[PATCH] opensm: Allow connectivity between switches in up down and fat tree

[PATCH] opensm: Allow connectivity between switches in up down and fat tree

[Eli Dorfman (Voltaire)]

Modify option connect_roots to connect_switches and allow connectivity
between all switches (not only root nodes) in up down and fat tree routing algorithms

Signed-off-by: Eli Dorfman <elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
---
 opensm/include/opensm/osm_subnet.h |    6 +++---
 opensm/man/opensm.8.in             |    6 +++---
 opensm/opensm/main.c               |    8 ++++----
 opensm/opensm/osm_subnet.c         |   10 +++++-----
 opensm/opensm/osm_ucast_ftree.c    |    2 +-
 opensm/opensm/osm_ucast_updn.c     |    6 ++----
 6 files changed, 18 insertions(+), 20 deletions(-)

diff --git a/opensm/include/opensm/osm_subnet.h b/opensm/include/opensm/osm_subnet.h
index 95a635c..9ef4c9e 100644
--- a/opensm/include/opensm/osm_subnet.h
+++ b/opensm/include/opensm/osm_subnet.h
@@ -193,7 +193,7 @@ typedef struct osm_subn_opt {
 	boolean_t sweep_on_trap;
 	char *routing_engine_names;
 	boolean_t use_ucast_cache;
-	boolean_t connect_roots;
+	boolean_t connect_switches;
 	char *lid_matrix_dump_file;
 	char *lfts_file;
 	char *root_guid_file;
@@ -388,8 +388,8 @@ typedef struct osm_subn_opt {
 *	routing_engine_names
 *		Name of routing engine(s) to use.
 *
-*	connect_roots
-*		The option which will enforce root to root connectivity with
+*	connect_switches
+*		The option which will enforce all switch connectivity with
 *		up/down and fat-tree routing engines (even if this violates
 *		"pure" deadlock free up/down or fat-tree algorithm)
 *
diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
index 9053611..e530320 100644
--- a/opensm/man/opensm.8.in
+++ b/opensm/man/opensm.8.in
@@ -18,7 +18,7 @@ opensm \- InfiniBand subnet manager and administration (SM/SA)
 [\-\-do_mesh_analysis]
 [\-\-lash_start_vl <vl number>]
 [\-A | \-\-ucast_cache]
-[\-z | \-\-connect_roots]
+[\-z | \-\-connect_switches]
 [\-M <file name> | \-\-lid_matrix_file <file name>]
 [\-U <file name> | \-\-lfts_file <file name>]
 [\-S | \-\-sadb_file <file name>]
@@ -172,9 +172,9 @@ is host reboot, which otherwise would cause two full routing
 recalculations: one when the host goes down, and the other when
 the host comes back online.
 .TP
-\fB\-z\fR, \fB\-\-connect_roots\fR
+\fB\-z\fR, \fB\-\-connect_switches\fR
 This option enforces routing engines (up/down and
-fat-tree) to make connectivity between root switches and in
+fat-tree) to make connectivity between all switches and in
 this way to be fully IBA complaint. In many cases this can
 violate "pure" deadlock free algorithm, so use it carefully.
 .TP
diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
index 0093aa7..95f4432 100644
--- a/opensm/opensm/main.c
+++ b/opensm/opensm/main.c
@@ -185,7 +185,7 @@ static void show_usage(void)
 	       "          Defaults to 0.\n");
 	printf("--sm_sl <sl number>\n"
 	       "          Sets the SL to use to communicate with the SM/SA. Defaults to 0.\n\n");
-	printf("--connect_roots, -z\n"
+	printf("--connect_switches, -z\n"
 	       "          This option enforces routing engines (up/down and \n"
 	       "          fat-tree) to make connectivity between root switches\n"
 	       "          and in this way be IBA compliant. In many cases,\n"
@@ -587,7 +587,7 @@ int main(int argc, char *argv[])
 		{"smkey", 1, NULL, 'k'},
 		{"routing_engine", 1, NULL, 'R'},
 		{"ucast_cache", 0, NULL, 'A'},
-		{"connect_roots", 0, NULL, 'z'},
+		{"connect_switches", 0, NULL, 'z'},
 		{"lid_matrix_file", 1, NULL, 'M'},
 		{"lfts_file", 1, NULL, 'U'},
 		{"sadb_file", 1, NULL, 'S'},
@@ -887,8 +887,8 @@ int main(int argc, char *argv[])
 			break;
 
