summaryrefslogtreecommitdiff
path: root/src/gw_backend.c
Commit message (Collapse)AuthorAgeFilesLines
* [core] return pid_t from fdevent_waitpid()Glenn Strauss2023-05-031-1/+2
|
* [tests] _WIN32 cygwin test supportGlenn Strauss2023-05-031-0/+13
|
* [multiple] _WIN32 use fdevent_kill()Glenn Strauss2023-05-031-13/+4
|
* [core] _WIN32 use log_serror() for WSAGetLastErrorGlenn Strauss2023-05-031-10/+15
|
* [core] _WIN32 check WSAGetLastError() w/ socketsGlenn Strauss2023-05-031-3/+30
| | | | | | | check WSAGetLastError() after socket operations return non-zero Notably, MS winsock2 returns WSAEWOULDBLOCK instead of WSAEINPROGRESS for connect() if socket is configured nonblocking
* [core] _WIN32 socket-compat, filesystem-compatGlenn Strauss2023-05-031-5/+13
| | | | | | | | | | | | | | | | | | | | _WIN32 is sufficiently different -- *different*; not better -- that isolating _WIN32 code is clearer than #ifdef _WIN32 in almost every func in fdevent.c _WIN32-specific fdevent_socket_* funcs _WIN32 SOCKET fds must be closed with closesocket(), not close() _WIN32 HANDLE_FLAG_INHERIT for FD_CLOEXEC _WIN32 use _sopen_s() without _O_TEMPORARY Use _sopen_s() without _O_TEMPORARY in fdevent_mkostemp(). _O_TEMPORARY would remove file once last handle to file is closed. Temporary files in chunkqueue may be closed for large request/response _WIN32 fdevent_rename() using MoveFileExA _WIN32 rename() fails if the target file already exists. Alternatives are MoveFileExA() or ReplaceFileA(). Both of the above fail if either oldfile or newfile are open, so - not atomic - may fail sporadically
* [core] _WIN32 fdevent_createprocess()Glenn Strauss2023-05-031-3/+31
|
* [core] _WIN32 sys-wait.h to wrap <sys/wait.h>Glenn Strauss2023-05-031-3/+1
|
* [core] _WIN32 sys-unistd.h to wrap <unistd.h>Glenn Strauss2023-05-031-1/+1
| | | | (selective implementations; not complete)
* [core] _WIN32 signal-related compatGlenn Strauss2023-05-031-0/+4
| | | | Note: behavior under bash.exe might be different than under cmd.exe
* [multiple] _WIN32 stat() compat sys-stat.hGlenn Strauss2023-05-031-1/+1
| | | | | | | no lstat() no S_IFSOCK no S_ISSOCK() ...
* [multiple] _WIN32 protect code w/ HAVE_SYS_UN_HGlenn Strauss2023-05-031-0/+9
|
* [multiple] _WIN32 protect code w/ HAVE_FORKGlenn Strauss2023-05-031-0/+8
| | | | wrap fork() and kill()
* [multiple] optimistic client read only if HTTP/1.xGlenn Strauss2023-05-031-1/+2
| | | | skip optimistic client read if http version >= HTTP/2
* [multiple] store ptrs to remote addr in request_st (#3192)Glenn Strauss2023-02-281-1/+1
| | | | | | | | | | | | | | | | | adds two pointers to (request_st *) (cost: 16 bytes in 64-bit builds) prepares for upcoming changes to mod_extforward to manage remote addr per request for HTTP/2 requests, rather than remote addr per connection. Modern load balancers often provide options to reuse connections for *different* clients, and therefore mod_extforward might change the remote addr per request. x-ref: "RFE: mod_extforward and multiplexed requests via HTTP/2" https://redmine.lighttpd.net/issues/3192 "Evaluation of remote_addr for mod_maxminddb for multiplexed connections" https://redmine.lighttpd.net/issues/3191
* [core] gw_backend more precise backend env allocGlenn Strauss2023-02-031-1/+1
| | | | remove excess +1 per env string
* [core] pass fdn to fdevent_sched_close,_unregisterGlenn Strauss2023-01-231-2/+2
| | | | remove issock flag; on _WIN32, select(), WSAPoll() work only on sockets
* [multiple] quiet some coverity false positivesGlenn Strauss2023-01-131-1/+1
|
* [multiple] employ ck_calloc, ck_malloc shared codeGlenn Strauss2022-12-101-24/+12
| | | | | employ ck_calloc(), ck_malloc() shared code to slightly reduce code size (centralize the ck_assert() to check that memory allocation succeeded)
* [core] mark gw_proc_free() coldGlenn Strauss2022-12-101-0/+2
|
* [multiple] employ ck_realloc_u32() shared codeGlenn Strauss2022-12-101-44/+15
| | | | employ ck_realloc_u32() shared code to slightly reduce code size
* [core] tweak streaming request body to backendsGlenn Strauss2022-09-111-21/+41
| | | | | | prefer to avoid tempfiles when streaming request body to fast backend keeping up with reading request body as request body is received from client
* [core] handle RDHUP when collecting chunked bodyGlenn Strauss2022-09-111-1/+1
| | | | | | | | handle RDHUP as soon as RDHUP detected when collecting HTTP/1.1 chunked request body (and when not streaming request body to backend) x-ref: https://github.com/lighttpd/lighttpd1.4/pull/115
* [multiple] rename status_counter -> plugin_statsGlenn Strauss2022-05-111-4/+2
|
* [core] reduce trace on Upgrade backend connectionGlenn Strauss2022-03-251-1/+2
| | | | | | | | reduce trace on Upgrade'd backend connection when ECONNRESET received, which, for example, apparently might occur if a backend calls close() on socket without first calling shutdown(fd, SHUT_WR) -- seen on Linux kernel 5.16.15 where lighttpd received ECONNRESET when trying to read() (instead of receiving EOF).
* [core] ensure socket ready before checking connectGlenn Strauss2022-03-251-0/+2
| | | | | | | ensure socket ready for writing before checking connect() status (sanity check in case request gets rescheduled for another reason before the socket is ready for writing)
* [multiple] immed connect to backend for streamingGlenn Strauss2022-03-251-6/+16
| | | | | | connect to backend (mod_cgi, mod_proxy, mod_sockproxy, mod_wstunnel) for streaming request body without waiting for initial data in request body. Useful for things like websockets when data starts on server-side
* [core] restart dead proc on connect error if localGlenn Strauss2021-10-271-3/+5
| | | | restart dead proc on connect error only if proc started by lighttpd
* [core] separate func for gw_authorizer_ok()Glenn Strauss2021-10-271-50/+53
|
* [core] disable server.graceful-restart-bg if spawnGlenn Strauss2021-10-011-0/+20
| | | | | disable server.graceful-restart-bg if spawning gateway backends via *.server "bin-path"
* [core] disable streaming response with authorizer (fixes #3106)Glenn Strauss2021-10-011-0/+3
| | | | | | | | | | | | | disable streaming response while processing "authorizer" mode until "authorizer" response 200 OK from the backend is complete (thx jefftharris) x-ref: "FastCGI authorizer hang with server.stream-response-body" https://redmine.lighttpd.net/boards/2/topics/9969 "FastCGI authorizer hang with server.stream-response-body" https://redmine.lighttpd.net/issues/3106
* [core] improve chunk buffer reuse from backendsGlenn Strauss2021-09-301-14/+18
| | | | | mark and temporarily defer event handling of backend events in order to handle at same time as the associated front-end connection events
* [multiple] internal control for backend read bytesGlenn Strauss2021-09-281-0/+7
| | | | | | | | | | | | | separate internal control for backend max_per_read When not streaming, large reads will be flushed to temp files on disk. When streaming, use a smaller buffer to help reduce memory usage. When not streaming, attempt to read and empty kernel socket bufs. (e.g. MAX_READ_LIMIT 256k) When writing to sockets (or pipes) attempt to fill kernel socket bufs. (e.g. MAX_WRITE_LIMIT 256k)
* [core] mark attr malloc, returns nonnullGlenn Strauss2021-09-171-0/+6
| | | | mark some core code with attributes malloc, returns nonnull
