Skip to content

sharedfp/sm and lockedfile: fix naming bug #5607

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
Sep 2, 2018
Merged

sharedfp/sm and lockedfile: fix naming bug #5607

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
23 changes: 19 additions & 4 deletions ompi/mca/sharedfp/lockedfile/sharedfp_lockedfile_file_open.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,7 @@
#include <sys/stat.h>
#endif
#include <fcntl.h>
#include <unistd.h>

int mca_sharedfp_lockedfile_file_open (struct ompi_communicator_t *comm,
const char* filename,
Expand All @@ -47,7 +48,9 @@ int mca_sharedfp_lockedfile_file_open (struct ompi_communicator_t *comm,
int handle;
struct mca_sharedfp_lockedfile_data * module_data = NULL;
struct mca_sharedfp_base_data_t* sh;

pid_t my_pid;
int int_pid;

/*Memory is allocated here for the sh structure*/
sh = (struct mca_sharedfp_base_data_t*)malloc(sizeof(struct mca_sharedfp_base_data_t));
if ( NULL == sh){
Expand Down Expand Up @@ -87,16 +90,28 @@ int mca_sharedfp_lockedfile_file_open (struct ompi_communicator_t *comm,
return err;
}

size_t filenamelen = strlen(filename) + 16;
if ( 0 == fh->f_rank ) {
my_pid = getpid();
int_pid = (int) my_pid;
}
err = comm->c_coll->coll_bcast (&int_pid, 1, MPI_INT, 0, comm, comm->c_coll->coll_bcast_module );
if ( OMPI_SUCCESS != err ) {
opal_output(0, "[%d]mca_sharedfp_lockedfile_file_open: Error in bcast operation\n", fh->f_rank);
free (sh);
free(module_data);
return err;
}

size_t filenamelen = strlen(filename) + 24;
lockedfilename = (char*)malloc(sizeof(char) * filenamelen);
if ( NULL == lockedfilename ) {
free (sh);
free (module_data);
return OMPI_ERR_OUT_OF_RESOURCE;
}
snprintf(lockedfilename, filenamelen, "%s-%u%s",filename,masterjobid,".lock");
snprintf(lockedfilename, filenamelen, "%s-%u-%d%s",filename,masterjobid,int_pid,".lock");
module_data->filename = lockedfilename;

/*-------------------------------------------------*/
/*Open the lockedfile without shared file pointer */
/*-------------------------------------------------*/
Expand Down
22 changes: 19 additions & 3 deletions ompi/mca/sharedfp/sm/sharedfp_sm_file_open.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -44,7 +44,7 @@
#include <semaphore.h>
#include <sys/mman.h>
#include <libgen.h>

#include <unistd.h>

int mca_sharedfp_sm_file_open (struct ompi_communicator_t *comm,
const char* filename,
Expand All @@ -62,6 +62,8 @@ int mca_sharedfp_sm_file_open (struct ompi_communicator_t *comm,
struct mca_sharedfp_sm_offset sm_offset;
int sm_fd;
uint32_t comm_cid;
int int_pid;
pid_t my_pid;

/*Memory is allocated here for the sh structure*/
if ( mca_sharedfp_sm_verbose ) {
Expand Down Expand Up @@ -100,7 +102,7 @@ int mca_sharedfp_sm_file_open (struct ompi_communicator_t *comm,
** For sharedfp we also want to put the file backed shared memory into the tmp directory
*/
filename_basename = basename((char*)filename);
/* format is "%s/%s_cid-%d.sm", see below */
/* format is "%s/%s_cid-%d-%d.sm", see below */
sm_filename_length = strlen(ompi_process_info.job_session_dir) + 1 + strlen(filename_basename) + 5 + (3*sizeof(uint32_t)+1) + 4;
sm_filename = (char*) malloc( sizeof(char) * sm_filename_length);
if (NULL == sm_filename) {
Expand All @@ -111,7 +113,21 @@ int mca_sharedfp_sm_file_open (struct ompi_communicator_t *comm,
}

comm_cid = ompi_comm_get_cid(comm);
sprintf(sm_filename, "%s/%s_cid-%d.sm", ompi_process_info.job_session_dir, filename_basename, comm_cid);
if ( 0 == fh->f_rank ) {
my_pid = getpid();
int_pid = (int) my_pid;
}
err = comm->c_coll->coll_bcast (&int_pid, 1, MPI_INT, 0, comm, comm->c_coll->coll_bcast_module );
if ( OMPI_SUCCESS != err ) {
opal_output(0,"mca_sharedfp_sm_file_open: Error in bcast operation \n");
free(sm_filename);
free(sm_data);
free(sh);
return err;
}

snprintf(sm_filename, sm_filename_length, "%s/%s_cid-%d-%d.sm", ompi_process_info.job_session_dir,
filename_basename, comm_cid, int_pid);
/* open shared memory file, initialize to 0, map into memory */
sm_fd = open(sm_filename, O_RDWR | O_CREAT,
S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH);
Expand Down