 		case 'z':
-			opt.connect_roots = TRUE;
-			printf(" Connect roots option is on\n");
+			opt.connect_switches = TRUE;
+			printf(" Connect switches option is on\n");
 			break;
 
 		case 'A':
diff --git a/opensm/opensm/osm_subnet.c b/opensm/opensm/osm_subnet.c
index d5c5ab2..054df50 100644
--- a/opensm/opensm/osm_subnet.c
+++ b/opensm/opensm/osm_subnet.c
@@ -330,7 +330,7 @@ static const opt_rec_t opt_tbl[] = {
 	{ "port_profile_switch_nodes", OPT_OFFSET(port_profile_switch_nodes), opts_parse_boolean, NULL, 1 },
 	{ "sweep_on_trap", OPT_OFFSET(sweep_on_trap), opts_parse_boolean, NULL, 1 },
 	{ "routing_engine", OPT_OFFSET(routing_engine_names), opts_parse_charp, NULL, 0 },
-	{ "connect_roots", OPT_OFFSET(connect_roots), opts_parse_boolean, NULL, 1 },
+	{ "connect_switches", OPT_OFFSET(connect_switches), opts_parse_boolean, NULL, 1 },
 	{ "use_ucast_cache", OPT_OFFSET(use_ucast_cache), opts_parse_boolean, NULL, 0 },
 	{ "log_file", OPT_OFFSET(log_file), opts_parse_charp, NULL, 0 },
 	{ "log_max_size", OPT_OFFSET(log_max_size), opts_parse_uint32, opts_setup_log_max_size, 1 },
@@ -741,7 +741,7 @@ void osm_subn_set_default_opt(IN osm_subn_opt_t * p_opt)
 	p_opt->sweep_on_trap = TRUE;
 	p_opt->use_ucast_cache = FALSE;
 	p_opt->routing_engine_names = NULL;
-	p_opt->connect_roots = FALSE;
+	p_opt->connect_switches = FALSE;
 	p_opt->lid_matrix_dump_file = NULL;
 	p_opt->lfts_file = NULL;
 	p_opt->root_guid_file = NULL;
@@ -1401,9 +1401,9 @@ int osm_subn_output_conf(FILE *out, IN osm_subn_opt_t * p_opts)
 		p_opts->routing_engine_names : null_str);
 
 	fprintf(out,
-		"# Connect roots (use FALSE if unsure)\n"
-		"connect_roots %s\n\n",
-		p_opts->connect_roots ? "TRUE" : "FALSE");
+		"# Connect switches (use FALSE if unsure)\n"
+		"connect_switches %s\n\n",
+		p_opts->connect_switches ? "TRUE" : "FALSE");
 
 	fprintf(out,
 		"# Use unicast routing cache (use FALSE if unsure)\n"
diff --git a/opensm/opensm/osm_ucast_ftree.c b/opensm/opensm/osm_ucast_ftree.c
index 88ea344..8842073 100644
--- a/opensm/opensm/osm_ucast_ftree.c
+++ b/opensm/opensm/osm_ucast_ftree.c
@@ -4061,7 +4061,7 @@ static int do_routing(IN void *context)
 		"Filling switch forwarding tables for switch-to-switch paths\n");
 	fabric_route_to_switches(p_ftree);
 
