Skip to content

opal/pmix: add blocking fence to SLURM components #1858

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jul 11, 2016
Merged

opal/pmix: add blocking fence to SLURM components #1858

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
33 changes: 32 additions & 1 deletion opal/mca/pmix/s1/pmix_s1.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -38,6 +38,7 @@ static int s1_abort(int flag, const char msg[],
static int s1_commit(void);
static int s1_fencenb(opal_list_t *procs, int collect_data,
opal_pmix_op_cbfunc_t cbfunc, void *cbdata);
static int s1_fence(opal_list_t *procs, int collect_data);
static int s1_put(opal_pmix_scope_t scope,
opal_value_t *kv);
static int s1_get(const opal_process_name_t *id,
Expand All @@ -61,6 +62,7 @@ const opal_pmix_base_module_t opal_pmix_s1_module = {
.abort = s1_abort,
.commit = s1_commit,
.fence_nb = s1_fencenb,
.fence = s1_fence,
.put = s1_put,
.get = s1_get,
.publish = s1_publish,
Expand Down Expand Up @@ -527,7 +529,7 @@ static int s1_commit(void)
static void fencenb(int sd, short args, void *cbdata)
{
pmi_opcaddy_t *op = (pmi_opcaddy_t*)cbdata;
int rc;
int rc = OPAL_SUCCESS;
int32_t i;
opal_value_t *kp, kvn;
opal_hwloc_locality_t locality;
Expand Down Expand Up @@ -617,6 +619,35 @@ static int s1_fencenb(opal_list_t *procs, int collect_data,
return OPAL_SUCCESS;
}

#define S1_WAIT_FOR_COMPLETION(a) \
do { \
while ((a)) { \
usleep(10); \
} \
} while (0)

struct fence_result {
volatile int flag;
int status;
};

static void fence_release(int status, void *cbdata)
{
struct fence_result *res = (struct fence_result*)cbdata;
res->status = status;
opal_atomic_wmb();
res->flag = 0;
}

static int s1_fence(opal_list_t *procs, int collect_data)
{
struct fence_result result = { 1, OPAL_SUCCESS };
s1_fencenb(procs, collect_data, fence_release, (void*)&result);
S1_WAIT_FOR_COMPLETION(result.flag);
return result.status;
}


static int s1_get(const opal_process_name_t *id,
const char *key, opal_list_t *info,
opal_value_t **kv)
Expand Down
33 changes: 32 additions & 1 deletion opal/mca/pmix/s2/pmix_s2.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,6 +45,7 @@ static int s2_abort(int flag, const char msg[],
static int s2_commit(void);
static int s2_fencenb(opal_list_t *procs, int collect_data,
opal_pmix_op_cbfunc_t cbfunc, void *cbdata);
static int s2_fence(opal_list_t *procs, int collect_data);
static int s2_put(opal_pmix_scope_t scope,
opal_value_t *kv);
static int s2_get(const opal_process_name_t *id,
Expand All @@ -68,6 +69,7 @@ const opal_pmix_base_module_t opal_pmix_s2_module = {
.abort = s2_abort,
.commit = s2_commit,
.fence_nb = s2_fencenb,
.fence = s2_fence,
.put = s2_put,
.get = s2_get,
.publish = s2_publish,
Expand Down Expand Up @@ -545,7 +547,7 @@ static int s2_commit(void)
static void fencenb(int sd, short args, void *cbdata)
{
pmi_opcaddy_t *op = (pmi_opcaddy_t*)cbdata;
int rc;
int rc = OPAL_SUCCESS;
int32_t i;
opal_value_t *kp, kvn;
opal_hwloc_locality_t locality;
Expand Down Expand Up @@ -635,6 +637,35 @@ static int s2_fencenb(opal_list_t *procs, int collect_data,
return OPAL_SUCCESS;
}

#define S2_WAIT_FOR_COMPLETION(a) \
do { \
while ((a)) { \
usleep(10); \
} \
} while (0)

struct fence_result {
volatile int flag;
int status;
};

static void fence_release(int status, void *cbdata)
{
struct fence_result *res = (struct fence_result*)cbdata;
res->status = status;
opal_atomic_wmb();
res->flag = 0;
}

static int s2_fence(opal_list_t *procs, int collect_data)
{
struct fence_result result = { 1, OPAL_SUCCESS };
s2_fencenb(procs, collect_data, fence_release, (void*)&result);
S2_WAIT_FOR_COMPLETION(result.flag);
return result.status;
}


static int s2_get(const opal_process_name_t *id,
const char *key, opal_list_t *info,
opal_value_t **kv)
Expand Down