lighttpd 1.4.x https://www.lighttpd.net/
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

2970 lines
105 KiB

/*
* gw_backend - gateway backend code shared by dynamic socket backends
*
* Copyright(c) 2017 Glenn Strauss gstrauss()gluelogic.com All rights reserved
* License: BSD 3-clause (same as lighttpd)
*/
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "first.h"
#include "gw_backend.h"
#include <sys/types.h>
#include <sys/stat.h>
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "sys-socket.h"
#ifdef HAVE_SYS_UIO_H
#include <sys/uio.h>
#endif
#ifdef HAVE_SYS_WAIT_H
#include <sys/wait.h>
#endif
#include <errno.h>
#include <fcntl.h>
#include <limits.h>
#include <stdint.h>
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include <stdlib.h>
#include <signal.h>
#include <string.h>
#include <unistd.h>
#include "base.h"
#include "algo_md.h"
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "array.h"
#include "buffer.h"
#include "chunk.h"
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "fdevent.h"
#include "http_header.h"
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "log.h"
#include "sock_addr.h"
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
#include "status_counter.h"
__attribute_noinline__
static int * gw_status_get_counter(gw_host *host, gw_proc *proc, const char *tag, size_t tlen) {
/*(At the cost of some memory, could prepare strings for host and for proc
* so that here we would copy ready made string for proc (or if NULL,
* for host), and then append tag to produce key)*/
char label[288];
size_t llen = sizeof("gw.backend.")-1, len;
memcpy(label, "gw.backend.", llen);
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
len = buffer_clen(host->id);
if (len) {
force_assert(len < sizeof(label) - llen);
memcpy(label+llen, host->id->ptr, len);
llen += len;
}
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (proc) {
force_assert(llen < sizeof(label) - (LI_ITOSTRING_LENGTH + 1));
label[llen++] = '.';
len = li_utostrn(label+llen, LI_ITOSTRING_LENGTH, proc->id);
llen += len;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
force_assert(tlen < sizeof(label) - llen);
memcpy(label+llen, tag, tlen);
llen += tlen;
label[llen] = '\0';
return status_counter_get_counter(label, llen);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_proc_tag_inc(gw_host *host, gw_proc *proc, const char *tag, size_t len) {
++(*gw_status_get_counter(host, proc, tag, len));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_proc_connected_inc(gw_host *host, gw_proc *proc) {
UNUSED(host);
++(*proc->stats_connected); /* "gw.backend...connected" */
}
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
static void gw_proc_load_inc(gw_host *host, gw_proc *proc) {
*proc->stats_load = ++proc->load; /* "gw.backend...load" */
++(*host->stats_global_active); /* "gw.active-requests" */
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_proc_load_dec(gw_host *host, gw_proc *proc) {
*proc->stats_load = --proc->load; /* "gw.backend...load" */
--(*host->stats_global_active); /* "gw.active-requests" */
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_host_assign(gw_host *host) {
*host->stats_load = ++host->load; /* "gw.backend...load" */
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_host_reset(gw_host *host) {
*host->stats_load = --host->load; /* "gw.backend...load" */
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_status_init_proc(gw_host *host, gw_proc *proc) {
*gw_status_get_counter(host, proc, CONST_STR_LEN(".disabled")) = 0;
*gw_status_get_counter(host, proc, CONST_STR_LEN(".died")) = 0;
*gw_status_get_counter(host, proc, CONST_STR_LEN(".overloaded")) = 0;
proc->stats_connected =
gw_status_get_counter(host, proc, CONST_STR_LEN(".connected"));
*proc->stats_connected = 0;
proc->stats_load =
gw_status_get_counter(host, proc, CONST_STR_LEN(".load"));
*proc->stats_load = 0;
}
static void gw_status_init_host(gw_host *host) {
host->stats_load =
gw_status_get_counter(host, NULL, CONST_STR_LEN(".load"));
*host->stats_load = 0;
host->stats_global_active =
status_counter_get_counter(CONST_STR_LEN("gw.active-requests"));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
__attribute_cold__
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
static void gw_proc_set_state(gw_host *host, gw_proc *proc, int state) {
if ((int)proc->state == state) return;
if (proc->state == PROC_STATE_RUNNING) {
--host->active_procs;
} else if (state == PROC_STATE_RUNNING) {
++host->active_procs;
}
proc->state = state;
}
__attribute_cold__
__attribute_noinline__
static void gw_proc_init_portpath(gw_host *host, gw_proc *proc) {