-	if (p_ftree->p_osm->subn.opt.connect_roots) {
+	if (p_ftree->p_osm->subn.opt.connect_switches) {
 		OSM_LOG(&p_ftree->p_osm->log, OSM_LOG_VERBOSE,
 			"Connecting switches that are unreachable within "
 			"Up/Down rules\n");
diff --git a/opensm/opensm/osm_ucast_updn.c b/opensm/opensm/osm_ucast_updn.c
index 164c6f4..7619711 100644
--- a/opensm/opensm/osm_ucast_updn.c
+++ b/opensm/opensm/osm_ucast_updn.c
@@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t * p_updn)
 	     item = cl_qmap_next(item)) {
 		p_sw = (osm_switch_t *)item;
 		/* Clear Min Hop Table */
-		if (p_subn->opt.connect_roots)
-			updn_clear_non_root_hops(p_updn, p_sw);
-		else
+		if (!p_subn->opt.connect_switches)
 			osm_switch_clear_hops(p_sw);
 	}
 
@@ -596,7 +594,7 @@ static int updn_lid_matrices(void *ctx)
 			OSM_LOG(&p_updn->p_osm->log, OSM_LOG_ERROR, "ERR : "
 				"cannot parse root guids file \'%s\'\n",
 				p_updn->p_osm->subn.opt.root_guid_file);
-		if (p_updn->p_osm->subn.opt.connect_roots &&
+		if (p_updn->p_osm->subn.opt.connect_switches &&
 		    p_updn->num_roots > 1)
 			osm_ucast_mgr_build_lid_matrices(&p_updn->p_osm->sm.ucast_mgr);
 	} else {
-- 
1.5.5

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH v2] opensm: Modify connect_roots to allow connectivity between all switches

[Eli Dorfman (Voltaire)]

After a second thought and in order not to break current configuration,
I send this modified patch that does not change connect_roots option but
changes its functionality in up-down (I think that in fat-tree it is already implemented)

Modify connect_roots option to allow connectivity between
all switches in up-down routing algorithm and in this way be
fully IBA compliant

Signed-off-by: Eli Dorfman <elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
---
 opensm/man/opensm.8.in         |    2 +-
 opensm/opensm/main.c           |    2 +-
 opensm/opensm/osm_ucast_updn.c |    4 +---
 3 files changed, 3 insertions(+), 5 deletions(-)

diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
index 9053611..c67126e 100644
--- a/opensm/man/opensm.8.in
+++ b/opensm/man/opensm.8.in
@@ -174,7 +174,7 @@ the host comes back online.
 .TP
 \fB\-z\fR, \fB\-\-connect_roots\fR
 This option enforces routing engines (up/down and
-fat-tree) to make connectivity between root switches and in
+fat-tree) to make connectivity between all switches and in
 this way to be fully IBA complaint. In many cases this can
 violate "pure" deadlock free algorithm, so use it carefully.
 .TP
diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
index 0093aa7..82ca78f 100644
--- a/opensm/opensm/main.c
+++ b/opensm/opensm/main.c
@@ -187,7 +187,7 @@ static void show_usage(void)
 	       "          Sets the SL to use to communicate with the SM/SA. Defaults to 0.\n\n");
 	printf("--connect_roots, -z\n"
 	       "          This option enforces routing engines (up/down and \n"
-	       "          fat-tree) to make connectivity between root switches\n"
+	       "          fat-tree) to make connectivity between all switches\n"
 	       "          and in this way be IBA compliant. In many cases,\n"
 	       "          this can violate \"pure\" deadlock free algorithm, so\n"
 	       "          use it carefully.\n\n");
diff --git a/opensm/opensm/osm_ucast_updn.c b/opensm/opensm/osm_ucast_updn.c
index 164c6f4..f44ca24 100644
--- a/opensm/opensm/osm_ucast_updn.c
+++ b/opensm/opensm/osm_ucast_updn.c
@@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t * p_updn)
 	     item = cl_qmap_next(item)) {
 		p_sw = (osm_switch_t *)item;
 		/* Clear Min Hop Table */
-		if (p_subn->opt.connect_roots)
-			updn_clear_non_root_hops(p_updn, p_sw);
-		else
+		if (!p_subn->opt.connect_roots)
 			osm_switch_clear_hops(p_sw);
 	}
 