* [core] default backend "connect-timeout" to 8 (#3086)Glenn Strauss2021-09-161-0/+1
| | | | | | | | | | | | | default backend "connect-timeout" to 8 seconds Though this is is a behavior change where there previously was no timeout, this is configurable by lighttpd.conf admin, and having a default connection timeout of a fairly large value (8 seconds) puts a (default) limit on resource usage waiting for socket connect(). x-ref: "sockets disabled, out-of-fds with proxy module" https://redmine.lighttpd.net/issues/3086
* [multiple] quiet coverity warningsGlenn Strauss2021-09-091-0/+3
|
* [core] remove redundant waitpid() on each backendGlenn Strauss2021-09-081-0/+6
| | | | | | | remove redundant waitpid() on each backend (performed each second during periodic maintenance) Doing so is redundant with handle_waitpid hook since lighttpd 1.4.46
* [core] reduce wait time in gw_spawn_connection()Glenn Strauss2021-09-081-1/+1
| | | | | | reduce wait time in gw_spawn_connection() from 10ms down to 1ms. That is still plenty of time to detect immediate execve()/clone() failures.
* [core] mark cold more gateway maintenance codeGlenn Strauss2021-09-081-8/+21
|
* [core] consolidate more gw_host, gw_proc init codeGlenn Strauss2021-09-081-73/+60
| | | | clearer code: host->num_procs as num gw_proc in host->first linked list
* [core] use mod name in trace instead of mod_gwGlenn Strauss2021-09-081-1/+2
|
* [core] remove HANDLER_WAIT_FOR_FDGlenn Strauss2021-09-081-9/+3
| | | | | | | | If a system call returns EMFILE, then admin should increase server.max-fds and check/increase rlimits for num files (ulimit -Hn) Alternatively, the admin might decrease server.max-connections to limit the number of connections served in parallel.
* [core] remove connection_list_append()Glenn Strauss2021-09-081-1/+0
| | | | | | | | | | | | merge connection_list_append() into connection_fdwaitqueue_append() (not converted to singly-linked-list since fdwaitqueue is not expected to be used during normal operation (except extreme overload condition), so use allocated list of pointers (allocated when needed) instead of adding ptr member to (every) allocated struct connection) remove connections-glue.c remove inclusion of connections.h by non-base files
* [core] clarify error message in gw_backend.cGlenn Strauss2021-09-081-1/+1
| | | | clarify error message in gw_backend.c if connect() to unix socket fails
* [core] connect, write, read timeouts on backends (fixes #3086)Glenn Strauss2021-09-081-10/+128
| | | | | | | | | | | | implement connect(), write(), read() timeouts on backends "connect-timeout" "write-timeout" "read-timeout" x-ref: "sockets disabled, out-of-fds with proxy module" https://redmine.lighttpd.net/issues/3086
* [core] gw_backend_error() shared codeGlenn Strauss2021-09-041-21/+44
|
* [core] store time for last r/w to a backend socketGlenn Strauss2021-09-041-4/+10
|
* [core] quiet compiler warningsGlenn Strauss2021-09-041-2/+2
| | | | | | | | | | cast away signedness warning in request_check_hostname() mod_gnutls https_add_ssl_client_entries crts allocated if crt_size != 0 (which is already checked earlier in routine) report from FaceBook Infer static analysis tool (https://fbinfer.com/) - quiet dead store warnings - check return != NULL from allocation funcs
* [multiple] Y2038 32-bit signed time_t mitigationsGlenn Strauss2021-09-041-3/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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/
* [multiple] inline struct in con->dst_addr_bufGlenn Strauss2021-08-271-1/+1
| | | | | (mod_extforward recently changed to use buffer_move() to save addr instead of swapping pointers)
View Lorry Controller Status