Rename and export lib kernel consumer
[lttng-tools.git] / liblttkconsumerd / lttkconsumerd.c
diff --git a/liblttkconsumerd/lttkconsumerd.c b/liblttkconsumerd/lttkconsumerd.c
deleted file mode 100644 (file)
index 29f1f95..0000000
+++ /dev/null
@@ -1,989 +0,0 @@
-/*
- * Copyright (C) 2011 - Julien Desfossez <julien.desfossez@polymtl.ca>
- *                      Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; only version 2
- * of the License.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.
- */
-
-#define _GNU_SOURCE
-#include <fcntl.h>
-#include <poll.h>
-#include <pthread.h>
-#include <stdlib.h>
-#include <string.h>
-#include <sys/mman.h>
-#include <sys/socket.h>
-#include <sys/types.h>
-#include <unistd.h>
-#include <urcu/list.h>
-#include <assert.h>
-
-#include "kernelctl.h"
-#include "lttkconsumerd.h"
-#include "lttngerr.h"
-
-static
-struct kconsumerd_global_data {
-       /*
-        * kconsumerd_data.lock protects kconsumerd_data.fd_list,
-        * kconsumerd_data.fds_count, and kconsumerd_data.need_update.  It
-        * ensures the count matches the number of items in the fd_list.
-        * It ensures the list updates *always* trigger an fd_array
-        * update (therefore need to make list update vs
-        * kconsumerd_data.need_update flag update atomic, and also flag
-        * read, fd array and flag clear atomic).
-        */
-       pthread_mutex_t lock;
-       /*
-        * Number of element for the list below.  Protected by
-        * kconsumerd_data.lock.
-        */
-       unsigned int fds_count;
-       /*
-        * List of FDs.  Protected by kconsumerd_data.lock.
-        */
-       struct kconsumerd_fd_list fd_list;
-       /*
-        * Flag specifying if the local array of FDs needs update in the
-        * poll function.  Protected by kconsumerd_data.lock.
-        */
-       unsigned int need_update;
-} kconsumerd_data = {
-       .fd_list.head = CDS_LIST_HEAD_INIT(kconsumerd_data.fd_list.head),
-       .fds_count = 0,
-       .need_update = 1,
-};
-
-/* timeout parameter, to control the polling thread grace period */
-static int kconsumerd_poll_timeout = -1;
-
-/*
- * flag to inform the polling thread to quit when all fd hung up.
- * Updated by the kconsumerd_thread_receive_fds when it notices that all
- * fds has hung up. Also updated by the signal handler
- * (kconsumerd_should_exit()). Read by the polling threads.
- */
-static volatile int kconsumerd_quit = 0;
-
-/*
- * kconsumerd_set_error_socket
- *
- * Set the error socket
- */
-void kconsumerd_set_error_socket(struct kconsumerd_local_data *ctx, int sock)
-{
-       ctx->kconsumerd_error_socket = sock;
-}
-
-/*
- * kconsumerd_set_command_socket_path
- *
- * Set the command socket path
- */
-void kconsumerd_set_command_socket_path(struct kconsumerd_local_data *ctx,
-               char *sock)
-{
-       ctx->kconsumerd_command_sock_path = sock;
-}
-
-/*
- * kconsumerd_find_session_fd
- *
- * Find a session fd in the global list.
