for env var is 4095
18:42:02 vsetenv ghprbPullLongDescription failed
18:43:43 [jenkins01][[36327,1],0][btl_openib_endpoint.c:115:mca_btl_openib_endpoint_post_send] error posting send request error 12: Cannot allocate memory. size = 14
18:43:43
18:43:43 [jenkins01][[36327,1],0][btl_openib_endpoint.c:115:mca_btl_openib_endpoint_post_send] error posting send request error 12: Cannot allocate memory. size = 14
18:43:43
18:43:43 [jenkins01:8970] *** An error occurred in MPI_Send
18:43:43 [jenkins01:8970] *** reported by process [2380726273,0]
18:43:43 [jenkins01:8970] *** on communicator MPI_COMM_WORLD
18:43:43 [jenkins01:8970] *** MPI_ERR_OTHER: known error not in list
18:43:43 [jenkins01:8970] *** MPI_ERRORS_ARE_FATAL (processes in this communicator will now abort,
18:43:43 [jenkins01:8970] *** and potentially your MPI job)
18:43:43 [jenkins01][[36327,1],0][btl_openib_endpoint.c:115:mca_btl_openib_endpoint_post_send] error posting send request error 12: Cannot allocate memory. size = 14
18:43:43
18:43:43 [jenkins01:08965] 1 more process has sent help message help-mpi-errors.txt / mpi_errors_are_fatal
18:43:43 [jenkins01:08965] Set MCA parameter "orte_base_help_aggregate" to 0 to see all help / error messages
18:43:43 Build step 'Execute shell' marked build as failure
18:43:43 [htmlpublisher] Archiving HTML reports...
18:43:43 [htmlpublisher] Archiving at BUILD level /var/lib/jenkins/jobs/gh-ompi-master-pr/workspace/cov_build to /var/lib/jenkins/jobs/gh-ompi-master-pr/builds/2164/htmlreports/Coverity_Report
18:43:45 Setting commit status on GitHub for 2cd1e8c
18:43:46 [BFA] Scanning build for known causes...
18:43:46 [BFA] No failure causes found
18:43:46 [BFA] Done. 0s
18:43:46 Setting status of b09ae94 to FAILURE with url http://bgate.mellanox.com/jenkins/job/gh-ompi-master-pr/2164/ and message: 'Build finished. '
18:43:46 Using context: Mellanox
18:43:46 Finished: FAILURE
...

Doesn't seem related to the ompi_info changes?

rhc54 · 2016-04-02T16:00:33Z

Completely unrelated - ignore them

ghost · 2016-04-02T16:01:01Z

That's what I thought. I did restart the testsuite though...

rhc54 · 2016-04-02T16:03:45Z

👍 Looks good to me...

hjelmn · 2016-04-02T16:21:09Z

Curious how this affects --parsable? Does this cause any duplicate lines there? Not sure we care but it would be good to know.

ghost · 2016-04-02T16:34:45Z

@hjelmn Good call.

% ompi_info --parsable --param btl all
mca:btl:vader:version:"mca:2.1.0"
mca:btl:vader:version:"api:3.0.0"
mca:btl:vader:version:"component:3.0.0"
mca:btl:sm:version:"mca:2.1.0"
mca:btl:sm:version:"api:3.0.0"
mca:btl:sm:version:"component:3.0.0"
mca:btl:self:version:"mca:2.1.0"
mca:btl:self:version:"api:3.0.0"
mca:btl:self:version:"component:3.0.0"
mca:btl:tcp:version:"mca:2.1.0"
mca:btl:tcp:version:"api:3.0.0"
mca:btl:tcp:version:"component:3.0.0"
mca:btl:tcp:param:btl_tcp_if_include:value:
mca:btl:tcp:param:btl_tcp_if_include:source:default
mca:btl:tcp:param:btl_tcp_if_include:status:writeable
mca:btl:tcp:param:btl_tcp_if_include:level:1
mca:btl:tcp:param:btl_tcp_if_include:help:Comma-delimited list of devices and/or CIDR notation of 
networks to use for MPI communication (e.g., "eth0,192.168.0.0/16").  Mutually exclusive with   
btl_tcp_if_exclude.
mca:btl:tcp:param:btl_tcp_if_include:deprecated:no
mca:btl:tcp:param:btl_tcp_if_include:type:string
mca:btl:tcp:param:btl_tcp_if_include:disabled:false
mca:btl:tcp:param:btl_tcp_if_exclude:value:127.0.0.1/8,sppp
mca:btl:tcp:param:btl_tcp_if_exclude:source:default
mca:btl:tcp:param:btl_tcp_if_exclude:status:writeable
mca:btl:tcp:param:btl_tcp_if_exclude:level:1
mca:btl:tcp:param:btl_tcp_if_exclude:help:Comma-delimited list of devices and/or CIDR notation 
of networks to NOT use for MPI communication -- all devices not matching these specifications 
will be used (e.g., "eth0,192.168.0.0/16").  If set to a non-default value, it is mutually exclusive 
with btl_tcp_if_include.
mca:btl:tcp:param:btl_tcp_if_exclude:deprecated:no
mca:btl:tcp:param:btl_tcp_if_exclude:type:string
mca:btl:tcp:param:btl_tcp_if_exclude:disabled:false
mca:btl:tcp:param:btl_tcp_progress_thread:value:0
mca:btl:tcp:param:btl_tcp_progress_thread:source:default
mca:btl:tcp:param:btl_tcp_progress_thread:status:writeable
mca:btl:tcp:param:btl_tcp_progress_thread:level:1
mca:btl:tcp:param:btl_tcp_progress_thread:deprecated:no
mca:btl:tcp:param:btl_tcp_progress_thread:type:int
mca:btl:tcp:param:btl_tcp_progress_thread:disabled:false