if (!host->unixsocket) {
proc->port = host->port + proc->id;
return;
}
if (!proc->unixsocket)
proc->unixsocket = buffer_init();
if (!host->bin_path)
buffer_copy_buffer(proc->unixsocket, host->unixsocket);
else {
buffer_clear(proc->unixsocket);
buffer_append_str2(proc->unixsocket, BUF_PTR_LEN(host->unixsocket),
CONST_STR_LEN("-"));
buffer_append_int(proc->unixsocket, proc->id);
}
}
__attribute_cold__
__attribute_noinline__
__attribute_returns_nonnull__
static gw_proc *gw_proc_init(gw_host *host) {
gw_proc *proc = calloc(1, sizeof(*proc));
force_assert(proc);
/*proc->unixsocket = buffer_init();*//*(init on demand)*/
proc->connection_name = buffer_init();
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
proc->prev = NULL;
proc->next = NULL;
proc->state = PROC_STATE_DIED;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
proc->id = host->max_id++;
gw_status_init_proc(host, proc); /*(proc->id must be set)*/
gw_proc_init_portpath(host, proc);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
return proc;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
static void gw_proc_free(gw_proc *proc) {
if (!proc) return;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_proc_free(proc->next);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
buffer_free(proc->unixsocket);
buffer_free(proc->connection_name);
free(proc->saddr);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
free(proc);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
__attribute_malloc__
__attribute_returns_nonnull__
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
static gw_host *gw_host_init(void) {
gw_host *f = calloc(1, sizeof(*f));
force_assert(f);
return f;
}
static void gw_host_free(gw_host *h) {
if (!h) return;
if (h->refcount) {
--h->refcount;
return;
}
gw_proc_free(h->first);
gw_proc_free(h->unused_procs);
for (uint32_t i = 0; i < h->args.used; ++i) free(h->args.ptr[i]);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
free(h->args.ptr);
free(h);
}
__attribute_malloc__
__attribute_returns_nonnull__
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
static gw_exts *gw_extensions_init(void) {
gw_exts *f = calloc(1, sizeof(*f));
force_assert(f);
return f;
}
static void gw_extensions_free(gw_exts *f) {
if (!f) return;
for (uint32_t i = 0; i < f->used; ++i) {
gw_extension *fe = f->exts+i;
for (uint32_t j = 0; j < fe->used; ++j) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_host_free(fe->hosts[j]);
}
free(fe->hosts);
}
free(f->exts);
free(f);
}
static int gw_extension_insert(gw_exts *ext, const buffer *key, gw_host *fh) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_extension *fe = NULL;
for (uint32_t i = 0; i < ext->used; ++i) {
if (buffer_is_equal(key, &ext->exts[i].key)) {
fe = ext->exts+i;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
break;
}
}
if (NULL == fe) {
if (ext->used == ext->size) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
ext->size += 8;
ext->exts = realloc(ext->exts, ext->size * sizeof(gw_extension));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
force_assert(ext->exts);
memset(ext->exts + ext->used, 0, 8 * sizeof(gw_extension));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
fe = ext->exts + ext->used++;
fe->last_used_ndx = -1;
buffer *b;
*(const buffer **)&b = &fe->key;
memcpy(b, key, sizeof(buffer)); /*(copy; not later free'd)*/
}
if (fe->size == fe->used) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
fe->size += 4;
fe->hosts = realloc(fe->hosts, fe->size * sizeof(*(fe->hosts)));
force_assert(fe->hosts);
}
fe->hosts[fe->used++] = fh;
return 0;
}
static void gw_proc_connect_success(gw_host *host, gw_proc *proc, int debug, request_st * const r) {
gw_proc_connected_inc(host, proc); /*(".connected")*/
proc->last_used = log_monotonic_secs;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (debug) {
log_error(r->conf.errh, __FILE__, __LINE__,
"got proc: pid: %d socket: %s load: %d",
proc->pid, proc->connection_name->ptr, proc->load);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
}
__attribute_cold__
static void gw_proc_connect_error(request_st * const r, gw_host *host, gw_proc *proc, pid_t pid, int errnum, int debug) {
[multiple] Y2038 32-bit signed time_t mitigations Most OS platforms have already provided solutions to Y2038 32-bit signed time_t 5 - 10 years ago (or more!) Notable exceptions are Linux i686 and FreeBSD i386. Since 32-bit systems tend to be embedded systems, and since many distros take years to pick up new software, this commit aims to provide Y2038 mitigations for lighttpd running on 32-bit systems with Y2038-unsafe 32-bit signed time_t * Y2038: lighttpd 1.4.60 and later report Y2038 safety $ lighttpd -V + Y2038 support # Y2038-SAFE $ lighttpd -V - Y2038 support (unsafe 32-bit signed time_t) # Y2038-UNSAFE * Y2038: general platform info * Y2038-SAFE: lighttpd 64-bit builds on platforms using 64-bit time_t - all major 64-bit platforms (known to this author) use 64-bit time_t * Y2038-SAFE: lighttpd 32-bit builds on platforms using 64-bit time_t - Linux x32 ABI (different from i686) - FreeBSD all 32-bit and 64-bit architectures *except* 32-bit i386 - NetBSD 6.0 (released Oct 2012) all 32-bit and 64-bit architectures - OpenBSD 5.5 (released May 2014) all 32-bit and 64-bit architectures - Microsoft Windows XP and Visual Studio 2005 (? unsure ?) Another reference suggests Visual Studio 2015 defaults to 64-bit time_t - MacOS 10.15 Catalina (released 2019) drops support for 32-bit apps * Y2038-SAFE: lighttpd 32-bit builds on platforms using 32-bit unsigned time_t - e.g. OpenVMS (unknown if lighttpd builds on this platform) * Y2038-UNSAFE: lighttpd 32-bit builds on platforms using 32-bit signed time_t - Linux 32-bit (including i686) - glibc 32-bit library support not yet available for 64-bit time_t - https://sourceware.org/glibc/wiki/Y2038ProofnessDesign - Linux kernel 5.6 on 32-bit platforms does support 64-bit time_t https://itsubuntu.com/linux-kernel-5-6-to-fix-the-year-2038-issue-unix-y2k/ - https://www.gnu.org/software/libc/manual/html_node/64_002dbit-time-symbol-handling.html "Note: at this point, 64-bit time support in dual-time configurations is work-in-progress, so for these configurations, the public API only makes the 32-bit time support available. In a later change, the public API will allow user code to choose the time size for a given compilation unit." - compiling with -D_TIME_BITS=64 currently has no effect - glibc recent (Jul 2021) mailing list discussion - https://public-inbox.org/bug-gnulib/878s2ozq70.fsf@oldenburg.str.redhat.com/T/ - FreeBSD i386 - DragonFlyBSD 32-bit * Y2038 mitigations attempted on Y2038-UNSAFE platforms (32-bit signed time_t) * lighttpd prefers system monotonic clock instead of realtime clock in places where realtime clock is not required * lighttpd treats negative time_t values as after 19 Jan 2038 03:14:07 GMT * (lighttpd presumes that lighttpd will not encounter dates before 1970 during normal operation.) * lighttpd casts struct stat st.st_mtime (and st.st_*time) through uint64_t to convert negative timestamps for comparisions with 64-bit timestamps (treating negative timestamp values as after 19 Jan 2038 03:14:07 GMT) * lighttpd provides unix_time64_t (int64_t) and * lighttpd provides struct unix_timespec64 (unix_timespec64_t) (struct timespec equivalent using unix_time64_t tv_sec member) * lighttpd provides gmtime64_r() and localtime64_r() wrappers for platforms 32-bit platforms using 32-bit time_t and lighttpd temporarily shifts the year in order to use gmtime_r() and localtime_r() (or gmtime() and localtime()) from standard libraries, before readjusting year and passing struct tm to formatting functions such as strftime() * lighttpd provides TIME64_CAST() macro to cast signed 32-bit time_t to unsigned 32-bit and then to unix_time64_t * Note: while lighttpd tries handle times past 19 Jan 2038 03:14:07 GMT on 32-bit platforms using 32-bit signed time_t, underlying libraries and underlying filesystems might not behave properly after 32-bit signed time_t overflows (19 Jan 2038 03:14:08 GMT). If a given 32-bit OS does not work properly using negative time_t values, then lighttpd likely will not work properly on that system. * Other references and blogs - https://en.wikipedia.org/wiki/Year_2038_problem - https://en.wikipedia.org/wiki/Time_formatting_and_storage_bugs - http://www.lieberbiber.de/2017/03/14/a-look-at-the-year-20362038-problems-and-time-proofness-in-various-systems/
5 months ago
const unix_time64_t cur_ts = log_monotonic_secs;
log_error_st * const errh = r->conf.errh;
errno = errnum; /*(for log_perror())*/
log_perror(errh, __FILE__, __LINE__,
"establishing connection failed: socket: %s", proc->connection_name->ptr);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (!proc->is_local) {
proc->disabled_until = cur_ts + host->disable_time;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_proc_set_state(host, proc, PROC_STATE_OVERLOADED);
}
else if (proc->pid == pid && proc->state == PROC_STATE_RUNNING) {
/* several requests from lighttpd might reference the same proc
*
* Only one of them should mark the proc
* and all other ones should just take a new one.
*
* If a new proc was started with the old struct, this might
* otherwise lead to marking a perfectly good proc as dead
*/
log_error(errh, __FILE__, __LINE__,
"backend error; we'll disable for %d"
"secs and send the request to another backend instead:"
"load: %d", host->disable_time, host->load);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (EAGAIN == errnum) {
/* - EAGAIN: cool down the backend; it is overloaded */
#ifdef __linux__
log_error(errh, __FILE__, __LINE__,
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
"If this happened on Linux: You have run out of local ports. "
"Check the manual, section Performance how to handle this.");
#endif
if (debug) {
log_error(errh, __FILE__, __LINE__,
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
"This means that you have more incoming requests than your "
"FastCGI backend can handle in parallel. It might help to "
"spawn more FastCGI backends or PHP children; if not, "
"decrease server.max-connections. The load for this FastCGI "
"backend %s is %d", proc->connection_name->ptr, proc->load);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
proc->disabled_until = cur_ts + host->disable_time;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_proc_set_state(host, proc, PROC_STATE_OVERLOADED);
}
else {
/* we got a hard error from the backend like
* - ECONNREFUSED for tcp-ip sockets
* - ENOENT for unix-domain-sockets
*/
#if 0
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_proc_set_state(host, proc, PROC_STATE_DIED_WAIT_FOR_PID);
#else /* treat as overloaded (future: unless we send kill() signal)*/
proc->disabled_until = cur_ts + host->disable_time;
gw_proc_set_state(host, proc, PROC_STATE_OVERLOADED);
#endif
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
}
if (EAGAIN == errnum) {
gw_proc_tag_inc(host, proc, CONST_STR_LEN(".overloaded"));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
else {
gw_proc_tag_inc(host, proc, CONST_STR_LEN(".died"));
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
}
static void gw_proc_release(gw_host *host, gw_proc *proc, int debug, log_error_st *errh) {
gw_proc_load_dec(host, proc);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (debug) {
log_error(errh, __FILE__, __LINE__,
"released proc: pid: %d socket: %s load: %u",
proc->pid, proc->connection_name->ptr, proc->load);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
}
__attribute_cold__
static void gw_proc_check_enable(gw_host * const host, gw_proc * const proc, log_error_st * const errh) {
if (log_monotonic_secs <= proc->disabled_until) return;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (proc->state != PROC_STATE_OVERLOADED) return;
gw_proc_set_state(host, proc, PROC_STATE_RUNNING);
log_error(errh, __FILE__, __LINE__,
"gw-server re-enabled: %s %s %hu %s",
proc->connection_name->ptr,
host->host ? host->host->ptr : "", host->port,
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
host->unixsocket ? host->unixsocket->ptr : "");
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
__attribute_cold__
static void gw_proc_waitpid_log(const gw_host * const host, const gw_proc * const proc, log_error_st * const errh, const int status) {
if (WIFEXITED(status)) {
if (proc->state != PROC_STATE_KILLED) {
log_error(errh, __FILE__, __LINE__,
"child exited: %d %s",
WEXITSTATUS(status), proc->connection_name->ptr);
}
} else if (WIFSIGNALED(status)) {
if (WTERMSIG(status) != SIGTERM && WTERMSIG(status) != SIGINT
&& WTERMSIG(status) != host->kill_signal) {
log_error(errh, __FILE__, __LINE__,
"child signalled: %d", WTERMSIG(status));
}
} else {
log_error(errh, __FILE__, __LINE__,
"child died somehow: %d", status);
}
}
static int gw_proc_waitpid(gw_host *host, gw_proc *proc, log_error_st *errh) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
int rc, status;
if (!proc->is_local) return 0;
if (proc->pid <= 0) return 0;
rc = fdevent_waitpid(proc->pid, &status, 1);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (0 == rc) return 0; /* child still running */
/* child terminated */
if (-1 == rc) {
/* EINVAL or ECHILD no child processes */
/* should not happen; someone else has cleaned up for us */
log_perror(errh, __FILE__, __LINE__,
"pid %d %d not found", proc->pid, proc->state);
}
else {
gw_proc_waitpid_log(host, proc, errh, status);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
proc->pid = 0;
if (proc->state != PROC_STATE_KILLED)
proc->disabled_until = log_monotonic_secs;
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
gw_proc_set_state(host, proc, PROC_STATE_DIED);
return 1;
}
__attribute_cold__
static int gw_proc_sockaddr_init(gw_host * const host, gw_proc * const proc, log_error_st * const errh) {
sock_addr addr;
socklen_t addrlen;
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
if (proc->unixsocket) {
if (1 != sock_addr_from_str_hints(&addr,&addrlen,proc->unixsocket->ptr,
AF_UNIX, 0, errh)) {
errno = EINVAL;
return -1;
}
buffer_clear(proc->connection_name);
buffer_append_str2(proc->connection_name,
CONST_STR_LEN("unix:"),
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
BUF_PTR_LEN(proc->unixsocket));
}
else {
#ifdef __COVERITY__
force_assert(host->host); /*(not NULL if !host->unixsocket)*/
#endif
/*(note: name resolution here is *blocking* if IP string not supplied)*/
if (1 != sock_addr_from_str_hints(&addr, &addrlen, host->host->ptr,
0, proc->port, errh)) {
errno = EINVAL;
return -1;
}
else if (host->host->size) {
/*(skip if constant string set in gw_set_defaults_backend())*/
/* overwrite host->host buffer with IP addr string so that
* any further use of gw_host does not block on DNS lookup */
buffer *h;
*(const buffer **)&h = host->host;
sock_addr_inet_ntop_copy_buffer(h, &addr);
host->family = sock_addr_get_family(&addr);
}
buffer_clear(proc->connection_name);
buffer_append_str3(proc->connection_name,
CONST_STR_LEN("tcp:"),
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
BUF_PTR_LEN(host->host),
CONST_STR_LEN(":"));
buffer_append_int(proc->connection_name, proc->port);
}
if (NULL != proc->saddr && proc->saddrlen < addrlen) {
free(proc->saddr);
proc->saddr = NULL;
}
if (NULL == proc->saddr) {
proc->saddr = (struct sockaddr *)malloc(addrlen);
force_assert(proc->saddr);
}
proc->saddrlen = addrlen;
memcpy(proc->saddr, &addr, addrlen);
return 0;
}
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
static int env_add(char_array *env, const char *key, size_t key_len, const char *val, size_t val_len) {
char *dst;
if (!key || !val) return -1;
dst = malloc(key_len + val_len + 3);
force_assert(dst);
memcpy(dst, key, key_len);
dst[key_len] = '=';
memcpy(dst + key_len + 1, val, val_len + 1); /* add the \0 from the value */
for (uint32_t i = 0; i < env->used; ++i) {
#ifdef __COVERITY__
force_assert(env->ptr); /*(non-NULL if env->used != 0)*/
#endif
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (0 == strncmp(dst, env->ptr[i], key_len + 1)) {
free(env->ptr[i]);
env->ptr[i] = dst;
return 0;
}
}
if (env->size <= env->used + 1) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
env->size += 16;
env->ptr = realloc(env->ptr, env->size * sizeof(*env->ptr));
force_assert(env->ptr);
}
#ifdef __COVERITY__
force_assert(env->ptr); /*(non-NULL if env->used != 0; guaranteed above)*/
#endif
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
env->ptr[env->used++] = dst;
return 0;
}
__attribute_cold__
static int gw_spawn_connection(gw_host * const host, gw_proc * const proc, log_error_st * const errh, int debug) {
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
int gw_fd;
int status;
if (debug) {
log_error(errh, __FILE__, __LINE__,
"new proc, socket: %hu %s",
[multiple] reduce redundant NULL buffer checks This commit is a large set of code changes and results in removal of hundreds, perhaps thousands, of CPU instructions, a portion of which are on hot code paths. Most (buffer *) used by lighttpd are not NULL, especially since buffers were inlined into numerous larger structs such as request_st and chunk. In the small number of instances where that is not the case, a NULL check is often performed earlier in a function where that buffer is later used with a buffer_* func. In the handful of cases that remained, a NULL check was added, e.g. with r->http_host and r->conf.server_tag. - check for empty strings at config time and set value to NULL if blank string will be ignored at runtime; at runtime, simple pointer check for NULL can be used to check for a value that has been set and is not blank ("") - use buffer_is_blank() instead of buffer_string_is_empty(), and use buffer_is_unset() instead of buffer_is_empty(), where buffer is known not to be NULL so that NULL check can be skipped - use buffer_clen() instead of buffer_string_length() when buffer is known not to be NULL (to avoid NULL check at runtime) - use buffer_truncate() instead of buffer_string_set_length() to truncate string, and use buffer_extend() to extend Examples where buffer known not to be NULL: - cpv->v.b from config_plugin_values_init is not NULL if T_CONFIG_BOOL (though we might set it to NULL if buffer_is_blank(cpv->v.b)) - address of buffer is arg (&foo) (compiler optimizer detects this in most, but not all, cases) - buffer is checked for NULL earlier in func - buffer is accessed in same scope without a NULL check (e.g. b->ptr) internal behavior change: callers must not pass a NULL buffer to some funcs. - buffer_init_buffer() requires non-null args - buffer_copy_buffer() requires non-null args - buffer_append_string_buffer() requires non-null args - buffer_string_space() requires non-null arg
6 months ago
proc->port, proc->unixsocket ? proc->unixsocket->ptr : "");
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
}
gw_fd = fdevent_socket_cloexec(proc->saddr->sa_family, SOCK_STREAM, 0);
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
if (-1 == gw_fd) {
log_perror(errh, __FILE__, __LINE__, "socket()");
[core] shared code for socket backends common codebase for socket backends, based off mod_fastcgi with some features added for mod_proxy (mostly intended to reduce code duplication and enhance code isolation) mod_fastcgi and mod_scgi can now use fastcgi.balance and scgi.balance for similar behavior as proxy.balance, but the balancing is per-host and not per-proc. proxy.balance is also per-host and not per-proc. mod_proxy and mod_scgi can now use proxy.map-extensions and scgi.map-extensions, similar to fastcgi.map-extensions. mod_fastcgi behavior change (affects only mod_status): - statistics tags have been renamed from "fastcgi.*" to "gw.*" "fastcgi.backend.*" -> "gw.backend.*" "fastcgi.active-requests" -> "gw.active-requests" ("fastcgi.requests" remains "fastcgi.requests") ("proxy.requests" is new) ("scgi.requests" is new) mod_scgi behavior change (likely minor): - removed scgi_proclist_sort_down() and scgi_proclist_sort_up(). procs now chosen based on load as measured by num socket connnections Note: modules using gw_backend.[ch] are currently still independent modules. If it had been written as a single module with fastcgi, scgi, proxy implementations, then there would have been a chance of breaking some existing user configurations where module ordering made a difference for which module handled a given request, though for most people, this would have made no difference. Details about mod_fastcgi code transformations: unsigned int debug -> int debug fastcgi_env member removed from plugin_config renamed "fcgi" and "fastcgi" to "gw", and "FCGI" to "GW" reorganize routines for high-level and lower-level interfaces some lower-level internal interfaces changed to use host,proc,debug args rather than knowing about higher-level (app) hctx and plugin_data tabs->spaces and reformatting
4 years ago
return -1;
}
do {
status = connect(gw_fd, proc->saddr, proc->saddrlen);