- * The kconsumerd_data.lock must be locked during this call
- *
- * Return 1 if found else 0
- */
-static int kconsumerd_find_session_fd(int fd)
-{
-       struct kconsumerd_fd *iter;
-
-       cds_list_for_each_entry(iter, &kconsumerd_data.fd_list.head, list) {
-               if (iter->sessiond_fd == fd) {
-                       DBG("Duplicate session fd %d", fd);
-                       return 1;
-               }
-       }
-
-       return 0;
-}
-
-/*
- * kconsumerd_del_fd
- *
- * Remove a fd from the global list protected by a mutex
- */
-static void kconsumerd_del_fd(struct kconsumerd_fd *lcf)
-{
-       int ret;
-       pthread_mutex_lock(&kconsumerd_data.lock);
-       cds_list_del(&lcf->list);
-       if (kconsumerd_data.fds_count > 0) {
-               kconsumerd_data.fds_count--;
-               if (lcf != NULL) {
-                       if (lcf->mmap_base != NULL) {
-                               ret = munmap(lcf->mmap_base, lcf->mmap_len);
-                               if (ret != 0) {
-                                       perror("munmap");
-                               }
-                       }
-                       if (lcf->out_fd != 0) {
-                               close(lcf->out_fd);
-                       }
-                       close(lcf->consumerd_fd);
-                       free(lcf);
-                       lcf = NULL;
-               }
-       }
-       kconsumerd_data.need_update = 1;
-       pthread_mutex_unlock(&kconsumerd_data.lock);
-}
-
-/*
- * kconsumerd_add_fd
- *
- * Add a fd to the global list protected by a mutex
- */
-static int kconsumerd_add_fd(struct lttcomm_kconsumerd_msg *buf, int consumerd_fd)
-{
-       struct kconsumerd_fd *tmp_fd;
-       int ret = 0;
-
-       pthread_mutex_lock(&kconsumerd_data.lock);
-       /* Check if already exist */
-       ret = kconsumerd_find_session_fd(buf->fd);
-       if (ret == 1) {
-               goto end;
-       }
-
-       tmp_fd = malloc(sizeof(struct kconsumerd_fd));
-       tmp_fd->sessiond_fd = buf->fd;
-       tmp_fd->consumerd_fd = consumerd_fd;
-       tmp_fd->state = buf->state;
-       tmp_fd->max_sb_size = buf->max_sb_size;
-       tmp_fd->out_fd = 0;
-       tmp_fd->out_fd_offset = 0;
-       tmp_fd->mmap_len = 0;
-       tmp_fd->mmap_base = NULL;
-       tmp_fd->output = buf->output;
-       strncpy(tmp_fd->path_name, buf->path_name, PATH_MAX);
-       tmp_fd->path_name[PATH_MAX - 1] = '\0';
-
-       /* Opening the tracefile in write mode */
-       if (tmp_fd->path_name != NULL) {
-               ret = open(tmp_fd->path_name,
-                               O_WRONLY|O_CREAT|O_TRUNC, S_IRWXU|S_IRWXG|S_IRWXO);
-               if (ret < 0) {
-                       ERR("Opening %s", tmp_fd->path_name);
-                       perror("open");
-                       goto end;
-               }
-               tmp_fd->out_fd = ret;
-               DBG("Adding %s (%d, %d, %d)", tmp_fd->path_name,
-                               tmp_fd->sessiond_fd, tmp_fd->consumerd_fd, tmp_fd->out_fd);
-       }
-
-       if (tmp_fd->output == LTTNG_EVENT_MMAP) {
-               /* get the len of the mmap region */
-               ret = kernctl_get_mmap_len(tmp_fd->consumerd_fd, &tmp_fd->mmap_len);
-               if (ret != 0) {
-                       ret = errno;
-                       perror("kernctl_get_mmap_len");
-                       goto end;
-               }
-
-               tmp_fd->mmap_base = mmap(NULL, tmp_fd->mmap_len,
-                               PROT_READ, MAP_PRIVATE, tmp_fd->consumerd_fd, 0);
-               if (tmp_fd->mmap_base == MAP_FAILED) {
-                       perror("Error mmaping");
-                       ret = -1;
-                       goto end;
-               }
-       }
-
-       cds_list_add(&tmp_fd->list, &kconsumerd_data.fd_list.head);
-       kconsumerd_data.fds_count++;
-       kconsumerd_data.need_update = 1;
-end:
-       pthread_mutex_unlock(&kconsumerd_data.lock);
-       return ret;
-}
-
-/*
- * kconsumerd_change_fd_state
- *
- * Update a fd according to what we just received
- */
-static void kconsumerd_change_fd_state(int sessiond_fd,
-               enum kconsumerd_fd_state state)
-{
-       struct kconsumerd_fd *iter;
-
-       pthread_mutex_lock(&kconsumerd_data.lock);
-       cds_list_for_each_entry(iter, &kconsumerd_data.fd_list.head, list) {
-               if (iter->sessiond_fd == sessiond_fd) {
-                       iter->state = state;
-                       break;
-               }
-       }
-       kconsumerd_data.need_update = 1;
-       pthread_mutex_unlock(&kconsumerd_data.lock);
-}
-
-/*
- * kconsumerd_update_poll_array
- *
- * Allocate the pollfd structure and the local view of the out fds
- * to avoid doing a lookup in the linked list and concurrency issues
- * when writing is needed.
- * Returns the number of fds in the structures
- * Called with kconsumerd_data.lock held.
- */
-static int kconsumerd_update_poll_array(struct kconsumerd_local_data *ctx,
-               struct pollfd **pollfd, struct kconsumerd_fd **local_kconsumerd_fd)
-{
-       struct kconsumerd_fd *iter;
-       int i = 0;
-
-       DBG("Updating poll fd array");
-       cds_list_for_each_entry(iter, &kconsumerd_data.fd_list.head, list) {
-               if (iter->state == ACTIVE_FD) {
-                       DBG("Active FD %d", iter->consumerd_fd);
-                       (*pollfd)[i].fd = iter->consumerd_fd;
-                       (*pollfd)[i].events = POLLIN | POLLPRI;
-                       local_kconsumerd_fd[i] = iter;
-                       i++;
-               }
-       }
-
-       /*
-        * insert the kconsumerd_poll_pipe at the end of the array and don't
-        * increment i so nb_fd is the number of real FD
-        */
-       (*pollfd)[i].fd = ctx->kconsumerd_poll_pipe[0];
-       (*pollfd)[i].events = POLLIN;
-       return i;
-}
-
-
-/*
- * kconsumerd_on_read_subbuffer_mmap
- *
- * mmap the ring buffer, read it and write the data to the tracefile.
- * Returns the number of bytes written
- */
-int kconsumerd_on_read_subbuffer_mmap(struct kconsumerd_local_data *ctx,
-               struct kconsumerd_fd *kconsumerd_fd, unsigned long len)
-{
-       unsigned long mmap_offset;
-       char *padding = NULL;
-       long ret = 0;
-       off_t orig_offset = kconsumerd_fd->out_fd_offset;
-       int fd = kconsumerd_fd->consumerd_fd;
-       int outfd = kconsumerd_fd->out_fd;
-
-       /* get the offset inside the fd to mmap */
-       ret = kernctl_get_mmap_read_offset(fd, &mmap_offset);
-       if (ret != 0) {
-               ret = errno;
-               perror("kernctl_get_mmap_read_offset");
-               goto end;
-       }
-
-       while (len > 0) {
-               ret = write(outfd, kconsumerd_fd->mmap_base + mmap_offset, len);
-               if (ret >= len) {
-                       len = 0;
-               } else if (ret < 0) {
-                       ret = errno;
-                       perror("Error in file write");
-                       goto end;
-               }
-               /* This won't block, but will start writeout asynchronously */
-               sync_file_range(outfd, kconsumerd_fd->out_fd_offset, ret,
-                               SYNC_FILE_RANGE_WRITE);
-               kconsumerd_fd->out_fd_offset += ret;
-       }
-
-       /*
-        * This does a blocking write-and-wait on any page that belongs to the
-        * subbuffer prior to the one we just wrote.
-        * Don't care about error values, as these are just hints and ways to
-        * limit the amount of page cache used.
-        */
-       if (orig_offset >= kconsumerd_fd->max_sb_size) {
-               sync_file_range(outfd, orig_offset - kconsumerd_fd->max_sb_size,
-                               kconsumerd_fd->max_sb_size,
-                               SYNC_FILE_RANGE_WAIT_BEFORE
-                               | SYNC_FILE_RANGE_WRITE
-                               | SYNC_FILE_RANGE_WAIT_AFTER);
-
-               /*
-                * Give hints to the kernel about how we access the file:
-                * POSIX_FADV_DONTNEED : we won't re-access data in a near future after
-                * we write it.
-                *
-                * We need to call fadvise again after the file grows because the
-                * kernel does not seem to apply fadvise to non-existing parts of the
-                * file.
-                *
-                * Call fadvise _after_ having waited for the page writeback to
-                * complete because the dirty page writeback semantic is not well
-                * defined. So it can be expected to lead to lower throughput in
-                * streaming.
-                */
-               posix_fadvise(outfd, orig_offset - kconsumerd_fd->max_sb_size,
-                               kconsumerd_fd->max_sb_size, POSIX_FADV_DONTNEED);
-       }
-       goto end;
-
-end:
-       if (padding != NULL) {
-               free(padding);
-       }
-       return ret;
-}
-
-/*
- * kconsumerd_on_read_subbuffer
- *
- * Splice the data from the ring buffer to the tracefile.
- * Returns the number of bytes spliced
- */
-int kconsumerd_on_read_subbuffer_splice(struct kconsumerd_local_data *ctx,
-               struct kconsumerd_fd *kconsumerd_fd, unsigned long len)
-{
-       long ret = 0;
-       loff_t offset = 0;
-       off_t orig_offset = kconsumerd_fd->out_fd_offset;
-       int fd = kconsumerd_fd->consumerd_fd;
-       int outfd = kconsumerd_fd->out_fd;
-
-       while (len > 0) {
-               DBG("splice chan to pipe offset %lu (fd : %d)",
-                               (unsigned long)offset, fd);
-               ret = splice(fd, &offset, ctx->kconsumerd_thread_pipe[1], NULL, len,
-                               SPLICE_F_MOVE | SPLICE_F_MORE);
-               DBG("splice chan to pipe ret %ld", ret);
-               if (ret < 0) {
-                       ret = errno;
-                       perror("Error in relay splice");
-                       goto splice_error;
-               }
-
-               ret = splice(ctx->kconsumerd_thread_pipe[0], NULL, outfd, NULL, ret,
-                               SPLICE_F_MOVE | SPLICE_F_MORE);
-               DBG("splice pipe to file %ld", ret);
-               if (ret < 0) {
-                       ret = errno;
-                       perror("Error in file splice");
-                       goto splice_error;
-               }
-               if (ret >= len) {
-                       len = 0;
-               }
-               /* This won't block, but will start writeout asynchronously */
-               sync_file_range(outfd, kconsumerd_fd->out_fd_offset, ret,
-                               SYNC_FILE_RANGE_WRITE);
-               kconsumerd_fd->out_fd_offset += ret;
-       }
-
-       /*
-        * This does a blocking write-and-wait on any page that belongs to the
-        * subbuffer prior to the one we just wrote.
-        * Don't care about error values, as these are just hints and ways to
-        * limit the amount of page cache used.
-        */
-       if (orig_offset >= kconsumerd_fd->max_sb_size) {
-               sync_file_range(outfd, orig_offset - kconsumerd_fd->max_sb_size,
-                               kconsumerd_fd->max_sb_size,
-                               SYNC_FILE_RANGE_WAIT_BEFORE
-                               | SYNC_FILE_RANGE_WRITE
-                               | SYNC_FILE_RANGE_WAIT_AFTER);
-               /*
-                * Give hints to the kernel about how we access the file:
-                * POSIX_FADV_DONTNEED : we won't re-access data in a near future after
-                * we write it.
-                *
-                * We need to call fadvise again after the file grows because the
-                * kernel does not seem to apply fadvise to non-existing parts of the
-                * file.
-                *
-                * Call fadvise _after_ having waited for the page writeback to
-                * complete because the dirty page writeback semantic is not well
-                * defined. So it can be expected to lead to lower throughput in
-                * streaming.
-                */
-               posix_fadvise(outfd, orig_offset - kconsumerd_fd->max_sb_size,
-                               kconsumerd_fd->max_sb_size, POSIX_FADV_DONTNEED);
-       }
-       goto end;
-
-splice_error:
-       /* send the appropriate error description to sessiond */
-       switch(ret) {
-       case EBADF:
-               kconsumerd_send_error(ctx, KCONSUMERD_SPLICE_EBADF);
-               break;
-       case EINVAL:
-               kconsumerd_send_error(ctx, KCONSUMERD_SPLICE_EINVAL);
-               break;
-       case ENOMEM:
-               kconsumerd_send_error(ctx, KCONSUMERD_SPLICE_ENOMEM);
-               break;
-       case ESPIPE:
-               kconsumerd_send_error(ctx, KCONSUMERD_SPLICE_ESPIPE);
-               break;
-       }
-
-end:
-       return ret;
-}
-
-/*
- * kconsumerd_poll_socket
- *
- * Poll on the should_quit pipe and the command socket
- * return -1 on error and should exit, 0 if data is
- * available on the command socket
- */
-int kconsumerd_poll_socket(struct pollfd *kconsumerd_sockpoll)
-{
-       int num_rdy;
-
-       num_rdy = poll(kconsumerd_sockpoll, 2, -1);
-       if (num_rdy == -1) {
-               perror("Poll error");
-               goto exit;
-       }
-       if (kconsumerd_sockpoll[0].revents == POLLIN) {
-               DBG("kconsumerd_should_quit wake up");
-               goto exit;
-       }
-       return 0;
-
-exit:
-       return -1;
-}
-
-/*
- * kconsumerd_consumerd_recv_fd
- *
- * Receives an array of file descriptors and the associated
- * structures describing each fd (path name).