% ompi_info --parsable --param btl tcp
mca:btl:tcp:version:"mca:2.1.0"
mca:btl:tcp:version:"api:3.0.0"
mca:btl:tcp:version:"component:3.0.0"
mca:btl:tcp:param:btl_tcp_if_include:value:
mca:btl:tcp:param:btl_tcp_if_include:source:default
mca:btl:tcp:param:btl_tcp_if_include:status:writeable
mca:btl:tcp:param:btl_tcp_if_include:level:1
[...]

% ompi_info -a --parsable
[...]
mca:backtrace:execinfo:version:"component:3.0.0"
mca:btl:vader:version:"mca:2.1.0"
mca:btl:vader:version:"api:3.0.0"
mca:btl:vader:version:"component:3.0.0"
mca:btl:sm:version:"mca:2.1.0"
mca:btl:sm:version:"api:3.0.0"
mca:btl:sm:version:"component:3.0.0"
mca:btl:self:version:"mca:2.1.0"
mca:btl:self:version:"api:3.0.0"
mca:btl:self:version:"component:3.0.0"
mca:btl:tcp:version:"mca:2.1.0"
mca:btl:tcp:version:"api:3.0.0"
mca:btl:tcp:version:"component:3.0.0"
mca:compress:bzip:version:"mca:2.1.0"
[...]

There don't appear to be any duplicate lines.

Pass component_map to opal_info_do_params(). It will be needed to output component versions. Signed-off-by: Karol Mroz <mroz.karol@gmail.com>

ghost · 2016-04-02T19:19:25Z

Tabs... fixing now... done.

When invoking, for example, `ompi_info` with: -a --params foo all --params foo bar it's useful to have the appropriate components and their versions be displayed, regardless of whether they have registered any parameters. Signed-off-by: Karol Mroz <mroz.karol@gmail.com>

ghost mentioned this pull request Apr 2, 2016

opal_info_support: fix api comments #1516

Merged

opal_info_support: pass component map when handling params

a468c3b

Pass component_map to opal_info_do_params(). It will be needed to output component versions. Signed-off-by: Karol Mroz <mroz.karol@gmail.com>

ghost changed the title ~~opal_info_support: output component versions~~ DNM opal_info_support: output component versions Apr 2, 2016

ghost changed the title ~~DNM opal_info_support: output component versions~~ opal_info_support: output component versions Apr 2, 2016

ghost mentioned this pull request Apr 3, 2016

opal_info_support: add component to param pretty output #1518

Merged

rhc54 merged commit d724d8a into open-mpi:master Apr 3, 2016

ghost mentioned this pull request Apr 3, 2016

ompi_info -a should show all built components #1396

Closed

hppritcha mentioned this pull request Apr 12, 2016

Topic/opal pretty output open-mpi/ompi-release#1068

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

opal_info_support: output component versions #1515

opal_info_support: output component versions #1515

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

hjelmn commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

opal_info_support: output component versions #1515

opal_info_support: output component versions #1515

Uh oh!

Conversation

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

rhc54 commented Apr 2, 2016

Uh oh!

hjelmn commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

ghost commented Apr 2, 2016

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants