diff options
Diffstat (limited to 'libglusterfs/src/graph.c')
| -rw-r--r-- | libglusterfs/src/graph.c | 1987 |
1 files changed, 1661 insertions, 326 deletions
diff --git a/libglusterfs/src/graph.c b/libglusterfs/src/graph.c index 830c8836300..13f298eb3bd 100644 --- a/libglusterfs/src/graph.c +++ b/libglusterfs/src/graph.c @@ -1,35 +1,39 @@ /* - Copyright (c) 2010-2011 Gluster, Inc. <http://www.gluster.com> + Copyright (c) 2008-2012 Red Hat, Inc. <http://www.redhat.com> This file is part of GlusterFS. - GlusterFS 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; either version 3 of the License, - or (at your option) any later version. - - GlusterFS 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, see - <http://www.gnu.org/licenses/>. + This file is licensed to you under your choice of the GNU Lesser + General Public License, version 3 or any later version (LGPLv3 or + later), or the GNU General Public License, version 2 (GPLv2), in all + cases as published by the Free Software Foundation. */ -#ifndef _CONFIG_H -#define _CONFIG_H -#include "config.h" -#endif - -#include "xlator.h" -#include <dlfcn.h> -#include <netdb.h> -#include <fnmatch.h> -#include "defaults.h" - - - +#include <stdint.h> // for uint32_t +#include <sys/time.h> // for timeval +#include <errno.h> // for EIO, errno, EINVAL, ENOMEM +#include <fnmatch.h> // for fnmatch, FNM_NOESCAPE +#include <openssl/sha.h> // for SHA256_DIGEST_LENGTH +#include <regex.h> // for regmatch_t, regcomp +#include <stdio.h> // for fclose, fopen, snprintf +#include <stdlib.h> // for NULL, atoi, mkstemp +#include <string.h> // for strcmp, strerror, memcpy +#include <strings.h> // for rindex +#include <sys/stat.h> // for stat +#include <sys/time.h> // for gettimeofday +#include <unistd.h> // for gethostname, getpid +#include "glusterfs/common-utils.h" // for gf_strncpy, gf_time_fmt +#include "glusterfs/defaults.h" +#include "glusterfs/dict.h" // for dict_foreach, dict_set_... +#include "glusterfs/globals.h" // for xlator_t, xlator_list_t +#include "glusterfs/glusterfs.h" // for glusterfs_graph_t, glus... +#include "glusterfs/glusterfs-fops.h" // for GF_EVENT_GRAPH_NEW, GF_... +#include "glusterfs/libglusterfs-messages.h" // for LG_MSG_GRAPH_ERROR, LG_... +#include "glusterfs/list.h" // for list_add, list_del_init +#include "glusterfs/logging.h" // for gf_msg, GF_LOG_ERROR +#include "glusterfs/mem-pool.h" // for GF_FREE, gf_strdup, GF_... +#include "glusterfs/mem-types.h" // for gf_common_mt_xlator_list_t +#include "glusterfs/options.h" // for xlator_tree_reconfigure +#include "glusterfs/syscall.h" // for sys_close, sys_stat #if 0 static void @@ -37,20 +41,17 @@ _gf_dump_details (int argc, char **argv) { extern FILE *gf_log_logfile; int i = 0; - char timestr[256]; + char timestr[GF_TIMESTR_SIZE]; time_t utime = 0; - struct tm *tm = NULL; pid_t mypid = 0; struct utsname uname_buf = {{0, }, }; int uname_ret = -1; - utime = time (NULL); - tm = localtime (&utime); mypid = getpid (); uname_ret = uname (&uname_buf); - /* Which git? What time? */ - strftime (timestr, 256, "%Y-%m-%d %H:%M:%S", tm); + utime = time (NULL); + gf_time_fmt (timestr, sizeof timestr, utime, gf_timefmt_FT); fprintf (gf_log_logfile, "========================================" "========================================\n"); @@ -83,463 +84,1797 @@ _gf_dump_details (int argc, char **argv) } #endif +int +glusterfs_read_secure_access_file(void) +{ + FILE *fp = NULL; + char line[100] = { + 0, + }; + int cert_depth = 1; /* Default SSL CERT DEPTH */ + regex_t regcmpl; + char *key = {"^option transport.socket.ssl-cert-depth"}; + char keyval[50] = { + 0, + }; + int start = 0, end = 0, copy_len = 0; + regmatch_t result[1] = {{0}}; + + fp = fopen(SECURE_ACCESS_FILE, "r"); + if (!fp) + goto out; + + /* Check if any line matches with key */ + while (fgets(line, sizeof(line), fp) != NULL) { + if (regcomp(®cmpl, key, REG_EXTENDED)) { + goto out; + } + if (!regexec(®cmpl, line, 1, result, 0)) { + start = result[0].rm_so; + end = result[0].rm_eo; + copy_len = end - start; + gf_strncpy(keyval, line + copy_len, sizeof(keyval)); + if (keyval[0]) { + cert_depth = atoi(keyval); + if (cert_depth == 0) + cert_depth = 1; /* Default SSL CERT DEPTH */ + break; + } + } + regfree(®cmpl); + } +out: + if (fp) + fclose(fp); + return cert_depth; +} -int -glusterfs_xlator_link (xlator_t *pxl, xlator_t *cxl) +xlator_t * +glusterfs_get_last_xlator(glusterfs_graph_t *graph) { - xlator_list_t *xlchild = NULL; - xlator_list_t *xlparent = NULL; - xlator_list_t **tmp = NULL; + xlator_t *trav = graph->first; + if (!trav) + return NULL; - xlparent = (void *) GF_CALLOC (1, sizeof (*xlparent), - gf_common_mt_xlator_list_t); - if (!xlparent) - return -1; + while (trav->next) + trav = trav->next; - xlchild = (void *) GF_CALLOC (1, sizeof (*xlchild), - gf_common_mt_xlator_list_t); - if (!xlchild) { - GF_FREE (xlparent); + return trav; +} - return -1; +xlator_t * +glusterfs_mux_xlator_unlink(xlator_t *pxl, xlator_t *cxl) +{ + xlator_list_t *unlink = NULL; + xlator_list_t *prev = NULL; + xlator_list_t **tmp = NULL; + xlator_t *next_child = NULL; + xlator_t *xl = NULL; + + for (tmp = &pxl->children; *tmp; tmp = &(*tmp)->next) { + if ((*tmp)->xlator == cxl) { + unlink = *tmp; + *tmp = (*tmp)->next; + if (*tmp) + next_child = (*tmp)->xlator; + break; } + prev = *tmp; + } - xlparent->xlator = pxl; - for (tmp = &cxl->parents; *tmp; tmp = &(*tmp)->next); - *tmp = xlparent; + if (!prev) + xl = pxl; + else if (prev->xlator) + xl = prev->xlator->graph->last_xl; - xlchild->xlator = cxl; - for (tmp = &pxl->children; *tmp; tmp = &(*tmp)->next); - *tmp = xlchild; + if (xl) + xl->next = next_child; + if (next_child) + next_child->prev = xl; - return 0; + GF_FREE(unlink); + return next_child; } +int +glusterfs_xlator_link(xlator_t *pxl, xlator_t *cxl) +{ + xlator_list_t *xlchild = NULL; + xlator_list_t *xlparent = NULL; + xlator_list_t **tmp = NULL; + + xlparent = (void *)GF_CALLOC(1, sizeof(*xlparent), + gf_common_mt_xlator_list_t); + if (!xlparent) + return -1; + + xlchild = (void *)GF_CALLOC(1, sizeof(*xlchild), + gf_common_mt_xlator_list_t); + if (!xlchild) { + GF_FREE(xlparent); + + return -1; + } + + xlparent->xlator = pxl; + for (tmp = &cxl->parents; *tmp; tmp = &(*tmp)->next) + ; + *tmp = xlparent; + + xlchild->xlator = cxl; + for (tmp = &pxl->children; *tmp; tmp = &(*tmp)->next) + ; + *tmp = xlchild; + + return 0; +} void -glusterfs_graph_set_first (glusterfs_graph_t *graph, xlator_t *xl) +glusterfs_graph_set_first(glusterfs_graph_t *graph, xlator_t *xl) { - xl->next = graph->first; - if (graph->first) - ((xlator_t *)graph->first)->prev = xl; - graph->first = xl; + xl->next = graph->first; + if (graph->first) + ((xlator_t *)graph->first)->prev = xl; + graph->first = xl; - graph->xl_count++; + graph->xl_count++; + xl->xl_id = graph->xl_count; } - int -glusterfs_graph_insert (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx, - const char *type, const char *name) +glusterfs_graph_insert(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx, + const char *type, const char *name, + gf_boolean_t autoload) { - xlator_t *ixl = NULL; + xlator_t *ixl = NULL; - if (!ctx->master) { - gf_log ("glusterfs", GF_LOG_ERROR, - "volume \"%s\" can be added from command line only " - "on client side", type); + if (!ctx->master) { + gf_msg("glusterfs", GF_LOG_ERROR, 0, LG_MSG_VOLUME_ERROR, + "volume \"%s\" can be added from command line only " + "on client side", + type); - return -1; - } + return -1; + } - ixl = GF_CALLOC (1, sizeof (*ixl), gf_common_mt_xlator_t); - if (!ixl) - return -1; - - ixl->ctx = ctx; - ixl->graph = graph; - ixl->options = get_new_dict (); - if (!ixl->options) - goto err; - - ixl->name = gf_strdup (name); - if (!ixl->name) - goto err; - - if (xlator_set_type (ixl, type) == -1) { - gf_log ("glusterfs", GF_LOG_ERROR, - "%s (%s) initialization failed", - name, type); - return -1; - } + ixl = GF_CALLOC(1, sizeof(*ixl), gf_common_mt_xlator_t); + if (!ixl) + return -1; - if (glusterfs_xlator_link (ixl, graph->top) == -1) - goto err; - glusterfs_graph_set_first (graph, ixl); - graph->top = ixl; + ixl->ctx = ctx; + ixl->graph = graph; + ixl->options = dict_new(); + if (!ixl->options) + goto err; - return 0; -err: - xlator_destroy (ixl); + ixl->name = gf_strdup(name); + if (!ixl->name) + goto err; + + ixl->is_autoloaded = autoload; + + if (xlator_set_type(ixl, type) == -1) { + gf_msg("glusterfs", GF_LOG_ERROR, 0, LG_MSG_INIT_FAILED, + "%s (%s) initialization failed", name, type); return -1; + } + + if (glusterfs_xlator_link(ixl, graph->top) == -1) + goto err; + glusterfs_graph_set_first(graph, ixl); + graph->top = ixl; + + return 0; +err: + xlator_destroy(ixl); + return -1; } int -glusterfs_graph_acl (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +glusterfs_graph_acl(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) { - int ret = 0; - cmd_args_t *cmd_args = NULL; + int ret = 0; + cmd_args_t *cmd_args = NULL; - cmd_args = &ctx->cmd_args; + cmd_args = &ctx->cmd_args; - if (!cmd_args->acl) - return 0; + if (!cmd_args->acl) + return 0; - ret = glusterfs_graph_insert (graph, ctx, "system/posix-acl", - "posix-acl-autoload"); - return ret; + ret = glusterfs_graph_insert(graph, ctx, "system/posix-acl", + "posix-acl-autoload", 1); + return ret; } int -glusterfs_graph_worm (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +glusterfs_graph_worm(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) { - int ret = 0; - cmd_args_t *cmd_args = NULL; + int ret = 0; + cmd_args_t *cmd_args = NULL; - cmd_args = &ctx->cmd_args; + cmd_args = &ctx->cmd_args; - if (!cmd_args->worm) - return 0; + if (!cmd_args->worm) + return 0; - ret = glusterfs_graph_insert (graph, ctx, "features/worm", - "worm-autoload"); - return ret; + ret = glusterfs_graph_insert(graph, ctx, "features/worm", "worm-autoload", + 1); + return ret; } int -glusterfs_graph_mac_compat (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +glusterfs_graph_meta(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) { - int ret = 0; - cmd_args_t *cmd_args = NULL; + int ret = 0; - cmd_args = &ctx->cmd_args; + if (!ctx->master) + return 0; - if (cmd_args->mac_compat == GF_OPTION_DISABLE) - return 0; + ret = glusterfs_graph_insert(graph, ctx, "meta", "meta-autoload", 1); + return ret; +} - ret = glusterfs_graph_insert (graph, ctx, "features/mac-compat", - "mac-compat-autoload"); +int +glusterfs_graph_mac_compat(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +{ + int ret = 0; + cmd_args_t *cmd_args = NULL; - return ret; + cmd_args = &ctx->cmd_args; + + if (cmd_args->mac_compat == GF_OPTION_DISABLE) + return 0; + + ret = glusterfs_graph_insert(graph, ctx, "features/mac-compat", + "mac-compat-autoload", 1); + + return ret; } +int +glusterfs_graph_gfid_access(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +{ + int ret = 0; + cmd_args_t *cmd_args = NULL; + + cmd_args = &ctx->cmd_args; + + if (!cmd_args->aux_gfid_mount) + return 0; + + ret = glusterfs_graph_insert(graph, ctx, "features/gfid-access", + "gfid-access-autoload", 1); + return ret; +} static void -gf_add_cmdline_options (glusterfs_graph_t *graph, cmd_args_t *cmd_args) -{ - int ret = 0; - xlator_t *trav = NULL; - xlator_cmdline_option_t *cmd_option = NULL; - - trav = graph->first; - - while (trav) { - list_for_each_entry (cmd_option, - &cmd_args->xlator_options, cmd_args) { - if (!fnmatch (cmd_option->volume, - trav->name, FNM_NOESCAPE)) { - ret = dict_set_str (trav->options, - cmd_option->key, - cmd_option->value); - if (ret == 0) { - gf_log (trav->name, GF_LOG_INFO, - "adding option '%s' for " - "volume '%s' with value '%s'", - cmd_option->key, trav->name, - cmd_option->value); - } else { - gf_log (trav->name, GF_LOG_WARNING, - "adding option '%s' for " - "volume '%s' failed: %s", - cmd_option->key, trav->name, - strerror (-ret)); - } - } +gf_add_cmdline_options(glusterfs_graph_t *graph, cmd_args_t *cmd_args) +{ + int ret = 0; + xlator_t *trav = NULL; + xlator_cmdline_option_t *cmd_option = NULL; + + trav = graph->first; + + while (trav) { + list_for_each_entry(cmd_option, &cmd_args->xlator_options, cmd_args) + { + if (!fnmatch(cmd_option->volume, trav->name, FNM_NOESCAPE)) { + ret = dict_set_str(trav->options, cmd_option->key, + cmd_option->value); + if (ret == 0) { + gf_msg(trav->name, GF_LOG_TRACE, 0, LG_MSG_VOL_OPTION_ADD, + "adding option '%s' for " + "volume '%s' with value '%s'", + cmd_option->key, trav->name, cmd_option->value); + } else { + gf_msg(trav->name, GF_LOG_WARNING, -ret, + LG_MSG_VOL_OPTION_ADD, + "adding option '%s' for " + "volume '%s' failed", + cmd_option->key, trav->name); } - trav = trav->next; + } } + trav = trav->next; + } } - int -glusterfs_graph_validate_options (glusterfs_graph_t *graph) +glusterfs_graph_validate_options(glusterfs_graph_t *graph) { - xlator_t *trav = NULL; - int ret = -1; - char *errstr = NULL; + xlator_t *trav = NULL; + int ret = -1; + char *errstr = NULL; - trav = graph->first; + trav = graph->first; - while (trav) { - if (list_empty (&trav->volume_options)) - continue; + while (trav) { + if (list_empty(&trav->volume_options)) { + trav = trav->next; + continue; + } - ret = xlator_options_validate (trav, trav->options, &errstr); - if (ret) { - gf_log (trav->name, GF_LOG_ERROR, - "validation failed: %s", errstr); - return ret; - } - trav = trav->next; + ret = xlator_options_validate(trav, trav->options, &errstr); + if (ret) { + gf_msg(trav->name, GF_LOG_ERROR, 0, LG_MSG_VALIDATION_FAILED, + "validation failed: " + "%s", + errstr); + return ret; } + trav = trav->next; + } - return 0; + return 0; } - int -glusterfs_graph_init (glusterfs_graph_t *graph) +glusterfs_graph_init(glusterfs_graph_t *graph) { - xlator_t *trav = NULL; - int ret = -1; + xlator_t *trav = NULL; + int ret = -1; - trav = graph->first; + trav = graph->first; - while (trav) { - ret = xlator_init (trav); - if (ret) { - gf_log (trav->name, GF_LOG_ERROR, - "initializing translator failed"); - return ret; - } - trav = trav->next; + while (trav) { + ret = xlator_init(trav); + if (ret) { + gf_msg(trav->name, GF_LOG_ERROR, 0, LG_MSG_TRANSLATOR_INIT_FAILED, + "initializing translator failed"); + return ret; } + trav = trav->next; + } - return 0; + return 0; } +int +glusterfs_graph_deactivate(glusterfs_graph_t *graph) +{ + xlator_t *top = NULL; -static void -_log_if_unknown_option (dict_t *dict, char *key, data_t *value, void *data) + if (graph == NULL) + goto out; + + top = graph->top; + xlator_tree_fini(top); +out: + return 0; +} + +static int +_log_if_unknown_option(dict_t *dict, char *key, data_t *value, void *data) { - volume_option_t *found = NULL; - xlator_t *xl = NULL; + volume_option_t *found = NULL; + xlator_t *xl = NULL; - xl = data; + xl = data; - found = xlator_volume_option_get (xl, key); + found = xlator_volume_option_get(xl, key); - if (!found) { - gf_log (xl->name, GF_LOG_WARNING, - "option '%s' is not recognized", key); - } + if (!found) { + gf_msg(xl->name, GF_LOG_DEBUG, 0, LG_MSG_XLATOR_OPTION_INVALID, + "option '%s' is not recognized", key); + } - return; + return 0; } - static void -_xlator_check_unknown_options (xlator_t *xl, void *data) +_xlator_check_unknown_options(xlator_t *xl, void *data) { - dict_foreach (xl->options, _log_if_unknown_option, xl); + dict_foreach(xl->options, _log_if_unknown_option, xl); } - -int -glusterfs_graph_unknown_options (glusterfs_graph_t *graph) +static int +glusterfs_graph_unknown_options(glusterfs_graph_t *graph) { - xlator_foreach (graph->first, _xlator_check_unknown_options, NULL); - return 0; + xlator_foreach(graph->first, _xlator_check_unknown_options, NULL); + return 0; } +static void +fill_uuid(char *uuid, int size, struct timeval tv) +{ + char hostname[50] = { + 0, + }; + char now_str[GF_TIMESTR_SIZE]; -void -fill_uuid (char *uuid, int size) + if (gethostname(hostname, sizeof(hostname) - 1) != 0) { + gf_msg("graph", GF_LOG_ERROR, errno, LG_MSG_GETHOSTNAME_FAILED, + "gethostname failed"); + hostname[sizeof(hostname) - 1] = '\0'; + } + + gf_time_fmt_tv(now_str, sizeof now_str, &tv, gf_timefmt_dirent); + snprintf(uuid, size, "%s-%d-%s", hostname, getpid(), now_str); + + return; +} + +static int +glusterfs_graph_settop(glusterfs_graph_t *graph, char *volume_name, + gf_boolean_t exact_match) { - char hostname[256] = {0,}; - struct timeval tv = {0,}; - struct tm now = {0, }; - char now_str[32]; + int ret = -1; + xlator_t *trav = NULL; - if (gettimeofday (&tv, NULL) == -1) { - gf_log ("graph", GF_LOG_ERROR, - "gettimeofday: failed %s", - strerror (errno)); + if (!volume_name || !exact_match) { + graph->top = graph->first; + ret = 0; + } else { + for (trav = graph->first; trav; trav = trav->next) { + if (strcmp(trav->name, volume_name) == 0) { + graph->top = trav; + ret = 0; + break; + } } + } + + return ret; +} + +int +glusterfs_graph_parent_up(glusterfs_graph_t *graph) +{ + xlator_t *trav = NULL; + int ret = -1; - if (gethostname (hostname, 256) == -1) { - gf_log ("graph", GF_LOG_ERROR, - "gethostname: failed %s", - strerror (errno)); + trav = graph->first; + + while (trav) { + if (!xlator_has_parent(trav)) { + ret = xlator_notify(trav, GF_EVENT_PARENT_UP, trav); } - localtime_r (&tv.tv_sec, &now); - strftime (now_str, 32, "%Y/%m/%d-%H:%M:%S", &now); - snprintf (uuid, size, "%s-%d-%s:%"GF_PRI_SUSECONDS, - hostname, getpid(), now_str, tv.tv_usec); + if (ret) + break; - return; -} + trav = trav->next; + } + return ret; +} int -glusterfs_graph_settop (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +glusterfs_graph_prepare(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx, + char *volume_name) { - const char *volume_name = NULL; - xlator_t *trav = NULL; + xlator_t *trav = NULL; + int ret = 0; + + /* XXX: CHECKSUM */ + + /* XXX: attach to -n volname */ + /* A '/' in the volume name suggests brick multiplexing is used, find + * the top of the (sub)graph. The volname MUST match the subvol in this + * case. In other cases (like for gfapi) the default top for the + * (sub)graph is ok. */ + if (!volume_name) { + /* GlusterD does not pass a volume_name */ + ret = glusterfs_graph_settop(graph, volume_name, _gf_false); + } else if (strncmp(volume_name, "/snaps/", 7) == 0) { + /* snap shots have their top xlator named like "/snaps/..." */ + ret = glusterfs_graph_settop(graph, volume_name, _gf_false); + } else if (volume_name[0] == '/') { + /* brick multiplexing passes the brick path */ + ret = glusterfs_graph_settop(graph, volume_name, _gf_true); + } else { + ret = glusterfs_graph_settop(graph, volume_name, _gf_false); + } + + if (ret) { + gf_msg("graph", GF_LOG_ERROR, EINVAL, LG_MSG_GRAPH_ERROR, + "glusterfs graph settop failed"); + errno = EINVAL; + return -1; + } - volume_name = ctx->cmd_args.volume_name; + /* XXX: WORM VOLUME */ + ret = glusterfs_graph_worm(graph, ctx); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_ERROR, + "glusterfs graph worm failed"); + return -1; + } + ret = glusterfs_graph_acl(graph, ctx); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_ERROR, + "glusterfs graph ACL failed"); + return -1; + } - if (!volume_name) { - graph->top = graph->first; - return 0; - } + /* XXX: MAC COMPAT */ + ret = glusterfs_graph_mac_compat(graph, ctx); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_ERROR, + "glusterfs graph mac compat failed"); + return -1; + } - for (trav = graph->first; trav; trav = trav->next) { - if (strcmp (trav->name, volume_name) == 0) { - graph->top = trav; - return 0; - } - } + /* XXX: gfid-access */ + ret = glusterfs_graph_gfid_access(graph, ctx); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_ERROR, + "glusterfs graph 'gfid-access' failed"); + return -1; + } + /* XXX: topmost xlator */ + ret = glusterfs_graph_meta(graph, ctx); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_ERROR, + "glusterfs graph meta failed"); return -1; + } + + /* XXX: this->ctx setting */ + for (trav = graph->first; trav; trav = trav->next) { + trav->ctx = ctx; + } + + /* XXX: DOB setting */ + gettimeofday(&graph->dob, NULL); + + fill_uuid(graph->graph_uuid, sizeof(graph->graph_uuid), graph->dob); + + graph->id = ctx->graph_id++; + + /* XXX: --xlator-option additions */ + gf_add_cmdline_options(graph, &ctx->cmd_args); + + return 0; +} + +static xlator_t * +glusterfs_root(glusterfs_graph_t *graph) +{ + return graph->first; +} + +static int +glusterfs_is_leaf(xlator_t *xl) +{ + int ret = 0; + + if (!xl->children) + ret = 1; + + return ret; +} + +static uint32_t +glusterfs_count_leaves(xlator_t *xl) +{ + int n = 0; + xlator_list_t *list = NULL; + + if (glusterfs_is_leaf(xl)) + n = 1; + else + for (list = xl->children; list; list = list->next) + n += glusterfs_count_leaves(list->xlator); + + return n; +} + +int +glusterfs_get_leaf_count(glusterfs_graph_t *graph) +{ + return graph->leaf_count; } +static int +_glusterfs_leaf_position(xlator_t *tgt, int *id, xlator_t *xl) +{ + xlator_list_t *list = NULL; + int found = 0; + + if (xl == tgt) + found = 1; + else if (glusterfs_is_leaf(xl)) + *id += 1; + else + for (list = xl->children; !found && list; list = list->next) + found = _glusterfs_leaf_position(tgt, id, list->xlator); + + return found; +} int -glusterfs_graph_parent_up (glusterfs_graph_t *graph) +glusterfs_leaf_position(xlator_t *tgt) { - xlator_t *trav = NULL; - int ret = -1; + xlator_t *root = NULL; + int pos = 0; - trav = graph->first; + root = glusterfs_root(tgt->graph); - while (trav) { - if (!xlator_has_parent (trav)) { - ret = xlator_notify (trav, GF_EVENT_PARENT_UP, trav); - } + if (!_glusterfs_leaf_position(tgt, &pos, root)) + pos = -1; + + return pos; +} + +static int +_glusterfs_reachable_leaves(xlator_t *base, xlator_t *xl, dict_t *leaves) +{ + xlator_list_t *list = NULL; + int err = 1; + int pos = 0; + char *strpos = NULL; + + if (glusterfs_is_leaf(xl)) { + pos = glusterfs_leaf_position(xl); + if (pos < 0) + goto out; - if (ret) - break; + err = gf_asprintf(&strpos, "%d", pos); - trav = trav->next; + if (err >= 0) { + err = dict_set_static_ptr(leaves, strpos, base); + GF_FREE(strpos); } + } else { + for (err = 0, list = xl->children; !err && list; list = list->next) + err = _glusterfs_reachable_leaves(base, list->xlator, leaves); + } - return ret; +out: + return err; } +/* + * This function determines which leaves are children (or grandchildren) + * of the given base. The base may have multiple sub volumes. Each sub + * volumes in turn may have sub volumes.. until the leaves are reached. + * Each leaf is numbered 1,2,3,...etc. + * + * The base translator calls this function to see which of *its* subvolumes + * it would forward an FOP to, to *get to* a particular leaf. + * That information is built into the "leaves" dictionary. + * key:destination leaf# -> value:base subvolume xlator. + */ + +int +glusterfs_reachable_leaves(xlator_t *base, dict_t *leaves) +{ + xlator_list_t *list = NULL; + int err = 0; + + for (list = base->children; !err && list; list = list->next) + err = _glusterfs_reachable_leaves(list->xlator, list->xlator, leaves); + + return err; +} int -glusterfs_graph_prepare (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +glusterfs_graph_activate(glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) { - xlator_t *trav = NULL; - int ret = 0; + int ret = 0; + xlator_t *root = NULL; - /* XXX: CHECKSUM */ + root = glusterfs_root(graph); - /* XXX: attach to -n volname */ - ret = glusterfs_graph_settop (graph, ctx); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "glusterfs graph settop failed"); - return -1; - } + graph->leaf_count = glusterfs_count_leaves(root); - /* XXX: WORM VOLUME */ - ret = glusterfs_graph_worm (graph, ctx); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "glusterfs graph worm failed"); - return -1; - } - ret = glusterfs_graph_acl (graph, ctx); + /* XXX: all xlator options validation */ + ret = glusterfs_graph_validate_options(graph); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_VALIDATION_FAILED, + "validate options failed"); + return ret; + } + + /* XXX: perform init () */ + ret = glusterfs_graph_init(graph); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_GRAPH_INIT_FAILED, + "init failed"); + return ret; + } + + ret = glusterfs_graph_unknown_options(graph); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_UNKNOWN_OPTIONS_FAILED, + "unknown options " + "failed"); + return ret; + } + + /* XXX: log full graph (_gf_dump_details) */ + + list_add(&graph->list, &ctx->graphs); + ctx->active = graph; + + /* XXX: attach to master and set active pointer */ + if (ctx->master) { + ret = xlator_notify(ctx->master, GF_EVENT_GRAPH_NEW, graph); if (ret) { - gf_log ("graph", GF_LOG_ERROR, "glusterfs graph ACL failed"); - return -1; + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_EVENT_NOTIFY_FAILED, + "graph new notification failed"); + return ret; } + ((xlator_t *)ctx->master)->next = graph->top; + } + + /* XXX: perform parent up */ + ret = glusterfs_graph_parent_up(graph); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_EVENT_NOTIFY_FAILED, + "parent up notification failed"); + return ret; + } + + return 0; +} + +int +xlator_equal_rec(xlator_t *xl1, xlator_t *xl2) +{ + xlator_list_t *trav1 = NULL; + xlator_list_t *trav2 = NULL; + int ret = 0; - /* XXX: MAC COMPAT */ - ret = glusterfs_graph_mac_compat (graph, ctx); + if (xl1 == NULL || xl2 == NULL) { + gf_msg_debug("xlator", 0, "invalid argument"); + return -1; + } + + trav1 = xl1->children; + trav2 = xl2->children; + + while (trav1 && trav2) { + ret = xlator_equal_rec(trav1->xlator, trav2->xlator); if (ret) { - gf_log ("graph", GF_LOG_ERROR, "glusterfs graph mac compat failed"); - return -1; + gf_msg_debug("glusterfsd-mgmt", 0, + "xlators children " + "not equal"); + goto out; } - /* XXX: this->ctx setting */ - for (trav = graph->first; trav; trav = trav->next) { - trav->ctx = ctx; - } + trav1 = trav1->next; + trav2 = trav2->next; + } + + if (trav1 || trav2) { + ret = -1; + goto out; + } + + if (strcmp(xl1->name, xl2->name)) { + ret = -1; + goto out; + } + + /* type could have changed even if xlator names match, + e.g cluster/distribute and cluster/nufa share the same + xlator name + */ + if (strcmp(xl1->type, xl2->type)) { + ret = -1; + goto out; + } +out: + return ret; +} - /* XXX: DOB setting */ - gettimeofday (&graph->dob, NULL); +gf_boolean_t +is_graph_topology_equal(glusterfs_graph_t *graph1, glusterfs_graph_t *graph2) +{ + xlator_t *trav1 = NULL; + xlator_t *trav2 = NULL; + gf_boolean_t ret = _gf_true; + xlator_list_t *ltrav; + + trav1 = graph1->first; + trav2 = graph2->first; + + if (strcmp(trav2->type, "protocol/server") == 0) { + trav2 = trav2->children->xlator; + for (ltrav = trav1->children; ltrav; ltrav = ltrav->next) { + trav1 = ltrav->xlator; + if (!trav1->cleanup_starting && !strcmp(trav1->name, trav2->name)) { + break; + } + } + if (!ltrav) { + return _gf_false; + } + } - fill_uuid (graph->graph_uuid, 128); + ret = xlator_equal_rec(trav1, trav2); - graph->id = ctx->graph_id++; + if (ret) { + gf_msg_debug("glusterfsd-mgmt", 0, "graphs are not equal"); + ret = _gf_false; + goto out; + } - /* XXX: --xlator-option additions */ - gf_add_cmdline_options (graph, &ctx->cmd_args); + ret = _gf_true; + gf_msg_debug("glusterfsd-mgmt", 0, "graphs are equal"); +out: + return ret; +} - return 0; +/* Function has 3types of return value 0, -ve , 1 + * return 0 =======> reconfiguration of options has succeeded + * return 1 =======> the graph has to be reconstructed and all the + * xlators should be inited return -1(or -ve) =======> Some Internal Error + * occurred during the operation + */ +int +glusterfs_volfile_reconfigure(FILE *newvolfile_fp, glusterfs_ctx_t *ctx) +{ + glusterfs_graph_t *oldvolfile_graph = NULL; + glusterfs_graph_t *newvolfile_graph = NULL; + + int ret = -1; + + if (!ctx) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, 0, LG_MSG_CTX_NULL, + "ctx is NULL"); + goto out; + } + + oldvolfile_graph = ctx->active; + if (!oldvolfile_graph) { + ret = 1; + goto out; + } + + newvolfile_graph = glusterfs_graph_construct(newvolfile_fp); + + if (!newvolfile_graph) { + goto out; + } + + glusterfs_graph_prepare(newvolfile_graph, ctx, ctx->cmd_args.volume_name); + + if (!is_graph_topology_equal(oldvolfile_graph, newvolfile_graph)) { + ret = 1; + gf_msg_debug("glusterfsd-mgmt", 0, + "Graph topology not " + "equal(should call INIT)"); + goto out; + } + + gf_msg_debug("glusterfsd-mgmt", 0, + "Only options have changed in the" + " new graph"); + + ret = glusterfs_graph_reconfigure(oldvolfile_graph, newvolfile_graph); + if (ret) { + gf_msg_debug("glusterfsd-mgmt", 0, + "Could not reconfigure " + "new options in old graph"); + goto out; + } + + ret = 0; +out: + + if (newvolfile_graph) + glusterfs_graph_destroy(newvolfile_graph); + + return ret; } +/* This function need to remove. This added to support gfapi volfile + * reconfigure. + */ int -glusterfs_graph_activate (glusterfs_graph_t *graph, glusterfs_ctx_t *ctx) +gf_volfile_reconfigure(int oldvollen, FILE *newvolfile_fp, glusterfs_ctx_t *ctx, + const char *oldvolfile) { - int ret = 0; + glusterfs_graph_t *oldvolfile_graph = NULL; + glusterfs_graph_t *newvolfile_graph = NULL; + FILE *oldvolfile_fp = NULL; + /*Since the function mkstemp() replaces XXXXXX, + * assigning it to a variable + */ + char temp_file[] = "/tmp/temp_vol_file_XXXXXX"; + gf_boolean_t active_graph_found = _gf_true; + + int ret = -1; + int u_ret = -1; + int file_desc = -1; + + if (!oldvollen) { + ret = 1; // Has to call INIT for the whole graph + goto out; + } + + if (!ctx) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, 0, LG_MSG_CTX_NULL, + "ctx is NULL"); + goto out; + } + + oldvolfile_graph = ctx->active; + if (!oldvolfile_graph) { + active_graph_found = _gf_false; + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, 0, LG_MSG_ACTIVE_GRAPH_NULL, + "glusterfs_ctx->active is NULL"); + + /* coverity[secure_temp] mkstemp uses 0600 as the mode and is safe */ + file_desc = mkstemp(temp_file); + if (file_desc < 0) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, errno, + LG_MSG_TMPFILE_CREATE_FAILED, + "Unable to " + "create temporary volfile"); + goto out; + } - /* XXX: all xlator options validation */ - ret = glusterfs_graph_validate_options (graph); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "validate options failed"); - return ret; + /*Calling unlink so that when the file is closed or program + *terminates the tempfile is deleted. + */ + u_ret = sys_unlink(temp_file); + + if (u_ret < 0) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, errno, + LG_MSG_TMPFILE_DELETE_FAILED, + "Temporary file" + " delete failed."); + sys_close(file_desc); + goto out; } - /* XXX: perform init () */ - ret = glusterfs_graph_init (graph); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "init failed"); - return ret; + oldvolfile_fp = fdopen(file_desc, "w+b"); + if (!oldvolfile_fp) + goto out; + + fwrite(oldvolfile, oldvollen, 1, oldvolfile_fp); + fflush(oldvolfile_fp); + if (ferror(oldvolfile_fp)) { + goto out; } - ret = glusterfs_graph_unknown_options (graph); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "unknown options failed"); - return ret; + oldvolfile_graph = glusterfs_graph_construct(oldvolfile_fp); + if (!oldvolfile_graph) + goto out; + } + + newvolfile_graph = glusterfs_graph_construct(newvolfile_fp); + if (!newvolfile_graph) { + goto out; + } + + glusterfs_graph_prepare(newvolfile_graph, ctx, ctx->cmd_args.volume_name); + + if (!is_graph_topology_equal(oldvolfile_graph, newvolfile_graph)) { + ret = 1; + gf_msg_debug("glusterfsd-mgmt", 0, + "Graph topology not " + "equal(should call INIT)"); + goto out; + } + + gf_msg_debug("glusterfsd-mgmt", 0, + "Only options have changed in the" + " new graph"); + + /* */ + ret = glusterfs_graph_reconfigure(oldvolfile_graph, newvolfile_graph); + if (ret) { + gf_msg_debug("glusterfsd-mgmt", 0, + "Could not reconfigure " + "new options in old graph"); + goto out; + } + + ret = 0; +out: + if (oldvolfile_fp) + fclose(oldvolfile_fp); + + /* Do not simply destroy the old graph here. If the oldgraph + is constructed here in this function itself instead of getting + it from ctx->active (which happens only of ctx->active is NULL), + then destroy the old graph. If some i/o is still happening in + the old graph and the old graph is obtained from ctx->active, + then destroying the graph will cause problems. + */ + if (!active_graph_found && oldvolfile_graph) + glusterfs_graph_destroy(oldvolfile_graph); + if (newvolfile_graph) + glusterfs_graph_destroy(newvolfile_graph); + + return ret; +} + +int +glusterfs_graph_reconfigure(glusterfs_graph_t *oldgraph, + glusterfs_graph_t *newgraph) +{ + xlator_t *old_xl = NULL; + xlator_t *new_xl = NULL; + xlator_list_t *trav; + + GF_ASSERT(oldgraph); + GF_ASSERT(newgraph); + + old_xl = oldgraph->first; + while (old_xl->is_autoloaded) { + old_xl = old_xl->children->xlator; + } + + new_xl = newgraph->first; + while (new_xl->is_autoloaded) { + new_xl = new_xl->children->xlator; + } + + if (strcmp(old_xl->type, "protocol/server") != 0) { + return xlator_tree_reconfigure(old_xl, new_xl); + } + + /* Some options still need to be handled by the server translator. */ + if (old_xl->reconfigure) { + old_xl->reconfigure(old_xl, new_xl->options); + } + + (void)copy_opts_to_child(new_xl, FIRST_CHILD(new_xl), "*auth*"); + new_xl = FIRST_CHILD(new_xl); + + for (trav = old_xl->children; trav; trav = trav->next) { + if (!trav->xlator->cleanup_starting && + !strcmp(trav->xlator->name, new_xl->name)) { + return xlator_tree_reconfigure(trav->xlator, new_xl); } + } - /* XXX: log full graph (_gf_dump_details) */ + return -1; +} - list_add (&graph->list, &ctx->graphs); - ctx->active = graph; +int +glusterfs_graph_destroy_residual(glusterfs_graph_t *graph) +{ + int ret = -1; - /* XXX: attach to master and set active pointer */ - if (ctx->master) - ret = xlator_notify (ctx->master, GF_EVENT_GRAPH_NEW, graph); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "graph new notification failed"); - return ret; + if (graph == NULL) + return ret; + + ret = xlator_tree_free_memacct(graph->first); + + list_del_init(&graph->list); + pthread_mutex_destroy(&graph->mutex); + pthread_cond_destroy(&graph->child_down_cond); + GF_FREE(graph); + + return ret; +} + +/* This function destroys all the xlator members except for the + * xlator strcuture and its mem accounting field. + * + * If otherwise, it would destroy the master xlator object as well + * its mem accounting, which would mean after calling glusterfs_graph_destroy() + * there cannot be any reference to GF_FREE() from the master xlator, this is + * not possible because of the following dependencies: + * - glusterfs_ctx_t will have mem pools allocated by the master xlators + * - xlator objects will have references to those mem pools(g: dict) + * + * Ordering the freeing in any of the order will also not solve the dependency: + * - Freeing xlator objects(including memory accounting) before mem pools + * destruction will mean not use GF_FREE while destroying mem pools. + * - Freeing mem pools and then destroying xlator objects would lead to crashes + * when xlator tries to unref dict or other mem pool objects. + * + * Hence the way chosen out of this interdependency is to split xlator object + * free into two stages: + * - Free all the xlator members excpet for its mem accounting structure + * - Free all the mem accouting structures of xlator along with the xlator + * object itself. + */ +int +glusterfs_graph_destroy(glusterfs_graph_t *graph) +{ + int ret = 0; + + GF_VALIDATE_OR_GOTO("graph", graph, out); + + ret = xlator_tree_free_members(graph->first); + + ret = glusterfs_graph_destroy_residual(graph); +out: + return ret; +} + +int +glusterfs_graph_fini(glusterfs_graph_t *graph) +{ + xlator_t *trav = NULL; + + trav = graph->first; + + while (trav) { + if (trav->init_succeeded) { + trav->cleanup_starting = 1; + trav->fini(trav); + if (trav->local_pool) { + mem_pool_destroy(trav->local_pool); + trav->local_pool = NULL; + } + if (trav->itable) { + inode_table_destroy(trav->itable); + trav->itable = NULL; + } + trav->init_succeeded = 0; } + trav = trav->next; + } - /* XXX: perform parent up */ - ret = glusterfs_graph_parent_up (graph); - if (ret) { - gf_log ("graph", GF_LOG_ERROR, "parent up notification failed"); - return ret; + return 0; +} + +int +glusterfs_graph_attach(glusterfs_graph_t *orig_graph, char *path, + glusterfs_graph_t **newgraph) +{ + xlator_t *this = THIS; + FILE *fp; + glusterfs_graph_t *graph; + xlator_t *xl; + char *volfile_id = NULL; + char *volfile_content = NULL; + struct stat stbuf = { + 0, + }; + size_t file_len = -1; + gf_volfile_t *volfile_obj = NULL; + int ret = -1; + char sha256_hash[SHA256_DIGEST_LENGTH] = { + 0, + }; + + if (!orig_graph) { + return -EINVAL; + } + + ret = sys_stat(path, &stbuf); + if (ret < 0) { + gf_log(THIS->name, GF_LOG_ERROR, "Unable to stat %s (%s)", path, + strerror(errno)); + return -EINVAL; + } + + file_len = stbuf.st_size; + volfile_content = GF_MALLOC(file_len + 1, gf_common_mt_char); + if (!volfile_content) + return -ENOMEM; + + fp = fopen(path, "r"); + if (!fp) { + gf_log(THIS->name, GF_LOG_WARNING, "oops, %s disappeared on us", path); + GF_FREE(volfile_content); + return -EIO; + } + + ret = fread(volfile_content, sizeof(char), file_len, fp); + if (ret == file_len) { + glusterfs_compute_sha256((const unsigned char *)volfile_content, + file_len, sha256_hash); + } else { + gf_log(THIS->name, GF_LOG_ERROR, + "read failed on path %s. File size=%" GF_PRI_SIZET + "read size=%d", + path, file_len, ret); + GF_FREE(volfile_content); + fclose(fp); + return -EIO; + } + + GF_FREE(volfile_content); + + graph = glusterfs_graph_construct(fp); + fclose(fp); + if (!graph) { + gf_log(this->name, GF_LOG_WARNING, "could not create graph from %s", + path); + return -EIO; + } + + /* + * If there's a server translator on top, we want whatever's below + * that. + */ + xl = graph->first; + if (strcmp(xl->type, "protocol/server") == 0) { + (void)copy_opts_to_child(xl, FIRST_CHILD(xl), "*auth*"); + xl = FIRST_CHILD(xl); + } + graph->first = xl; + *newgraph = graph; + + volfile_id = strstr(path, "/snaps/"); + if (!volfile_id) { + volfile_id = rindex(path, '/'); + if (volfile_id) { + ++volfile_id; + } + } + if (volfile_id) { + xl->volfile_id = gf_strdup(volfile_id); + /* There's a stray ".vol" at the end. */ + xl->volfile_id[strlen(xl->volfile_id) - 4] = '\0'; + } + + /* TODO memory leaks everywhere need to free graph in case of error */ + if (glusterfs_graph_prepare(graph, this->ctx, xl->name)) { + gf_log(this->name, GF_LOG_WARNING, + "failed to prepare graph for xlator %s", xl->name); + return -EIO; + } else if (glusterfs_graph_init(graph)) { + gf_log(this->name, GF_LOG_WARNING, + "failed to initialize graph for xlator %s", xl->name); + return -EIO; + } else if (glusterfs_xlator_link(orig_graph->top, graph->top)) { + gf_log(this->name, GF_LOG_WARNING, + "failed to link the graphs for xlator %s ", xl->name); + return -EIO; + } + + if (!volfile_obj) { + volfile_obj = GF_CALLOC(1, sizeof(gf_volfile_t), gf_common_volfile_t); + if (!volfile_obj) { + return -EIO; } + } + + INIT_LIST_HEAD(&volfile_obj->volfile_list); + snprintf(volfile_obj->vol_id, sizeof(volfile_obj->vol_id), "%s", + xl->volfile_id); + memcpy(volfile_obj->volfile_checksum, sha256_hash, + sizeof(volfile_obj->volfile_checksum)); + list_add(&volfile_obj->volfile_list, &this->ctx->volfile_list); + + return 0; +} +int +glusterfs_muxsvc_cleanup_parent(glusterfs_ctx_t *ctx, + glusterfs_graph_t *parent_graph) +{ + if (parent_graph) { + if (parent_graph->first) { + xlator_destroy(parent_graph->first); + } + ctx->active = NULL; + GF_FREE(parent_graph); + parent_graph = NULL; + } + return 0; +} +void * +glusterfs_graph_cleanup(void *arg) +{ + glusterfs_graph_t *graph = NULL; + glusterfs_ctx_t *ctx = THIS->ctx; + int ret = -1; + graph = arg; + + if (!graph) + return NULL; + + /* To destroy the graph, fitst sent a GF_EVENT_PARENT_DOWN + * Then wait for GF_EVENT_CHILD_DOWN to get on the top + * xl. Once we have GF_EVENT_CHILD_DOWN event, then proceed + * to fini. + * + * During fini call, this will take a last unref on rpc and + * rpc_transport_object. + */ + if (graph->first) + default_notify(graph->first, GF_EVENT_PARENT_DOWN, graph->first); + + ret = pthread_mutex_lock(&graph->mutex); + if (ret != 0) { + gf_msg("glusterfs", GF_LOG_ERROR, EAGAIN, LG_MSG_GRAPH_CLEANUP_FAILED, + "Failed to acquire a lock"); + goto out; + } + /* check and wait for CHILD_DOWN for top xlator*/ + while (graph->used) { + ret = pthread_cond_wait(&graph->child_down_cond, &graph->mutex); + if (ret != 0) + gf_msg("glusterfs", GF_LOG_INFO, 0, LG_MSG_GRAPH_CLEANUP_FAILED, + "cond wait failed "); + } + + ret = pthread_mutex_unlock(&graph->mutex); + if (ret != 0) { + gf_msg("glusterfs", GF_LOG_ERROR, EAGAIN, LG_MSG_GRAPH_CLEANUP_FAILED, + "Failed to release a lock"); + } + + /* Though we got a child down on top xlator, we have to wait until + * all the notifier to exit. Because there should not be any threads + * that access xl variables. + */ + pthread_mutex_lock(&ctx->notify_lock); + { + while (ctx->notifying) + pthread_cond_wait(&ctx->notify_cond, &ctx->notify_lock); + } + pthread_mutex_unlock(&ctx->notify_lock); + + pthread_mutex_lock(&ctx->cleanup_lock); + { + glusterfs_graph_fini(graph); + glusterfs_graph_destroy(graph); + } + pthread_mutex_unlock(&ctx->cleanup_lock); +out: + return NULL; +} + +glusterfs_graph_t * +glusterfs_muxsvc_setup_parent_graph(glusterfs_ctx_t *ctx, char *name, + char *type) +{ + glusterfs_graph_t *parent_graph = NULL; + xlator_t *ixl = NULL; + int ret = -1; + parent_graph = GF_CALLOC(1, sizeof(*parent_graph), + gf_common_mt_glusterfs_graph_t); + if (!parent_graph) + goto out; + + INIT_LIST_HEAD(&parent_graph->list); + + ctx->active = parent_graph; + ixl = GF_CALLOC(1, sizeof(*ixl), gf_common_mt_xlator_t); + if (!ixl) + goto out; + + ixl->ctx = ctx; + ixl->graph = parent_graph; + ixl->options = dict_new(); + if (!ixl->options) + goto out; + + ixl->name = gf_strdup(name); + if (!ixl->name) + goto out; + + ixl->is_autoloaded = 1; + + if (xlator_set_type(ixl, type) == -1) { + gf_msg("glusterfs", GF_LOG_ERROR, EINVAL, LG_MSG_GRAPH_SETUP_FAILED, + "%s (%s) set type failed", name, type); + goto out; + } + + glusterfs_graph_set_first(parent_graph, ixl); + parent_graph->top = ixl; + ixl = NULL; + + gettimeofday(&parent_graph->dob, NULL); + fill_uuid(parent_graph->graph_uuid, 128, parent_graph->dob); + parent_graph->id = ctx->graph_id++; + ret = 0; +out: + if (ixl) + xlator_destroy(ixl); + + if (ret) { + glusterfs_muxsvc_cleanup_parent(ctx, parent_graph); + parent_graph = NULL; + } + return parent_graph; +} + +int +glusterfs_svc_mux_pidfile_cleanup(gf_volfile_t *volfile_obj) +{ + if (!volfile_obj || !volfile_obj->pidfp) return 0; + + gf_msg_trace("glusterfsd", 0, "pidfile %s cleanup", volfile_obj->vol_id); + + lockf(fileno(volfile_obj->pidfp), F_ULOCK, 0); + fclose(volfile_obj->pidfp); + volfile_obj->pidfp = NULL; + + return 0; } int -glusterfs_graph_reconfigure (glusterfs_graph_t *oldgraph, - glusterfs_graph_t *newgraph) +glusterfs_process_svc_detach(glusterfs_ctx_t *ctx, gf_volfile_t *volfile_obj) { - xlator_t *old_xl = NULL; - xlator_t *new_xl = NULL; + xlator_t *last_xl = NULL; + glusterfs_graph_t *graph = NULL; + glusterfs_graph_t *parent_graph = NULL; + pthread_t clean_graph = { + 0, + }; + int ret = -1; + xlator_t *xl = NULL; + + if (!ctx || !ctx->active || !volfile_obj) + goto out; + + pthread_mutex_lock(&ctx->cleanup_lock); + { + parent_graph = ctx->active; + graph = volfile_obj->graph; + if (!graph) + goto unlock; + if (graph->first) + xl = graph->first; + + last_xl = graph->last_xl; + if (last_xl) + last_xl->next = NULL; + if (!xl || xl->cleanup_starting) + goto unlock; + + xl->cleanup_starting = 1; + gf_msg("mgmt", GF_LOG_INFO, 0, LG_MSG_GRAPH_DETACH_STARTED, + "detaching child %s", volfile_obj->vol_id); + + list_del_init(&volfile_obj->volfile_list); + glusterfs_mux_xlator_unlink(parent_graph->top, xl); + glusterfs_svc_mux_pidfile_cleanup(volfile_obj); + parent_graph->last_xl = glusterfs_get_last_xlator(parent_graph); + parent_graph->xl_count -= graph->xl_count; + parent_graph->leaf_count -= graph->leaf_count; + parent_graph->id++; + ret = 0; + } +unlock: + pthread_mutex_unlock(&ctx->cleanup_lock); +out: + if (!ret) { + list_del_init(&volfile_obj->volfile_list); + if (graph) { + ret = gf_thread_create_detached( + &clean_graph, glusterfs_graph_cleanup, graph, "graph_clean"); + if (ret) { + gf_msg("glusterfs", GF_LOG_ERROR, EINVAL, + LG_MSG_GRAPH_CLEANUP_FAILED, + "%s failed to create clean " + "up thread", + volfile_obj->vol_id); + ret = 0; + } + } + GF_FREE(volfile_obj); + } + return ret; +} - GF_ASSERT (oldgraph); - GF_ASSERT (newgraph); +int +glusterfs_svc_mux_pidfile_setup(gf_volfile_t *volfile_obj, const char *pid_file) +{ + int ret = -1; + FILE *pidfp = NULL; + + if (!pid_file || !volfile_obj) + goto out; + + if (volfile_obj->pidfp) { + ret = 0; + goto out; + } + pidfp = fopen(pid_file, "a+"); + if (!pidfp) { + goto out; + } + volfile_obj->pidfp = pidfp; + + ret = lockf(fileno(pidfp), F_TLOCK, 0); + if (ret) { + ret = 0; + goto out; + } +out: + return ret; +} - old_xl = oldgraph->first; - new_xl = newgraph->first; +int +glusterfs_svc_mux_pidfile_update(gf_volfile_t *volfile_obj, + const char *pid_file, pid_t pid) +{ + int ret = 0; + FILE *pidfp = NULL; + int old_pid; + + if (!volfile_obj->pidfp) { + ret = glusterfs_svc_mux_pidfile_setup(volfile_obj, pid_file); + if (ret == -1) + goto out; + } + pidfp = volfile_obj->pidfp; + ret = fscanf(pidfp, "%d", &old_pid); + if (ret <= 0) { + goto update; + } + if (old_pid == pid) { + ret = 0; + goto out; + } else { + gf_msg("mgmt", GF_LOG_INFO, 0, LG_MSG_GRAPH_ATTACH_PID_FILE_UPDATED, + "Old pid=%d found in pidfile %s. Cleaning the old pid and " + "Updating new pid=%d", + old_pid, pid_file, pid); + } +update: + ret = sys_ftruncate(fileno(pidfp), 0); + if (ret) { + gf_msg("glusterfsd", GF_LOG_ERROR, errno, + LG_MSG_GRAPH_ATTACH_PID_FILE_UPDATED, + "pidfile %s truncation failed", pid_file); + goto out; + } + + ret = fprintf(pidfp, "%d\n", pid); + if (ret <= 0) { + gf_msg("glusterfsd", GF_LOG_ERROR, errno, + LG_MSG_GRAPH_ATTACH_PID_FILE_UPDATED, "pidfile %s write failed", + pid_file); + goto out; + } + + ret = fflush(pidfp); + if (ret) { + gf_msg("glusterfsd", GF_LOG_ERROR, errno, + LG_MSG_GRAPH_ATTACH_PID_FILE_UPDATED, "pidfile %s write