- * Returns the size of received data
- */
-static int kconsumerd_consumerd_recv_fd(struct kconsumerd_local_data *ctx,
-               int sfd, struct pollfd *kconsumerd_sockpoll, int size,
-               enum kconsumerd_command cmd_type)
-{
-       struct iovec iov[1];
-       int ret = 0, i, tmp2;
-       struct cmsghdr *cmsg;
-       int nb_fd;
-       char recv_fd[CMSG_SPACE(sizeof(int))];
-       struct lttcomm_kconsumerd_msg lkm;
-
-       /* the number of fds we are about to receive */
-       nb_fd = size / sizeof(struct lttcomm_kconsumerd_msg);
-
-       /*
-        * nb_fd is the number of fds we receive. One fd per recvmsg.
-        */
-       for (i = 0; i < nb_fd; i++) {
-               struct msghdr msg = { 0 };
-
-               /* Prepare to receive the structures */
-               iov[0].iov_base = &lkm;
-               iov[0].iov_len = sizeof(lkm);
-               msg.msg_iov = iov;
-               msg.msg_iovlen = 1;
-
-               msg.msg_control = recv_fd;
-               msg.msg_controllen = sizeof(recv_fd);
-
-               DBG("Waiting to receive fd");
-               if (kconsumerd_poll_socket(kconsumerd_sockpoll) < 0) {
-                       goto end;
-               }
-
-               if ((ret = recvmsg(sfd, &msg, 0)) < 0) {
-                       perror("recvmsg");
-                       continue;
-               }
-
-               if (ret != (size / nb_fd)) {
-                       ERR("Received only %d, expected %d", ret, size);
-                       kconsumerd_send_error(ctx, KCONSUMERD_ERROR_RECV_FD);
-                       goto end;
-               }
-
-               cmsg = CMSG_FIRSTHDR(&msg);
-               if (!cmsg) {
-                       ERR("Invalid control message header");
-                       ret = -1;
-                       kconsumerd_send_error(ctx, KCONSUMERD_ERROR_RECV_FD);
-                       goto end;
-               }
-
-               /* if we received fds */
-               if (cmsg->cmsg_level == SOL_SOCKET && cmsg->cmsg_type == SCM_RIGHTS) {
-                       switch (cmd_type) {
-                       case ADD_STREAM:
-                               DBG("kconsumerd_add_fd %s (%d)", lkm.path_name, ((int *) CMSG_DATA(cmsg))[0]);
-                               ret = kconsumerd_add_fd(&lkm, ((int *) CMSG_DATA(cmsg))[0]);
-                               if (ret < 0) {
-                                       kconsumerd_send_error(ctx, KCONSUMERD_OUTFD_ERROR);
-                                       goto end;
-                               }
-                               break;
-                       case UPDATE_STREAM:
-                               kconsumerd_change_fd_state(lkm.fd, lkm.state);
-                               break;
-                       default:
-                               break;
-                       }
-                       /* signal the poll thread */
-                       tmp2 = write(ctx->kconsumerd_poll_pipe[1], "4", 1);
-                       if (tmp2 < 0) {
-                               perror("write kconsumerd poll");
-                       }
-               } else {
-                       ERR("Didn't received any fd");
-                       kconsumerd_send_error(ctx, KCONSUMERD_ERROR_RECV_FD);
-                       ret = -1;
-                       goto end;
-               }
-       }
-
-end:
-       return ret;
-}
-
-/*
- *  kconsumerd_thread_poll_fds
- *
- *  This thread polls the fds in the ltt_fd_list to consume the data
- *  and write it to tracefile if necessary.
