Mirrors/libwebsockets
1
0
Fork 0
mirror of https://github.com/warmcat/libwebsockets.git synced 2025-03-09 00:00:04 +01:00
Code Issues Releases Wiki Activity
libwebsockets/CMakeLists.txt

1077 lines
39 KiB
Text
Raw Permalink Normal View History

cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
#
# libwebsockets - small server side websockets and web server implementation
#
# Copyright (C) 2010 - 2020 Andy Green <andy@warmcat.com>
#
# Permission is hereby granted, free of charge, to any person obtaining a copy
# of this software and associated documentation files (the "Software"), to
# deal in the Software without restriction, including without limitation the
# rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
# sell copies of the Software, and to permit persons to whom the Software is
# furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice shall be included in
# all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
# AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
#
cmake: bump minimum version Next release of cmake will choke by default if minimum version < 2.8.12
2020-09-28 07:34:55 +01:00
cmake_minimum_required(VERSION 2.8.12)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
include(CheckFunctionExists)
include(CheckSymbolExists)
include(CheckIncludeFile)
include(CheckIncludeFiles)
include(CheckLibraryExists)
include(CheckTypeSize)
include(CheckCSourceCompiles)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
cmakelists: clean out pthreads and normalize libmount args
2020-06-03 17:14:03 +01:00
if (POLICY CMP0048)
CMake: Silence CMP0048 warnings When libwebsockets is included as a subdirectory in other projects that rely on a minimum CMake version of 3.x, a CMP0048 policy warning will be raised due to the project not specifying a version in the project call. This patch silences the warning by explicitly setting the policy within libwebsockets to NEW if it has already been forwarded as such, without any further impact on the behavior of CMake.
2020-04-21 01:33:33 +02:00
cmake_policy(SET CMP0048 NEW)
endif()
cmakelists: clean out pthreads and normalize libmount args
2020-06-03 17:14:03 +01:00
#if (POLICY CMP0024)
# cmake_policy(SET CMP0024 NEW)
#endif()
if (POLICY CMP0075)
cmake_policy(SET CMP0075 NEW)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
hpack: pseudoheader check improvement
2017-11-10 08:56:44 +08:00
# General Advice
#
# For selecting between DEBUG / RELEASE, use -DCMAKE_BUILD_TYPE=DEBUG or =RELEASE
# debug builds include source level debug info and extra logging
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
set(LWS_WITH_BUNDLED_ZLIB_DEFAULT OFF)
if(WIN32)
set(LWS_WITH_BUNDLED_ZLIB_DEFAULT ON)
endif()
refactor: move all ah and http specific vars to http private and conditional inclusion from there
2018-04-27 15:20:56 +08:00
cmake: allow setting LWS_ROLE_WS You can build lws without support for ws, with -DLWS_ROLE_WS=0. This is thanks to the role ops isolating all the ws-specific business in the ws role. Also retire more test apps replaced by minmal-examples.
2018-04-25 08:42:18 +08:00
set(LWS_ROLE_RAW 1)
refactor: apply ops structs to event loop handlers
2018-04-29 10:44:36 +08:00
set(LWS_WITH_POLL 1)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (ESP_PLATFORM)
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
set(LWS_ESP_PLATFORM 1)
esp32: update build arrangements
2020-12-06 17:29:36 +00:00
#set(CMAKE_TOOLCHAIN_FILE contrib/cross-esp32.cmake)
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
set(LWIP_PROVIDE_ERRNO 1)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
cmake: fix situation like mingw where toolchain file sets platform defines https://github.com/warmcat/libwebsockets/issues/1831
2020-01-20 18:13:53 +00:00
# it's at this point any toolchain file is brought in
wip: ss c++ classes C++ APIs wrapping SS client These are intended to provide an experimental protocol-independent c++ api even more abstracted than secure streams, along the lines of "wget -Omyfile https://example.com/thing" WIP
2020-09-28 10:13:39 +01:00
project(libwebsockets C CXX)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
include(CTest)
cmake: fix situation like mingw where toolchain file sets platform defines https://github.com/warmcat/libwebsockets/issues/1831
2020-01-20 18:13:53 +00:00
esp32: update build arrangements
2020-12-06 17:29:36 +00:00
if (ESP_PLATFORM)
include_directories(
esp-idf: update againt 2021-08-06 head
2021-08-26 12:31:06 +01:00
$ENV{IDF_PATH}/components/esp_hw_support/include/soc/
$ENV{IDF_PATH}/components/freertos/include/
$ENV{IDF_PATH}/components/xtensa/${CONFIG_IDF_TARGET}/include/
$ENV{IDF_PATH}/components/freertos/include/esp_additions
$ENV{IDF_PATH}/components/hal/include
$ENV{IDF_PATH}/components/soc/${CONFIG_IDF_TARGET}/include/
$ENV{IDF_PATH}/components/soc/include/
$ENV{IDF_PATH}/components/esp_hw_support/include
$ENV{IDF_PATH}/components/hal/${CONFIG_IDF_TARGET}/include/
)
if (CONFIG_IDF_TARGET_ARCH_RISCV)
include_directories(
$ENV{IDF_PATH}/components/freertos/port/riscv/include
$ENV{IDF_PATH}/components/riscv/include)
else()
include_directories(
$ENV{IDF_PATH}/components/freertos/port/xtensa/include
$ENV{IDF_PATH}/components/xtensa/include)
endif()
esp32: update build arrangements
2020-12-06 17:29:36 +00:00
endif()
LWS_WITH_DISTRO_RECOMMENDED Idea from Per Bothner https://libwebsockets.org/pipermail/libwebsockets/2018-January/003635.html
2018-01-27 08:01:33 +08:00
#
# Select features recommended for PC distro packaging
#
option(LWS_WITH_DISTRO_RECOMMENDED "Enable features recommended for distro packaging" OFF)
lws_diskcache: split generic parts from gitohashi into lws
2018-11-12 15:24:42 +08:00
option(LWS_FOR_GITOHASHI "Enable features recommended for use with gitohashi" OFF)
LWS_WITH_DISTRO_RECOMMENDED Idea from Per Bothner https://libwebsockets.org/pipermail/libwebsockets/2018-January/003635.html
2018-01-27 08:01:33 +08:00
CMakeLists: add the option to disable Werror https://github.com/warmcat/libwebsockets/pull/1926 Signed-off-by: Jakov Smolic <jakov.smolic@sartura.hr>
2020-05-26 16:35:16 +02:00
#
# Compiler features
#
option(DISABLE_WERROR "Avoid treating compiler warnings as fatal errors" OFF)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Major individual features
#
LWS_WITH_NETWORK: cmake option for no network code
2019-01-13 06:58:21 +08:00
option(LWS_WITH_NETWORK "Compile with network-related code" ON)
refactor: move all ah and http specific vars to http private and conditional inclusion from there
2018-04-27 15:20:56 +08:00
option(LWS_ROLE_H1 "Compile with support for http/1 (needed for ws)" ON)
cmake: allow setting LWS_ROLE_WS You can build lws without support for ws, with -DLWS_ROLE_WS=0. This is thanks to the role ops isolating all the ws-specific business in the ws role. Also retire more test apps replaced by minmal-examples.
2018-04-25 08:42:18 +08:00
option(LWS_ROLE_WS "Compile with support for websockets" ON)
client: MQTT Adds client support for MQTT QoS0 and QoS1, compatible with AWS IoT Supports stream binding where independent client connections to the same endpoint can mux on a single tcp + tls connection with topic routing managed internally.
2020-02-25 13:28:25 +00:00
option(LWS_ROLE_MQTT "Build with support for MQTT client" OFF)
role: dbus This adds support for the integrating libdbus into the lws event loop. Unlike the other roles, lws doesn't completely adopt the fd and libdbus insists to retain control over the fd lifecycle. However libdbus provides apis for foreign code (lws) to provide event loop services to libdbus for the fd. Accordingly, unlike the other roles rx and writeable are not subsumed into lws callback messages and the events remain the property of libdbus. A context struct wrapper is provided that is available in the libdbus callbacks to bridge between the lws and dbus worlds, along with a minimal example dbus client and server.
2018-10-02 10:50:24 +08:00
option(LWS_ROLE_DBUS "Compile with support for DBUS" OFF)
role: raw-proxy
2018-11-29 08:29:48 +08:00
option(LWS_ROLE_RAW_PROXY "Raw packet proxy" OFF)
rtos diet: make raw_file role optional
2020-02-28 15:50:15 +00:00
option(LWS_ROLE_RAW_FILE "Compile with support for raw files" ON)
lws_diskcache: split generic parts from gitohashi into lws
2018-11-12 15:24:42 +08:00
option(LWS_WITH_HTTP2 "Compile with server support for HTTP/2" ON)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
option(LWS_WITH_LWSWS "Libwebsockets Webserver" OFF)
option(LWS_WITH_CGI "Include CGI (spawn process with network-connected stdin/out/err) APIs" OFF)
option(LWS_IPV6 "Compile with support for ipv6" OFF)
windows: recent win10 support Unix Domain Allow selection of Unix Domain Sockets on windows since it is supported for the last couple of years on windows 10 https://devblogs.microsoft.com/commandline/af_unix-comes-to-windows/ ... if only they could add a full set of posix pieces to go with it (and abstract namespace UDS which doesn't work apparently) so that the parts dealing with uid / gid don't have to be disabled.
2020-05-24 13:57:21 +01:00
option(LWS_UNIX_SOCK "Compile with support for UNIX domain socket if OS supports it" ON)
lws_plugins
2020-08-30 07:39:27 +01:00
option(LWS_WITH_PLUGINS "Support plugins for protocols and extensions (implies LWS_WITH_PLUGINS_API)" OFF)
plugins: add option LWS_WITH_PLUGINS_BUILTIN Add the ability to just build plugins into the main library. They are already designed to have a pinhole export for when they are used as dynamic lib plugins so their namespace does not conflict.
2021-01-31 05:50:12 +00:00
option(LWS_WITH_PLUGINS_BUILTIN "Build the plugin protocols directly into lws library" OFF)
client: make external http proxying optional Add LWS_CLIENT_HTTP_PROXYING on by default. Removing it saves a few hundred bytes of code and 128 bytes per vhost in heap.
2019-08-17 21:20:55 +01:00
option(LWS_WITH_HTTP_PROXY "Support for active HTTP proxying" OFF)
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
option(LWS_WITH_ZIP_FOPS "Support serving pre-zipped files" OFF)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
option(LWS_WITH_SOCKS5 "Allow use of SOCKS5 proxy on client connections" OFF)
option(LWS_WITH_PEER_LIMITS "Track peers and restrict resources a single peer can allocate" OFF)
option(LWS_WITH_ACCESS_LOG "Support generating Apache-compatible access logs" OFF)
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
option(LWS_WITH_RANGES "Support http ranges (RFC7233)" OFF)
lws_diskcache: split generic parts from gitohashi into lws
2018-11-12 15:24:42 +08:00
option(LWS_WITH_THREADPOOL "Managed worker thread pool support (relies on pthreads)" OFF)
http: compression methods Add generic http compression layer eanbled at cmake with LWS_WITH_HTTP_STREAM_COMPRESSION. This is wholly a feature of the HTTP role (used by h1 and h2 roles) and doesn't exist outside that context. Currently provides 'deflate' and 'br' compression methods for server side only. 'br' requires also -DLWS_WITH_HTTP_BROTLI=1 at cmake and the brotli libraries (available in your distro already) and dev package. Other compression methods can be added nicely using an ops struct. The built-in file serving stuff will use this is the client says he can handle it, and the mimetype of the file either starts with "text/" (html and css etc) or is the mimetype of Javascript. zlib allocates quite a bit while in use, it seems to be around 256KiB per stream. So this is only useful on relatively strong servers with lots of memory. However for some usecases where you are serving a lot of css and js assets, it's a nice help. The patch performs special treatment for http/1.1 pipelining, since the compression is performed on the fly the compressed content-length is not known until the end. So for h1 only, chunked transfer-encoding is automatically added so pipelining can continue of the connection. For h2 the chunking is neither supported nor required, so it "just works". User code can also request to add a compression transform before the reply headers were sent using the new api LWS_VISIBLE int lws_http_compression_apply(struct lws *wsi, const char *name, unsigned char **p, unsigned char *end, char decomp); ... this allows transparent compression of dynamically generated HTTP. The requested compression (eg, "deflate") is only applied if the client headers indicated it was supported, otherwise it's a NOP. Name may be NULL in which case the first compression method in the internal table at stream.c that is mentioned as acceptable by the client will be used. NOTE: the compression translation, same as h2 support, relies on the user code using LWS_WRITE_HTTP and then LWS_WRITE_HTTP_FINAL on the last part written. The internal lws fileserving code already does this.
2018-09-02 14:43:05 +08:00
option(LWS_WITH_HTTP_STREAM_COMPRESSION "Support HTTP stream compression" OFF)
option(LWS_WITH_HTTP_BROTLI "Also offer brotli http stream compression (requires LWS_WITH_HTTP_STREAM_COMPRESSION)" OFF)
ACME client plugin This adds support for a plugin that can be attached to a vhost to acquire and maintain its TLS cert automatically. It works the same with both OpenSSL and mbedTLS backends, but they can't share auth keys, delete the 'auth.jwk' file as it is in the example JSON when switching between libs
2017-10-21 06:36:35 +08:00
option(LWS_WITH_ACME "Enable support for ACME automatic cert acquisition + maintenance (letsencrypt etc)" OFF)
http proxy: client: unix socket support This allows the client stuff to understand that addresses beginning with '+' represent unix sockets. If the first character after the '+' is '@', it understands that the '@' should be read as '\0', in order to use Linux "abstract namespace" sockets. Further the lws_parse_uri() helper is extended to understand the convention that an address starting with + is a unix socket, and treats the socket path as delimited by ':', eg http://+/var/run/mysocket:/my/path HTTP Proxy is updated to allow mounts to these unix socket paths. Proxy connections go out on h1, but are dynamically translated to h1 or h2 on the incoming side. Proxy usage of libhubbub is separated out... LWS_WITH_HTTP_PROXY is on by default, and LWS_WITH_HUBBUB is off by default.
2018-09-04 08:06:46 +08:00
option(LWS_WITH_HUBBUB "Enable libhubbub rewriting support" OFF)
minimal-raw-audio Shows how to bring alsa's libasound into the lws event loop
2019-10-14 16:46:48 +01:00
option(LWS_WITH_ALSA "Enable alsa audio example" OFF)
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
option(LWS_WITH_GTK "Enable gtk example" OFF)
lws_diskcache: split generic parts from gitohashi into lws
2018-11-12 15:24:42 +08:00
option(LWS_WITH_FTS "Full Text Search support" OFF)
lws_system: ntpclient
2019-09-22 07:25:58 -07:00
option(LWS_WITH_SYS_ASYNC_DNS "Nonblocking internal IPv4 + IPv6 DNS resolver" OFF)
option(LWS_WITH_SYS_NTPCLIENT "Build in tiny ntpclient good for tls date validation and run via lws_system" OFF)
lws_system: dhcpclient Generic lws_system IPv4 DHCP client - netif and route control via lib/plat apis - linux plat pieces implemented - Uses raw ip socket for UDP broadcast and rx - security-aware - usual stuff plus up to 4 x dns server If it's enabled for build, it holds the system state at DHCP until at least one registered interface has acquired a set of IP / mask / router / DNS server It uses PF_PACKET which is Linux-only atm. But those areas are isolated into plat code. TODOs - lease timing and reacquire - plat pieces for other than Linux
2019-09-27 14:32:59 -07:00
option(LWS_WITH_SYS_DHCP_CLIENT "Build in tiny DHCP client" OFF)
rtos diet: make basic auth support a config option on by default Remove wrapper content for < tls v1.2
2020-02-28 09:29:25 +00:00
option(LWS_WITH_HTTP_BASIC_AUTH "Support Basic Auth" ON)
rtos diet: http: remove headers at buildtime according to config Headers related to ws or h2 are now elided if the ws or h2 role is not enabled for build. In addition, a new build-time option LWS_WITH_HTTP_UNCOMMON_HEADERS on by default allows removal of less-common http headers to shrink the parser footprint. Minilex is adapted to produce 8 different versions of the lex table, chosen at build-time according to which headers are included in the build. If you don't need the unusual headers, or aren't using h2 or ws, this chops down the size of the ah and the rodata needed to hold the parsing table from 87 strings / pointers to 49, and the parsing table from 1177 to 696 bytes.
2020-02-28 10:31:04 +00:00
option(LWS_WITH_HTTP_UNCOMMON_HEADERS "Include less common http header support" ON)
context: option to disable system state management
2020-06-23 13:19:30 +01:00
option(LWS_WITH_SYS_STATE "lws_system state support" ON)
lws_smd: system message distribution - Add low level system message distibution framework - Add support for local Secure Streams to participate using _lws_smd streamtype - Add apit test and minimal example - Add SS proxy support for _lws_smd See minimal-secure-streams-smd README.md
2020-06-24 20:15:46 +01:00
option(LWS_WITH_SYS_SMD "Lws System Message Distribution" ON)
fault injection
2021-02-17 10:31:22 +00:00
option(LWS_WITH_SYS_FAULT_INJECTION "Enable fault injection support" OFF)
lws_metrics There are a few build options that are trying to keep and report various statistics - DETAILED_LATENCY - SERVER_STATUS - WITH_STATS remove all those and establish a generic rplacement, lws_metrics. lws_metrics makes its stats available via an lws_system ops function pointer that the user code can set. Openmetrics export is supported, for, eg, prometheus scraping.
2021-01-06 15:08:22 +00:00
option(LWS_WITH_SYS_METRICS "Lws Metrics API" OFF)
rtos diet: make basic auth support a config option on by default Remove wrapper content for < tls v1.2
2020-02-28 09:29:25 +00:00
client: secure streams Secure Streams is an optional layer on top of lws that separates policy like endpoint selection and tls cert validation into a device JSON policy document. Code that wants to open a client connection just specifies a streamtype name, and no longer deals with details like the endpoint, the protocol (!) or anything else other than payloads and optionally generic metadata; the JSON policy contains all the details for each streamtype. h1, h2, ws and mqtt client connections are supported. Logical secure streams outlive any particular connection and supports "nailed-up" connectivity regardless of underlying connection stability.
2020-02-29 12:37:24 +00:00
#
# Secure Streams
#
option(LWS_WITH_SECURE_STREAMS "Secure Streams protocol-agnostic API" OFF)
wip: ss c++ classes C++ APIs wrapping SS client These are intended to provide an experimental protocol-independent c++ api even more abstracted than secure streams, along the lines of "wget -Omyfile https://example.com/thing" WIP
2020-09-28 10:13:39 +01:00
option(LWS_WITH_SECURE_STREAMS_CPP "Secure Streams C++ classes" OFF)
client: secure streams Secure Streams is an optional layer on top of lws that separates policy like endpoint selection and tls cert validation into a device JSON policy document. Code that wants to open a client connection just specifies a streamtype name, and no longer deals with details like the endpoint, the protocol (!) or anything else other than payloads and optionally generic metadata; the JSON policy contains all the details for each streamtype. h1, h2, ws and mqtt client connections are supported. Logical secure streams outlive any particular connection and supports "nailed-up" connectivity regardless of underlying connection stability.
2020-02-29 12:37:24 +00:00
option(LWS_WITH_SECURE_STREAMS_PROXY_API "Secure Streams support to work across processes" OFF)
option(LWS_WITH_SECURE_STREAMS_SYS_AUTH_API_AMAZON_COM "Auth support for api.amazon.com" OFF)
ss: add static policy as a build option In some cases devices may be too constrained to handle JSON policies but still want to use SS apis and methodology. This introduces an off-by-default cmake option LWS_WITH_SECURE_STREAMS_STATIC_POLICY_ONLY, if enabled the JSON parsing part is excluded and it's assumed the user code provides its policy as hardcoded policy structs.
2020-03-26 06:48:34 +00:00
option(LWS_WITH_SECURE_STREAMS_STATIC_POLICY_ONLY "Secure Streams Policy is hardcoded only" OFF)
ss: auth: sigv4 Add SS pieces for Sigv4 auth support
2020-12-29 16:42:25 -08:00
option(LWS_WITH_SECURE_STREAMS_AUTH_SIGV4 "Secure Streams Auth support for AWS Sigv4" OFF)
client: secure streams Secure Streams is an optional layer on top of lws that separates policy like endpoint selection and tls cert validation into a device JSON policy document. Code that wants to open a client connection just specifies a streamtype name, and no longer deals with details like the endpoint, the protocol (!) or anything else other than payloads and optionally generic metadata; the JSON policy contains all the details for each streamtype. h1, h2, ws and mqtt client connections are supported. Logical secure streams outlive any particular connection and supports "nailed-up" connectivity regardless of underlying connection stability.
2020-02-29 12:37:24 +00:00
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
#
# CTest options
#
#
# If you build with LWS_WITH_MINIMAL_EXAMPLES, you can use CTest / make test to run
# examples that can give a pass/fail response. By default it runs tests both against
# a local server peer and warmcat.com, if your CI wants to do the tests but does not
# have internet routing, then you can still run a subset of tests with CTest / make
# test that only does local tests by disabling this option.
#
option(LWS_CTEST_INTERNET_AVAILABLE "CTest will performs tests that need the Internet" ON)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# TLS library options... all except mbedTLS are basically OpenSSL variants.
#
option(LWS_WITH_SSL "Include SSL support (defaults to OpenSSL or similar, mbedTLS if LWS_WITH_MBEDTLS is set)" ON)
option(LWS_WITH_MBEDTLS "Use mbedTLS (>=2.0) replacement for OpenSSL. When setting this, you also may need to specify LWS_MBEDTLS_LIBRARIES and LWS_MBEDTLS_INCLUDE_DIRS" OFF)
option(LWS_WITH_BORINGSSL "Use BoringSSL replacement for OpenSSL" OFF)
option(LWS_WITH_CYASSL "Use CyaSSL replacement for OpenSSL. When setting this, you also need to specify LWS_CYASSL_LIBRARIES and LWS_CYASSL_INCLUDE_DIRS" OFF)
option(LWS_WITH_WOLFSSL "Use wolfSSL replacement for OpenSSL. When setting this, you also need to specify LWS_WOLFSSL_LIBRARIES and LWS_WOLFSSL_INCLUDE_DIRS" OFF)
option(LWS_SSL_CLIENT_USE_OS_CA_CERTS "SSL support should make use of the OS-installed CA root certs" ON)
tls: add cmake options to dump tls
2021-01-23 20:56:01 +00:00
option(LWS_TLS_LOG_PLAINTEXT_RX "For debugging log the received plaintext as soon as decrypted" OFF)
option(LWS_TLS_LOG_PLAINTEXT_TX "For debugging log the transmitted plaintext just before encryption" OFF)
tls: openssl-only: client-only: session caching
2021-03-25 08:37:36 +00:00
option(LWS_WITH_TLS_SESSIONS "Enable persistent, resumable TLS sessions" ON)
tls: add cmake options to dump tls
2021-01-23 20:56:01 +00:00
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Event library options (may select multiple, or none for default poll()
#
option(LWS_WITH_LIBEV "Compile with support for libev" OFF)
option(LWS_WITH_LIBUV "Compile with support for libuv" OFF)
option(LWS_WITH_LIBEVENT "Compile with support for libevent" OFF)
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
option(LWS_WITH_GLIB "Compile with support for glib event loop" OFF)
eventlib: add sd-event support
2021-01-06 14:15:53 +01:00
option(LWS_WITH_SDEVENT "Compile with support for sd-event loop" OFF)
uloop
2021-03-15 19:33:37 +00:00
option(LWS_WITH_ULOOP "Compile with support for uloop" OFF)
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
if (UNIX)
# since v4.1, on unix platforms default is build any event libs as runtime plugins
option(LWS_WITH_EVLIB_PLUGINS "Compile event lib support into runtime-selected plugins" ON)
else()
# otherwise default to linking the event lib(s) to libwebsockets.so
option(LWS_WITH_EVLIB_PLUGINS "Compile event lib support into runtime-selected plugins" OFF)
endif()
drivers: initial generic gpio and i2c plus bitbang Make a start on generic peripheral and bus drivers to provide meta-functionality regardless of platform. On the one hand this simply provides... - bitbang i2c on top of esp-idf gpio apis - ssd1306 oled chip driver as found on Heltec WB32 - modifications to the minimal example test for esp32 to use that ... on the other hand, those capabilities are provided by creating: - an abstract i2c class object - an abstract gpio class object - i2c class implementation using the abstract gpio for bitbang - an abstract display class object - an abstract display state (brightness, animated change, on/off/init tracking, autodim after inactive, auto-off / blanking after inactive) ... with the intention, eg, you only have to add a platform implementation for the gpio to be able to use the i2c-based display drivers and state handling, and i2c bitbang, without any other modifications.
2020-06-10 19:17:08 +01:00
#
# LWS Drivers
#
option(LWS_WITH_DRIVERS "With generic drivers for gpio, i2c, display etc" OFF)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Static / Dynamic build options
#
option(LWS_WITH_STATIC "Build the static version of the library" ON)
option(LWS_WITH_SHARED "Build the shared version of the library" ON)
option(LWS_LINK_TESTAPPS_DYNAMIC "Link the test apps to the shared version of the library. Default is to link statically" OFF)
option(LWS_STATIC_PIC "Build the static version of the library with position-independent code" OFF)
opensslv3: handle deprecation of EC apis openssl v3-alpha11 has marked EC_KEY pieces as deprecated... we use it in LWS_WITH_GENCRYPTO but the related RSA etc pieces were already deprecated for that. We use EC_KEY pieces in vhost init... The apis are not removed but deprecated, we should have a way to keep trucking, but as it is the deprecation warning is promoted to an error. Let's add LWS_SUPPRESS_DEPRECATED_API_WARNINGS option off by default. If enabled at cmake, external deprecated api warnings are suppressed. This gives a general workaround for now for opensslv3. In addition, even if you don't do that, let's notice we are on openssl v3 and don't build the EC curve selection stuff, I don't think anyone is actually using it anyway.
2021-02-14 06:05:16 +00:00
option(LWS_SUPPRESS_DEPRECATED_API_WARNINGS "Turn off complaints about, eg, openssl deprecated api usage" OFF)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Specific platforms
#
option(LWS_WITH_ESP32 "Build for ESP32" OFF)
option(LWS_PLAT_OPTEE "Build for OPTEE" OFF)
freertos: rename esp32 plat to freertos
2019-08-18 10:35:43 +01:00
option(LWS_PLAT_FREERTOS "Build for FreeRTOS" OFF)
asynchronous dns for ipv4 and ipv6 This adds the option to have lws do its own dns resolution on the event loop, without blocking. Existing implementations get the name resolution done by the libc, which is blocking. In the case you are opening client connections but need to carefully manage latency, another connection opening and doing the name resolution becomes a big problem. Currently it supports - ipv4 / A records - ipv6 / AAAA records - ipv4-over-ipv6 ::ffff:1.2.3.4 A record promotion for ipv6 - only one server supported over UDP :53 - nameserver discovery on linux, windows, freertos It also has some nice advantages - lws-style paranoid response parsing - random unique tid generation to increase difficulty of poisoning - it's really integrated with the lws event loop, it does not spawn threads or use the libc resolver, and of course no blocking at all - platform-specific server address capturing (from /etc/resolv.conf on linux, windows apis on windows) - it has LRU caching - piggybacking (multiple requests before the first completes go on a list on the first request, not spawn multiple requests) - observes TTL in cache - TTL and timeout use lws_sul timers on the event loop - ipv6 pieces only built if cmake LWS_IPV6 enabled
2019-09-19 06:54:53 +01:00
option(LWS_PLAT_ANDROID "Android flavour of unix platform" OFF)
freertos: rename esp32 plat to freertos
2019-08-18 10:35:43 +01:00
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Client / Server / Test Apps build control
#
option(LWS_WITHOUT_CLIENT "Don't build the client part of the library" OFF)
option(LWS_WITHOUT_SERVER "Don't build the server part of the library" OFF)
option(LWS_WITHOUT_TESTAPPS "Don't build the libwebsocket-test-apps" OFF)
option(LWS_WITHOUT_TEST_SERVER "Don't build the test server" OFF)
option(LWS_WITHOUT_TEST_SERVER_EXTPOLL "Don't build the test server version that uses external poll" OFF)
option(LWS_WITHOUT_TEST_PING "Don't build the ping test application" OFF)
option(LWS_WITHOUT_TEST_CLIENT "Don't build the client test application" OFF)
#
# Extensions (permessage-deflate)
#
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
option(LWS_WITHOUT_EXTENSIONS "Don't compile with extensions" ON)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# Helpers + misc
#
option(LWS_WITHOUT_BUILTIN_GETIFADDRS "Don't use the BSD getifaddrs implementation from libwebsockets if it is missing (this will result in a compilation error) ... The default is to assume that your libc provides it. On some systems such as uclibc it doesn't exist." OFF)
option(LWS_FALLBACK_GETHOSTBYNAME "Also try to do dns resolution using gethostbyname if getaddrinfo fails" OFF)
option(LWS_WITHOUT_BUILTIN_SHA1 "Don't build the lws sha-1 (eg, because openssl will provide it" OFF)
option(LWS_WITHOUT_DAEMONIZE "Don't build the daemonization api" ON)
option(LWS_SSL_SERVER_WITH_ECDH_CERT "Include SSL server use ECDH certificate" OFF)
cmake: LEJF and LEJP_CONF now on by default
2018-08-05 18:18:40 +08:00
option(LWS_WITH_LEJP "With the Lightweight JSON Parser" ON)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
option(LWS_WITH_SQLITE3 "Require SQLITE3 support" OFF)
client: do client stash in a single alloc Improve the code around stash, getting rid of the strdups for a net code reduction. Remove the special destroy helper for stash since it becomes a one-liner. Trade several stack allocs in the client reset function for a single sized brief heap alloc to reduce peak stack alloc by around 700 bytes.
2019-08-17 20:46:53 +01:00
option(LWS_WITH_STRUCT_JSON "Generic struct serialization to and from JSON" OFF)
lws_struct lws_struct JSON + sqlite3 serializer and deserializer See READMEs/README.lws_struct.md
2019-03-30 22:14:15 +08:00
option(LWS_WITH_STRUCT_SQLITE3 "Generic struct serialization to and from SQLITE3" OFF)
cleaning
2020-01-02 08:32:23 +00:00
# broken atm
#option(LWS_WITH_SMTP "Provide SMTP support" OFF)
windows: import tronkko's dirent.h Win32 compatible version of dirent.h microsoft just don't seem to be able to include themselves. MIT license, same as lws, link to original github project in the header
2020-04-08 10:51:50 +01:00
if (LWS_WITH_ESP32)
lws_dir: wrap dir scanning backend and convert lejp-conf We use POSIX dir scanning apis normally, but for windows, we require libuv to do it for us. Formalize that into a wrapper lws_dir() that hides the backend code. Make it configurable, ON by default and forced on with lejp-conf that depends on it.