-- 
1.5.5

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] opensm: Modify connect_roots to allow connectivity between all switches

[Yevgeny Kliteynik]

Eli,

On 13-Jun-10 6:10 PM, Eli Dorfman (Voltaire) wrote:
> After a second thought and in order not to break current configuration,
> I send this modified patch that does not change connect_roots option but
> changes its functionality in up-down (I think that in fat-tree it is already implemented)
>
> Modify connect_roots option to allow connectivity between
> all switches in up-down routing algorithm and in this way be
> fully IBA compliant
>
> Signed-off-by: Eli Dorfman<elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
> ---
>   opensm/man/opensm.8.in         |    2 +-
>   opensm/opensm/main.c           |    2 +-
>   opensm/opensm/osm_ucast_updn.c |    4 +---
>   3 files changed, 3 insertions(+), 5 deletions(-)
>
> diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
> index 9053611..c67126e 100644
> --- a/opensm/man/opensm.8.in
> +++ b/opensm/man/opensm.8.in
> @@ -174,7 +174,7 @@ the host comes back online.
>   .TP
>   \fB\-z\fR, \fB\-\-connect_roots\fR
>   This option enforces routing engines (up/down and
> -fat-tree) to make connectivity between root switches and in
> +fat-tree) to make connectivity between all switches and in
>   this way to be fully IBA complaint. In many cases this can
>   violate "pure" deadlock free algorithm, so use it carefully.
>   .TP
> diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
> index 0093aa7..82ca78f 100644
> --- a/opensm/opensm/main.c
> +++ b/opensm/opensm/main.c
> @@ -187,7 +187,7 @@ static void show_usage(void)
>   	       "          Sets the SL to use to communicate with the SM/SA. Defaults to 0.\n\n");
>   	printf("--connect_roots, -z\n"
>   	       "          This option enforces routing engines (up/down and \n"
> -	       "          fat-tree) to make connectivity between root switches\n"
> +	       "          fat-tree) to make connectivity between all switches\n"
>   	       "          and in this way be IBA compliant. In many cases,\n"
>   	       "          this can violate \"pure\" deadlock free algorithm, so\n"
>   	       "          use it carefully.\n\n");
> diff --git a/opensm/opensm/osm_ucast_updn.c b/opensm/opensm/osm_ucast_updn.c
> index 164c6f4..f44ca24 100644
> --- a/opensm/opensm/osm_ucast_updn.c
> +++ b/opensm/opensm/osm_ucast_updn.c
> @@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t * p_updn)
>   	     item = cl_qmap_next(item)) {
>   		p_sw = (osm_switch_t *)item;
>   		/* Clear Min Hop Table */
> -		if (p_subn->opt.connect_roots)
> -			updn_clear_non_root_hops(p_updn, p_sw);
> -		else
> +		if (!p_subn->opt.connect_roots)
>   			osm_switch_clear_hops(p_sw);
>   	}
>

What kind of testing was done for this?
I have a strong feeling that it will break up/down.
If the connect_roots option is on, you will not clear
the lid matrix at all, and it will contain also the
down/up routes.