- */
-void *kconsumerd_thread_poll_fds(void *data)
-{
-       int num_rdy, num_hup, high_prio, ret, i;
-       struct pollfd *pollfd = NULL;
-       /* local view of the fds */
-       struct kconsumerd_fd **local_kconsumerd_fd = NULL;
-       /* local view of kconsumerd_data.fds_count */
-       int nb_fd = 0;
-       char tmp;
-       int tmp2;
-       struct kconsumerd_local_data *ctx = data;
-
-
-       local_kconsumerd_fd = malloc(sizeof(struct kconsumerd_fd));
-
-       while (1) {
-               high_prio = 0;
-               num_hup = 0;
-
-               /*
-                * the ltt_fd_list has been updated, we need to update our
-                * local array as well
-                */
-               pthread_mutex_lock(&kconsumerd_data.lock);
-               if (kconsumerd_data.need_update) {
-                       if (pollfd != NULL) {
-                               free(pollfd);
-                               pollfd = NULL;
-                       }
-                       if (local_kconsumerd_fd != NULL) {
-                               free(local_kconsumerd_fd);
-                               local_kconsumerd_fd = NULL;
-                       }
-
-                       /* allocate for all fds + 1 for the kconsumerd_poll_pipe */
-                       pollfd = malloc((kconsumerd_data.fds_count + 1) * sizeof(struct pollfd));
-                       if (pollfd == NULL) {
-                               perror("pollfd malloc");
-                               pthread_mutex_unlock(&kconsumerd_data.lock);
-                               goto end;
-                       }
-
-                       /* allocate for all fds + 1 for the kconsumerd_poll_pipe */
-                       local_kconsumerd_fd = malloc((kconsumerd_data.fds_count + 1) *
-                                       sizeof(struct kconsumerd_fd));
-                       if (local_kconsumerd_fd == NULL) {
-                               perror("local_kconsumerd_fd malloc");
-                               pthread_mutex_unlock(&kconsumerd_data.lock);
-                               goto end;
-                       }
-                       ret = kconsumerd_update_poll_array(ctx, &pollfd, local_kconsumerd_fd);
-                       if (ret < 0) {
-                               ERR("Error in allocating pollfd or local_outfds");
-                               kconsumerd_send_error(ctx, KCONSUMERD_POLL_ERROR);
-                               pthread_mutex_unlock(&kconsumerd_data.lock);
-                               goto end;
-                       }
-                       nb_fd = ret;
-                       kconsumerd_data.need_update = 0;
-               }
-               pthread_mutex_unlock(&kconsumerd_data.lock);
-
-               /* poll on the array of fds */
-               DBG("polling on %d fd", nb_fd + 1);
-               num_rdy = poll(pollfd, nb_fd + 1, kconsumerd_poll_timeout);
-               DBG("poll num_rdy : %d", num_rdy);
-               if (num_rdy == -1) {
-                       perror("Poll error");
-                       kconsumerd_send_error(ctx, KCONSUMERD_POLL_ERROR);
-                       goto end;
-               } else if (num_rdy == 0) {
-                       DBG("Polling thread timed out");
-                       goto end;
-               }
-
-               /* No FDs and kconsumerd_quit, kconsumerd_cleanup the thread */
-               if (nb_fd == 0 && kconsumerd_quit == 1) {
-                       goto end;
-               }
-
-               /*
-                * If the kconsumerd_poll_pipe triggered poll go
-                * directly to the beginning of the loop to update the
-                * array. We want to prioritize array update over
-                * low-priority reads.
-                */
-               if (pollfd[nb_fd].revents == POLLIN) {
-                       DBG("kconsumerd_poll_pipe wake up");
-                       tmp2 = read(ctx->kconsumerd_poll_pipe[0], &tmp, 1);
-                       if (tmp2 < 0) {
-                               perror("read kconsumerd poll");
-                       }
-                       continue;
-               }
-
-               /* Take care of high priority channels first. */
-               for (i = 0; i < nb_fd; i++) {
-                       switch(pollfd[i].revents) {
-                       case POLLERR:
-                               ERR("Error returned in polling fd %d.", pollfd[i].fd);
-                               kconsumerd_del_fd(local_kconsumerd_fd[i]);
-                               num_hup++;
-                               break;
-                       case POLLHUP:
-                               DBG("Polling fd %d tells it has hung up.", pollfd[i].fd);
-                               kconsumerd_del_fd(local_kconsumerd_fd[i]);
-                               num_hup++;
-                               break;
-                       case POLLNVAL:
-                               ERR("Polling fd %d tells fd is not open.", pollfd[i].fd);
-                               kconsumerd_del_fd(local_kconsumerd_fd[i]);
-                               num_hup++;
-                               break;
-                       case POLLPRI:
-                               DBG("Urgent read on fd %d", pollfd[i].fd);
-                               high_prio = 1;
-                               ret = ctx->on_buffer_ready(local_kconsumerd_fd[i]);
-                               /* it's ok to have an unavailable sub-buffer */
-                               if (ret == EAGAIN) {
-                                       ret = 0;
-                               }
-                               break;
-                       }
-               }
-
-               /* If every buffer FD has hung up, we end the read loop here */
-               if (nb_fd > 0 && num_hup == nb_fd) {
-                       DBG("every buffer FD has hung up\n");
-                       if (kconsumerd_quit == 1) {
-                               goto end;
-                       }
-                       continue;
-               }
-
-               /* Take care of low priority channels. */
-               if (high_prio == 0) {
-                       for (i = 0; i < nb_fd; i++) {
-                               if (pollfd[i].revents == POLLIN) {
-                                       DBG("Normal read on fd %d", pollfd[i].fd);
-                                       ret = ctx->on_buffer_ready(local_kconsumerd_fd[i]);
-                                       /* it's ok to have an unavailable subbuffer */
-                                       if (ret == EAGAIN) {
-                                               ret = 0;
-                                       }
-                               }
-                       }
-               }
-       }
-end:
-       DBG("polling thread exiting");
-       if (pollfd != NULL) {
-               free(pollfd);
-               pollfd = NULL;
-       }
-       if (local_kconsumerd_fd != NULL) {
-               free(local_kconsumerd_fd);
-               local_kconsumerd_fd = NULL;
-       }
-       return NULL;
-}
-
-/*
- * kconsumerd_create
- *
- * initialise the necessary environnement :
- * - create a new context
- * - create the poll_pipe
- * - create the should_quit pipe (for signal handler)
- * - create the thread pipe (for splice)
- * Takes a function pointer as argument, this function is called when data is
- * available on a buffer. This function is responsible to do the
- * kernctl_get_next_subbuf, read the data with mmap or splice depending on the
- * buffer configuration and then kernctl_put_next_subbuf at the end.