failed", + pid_file); + goto out; + } +out: + return ret; +} - return xlator_tree_reconfigure (old_xl, new_xl); +int +glusterfs_update_mux_pid(dict_t *dict, gf_volfile_t *volfile_obj) +{ + char *file = NULL; + int ret = -1; + + GF_VALIDATE_OR_GOTO("graph", dict, out); + GF_VALIDATE_OR_GOTO("graph", volfile_obj, out); + + ret = dict_get_str(dict, "pidfile", &file); + if (ret < 0) { + gf_msg("mgmt", GF_LOG_ERROR, EINVAL, LG_MSG_GRAPH_SETUP_FAILED, + "Failed to get pidfile from dict for volfile_id=%s", + volfile_obj->vol_id); + } + + ret = glusterfs_svc_mux_pidfile_update(volfile_obj, file, getpid()); + if (ret < 0) { + ret = -1; + gf_msg("mgmt", GF_LOG_ERROR, EINVAL, LG_MSG_GRAPH_SETUP_FAILED, + "Failed to update " + "the pidfile for volfile_id=%s", + volfile_obj->vol_id); + + goto out; + } + + if (ret == 1) + gf_msg("mgmt", GF_LOG_INFO, 0, LG_MSG_GRAPH_ATTACH_PID_FILE_UPDATED, + "PID %d updated in pidfile=%s", getpid(), file); + ret = 0; +out: + return ret; +} +int +glusterfs_process_svc_attach_volfp(glusterfs_ctx_t *ctx, FILE *fp, + char *volfile_id, char *checksum, + dict_t *dict) +{ + glusterfs_graph_t *graph = NULL; + glusterfs_graph_t *parent_graph = NULL; + glusterfs_graph_t *clean_graph = NULL; + int ret = -1; + xlator_t *xl = NULL; + xlator_t *last_xl = NULL; + gf_volfile_t *volfile_obj = NULL; + pthread_t thread_id = { + 0, + }; + + if (!ctx) + goto out; + parent_graph = ctx->active; + graph = glusterfs_graph_construct(fp); + if (!graph) { + gf_msg("glusterfsd", GF_LOG_ERROR, EINVAL, LG_MSG_GRAPH_ATTACH_FAILED, + "failed to construct the graph"); + goto out; + } + graph->parent_down = 0; + graph->last_xl = glusterfs_get_last_xlator(graph); + + for (xl = graph->first; xl; xl = xl->next) { + if (strcmp(xl->type, "mount/fuse") == 0) { + gf_msg("glusterfsd", GF_LOG_ERROR, EINVAL, + LG_MSG_GRAPH_ATTACH_FAILED, + "fuse xlator cannot be specified in volume file"); + goto out; + } + } + + graph->leaf_count = glusterfs_count_leaves(glusterfs_root(graph)); + xl = graph->first; + /* TODO memory leaks everywhere need to free graph in case of error */ + if (glusterfs_graph_prepare(graph, ctx, xl->name)) { + gf_msg("glusterfsd", GF_LOG_WARNING, EINVAL, LG_MSG_GRAPH_ATTACH_FAILED, + "failed to prepare graph for xlator %s", xl->name); + ret = -1; + goto out; + } else if (glusterfs_graph_init(graph)) { + gf_msg("glusterfsd", GF_LOG_WARNING, EINVAL, LG_MSG_GRAPH_ATTACH_FAILED, + "failed to initialize graph for xlator %s", xl->name); + ret = -1; + goto out; + } else if (glusterfs_graph_parent_up(graph)) { + gf_msg("glusterfsd", GF_LOG_WARNING, EINVAL, LG_MSG_GRAPH_ATTACH_FAILED, + "failed to link the graphs for xlator %s ", xl->name); + ret = -1; + goto out; + } + + if (!parent_graph) { + parent_graph = glusterfs_muxsvc_setup_parent_graph(ctx, "glustershd", + "debug/io-stats"); + if (!parent_graph) + goto out; + ((xlator_t *)parent_graph->top)->next = xl; + clean_graph = parent_graph; + } else { + last_xl = parent_graph->last_xl; + if (last_xl) + last_xl->next = xl; + xl->prev = last_xl; + } + parent_graph->last_xl = graph->last_xl; + + ret = glusterfs_xlator_link(parent_graph->top, xl); + if (ret) { + gf_msg("graph", GF_LOG_ERROR, 0, LG_MSG_EVENT_NOTIFY_FAILED, + "parent up notification failed"); + goto out; + } + parent_graph->xl_count += graph->xl_count; + parent_graph->leaf_count += graph->leaf_count; + parent_graph->id++; + + volfile_obj = GF_CALLOC(1, sizeof(gf_volfile_t), gf_common_volfile_t); + if (!volfile_obj) { + ret = -1; + goto out; + } + volfile_obj->pidfp = NULL; + snprintf(volfile_obj->vol_id, sizeof(volfile_obj->vol_id), "%s", + volfile_id); + + if (strcmp(ctx->cmd_args.process_name, "glustershd") == 0) { + ret = glusterfs_update_mux_pid(dict, volfile_obj); + if (ret == -1) { + GF_FREE(volfile_obj); + goto out; + } + } + + graph->used = 1; + parent_graph->id++; + list_add(&graph->list, &ctx->graphs); + INIT_LIST_HEAD(&volfile_obj->volfile_list); + volfile_obj->graph = graph; + memcpy(volfile_obj->volfile_checksum, checksum, + sizeof(volfile_obj->volfile_checksum)); + list_add_tail(&volfile_obj->volfile_list, &ctx->volfile_list); + gf_log_dump_graph(fp, graph); + graph = NULL; + + ret = 0; +out: + if (ret) { + if (graph) { + gluster_graph_take_reference(graph->first); + ret = gf_thread_create_detached(&thread_id, glusterfs_graph_cleanup, + graph, "graph_clean"); + if (ret) { + gf_msg("glusterfs", GF_LOG_ERROR, EINVAL, + LG_MSG_GRAPH_CLEANUP_FAILED, + "%s failed to create clean " + "up thread", + volfile_id); + ret = 0; + } + } + if (clean_graph) + glusterfs_muxsvc_cleanup_parent(ctx, clean_graph); + } + return ret; } int -glusterfs_graph_destroy (glusterfs_graph_t *graph) +glusterfs_mux_volfile_reconfigure(FILE *newvolfile_fp, glusterfs_ctx_t *ctx, + gf_volfile_t *volfile_obj, char *checksum, + dict_t *dict) { - return 0; + glusterfs_graph_t *oldvolfile_graph = NULL; + glusterfs_graph_t *newvolfile_graph = NULL; + char vol_id[NAME_MAX + 1]; + + int ret = -1; + + if (!ctx) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, 0, LG_MSG_CTX_NULL, + "ctx is NULL"); + goto out; + } + + /* Change the message id */ + if (!volfile_obj) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, 0, LG_MSG_CTX_NULL, + "failed to get volfile object"); + goto out; + } + + oldvolfile_graph = volfile_obj->graph; + if (!oldvolfile_graph) { + goto out; + } + + newvolfile_graph = glusterfs_graph_construct(newvolfile_fp); + + if (!newvolfile_graph) { + goto out; + } + newvolfile_graph->last_xl = glusterfs_get_last_xlator(newvolfile_graph); + + glusterfs_graph_prepare(newvolfile_graph, ctx, newvolfile_graph->first); + + if (!is_graph_topology_equal(oldvolfile_graph, newvolfile_graph)) { + ret = snprintf(vol_id, sizeof(vol_id), "%s", volfile_obj->vol_id); + if (ret < 0) + goto out; + ret = glusterfs_process_svc_detach(ctx, volfile_obj); + if (ret) { + gf_msg("glusterfsd-mgmt", GF_LOG_ERROR, EINVAL, + LG_MSG_GRAPH_CLEANUP_FAILED, + "Could not detach " + "old graph. Aborting the reconfiguration operation"); + goto out; + } + volfile_obj = NULL; + ret = glusterfs_process_svc_attach_volfp(ctx, newvolfile_fp, vol_id, + checksum, dict); + goto out; + } + + gf_msg_debug("glusterfsd-mgmt", 0, + "Only options have changed in the" + " new graph"); + + ret = glusterfs_graph_reconfigure(oldvolfile_graph, newvolfile_graph); + if (ret) { + gf_msg_debug("glusterfsd-mgmt", 0, + "Could not reconfigure " + "new options in old graph"); + goto out; + } + memcpy(volfile_obj->volfile_checksum, checksum, + sizeof(volfile_obj->volfile_checksum)); + + ret = 0; +out: + + if (newvolfile_graph) + glusterfs_graph_destroy(newvolfile_graph); + + return ret; } |