-- Yevgeny
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] opensm: Modify connect_roots to allow connectivity between all switches

[Yevgeny Kliteynik]

On 15-Jun-10 10:09 AM, Eli Dorfman (Voltaire) wrote:
> Yevgeny Kliteynik wrote:
>> Eli,
>>
>> On 13-Jun-10 6:10 PM, Eli Dorfman (Voltaire) wrote:
>>> After a second thought and in order not to break current configuration,
>>> I send this modified patch that does not change connect_roots option but
>>> changes its functionality in up-down (I think that in fat-tree it is
>>> already implemented)
>>>
>>> Modify connect_roots option to allow connectivity between
>>> all switches in up-down routing algorithm and in this way be
>>> fully IBA compliant
>>>
>>> Signed-off-by: Eli Dorfman<elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
>>> ---
>>>    opensm/man/opensm.8.in         |    2 +-
>>>    opensm/opensm/main.c           |    2 +-
>>>    opensm/opensm/osm_ucast_updn.c |    4 +---
>>>    3 files changed, 3 insertions(+), 5 deletions(-)
>>>
>>> diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
>>> index 9053611..c67126e 100644
>>> --- a/opensm/man/opensm.8.in
>>> +++ b/opensm/man/opensm.8.in
>>> @@ -174,7 +174,7 @@ the host comes back online.
>>>    .TP
>>>    \fB\-z\fR, \fB\-\-connect_roots\fR
>>>    This option enforces routing engines (up/down and
>>> -fat-tree) to make connectivity between root switches and in
>>> +fat-tree) to make connectivity between all switches and in
>>>    this way to be fully IBA complaint. In many cases this can
>>>    violate "pure" deadlock free algorithm, so use it carefully.
>>>    .TP
>>> diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
>>> index 0093aa7..82ca78f 100644
>>> --- a/opensm/opensm/main.c
>>> +++ b/opensm/opensm/main.c
>>> @@ -187,7 +187,7 @@ static void show_usage(void)
>>>               "          Sets the SL to use to communicate with the
>>> SM/SA. Defaults to 0.\n\n");
>>>        printf("--connect_roots, -z\n"
>>>               "          This option enforces routing engines (up/down
>>> and \n"
>>> -           "          fat-tree) to make connectivity between root
>>> switches\n"
>>> +           "          fat-tree) to make connectivity between all
>>> switches\n"
>>>               "          and in this way be IBA compliant. In many
>>> cases,\n"
>>>               "          this can violate \"pure\" deadlock free
>>> algorithm, so\n"
>>>               "          use it carefully.\n\n");
>>> diff --git a/opensm/opensm/osm_ucast_updn.c
>>> b/opensm/opensm/osm_ucast_updn.c
>>> index 164c6f4..f44ca24 100644
>>> --- a/opensm/opensm/osm_ucast_updn.c
>>> +++ b/opensm/opensm/osm_ucast_updn.c
>>> @@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t * p_updn)
>>>             item = cl_qmap_next(item)) {
>>>            p_sw = (osm_switch_t *)item;
>>>            /* Clear Min Hop Table */
>>> -        if (p_subn->opt.connect_roots)
>>> -            updn_clear_non_root_hops(p_updn, p_sw);
>>> -        else
>>> +        if (!p_subn->opt.connect_roots)
>>>                osm_switch_clear_hops(p_sw);
>>>        }
>>>
>>
>> What kind of testing was done for this?
>> I have a strong feeling that it will break up/down.
>> If the connect_roots option is on, you will not clear
>> the lid matrix at all, and it will contain also the
>> down/up routes.
>
> that is the idea.
> instead of leaving only up/dpwn routes we want to keep routes between all switches.
> the routes between host nodes will still be up/down.

Lid matrix might contain down/up routes between many
kinds of switches, not only between roots. This means
that you might have down/up paths between leaf switches,
even though you could have found up/down path for them.
This also means that you might end up with down/up routes
between HCAs that are connected to these leafs.