- * Returns a pointer to the new context or NULL on error.
- */
-struct kconsumerd_local_data *kconsumerd_create(
-               int (*buffer_ready)(struct kconsumerd_fd *kconsumerd_fd))
-{
-       int ret;
-       struct kconsumerd_local_data *ctx;
-
-       ctx = malloc(sizeof(struct kconsumerd_local_data));
-       if (ctx == NULL) {
-               perror("allocating context");
-               goto end;
-       }
-
-       ctx->on_buffer_ready = buffer_ready;
-
-       ret = pipe(ctx->kconsumerd_poll_pipe);
-       if (ret < 0) {
-               perror("Error creating poll pipe");
-               ctx = NULL;
-               goto end;
-       }
-
-       ret = pipe(ctx->kconsumerd_should_quit);
-       if (ret < 0) {
-               perror("Error creating recv pipe");
-               ctx = NULL;
-               goto end;
-       }
-
-       ret = pipe(ctx->kconsumerd_thread_pipe);
-       if (ret < 0) {
-               perror("Error creating thread pipe");
-               ctx = NULL;
-               goto end;
-       }
-
-end:
-       return ctx;
-}
-
-/*
- * kconsumerd_destroy
- *
- * Close all fds associated with the instance and free the context
- */
-void kconsumerd_destroy(struct kconsumerd_local_data *ctx)
-{
-       close(ctx->kconsumerd_error_socket);
-       close(ctx->kconsumerd_thread_pipe[0]);
-       close(ctx->kconsumerd_thread_pipe[1]);
-       close(ctx->kconsumerd_poll_pipe[0]);
-       close(ctx->kconsumerd_poll_pipe[1]);
-       close(ctx->kconsumerd_should_quit[0]);
-       close(ctx->kconsumerd_should_quit[1]);
-       unlink(ctx->kconsumerd_command_sock_path);
-       free(ctx);
-       ctx = NULL;
-}
-
-/*
- *  kconsumerd_thread_receive_fds
- *
- *  This thread listens on the consumerd socket and
- *  receives the file descriptors from ltt-sessiond
- */
-void *kconsumerd_thread_receive_fds(void *data)
-{
-       int sock, client_socket, ret;
-       struct lttcomm_kconsumerd_header tmp;
-       /*
-        * structure to poll for incoming data on communication socket
-        * avoids making blocking sockets
-        */
-       struct pollfd kconsumerd_sockpoll[2];
-       struct kconsumerd_local_data *ctx = data;
-
-
-       DBG("Creating command socket %s", ctx->kconsumerd_command_sock_path);
-       unlink(ctx->kconsumerd_command_sock_path);
-       client_socket = lttcomm_create_unix_sock(ctx->kconsumerd_command_sock_path);
-       if (client_socket < 0) {
-               ERR("Cannot create command socket");
-               goto end;
-       }
-
-       ret = lttcomm_listen_unix_sock(client_socket);
-       if (ret < 0) {
-               goto end;
-       }
-
-       DBG("Sending ready command to ltt-sessiond");
-       ret = kconsumerd_send_error(ctx, KCONSUMERD_COMMAND_SOCK_READY);
-       if (ret < 0) {
-               ERR("Error sending ready command to ltt-sessiond");
-               goto end;
-       }
-
-       ret = fcntl(client_socket, F_SETFL, O_NONBLOCK);
-       if (ret < 0) {
-               perror("fcntl O_NONBLOCK");
-               goto end;
-       }
-
-       /* prepare the FDs to poll : to client socket and the should_quit pipe */
-       kconsumerd_sockpoll[0].fd = ctx->kconsumerd_should_quit[0];
-       kconsumerd_sockpoll[0].events = POLLIN | POLLPRI;
-       kconsumerd_sockpoll[1].fd = client_socket;
-       kconsumerd_sockpoll[1].events = POLLIN | POLLPRI;
-
-       if (kconsumerd_poll_socket(kconsumerd_sockpoll) < 0) {
-               goto end;
-       }
-       DBG("Connection on client_socket");
-
-       /* Blocking call, waiting for transmission */
-       sock = lttcomm_accept_unix_sock(client_socket);
-       if (sock <= 0) {
-               WARN("On accept");
-               goto end;