2019-03-14 08:24:40 +08:00
option(LWS_WITH_DIR "Directory scanning api support" OFF)
jwk: remove unistd.h include
2019-03-14 21:22:17 +08:00
option(LWS_WITH_LEJP_CONF "With LEJP configuration parser as used by lwsws" OFF)
lws_dir: wrap dir scanning backend and convert lejp-conf We use POSIX dir scanning apis normally, but for windows, we require libuv to do it for us. Formalize that into a wrapper lws_dir() that hides the backend code. Make it configurable, ON by default and forced on with lejp-conf that depends on it.
2019-03-14 08:24:40 +08:00
else()
option(LWS_WITH_DIR "Directory scanning api support" ON)
jwk: remove unistd.h include
2019-03-14 21:22:17 +08:00
option(LWS_WITH_LEJP_CONF "With LEJP configuration parser as used by lwsws" ON)
lws_dir: wrap dir scanning backend and convert lejp-conf We use POSIX dir scanning apis normally, but for windows, we require libuv to do it for us. Formalize that into a wrapper lws_dir() that hides the backend code. Make it configurable, ON by default and forced on with lejp-conf that depends on it.
2019-03-14 08:24:40 +08:00
endif()
logs: allow giving log bitfields from cmake to force build or exclusion By default this doesn't change any existing logging behaviour at all. But it allows you to define cmake options to force or force-disable the build of individual log levels using new cmake option bitfields LWS_LOGGING_BITFIELD_SET and LWS_LOGGING_BITFIELD_CLEAR. Eg, -DLWS_LOGGING_BITFIELD_SET="(LLL_INFO)" can force INFO log level built even in release mode. -DLWS_LOGGING_BITFIELD_CLEAR="(LLL_NOTICE)" will likewise remove NOTICE logging from the build regardless of DEBUG or RELEASE mode.
2020-04-22 06:32:15 +01:00
option(LWS_WITH_NO_LOGS "Disable all logging other than _err and _user from being compiled in" OFF)
set(LWS_LOGGING_BITFIELD_SET 0 CACHE STRING "Bitfield describing which log levels to force included into the build")
set(LWS_LOGGING_BITFIELD_CLEAR 0 CACHE STRING "Bitfield describing which log levels to force removed from the build")
rtos diet: http: remove headers at buildtime according to config Headers related to ws or h2 are now elided if the ws or h2 role is not enabled for build. In addition, a new build-time option LWS_WITH_HTTP_UNCOMMON_HEADERS on by default allows removal of less-common http headers to shrink the parser footprint. Minilex is adapted to produce 8 different versions of the lex table, chosen at build-time according to which headers are included in the build. If you don't need the unusual headers, or aren't using h2 or ws, this chops down the size of the ah and the rodata needed to hold the parsing table from 87 strings / pointers to 49, and the parsing table from 1177 to 696 bytes.
2020-02-28 10:31:04 +00:00
option(LWS_LOGS_TIMESTAMP "Timestamp at start of logs" ON)
logs: make tag lifecycle optional default on The elaborated tags for ls owsi, vh, and ss objects are very useful signposts to understand what's happening in the logs. But for busy h1 servers, they're just a lot of noise. This lets you stop the logging of tagged object lifecycle by cmake -DLWS_LOG_TAG_LIFECYCLE=0
2021-02-01 10:45:23 +00:00
option(LWS_LOG_TAG_LIFECYCLE "Log tagged object lifecycle as NOTICE" ON)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
option(LWS_AVOID_SIGPIPE_IGN "Android 7+ reportedly needs this" OFF)
jwe
2018-12-13 20:05:12 +08:00
option(LWS_WITH_JOSE "JSON Web Signature / Encryption / Keys (RFC7515/6/) API" OFF)
option(LWS_WITH_GENCRYPTO "Enable support for Generic Crypto apis independent of TLS backend" OFF)
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
option(LWS_WITH_SELFTESTS "Selftests run at context creation" OFF)
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
option(LWS_WITH_GCOV "Build with gcc gcov coverage instrumentation" OFF)
Add a flag to disable CMake exports Add LWS_EXPORT_LWSTARGETS (defaults to ON) that can be used to disable CMake export targets as well as install export for libwebsockets. AG: add _WITH_
2017-11-20 12:41:29 +01:00
option(LWS_WITH_EXPORT_LWSTARGETS "Export libwebsockets CMake targets. Disable if they conflict with an outer cmake project." ON)
CMakeLists: build reproducible by default Using the build user and build hostname as part of the git hash breaks reproducible builds. Make this part optional, but build reproducible by default.
2017-12-10 00:54:06 +01:00
option(LWS_REPRODUCIBLE "Build libwebsockets reproducible. It removes the build user and hostname from the build" ON)
LWS_WITH_MINIMAL_EXAMPLES: default-off build all minimal samples after lws for QA This adds an lws cmake option that builds all the minimal examples as part of lws, it's useful for QA. It adds a macro to examples that depend on a particular lws configuration to understand they should just null out their project definition in builds where the lws configuration requirement is not met, and we are building as part of lws. It also adapts all the example library additions to select the just-built-but-not-yet-installed library in the case it is built as part of lws. If built standalone, it now uses the cmake platform-abstracted way to add the library requirement too.
2018-03-17 06:43:12 +08:00
option(LWS_WITH_MINIMAL_EXAMPLES "Also build the normally standalone minimal examples, for QA" OFF)
lwsac Introduce an api for efficiently dealing with allocations for large, unknown amounts of objects.
2018-10-14 06:12:27 +08:00
option(LWS_WITH_LWSAC "lwsac Chunk Allocation api" ON)
ah: custom headers for h1 Until now lws only parses headers it knows at build-time from its prebuilt lexical analyzer. This adds an on-by-default cmake option and a couple of apis to also store and query "custom", ie, unknown-to-lws headers. A minimal example is also provided. At the moment it only works on h1, h2 support needs improvements to the hpack implementation. Since it bloats ah memory usage compared to without it if custom headers are present, the related code and ah footprint can be disabled with the cmake option LWS_WITH_CUSTOM_HEADERS, but it's on by default normally. ESP32 platform disables it. https://github.com/warmcat/libwebsockets/pull/1499
2019-02-26 17:18:24 +08:00
option(LWS_WITH_CUSTOM_HEADERS "Store and allow querying custom HTTP headers (H1 only)" ON)
lws_diskcache: split generic parts from gitohashi into lws
2018-11-12 15:24:42 +08:00
option(LWS_WITH_DISKCACHE "Hashed cache directory with lazy LRU deletion to size limit" OFF)
LWS_WITH_ASAN This lets you build using the runtime Address Sanitizer in gcc. LWS is heavily tested with valgrind routinely during development. But ASAN did find some theoretical-only issues with shifting, strictly ~(1 << 31) is a signed int, it should be ~(1u << 31). Gcc does the same for both, but it's good to have the ability to find these.
2018-11-03 14:47:48 +08:00
option(LWS_WITH_ASAN "Build with gcc runtime sanitizer options enabled (needs libasan)" OFF)
esp32: do not compile directory scanning stuff
2019-07-01 14:54:05 +01:00
option(LWS_WITH_LEJP_CONF "With LEJP configuration parser as used by lwsws" OFF)
cmake: miniz: add as option
2019-07-03 19:46:23 +01:00
option(LWS_WITH_ZLIB "Include zlib support (required for extensions)" OFF)
option(LWS_WITH_BUNDLED_ZLIB "Use bundled zlib version (Windows only)" ${LWS_WITH_BUNDLED_ZLIB_DEFAULT})
option(LWS_WITH_MINIZ "Use miniz instead of zlib" OFF)
sul: multiple timer domains Adapt the pt sul owner list to be an array, and define two different lists, one that acts like before and is the default for existing users, and another that has the ability to cooperate with systemwide suspend to restrict the interval spent suspended so that it will wake in time for the earliest thing on this wake-suspend sul list. Clean the api a bit and add lws_sul_cancel() that only needs the sul as the argument. Add a flag for client creation info to indicate that this client connection is important enough that, eg, validity checking it to detect silently dead connections should go on the wake-suspend sul list. That flag is exposed in secure streams policy so it can be added to a streamtype with "swake_validity": true Deprecate out the old vhost timer stuff that predates sul. Add a flag LWS_WITH_DEPRECATED_THINGS in cmake so users can get it back temporarily before it will be removed in a v4.2. Adapt all remaining in-tree users of it to use explicit suls.
2020-05-28 12:48:17 +01:00
option(LWS_WITH_DEPRECATED_THINGS "Temporary workaround for deprecated apis" OFF)
sequencer: disable by default
2020-09-29 09:29:18 +01:00
option(LWS_WITH_SEQUENCER "lws_seq_t support" OFF)
external poll: add default-off cmake option External poll support generates a lot of messages on a busy system for no value unless you're one of the few people using it. It's not recommended for new users and is there for backwards compatibility. Make it not built by default and selectable by cmake option.
2019-08-10 09:33:07 +01:00
option(LWS_WITH_EXTERNAL_POLL "Support external POLL integration using callback messages (not recommended)" OFF)
lws_dsh
2019-08-11 06:47:19 +01:00
option(LWS_WITH_LWS_DSH "Support lws_dsh_t Disordered Shared Heap" OFF)
client: make external http proxying optional Add LWS_CLIENT_HTTP_PROXYING on by default. Removing it saves a few hundred bytes of code and 128 bytes per vhost in heap.
2019-08-17 21:20:55 +01:00
option(LWS_CLIENT_HTTP_PROXYING "Support external http proxies for client connections" ON)
debloat: remove things from being built by default that should be conditional
2019-08-18 06:29:34 +01:00
option(LWS_WITH_FILE_OPS "Support file operations vfs" ON)
LWS_WITH_UDP
2019-09-30 09:42:38 -07:00
option(LWS_WITH_UDP "Platform supports UDP" ON)
lws_spawn_piped: break out from cgi The vfork optimized spawn, stdxxx and terminal handling in the cgi implementation is quite mature and sophisticated, and useful for other things unrelated to cgi. Break it out into its own public api under LWS_WITH_SPAWN, off by default. Expand it so the parent wsi is optional, and the role and protocol bindings for stdxxx pipes can be set. Allow optional sul timeout and external lws_dll2 owner for extant children. Remove inline style from minimal http-server-cgi
2020-02-12 10:12:39 +00:00
option(LWS_WITH_SPAWN "Spawn subprocesses with piped stdin/out/stderr" OFF)
lws_fsmount: overlayfs apis
2020-02-15 10:09:50 +00:00
option(LWS_WITH_FSMOUNT "Overlayfs and fallback mounting apis" OFF)
gcc: support -fanalyzer static analysis Starting with gcc 10 (in fedora 32) there's a new static analyzer built into gcc you can enable with -fanalyzer. It doesn't slow compilation much, but it's a bit hit and miss, it only analyzes each compilation unit standalone so it reports issues that can never happen. Enable it if the option LWS_WITH_FANALYZER is enabled and cmake can see the actual compiler supports it.
2020-03-27 18:35:23 +00:00
option(LWS_WITH_FANALYZER "Enable gcc -fanalyzer if compiler supports" OFF)
http: add cmake option for header reduction defeat Lws now strips out http headers releated to h2, ws and unusual headers based on cmake config settings for those features... it saves some heap for the ah and reduces the table size in .rodata. It's possible code might have some external dependency on the original header indexes, but, eg, you don't enable h2 so those indexes are optimized with the h2 ones taken out. This introduces a cmake option "LWS_HTTP_HEADERS_ALL", default-OFF, that defeats the header table optimization for compatibility with older versions in the case the client software can't be adapted to use the lws-exported matching header enums. You probably don't need this.
2020-04-23 15:12:17 +01:00
option(LWS_HTTP_HEADERS_ALL "Override header reduction optimization and include all like older lws versions" OFF)
sul: LWS_WITH_SUL_DEBUGGING
2020-07-15 10:25:18 +01:00
option(LWS_WITH_SUL_DEBUGGING "Enable zombie lws_sul checking on object deletion" OFF)
lws_plugins
2020-08-30 07:39:27 +01:00
option(LWS_WITH_PLUGINS_API "Build generic lws_plugins apis (see LWS_WITH_PLUGINS to also build protocol plugins)" OFF)
lws_conmon: connection monitoring and stats generation This provides a build option LWS_WITH_CONMON that lets user code recover detailed connection stats on client connections with the LCCSCF_CONMON flag. In addition to latencies for dns, socket connection, tls and first protocol response where possible, it also provides the user code an unfiltered list of DNS responses that the client received, and the peer it actually succeded to connect to.
2021-03-15 05:20:15 +00:00
option(LWS_WITH_CONMON "Collect introspectable connection latency stats on individual client connections" ON)
cmake: LWS_WITHOUT_EVENTFD
2021-04-20 14:58:20 +01:00
option(LWS_WITHOUT_EVENTFD "Force using pipe instead of eventfd" OFF)
LWS_WITH_UDP
2019-09-30 09:42:38 -07:00
roles: netlink This creates a role for RFC3549 Netlink monitoring. If the OS supports it (currently, linux) then each pt creates a wsi with the netlink role and dumps the current routing table at pt init. It then maintains a cache of the routing table in each pt. Upon routing table changes an SMD message is issued as an event, and Captive Portal Detection is triggered. All of the pt's current connections are reassessed for routability under the changed routing table, those that no longer have a valid route or gateway are closed.
2020-10-04 07:28:41 +01:00
if (${CMAKE_SYSTEM_NAME} MATCHES "Linux")
option(LWS_WITH_NETLINK "Monitor Netlink for Routing Table changes" ON)
else()
set(LWS_WITH_NETLINK 0)
endif()
solaris: openssl: defeat openssl deprecation of md5
2021-03-08 10:52:06 +00:00
if (${CMAKE_SYSTEM_NAME} MATCHES "SunOS")
# its openssl has md5 deprecated
set(LWS_SUPPRESS_DEPRECATED_API_WARNINGS 1)
endif()
roles: netlink This creates a role for RFC3549 Netlink monitoring. If the OS supports it (currently, linux) then each pt creates a wsi with the netlink role and dumps the current routing table at pt init. It then maintains a cache of the routing table in each pt. Upon routing table changes an SMD message is issued as an event, and Captive Portal Detection is triggered. All of the pt's current connections are reassessed for routability under the changed routing table, those that no longer have a valid route or gateway are closed.
2020-10-04 07:28:41 +01:00
cmake: miniz: add as option
2019-07-03 19:46:23 +01:00
#
# to use miniz, enable both LWS_WITH_ZLIB and LWS_WITH_MINIZ
cmake: refactor options into groups and move to top of file
2017-10-21 09:03:42 +08:00
#
# End of user settings
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# sets of sub-options implied by other options
#
ss: static policy: dynamic vhost instantiation Presently a vh is allocated per trust store at policy parsing-time, this is no problem on a linux-class device or if you decide you need a dynamic policy for functionality reasons. However if you're in a constrained enough situation that the static policy makes sense, in the case your trust stores do not have 100% duty cycle, ie, are anyway always in use, the currently-unused vhosts and their x.509 stack are sitting there taking up heap for no immediate benefit. This patch modifies behaviour in ..._STATIC_POLICY_ONLY so that vhosts and associated x.509 tls contexts are not instantiated until a secure stream using them is created; they are refcounted, and when the last logical secure stream using a vhost is destroyed, the vhost and its tls context is also destroyed. If another ss connection is created that wants to use the trust store, the vhost and x.509 context is regenerated again as needed. Currently the refcounting is by ss, it's also possible to move the refcounting to be by connection. The choice is between the delay to generate the vh being visisble at logical ss creation-time, or at connection-time. It's anyway not preferable to have ss instantiated and taking up space with no associated connection or connection attempt underway. NB you will need to reprocess any static policies after this patch so they conform to the trust_store changes.
2020-07-20 07:28:28 +01:00
set(LIB_LIST "")
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
set(LIB_LIST_AT_END)
set(LWS_LIBRARIES)
set(LWS_OPENSSL_SUPPORT 0)
include(CMakeLists-implied-options.txt)
dbus-ws-proxy This builds on the new dbus role support to provide a minimal example proxy between ws and dbus. A client app is provided that asks the proxy to connect to libwebsockets.org and proxy the drawing data from the mirror example there back to the dbus client using dbus messages.
2018-10-03 14:14:29 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
#
# Structural helpers for cmake in subdirs
#
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
macro(add_subdir_include_directories arg1)
add_subdirectory(${arg1})
include_directories(${_CMAKE_INC_LIST})
endmacro()
http proxy: client: unix socket support This allows the client stuff to understand that addresses beginning with '+' represent unix sockets. If the first character after the '+' is '@', it understands that the '@' should be read as '\0', in order to use Linux "abstract namespace" sockets. Further the lws_parse_uri() helper is extended to understand the convention that an address starting with + is a unix socket, and treats the socket path as delimited by ':', eg http://+/var/run/mysocket:/my/path HTTP Proxy is updated to allow mounts to these unix socket paths. Proxy connections go out on h1, but are dynamically translated to h1 or h2 on the incoming side. Proxy usage of libhubbub is separated out... LWS_WITH_HTTP_PROXY is on by default, and LWS_WITH_HUBBUB is off by default.
2018-09-04 08:06:46 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
macro(exports_to_parent_scope)
set(SOURCES ${SOURCES} PARENT_SCOPE)
windows: for data exports use explicit extern LWS_EXTERN needs to be empty for windows when declaring functions in the headers. But for data, it needs the explicit extern otherwise on windows or mingw based builds, it thinks we are redeclaring the data each time.
2020-06-07 07:53:36 +01:00
if (LIB_LIST)
set(LIB_LIST ${LIB_LIST} PARENT_SCOPE)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
get_property(_CURR DIRECTORY PROPERTY INCLUDE_DIRECTORIES)
set(_CMAKE_INC_LIST ${_CURR} PARENT_SCOPE)
windows: for data exports use explicit extern LWS_EXTERN needs to be empty for windows when declaring functions in the headers. But for data, it needs the explicit extern otherwise on windows or mingw based builds, it thinks we are redeclaring the data each time.
2020-06-07 07:53:36 +01:00
if (LWS_LIB_BUILD_INC_PATHS)
set(LWS_LIB_BUILD_INC_PATHS ${LWS_LIB_BUILD_INC_PATHS} PARENT_SCOPE)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
endmacro()
http proxy: client: unix socket support This allows the client stuff to understand that addresses beginning with '+' represent unix sockets. If the first character after the '+' is '@', it understands that the '@' should be read as '\0', in order to use Linux "abstract namespace" sockets. Further the lws_parse_uri() helper is extended to understand the convention that an address starting with + is a unix socket, and treats the socket path as delimited by ':', eg http://+/var/run/mysocket:/my/path HTTP Proxy is updated to allow mounts to these unix socket paths. Proxy connections go out on h1, but are dynamically translated to h1 or h2 on the incoming side. Proxy usage of libhubbub is separated out... LWS_WITH_HTTP_PROXY is on by default, and LWS_WITH_HUBBUB is off by default.
2018-09-04 08:06:46 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
macro(export_to_parent_intermediate)
set(SOURCES ${SOURCES} PARENT_SCOPE)
windows: for data exports use explicit extern LWS_EXTERN needs to be empty for windows when declaring functions in the headers. But for data, it needs the explicit extern otherwise on windows or mingw based builds, it thinks we are redeclaring the data each time.
2020-06-07 07:53:36 +01:00
if (LIB_LIST)
set(LIB_LIST ${LIB_LIST} PARENT_SCOPE)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
set(_CMAKE_INC_LIST ${_CMAKE_INC_LIST} PARENT_SCOPE)
windows: for data exports use explicit extern LWS_EXTERN needs to be empty for windows when declaring functions in the headers. But for data, it needs the explicit extern otherwise on windows or mingw based builds, it thinks we are redeclaring the data each time.
2020-06-07 07:53:36 +01:00
if (LWS_LIB_BUILD_INC_PATHS)
set(LWS_LIB_BUILD_INC_PATHS ${LWS_LIB_BUILD_INC_PATHS} PARENT_SCOPE)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
endmacro()
lwsac Introduce an api for efficiently dealing with allocations for large, unknown amounts of objects.
2018-10-14 06:12:27 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
#
# Try to find the current Git hash
#
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
find_package(Git)
if(GIT_EXECUTABLE)
execute_process(
WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
COMMAND "${GIT_EXECUTABLE}" describe --tags --always
OUTPUT_VARIABLE GIT_HASH
OUTPUT_STRIP_TRAILING_WHITESPACE
)
set(LWS_BUILD_HASH ${GIT_HASH})
# append the build user and hostname
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (NOT LWS_REPRODUCIBLE)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
execute_process(
WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
COMMAND "whoami"
OUTPUT_VARIABLE GIT_USER
OUTPUT_STRIP_TRAILING_WHITESPACE
)
execute_process(
WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
COMMAND "hostname"
OUTPUT_VARIABLE GIT_HOST
OUTPUT_STRIP_TRAILING_WHITESPACE
)
string(REGEX REPLACE "([^\\])[\\]([^\\])" "\\1\\\\\\\\\\2" GIT_USER ${GIT_USER})
set(LWS_BUILD_HASH ${GIT_USER}@${GIT_HOST}-${GIT_HASH})
endif()
message("Git commit hash: ${LWS_BUILD_HASH}")
endif()
cmake: also set LWS_BUILD_HASH to unknown if no git is found
2021-04-20 08:30:18 -04:00
if ("${LWS_BUILD_HASH}" STREQUAL "")
set(LWS_BUILD_HASH "unknown")
endif()
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
set(PACKAGE "libwebsockets")
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
set(CPACK_RPM_PACKAGE_LICENSE "MIT")
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
set(CPACK_PACKAGE_NAME "${PACKAGE}")
v4.0.0
2020-03-04 10:01:28 +00:00
set(CPACK_PACKAGE_VERSION_MAJOR "4")
v4.2.0 release
2021-04-15 16:55:07 +00:00
set(CPACK_PACKAGE_VERSION_MINOR "2")
v4.2.1
2021-07-13 07:22:14 +01:00
set(CPACK_PACKAGE_VERSION_PATCH_NUMBER "1")
cpack: keep elaborated patch string but use just number in lws
2020-05-29 08:24:07 +01:00
set(CPACK_PACKAGE_VERSION_PATCH "${CPACK_PACKAGE_VERSION_PATCH_NUMBER}-${LWS_BUILD_HASH}")
v3.1.0
2018-11-23 11:30:56 +08:00
set(CPACK_PACKAGE_RELEASE 1)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
Subject: [PATCH] Add LWS_LIBRARY_VERSION_NUMBER to lws_config.h. This changeset adds a few preprocessor macros to lws_config.h to allow a user of libwebsockets to determine at compile time which version of lws they are compiling against. This exposes the already existing LWS_LIBRARY_VERSION_MAJOR and _MINOR values, and adds LWS_LIBRARY_VERSION_PATCH. This suggests that future minor bugfix release versions of lws would be e.g. 1.6.0 -> 1.6.1 rather than the style used previously: 1.2 -> 1.21. The way this is currently set up means new minor revisions (with _PATCH==0) always end with .0 but I could change this if preferred. The most important addition is LWS_LIBRARY_VERSION_NUMBER, which produces a number of the form 1005001 for version 1.5.1 - i.e. each part major, minor, patch can extend from 0-999. This macro allows a very easy compile time comparison of version numbers.