-- Yevgeny

> Eli
>
>>
>> -- Yevgeny
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] opensm: Modify connect_roots to allow connectivity between all switches

[Eli Dorfman (Voltaire)]

Yevgeny Kliteynik wrote:
> Eli,
> 
> On 13-Jun-10 6:10 PM, Eli Dorfman (Voltaire) wrote:
>> After a second thought and in order not to break current configuration,
>> I send this modified patch that does not change connect_roots option but
>> changes its functionality in up-down (I think that in fat-tree it is
>> already implemented)
>>
>> Modify connect_roots option to allow connectivity between
>> all switches in up-down routing algorithm and in this way be
>> fully IBA compliant
>>
>> Signed-off-by: Eli Dorfman<elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
>> ---
>>   opensm/man/opensm.8.in         |    2 +-
>>   opensm/opensm/main.c           |    2 +-
>>   opensm/opensm/osm_ucast_updn.c |    4 +---
>>   3 files changed, 3 insertions(+), 5 deletions(-)
>>
>> diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
>> index 9053611..c67126e 100644
>> --- a/opensm/man/opensm.8.in
>> +++ b/opensm/man/opensm.8.in
>> @@ -174,7 +174,7 @@ the host comes back online.
>>   .TP
>>   \fB\-z\fR, \fB\-\-connect_roots\fR
>>   This option enforces routing engines (up/down and
>> -fat-tree) to make connectivity between root switches and in
>> +fat-tree) to make connectivity between all switches and in
>>   this way to be fully IBA complaint. In many cases this can
>>   violate "pure" deadlock free algorithm, so use it carefully.
>>   .TP
>> diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
>> index 0093aa7..82ca78f 100644
>> --- a/opensm/opensm/main.c
>> +++ b/opensm/opensm/main.c
>> @@ -187,7 +187,7 @@ static void show_usage(void)
>>              "          Sets the SL to use to communicate with the
>> SM/SA. Defaults to 0.\n\n");
>>       printf("--connect_roots, -z\n"
>>              "          This option enforces routing engines (up/down
>> and \n"
>> -           "          fat-tree) to make connectivity between root
>> switches\n"
>> +           "          fat-tree) to make connectivity between all
>> switches\n"
>>              "          and in this way be IBA compliant. In many
>> cases,\n"
>>              "          this can violate \"pure\" deadlock free
>> algorithm, so\n"
>>              "          use it carefully.\n\n");
>> diff --git a/opensm/opensm/osm_ucast_updn.c
>> b/opensm/opensm/osm_ucast_updn.c
>> index 164c6f4..f44ca24 100644
>> --- a/opensm/opensm/osm_ucast_updn.c
>> +++ b/opensm/opensm/osm_ucast_updn.c
>> @@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t * p_updn)
>>            item = cl_qmap_next(item)) {
>>           p_sw = (osm_switch_t *)item;
>>           /* Clear Min Hop Table */
>> -        if (p_subn->opt.connect_roots)
>> -            updn_clear_non_root_hops(p_updn, p_sw);
>> -        else
>> +        if (!p_subn->opt.connect_roots)
>>               osm_switch_clear_hops(p_sw);
>>       }
>>
> 
> What kind of testing was done for this?
> I have a strong feeling that it will break up/down.
> If the connect_roots option is on, you will not clear
> the lid matrix at all, and it will contain also the
> down/up routes.

that is the idea.
instead of leaving only up/dpwn routes we want to keep routes between all switches.
the routes between host nodes will still be up/down.

Eli

> 
> -- Yevgeny

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH v2] opensm: Modify connect_roots to allow connectivity between all switches

[Eli Dorfman (Voltaire)]