-       }
-       ret = fcntl(sock, F_SETFL, O_NONBLOCK);
-       if (ret < 0) {
-               perror("fcntl O_NONBLOCK");
-               goto end;
-       }
-
-       /* update the polling structure to poll on the established socket */
-       kconsumerd_sockpoll[1].fd = sock;
-       kconsumerd_sockpoll[1].events = POLLIN | POLLPRI;
-
-       while (1) {
-               if (kconsumerd_poll_socket(kconsumerd_sockpoll) < 0) {
-                       goto end;
-               }
-               DBG("Incoming fds on sock");
-
-               /* We first get the number of fd we are about to receive */
-               ret = lttcomm_recv_unix_sock(sock, &tmp,
-                               sizeof(struct lttcomm_kconsumerd_header));
-               if (ret <= 0) {
-                       ERR("Communication interrupted on command socket");
-                       goto end;
-               }
-               if (tmp.cmd_type == STOP) {
-                       DBG("Received STOP command");
-                       goto end;
-               }
-               if (kconsumerd_quit) {
-                       DBG("kconsumerd_thread_receive_fds received quit from signal");
-                       goto end;
-               }
-
-               /* we received a command to add or update fds */
-               ret = kconsumerd_consumerd_recv_fd(ctx, sock, kconsumerd_sockpoll,
-                               tmp.payload_size, tmp.cmd_type);
-               if (ret < 0) {
-                       ERR("Receiving the FD, exiting");
-                       goto end;
-               }
-               DBG("received fds on sock");
-       }
-
-end:
-       DBG("kconsumerd_thread_receive_fds exiting");
-
-       /*
-        * when all fds have hung up, the polling thread
-        * can exit cleanly
-        */
-       kconsumerd_quit = 1;
-
-       /*
-        * 2s of grace period, if no polling events occur during
-        * this period, the polling thread will exit even if there
-        * are still open FDs (should not happen, but safety mechanism).
-        */
-       kconsumerd_poll_timeout = KCONSUMERD_POLL_GRACE_PERIOD;
-
-       /* wake up the polling thread */
-       ret = write(ctx->kconsumerd_poll_pipe[1], "4", 1);
-       if (ret < 0) {
-               perror("poll pipe write");
-       }
-       return NULL;
-}
-
-/*
- *  kconsumerd_cleanup
- *
- *  Close all the tracefiles and stream fds, should be called when all
- *  instances are destroyed.
- */
-void kconsumerd_cleanup(void)
-{
-       struct kconsumerd_fd *iter, *tmp;
-
-       /*
-        * close all outfd. Called when there are no more threads
-        * running (after joining on the threads), no need to protect
-        * list iteration with mutex.
-        */
-       cds_list_for_each_entry_safe(iter, tmp, &kconsumerd_data.fd_list.head, list) {
-               kconsumerd_del_fd(iter);
-       }
-}
-
-/*
- * kconsumerd_should_exit
- *
- * Called from signal handler.
- */
-void kconsumerd_should_exit(struct kconsumerd_local_data *ctx)
-{
-       int ret;
-       kconsumerd_quit = 1;
-       ret = write(ctx->kconsumerd_should_quit[1], "4", 1);
-       if (ret < 0) {
-               perror("write kconsumerd quit");
-       }
-}
-
-/*
- * kconsumerd_send_error
- *
- * send return code to ltt-sessiond
- */
-int kconsumerd_send_error(struct kconsumerd_local_data *ctx, enum lttcomm_return_code cmd)
-{
-       if (ctx->kconsumerd_error_socket > 0) {
-               return lttcomm_send_unix_sock(ctx->kconsumerd_error_socket, &cmd,
-                               sizeof(enum lttcomm_sessiond_command));
-       }
-
-       return 0;
-}
This page took 0.031944 seconds and 4 git commands to generate.