summaryrefslogtreecommitdiff
path: root/src/lib/shm_flow_set.c
diff options
context:
space:
mode:
authordimitri staessens <dimitri.staessens@intec.ugent.be>2016-12-03 13:50:53 +0100
committerdimitri staessens <dimitri.staessens@intec.ugent.be>2016-12-03 13:50:53 +0100
commitef2e50b94fbc7ce1c7dbe1f2ea860e6ab0724726 (patch)
tree0e26e0987aeaf50f9fd2f4f8a48fe3bc562b7cb1 /src/lib/shm_flow_set.c
parentf1d67740ee202c6c3af2061df8cafdd265061b59 (diff)
downloadouroboros-ef2e50b94fbc7ce1c7dbe1f2ea860e6ab0724726.tar.gz
ouroboros-ef2e50b94fbc7ce1c7dbe1f2ea860e6ab0724726.zip
lib: Remove logs from stable sources
This removes log output in applications and some size_t printf errors when compiling on 32 bit machines.
Diffstat (limited to 'src/lib/shm_flow_set.c')
-rw-r--r--src/lib/shm_flow_set.c148
1 files changed, 60 insertions, 88 deletions
diff --git a/src/lib/shm_flow_set.c b/src/lib/shm_flow_set.c
index 6cc94573..4a1bff83 100644
--- a/src/lib/shm_flow_set.c
+++ b/src/lib/shm_flow_set.c
@@ -27,14 +27,11 @@
#include <ouroboros/fqueue.h>
#include <ouroboros/errno.h>
-#define OUROBOROS_PREFIX "shm_flow_set"
-
-#include <ouroboros/logs.h>
-
#include <pthread.h>
#include <sys/mman.h>
#include <fcntl.h>
#include <stdlib.h>
+#include <stdio.h>
#include <unistd.h>
#include <signal.h>
#include <sys/stat.h>
@@ -77,16 +74,13 @@ struct shm_flow_set * shm_flow_set_create()
sprintf(fn, SHM_FLOW_SET_PREFIX "%d", getpid());
set = malloc(sizeof(*set));
- if (set == NULL) {
- LOG_DBG("Could not allocate struct.");
+ if (set == NULL)
return NULL;
- }
mask = umask(0);
shm_fd = shm_open(fn, O_CREAT | O_EXCL | O_RDWR, 0666);
if (shm_fd == -1) {
- LOG_DBG("Failed creating flag file.");
free(set);
return NULL;
}
@@ -94,7 +88,6 @@ struct shm_flow_set * shm_flow_set_create()
umask(mask);
if (ftruncate(shm_fd, SHM_FLOW_SET_FILE_SIZE - 1) < 0) {
- LOG_DBG("Failed to extend flag file.");
free(set);
close(shm_fd);
return NULL;
@@ -110,10 +103,7 @@ struct shm_flow_set * shm_flow_set_create()
close(shm_fd);
if (shm_base == MAP_FAILED) {
- LOG_DBG("Failed to map shared memory.");
- if (shm_unlink(fn) == -1)
- LOG_DBG("Failed to remove invalid shm.");
-
+ shm_unlink(fn);
free(set);
return NULL;
}
@@ -160,14 +150,11 @@ struct shm_flow_set * shm_flow_set_open(pid_t api)
sprintf(fn, SHM_FLOW_SET_PREFIX "%d", api);
set = malloc(sizeof(*set));
- if (set == NULL) {
- LOG_DBG("Could not allocate struct.");
+ if (set == NULL)
return NULL;
- }
shm_fd = shm_open(fn, O_RDWR, 0666);
if (shm_fd == -1) {
- LOG_DBG("%d failed opening shared memory %s.", getpid(), fn);
free(set);
return NULL;
}
@@ -182,9 +169,7 @@ struct shm_flow_set * shm_flow_set_open(pid_t api)
close(shm_fd);
if (shm_base == MAP_FAILED) {
- LOG_DBG("Failed to map shared memory.");
- if (shm_unlink(fn) == -1)
- LOG_DBG("Failed to remove invalid shm.");
+ shm_unlink(fn);
free(set);
return NULL;
}
@@ -210,18 +195,12 @@ void shm_flow_set_destroy(struct shm_flow_set * set)
if (set->api != getpid()) {
lf = lockfile_open();
- if (lf == NULL) {
- LOG_ERR("Failed to open lockfile.");
+ if (lf == NULL)
return;
- }
if (lockfile_owner(lf) == getpid()) {
- LOG_DBG("Flow set %d destroyed by IRMd %d.",
- set->api, getpid());
lockfile_close(lf);
} else {
- LOG_ERR("AP-I %d tried to destroy flowset owned by %d.",
- getpid(), set->api);
lockfile_close(lf);
return;
}
@@ -229,11 +208,8 @@ void shm_flow_set_destroy(struct shm_flow_set * set)
sprintf(fn, SHM_FLOW_SET_PREFIX "%d", set->api);
- if (munmap(set->mtable, SHM_FLOW_SET_FILE_SIZE) == -1)
- LOG_DBG("Couldn't unmap shared memory.");
-
- if (shm_unlink(fn) == -1)
- LOG_DBG("Failed to unlink shm.");
+ munmap(set->mtable, SHM_FLOW_SET_FILE_SIZE);
+ shm_unlink(fn);
free(set);
}
@@ -242,111 +218,111 @@ void shm_flow_set_close(struct shm_flow_set * set)
{
assert(set);
- if (munmap(set->mtable, SHM_FLOW_SET_FILE_SIZE) == -1)
- LOG_DBG("Couldn't unmap shared memory.");
+ munmap(set->mtable, SHM_FLOW_SET_FILE_SIZE);
free(set);
}
-void shm_flow_set_zero(struct shm_flow_set * shm_set,
+void shm_flow_set_zero(struct shm_flow_set * set,
size_t idx)
{
ssize_t i = 0;
+ assert(set);
assert(idx < AP_MAX_FQUEUES);
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
for (i = 0; i < IRMD_MAX_FLOWS; ++i)
- if (shm_set->mtable[i] == (ssize_t) idx)
- shm_set->mtable[i] = -1;
+ if (set->mtable[i] == (ssize_t) idx)
+ set->mtable[i] = -1;
- shm_set->heads[idx] = 0;
+ set->heads[idx] = 0;
- pthread_mutex_unlock(shm_set->lock);
+ pthread_mutex_unlock(set->lock);
}
-int shm_flow_set_add(struct shm_flow_set * shm_set,
+int shm_flow_set_add(struct shm_flow_set * set,
size_t idx,
int port_id)
{
- assert(shm_set);
+ assert(set);
assert(!(port_id < 0) && port_id < IRMD_MAX_FLOWS);
assert(idx < AP_MAX_FQUEUES);
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
- if (shm_set->mtable[port_id] != -1) {
- pthread_mutex_unlock(shm_set->lock);
+ if (set->mtable[port_id] != -1) {
+ pthread_mutex_unlock(set->lock);
return -EPERM;
}
- shm_set->mtable[port_id] = idx;
+ set->mtable[port_id] = idx;
- pthread_mutex_unlock(shm_set->lock);
+ pthread_mutex_unlock(set->lock);
return 0;
}
-void shm_flow_set_del(struct shm_flow_set * shm_set,
+void shm_flow_set_del(struct shm_flow_set * set,
size_t idx,
int port_id)
{
- assert(shm_set);
+ assert(set);
assert(!(port_id < 0) && port_id < IRMD_MAX_FLOWS);
assert(idx < AP_MAX_FQUEUES);
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
- if (shm_set->mtable[port_id] == (ssize_t) idx)
- shm_set->mtable[port_id] = -1;
+ if (set->mtable[port_id] == (ssize_t) idx)
+ set->mtable[port_id] = -1;
- pthread_mutex_unlock(shm_set->lock);
+ pthread_mutex_unlock(set->lock);
}
-int shm_flow_set_has(struct shm_flow_set * shm_set,
+int shm_flow_set_has(struct shm_flow_set * set,
size_t idx,
int port_id)
{
int ret = 0;
- assert(shm_set);
+ assert(set);
assert(!(port_id < 0) && port_id < IRMD_MAX_FLOWS);
assert(idx < AP_MAX_FQUEUES);
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
- if (shm_set->mtable[port_id] == (ssize_t) idx)
+ if (set->mtable[port_id] == (ssize_t) idx)
ret = 1;
- pthread_mutex_unlock(shm_set->lock);
+ pthread_mutex_unlock(set->lock);
return ret;
}
-void shm_flow_set_notify(struct shm_flow_set * shm_set, int port_id)
+void shm_flow_set_notify(struct shm_flow_set * set, int port_id)
{
- assert(shm_set);
+ assert(set);
assert(!(port_id < 0) && port_id < IRMD_MAX_FLOWS);
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
- if (shm_set->mtable[port_id] == -1) {
- pthread_mutex_unlock(shm_set->lock);
+ if (set->mtable[port_id] == -1) {
+ pthread_mutex_unlock(set->lock);
return;
}
- *(fqueue_ptr(shm_set, shm_set->mtable[port_id]) +
- (shm_set->heads[shm_set->mtable[port_id]])++) = port_id;
+ *(fqueue_ptr(set, set->mtable[port_id]) +
+ (set->heads[set->mtable[port_id]])++) = port_id;
- pthread_cond_signal(&shm_set->conds[shm_set->mtable[port_id]]);
+ pthread_cond_signal(&set->conds[set->mtable[port_id]]);
- pthread_mutex_unlock(shm_set->lock);
+ pthread_mutex_unlock(set->lock);
}
-ssize_t shm_flow_set_wait(const struct shm_flow_set * shm_set,
+ssize_t shm_flow_set_wait(const struct shm_flow_set * set,
size_t idx,
int * fqueue,
const struct timespec * timeout)
@@ -354,17 +330,15 @@ ssize_t shm_flow_set_wait(const struct shm_flow_set * shm_set,
ssize_t ret = 0;
struct timespec abstime;
- assert(shm_set);
+ assert(set);
assert(idx < AP_MAX_FQUEUES);
assert(fqueue);
#ifdef __APPLE__
- pthread_mutex_lock(shm_set->lock);
+ pthread_mutex_lock(set->lock);
#else
- if (pthread_mutex_lock(shm_set->lock) == EOWNERDEAD) {
- LOG_DBG("Recovering dead mutex.");
- pthread_mutex_consistent(shm_set->lock);
- }
+ if (pthread_mutex_lock(set->lock) == EOWNERDEAD)
+ pthread_mutex_consistent(set->lock);
#endif
if (timeout != NULL) {
clock_gettime(PTHREAD_COND_CLOCK, &abstime);
@@ -372,21 +346,19 @@ ssize_t shm_flow_set_wait(const struct shm_flow_set * shm_set,
}
pthread_cleanup_push((void(*)(void *))pthread_mutex_unlock,
- (void *) shm_set->lock);
+ (void *) set->lock);
- while (shm_set->heads[idx] == 0 && ret != -ETIMEDOUT) {
+ while (set->heads[idx] == 0 && ret != -ETIMEDOUT) {
if (timeout != NULL)
- ret = -pthread_cond_timedwait(shm_set->conds + idx,
- shm_set->lock,
+ ret = -pthread_cond_timedwait(set->conds + idx,
+ set->lock,
&abstime);
else
- ret = -pthread_cond_wait(shm_set->conds + idx,
- shm_set->lock);
+ ret = -pthread_cond_wait(set->conds + idx,
+ set->lock);
#ifndef __APPLE__
- if (ret == -EOWNERDEAD) {
- LOG_DBG("Recovering dead mutex.");
- pthread_mutex_consistent(shm_set->lock);
- }
+ if (ret == -EOWNERDEAD)
+ pthread_mutex_consistent(set->lock);
#endif
if (ret == -ETIMEDOUT)
break;
@@ -394,10 +366,10 @@ ssize_t shm_flow_set_wait(const struct shm_flow_set * shm_set,
if (ret != -ETIMEDOUT) {
memcpy(fqueue,
- fqueue_ptr(shm_set, idx),
- shm_set->heads[idx] * sizeof(int));
- ret = shm_set->heads[idx];
- shm_set->heads[idx] = 0;
+ fqueue_ptr(set, idx),
+ set->heads[idx] * sizeof(int));
+ ret = set->heads[idx];
+ set->heads[idx] = 0;
}
pthread_cleanup_pop(true);