Yevgeny Kliteynik wrote:
> On 15-Jun-10 10:09 AM, Eli Dorfman (Voltaire) wrote:
>> Yevgeny Kliteynik wrote:
>>> Eli,
>>>
>>> On 13-Jun-10 6:10 PM, Eli Dorfman (Voltaire) wrote:
>>>> After a second thought and in order not to break current configuration,
>>>> I send this modified patch that does not change connect_roots option
>>>> but
>>>> changes its functionality in up-down (I think that in fat-tree it is
>>>> already implemented)
>>>>
>>>> Modify connect_roots option to allow connectivity between
>>>> all switches in up-down routing algorithm and in this way be
>>>> fully IBA compliant
>>>>
>>>> Signed-off-by: Eli Dorfman<elid-smomgflXvOZWk0Htik3J/w@public.gmane.org>
>>>> ---
>>>>    opensm/man/opensm.8.in         |    2 +-
>>>>    opensm/opensm/main.c           |    2 +-
>>>>    opensm/opensm/osm_ucast_updn.c |    4 +---
>>>>    3 files changed, 3 insertions(+), 5 deletions(-)
>>>>
>>>> diff --git a/opensm/man/opensm.8.in b/opensm/man/opensm.8.in
>>>> index 9053611..c67126e 100644
>>>> --- a/opensm/man/opensm.8.in
>>>> +++ b/opensm/man/opensm.8.in
>>>> @@ -174,7 +174,7 @@ the host comes back online.
>>>>    .TP
>>>>    \fB\-z\fR, \fB\-\-connect_roots\fR
>>>>    This option enforces routing engines (up/down and
>>>> -fat-tree) to make connectivity between root switches and in
>>>> +fat-tree) to make connectivity between all switches and in
>>>>    this way to be fully IBA complaint. In many cases this can
>>>>    violate "pure" deadlock free algorithm, so use it carefully.
>>>>    .TP
>>>> diff --git a/opensm/opensm/main.c b/opensm/opensm/main.c
>>>> index 0093aa7..82ca78f 100644
>>>> --- a/opensm/opensm/main.c
>>>> +++ b/opensm/opensm/main.c
>>>> @@ -187,7 +187,7 @@ static void show_usage(void)
>>>>               "          Sets the SL to use to communicate with the
>>>> SM/SA. Defaults to 0.\n\n");
>>>>        printf("--connect_roots, -z\n"
>>>>               "          This option enforces routing engines (up/down
>>>> and \n"
>>>> -           "          fat-tree) to make connectivity between root
>>>> switches\n"
>>>> +           "          fat-tree) to make connectivity between all
>>>> switches\n"
>>>>               "          and in this way be IBA compliant. In many
>>>> cases,\n"
>>>>               "          this can violate \"pure\" deadlock free
>>>> algorithm, so\n"
>>>>               "          use it carefully.\n\n");
>>>> diff --git a/opensm/opensm/osm_ucast_updn.c
>>>> b/opensm/opensm/osm_ucast_updn.c
>>>> index 164c6f4..f44ca24 100644
>>>> --- a/opensm/opensm/osm_ucast_updn.c
>>>> +++ b/opensm/opensm/osm_ucast_updn.c
>>>> @@ -314,9 +314,7 @@ static int updn_set_min_hop_table(IN updn_t *
>>>> p_updn)
>>>>             item = cl_qmap_next(item)) {
>>>>            p_sw = (osm_switch_t *)item;
>>>>            /* Clear Min Hop Table */
>>>> -        if (p_subn->opt.connect_roots)
>>>> -            updn_clear_non_root_hops(p_updn, p_sw);
>>>> -        else
>>>> +        if (!p_subn->opt.connect_roots)
>>>>                osm_switch_clear_hops(p_sw);
>>>>        }
>>>>
>>>
>>> What kind of testing was done for this?
>>> I have a strong feeling that it will break up/down.
>>> If the connect_roots option is on, you will not clear
>>> the lid matrix at all, and it will contain also the
>>> down/up routes.
>>
>> that is the idea.
>> instead of leaving only up/dpwn routes we want to keep routes between
>> all switches.
>> the routes between host nodes will still be up/down.
> 
> Lid matrix might contain down/up routes between many
> kinds of switches, not only between roots. This means
> that you might have down/up paths between leaf switches,
> even though you could have found up/down path for them.
> This also means that you might end up with down/up routes
> between HCAs that are connected to these leafs.

ok I'll check this again.

Thanks,
Eli

> 
> -- Yevgeny
> 
>> Eli
>>
>>>
>>> -- Yevgeny
>>
>> -- 
>> To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
>> the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html