Skip to content

sharedfp/sm and lockedfile: fix naming bug #5609

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 11, 2018
Merged

sharedfp/sm and lockedfile: fix naming bug #5609

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
20 changes: 18 additions & 2 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 @@ -50,6 +51,9 @@ int mca_sharedfp_lockedfile_file_open (struct ompi_communicator_t *comm,
mca_io_ompio_file_t * shfileHandle, *ompio_fh;
mca_io_ompio_data_t *data;

pid_t my_pid;
int int_pid;

/*------------------------------------------------------------*/
/*Open the same file again without shared file pointer support*/
/*------------------------------------------------------------*/
Expand Down Expand Up @@ -110,15 +114,27 @@ int mca_sharedfp_lockedfile_file_open (struct ompi_communicator_t *comm,
comm->c_coll->coll_bcast ( &masterjobid, 1, MPI_UNSIGNED, 0, comm,
comm->c_coll->coll_bcast_module );

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 (shfileHandle);
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;

/*-------------------------------------------------*/
Expand Down
24 changes: 20 additions & 4 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 @@ -65,7 +65,9 @@ int mca_sharedfp_sm_file_open (struct ompi_communicator_t *comm,
int sm_fd;
int rank;
uint32_t comm_cid;

int int_pid;
pid_t my_pid;

/*----------------------------------------------------*/
/*Open the same file again without shared file pointer*/
/*----------------------------------------------------*/
Expand Down Expand Up @@ -134,7 +136,7 @@ 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(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 @@ -146,7 +148,21 @@ 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