2015-12-14 17:25:29 +00:00
set(CPACK_PACKAGE_VERSION "${CPACK_PACKAGE_VERSION_MAJOR}.${CPACK_PACKAGE_VERSION_MINOR}.${CPACK_PACKAGE_VERSION_PATCH}")
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
set(CPACK_PACKAGE_VENDOR "andy@warmcat.com")
v3.1.0
2018-11-23 11:30:56 +08:00
set(CPACK_PACKAGE_CONTACT "andy@warmcat.com")
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
set(CPACK_PACKAGE_DESCRIPTION_SUMMARY "${PACKAGE} ${CPACK_PACKAGE_VERSION}")
v4.2.0 release
2021-04-15 16:55:07 +00:00
set(SOVERSION "18")
Better support for MINW32
2016-11-07 21:36:05 +01:00
if(NOT CPACK_GENERATOR)
if(UNIX)
set(CPACK_GENERATOR "TGZ")
else()
set(CPACK_GENERATOR "ZIP")
endif()
endif()
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
set(CPACK_SOURCE_GENERATOR "TGZ")
set(CPACK_SOURCE_PACKAGE_FILE_NAME "${CPACK_PACKAGE_NAME}-${CPACK_PACKAGE_VERSION}")
set(VERSION "${CPACK_PACKAGE_VERSION}")
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
set(CPACK_RPM_PACKAGE_RELEASE_DIST ON)
set(CPACK_RPM_FILE_NAME "RPM-DEFAULT")
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
# below makes path length problems in CI
set(CPACK_RPM_DEBUGINFO_PACKAGE OFF)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
# below makes some kind of chimera rpm with binaries and sources
set(CPACK_RPM_PACKAGE_SOURCES OFF)
set(CPACK_RPM_INSTALL_WITH_EXEC ON)
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
set(CPACK_RPM_COMPONENT_INSTALL ON)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
set(CPACK_DEBIAN_FILE_NAME "DEB-DEFAULT")
set(CPACK_DEBIAN_PACKAGE_SHLIBDEPS ON)
set(CPACK_DEBIAN_DEBUGINFO_PACKAGE ON)
set(CPACK_DEBIAN_PACKAGE_SOURCE ON)
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
set(CPACK_DEBIAN_COMPONENT_INSTALL ON)
Refresh of ctest-migrate-and-deprecate
2020-05-01 05:45:18 +01:00
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
set(LWS_LIBRARY_VERSION ${CPACK_PACKAGE_VERSION})
Subject: [PATCH] Add LWS_LIBRARY_VERSION_NUMBER to lws_config.h. This changeset adds a few preprocessor macros to lws_config.h to allow a user of libwebsockets to determine at compile time which version of lws they are compiling against. This exposes the already existing LWS_LIBRARY_VERSION_MAJOR and _MINOR values, and adds LWS_LIBRARY_VERSION_PATCH. This suggests that future minor bugfix release versions of lws would be e.g. 1.6.0 -> 1.6.1 rather than the style used previously: 1.2 -> 1.21. The way this is currently set up means new minor revisions (with _PATCH==0) always end with .0 but I could change this if preferred. The most important addition is LWS_LIBRARY_VERSION_NUMBER, which produces a number of the form 1005001 for version 1.5.1 - i.e. each part major, minor, patch can extend from 0-999. This macro allows a very easy compile time comparison of version numbers.
2015-12-14 17:25:29 +00:00
set(LWS_LIBRARY_VERSION_MAJOR ${CPACK_PACKAGE_VERSION_MAJOR})
set(LWS_LIBRARY_VERSION_MINOR ${CPACK_PACKAGE_VERSION_MINOR})
cpack: keep elaborated patch string but use just number in lws
2020-05-29 08:24:07 +01:00
set(LWS_LIBRARY_VERSION_PATCH ${CPACK_PACKAGE_VERSION_PATCH_NUMBER})
set(LWS_LIBRARY_VERSION_PATCH_ELABORATED ${CPACK_PACKAGE_VERSION_PATCH})
Subject: [PATCH] Add LWS_LIBRARY_VERSION_NUMBER to lws_config.h. This changeset adds a few preprocessor macros to lws_config.h to allow a user of libwebsockets to determine at compile time which version of lws they are compiling against. This exposes the already existing LWS_LIBRARY_VERSION_MAJOR and _MINOR values, and adds LWS_LIBRARY_VERSION_PATCH. This suggests that future minor bugfix release versions of lws would be e.g. 1.6.0 -> 1.6.1 rather than the style used previously: 1.2 -> 1.21. The way this is currently set up means new minor revisions (with _PATCH==0) always end with .0 but I could change this if preferred. The most important addition is LWS_LIBRARY_VERSION_NUMBER, which produces a number of the form 1005001 for version 1.5.1 - i.e. each part major, minor, patch can extend from 0-999. This macro allows a very easy compile time comparison of version numbers.
2015-12-14 17:25:29 +00:00
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
if (NOT CMAKE_MODULE_PATH)
set(CMAKE_MODULE_PATH "")
endif()
move cms cmake addition to cmake module paths earlier Signed-off-by: Edwin van den Oetelaar <oetelaar.automatisering@gmail.com>
2013-02-18 20:20:05 +08:00
set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} "${PROJECT_SOURCE_DIR}/cmake/")
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
v3.1.0
2018-11-23 11:30:56 +08:00
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
if (CMAKE_TOOLCHAIN_FILE)
message(STATUS "CMAKE_TOOLCHAIN_FILE='${CMAKE_TOOLCHAIN_FILE}'")
endif()
if (NOT LIB_SUFFIX)
set(LIB_SUFFIX "")
endif()
cmake improve cross documentation Signed-off-by: Andy Green <andy.green@linaro.org>
2013-03-30 10:30:03 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (WIN32)
windows: add version resource
2018-02-05 15:33:36 +01:00
configure_file(${CMAKE_CURRENT_SOURCE_DIR}/win32port/version.rc.in ${CMAKE_CURRENT_BINARY_DIR}/win32port/version.rc @ONLY)
set(RESOURCES ${CMAKE_CURRENT_BINARY_DIR}/win32port/version.rc)
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
include_directories(include)
cmake and other fixes
2014-02-15 13:48:56 +08:00
# Allow the user to override installation directories.
set(LWS_INSTALL_LIB_DIR lib CACHE PATH "Installation directory for libraries")
set(LWS_INSTALL_BIN_DIR bin CACHE PATH "Installation directory for executables")
set(LWS_INSTALL_INCLUDE_DIR include CACHE PATH "Installation directory for header files")
set(LWS_INSTALL_EXAMPLES_DIR bin CACHE PATH "Installation directory for example files")
cmake: miniz: add as option
2019-07-03 19:46:23 +01:00
# if you gave LWS_WITH_MINIZ, point to MINIZ here if not found
# automatically
set(LWS_ZLIB_LIBRARIES CACHE PATH "Path to the zlib/miniz library")
set(LWS_ZLIB_INCLUDE_DIRS CACHE PATH "Path to the zlib/miniz include directory")
Fix typo
2017-02-18 15:51:34 +08:00
set(LWS_SQLITE3_LIBRARIES CACHE PATH "Path to the sqlite3 library")
set(LWS_SQLITE3_INCLUDE_DIRS CACHE PATH "Path to the sqlite3 include directory")
lws_fsmount: overlayfs apis
2020-02-15 10:09:50 +00:00
set(LWS_LIBMOUNT_INCLUDE_DIRS CACHE PATH "Path to the libmount include directory")
set(LWS_LIBMOUNT_LIBRARIES CACHE PATH "Path to the libmount library")
windows: pthreads Add support for external pthreads lib on windows and some docs about how to do. It can build with LWS_WITH_THREADPOOL and LWS_WITH_MINIMAL_EXAMPLES including the pthreads-dependent ones without warnings or errors on windows platform as well with this. pthreads_t can be anything, including a struct - not a pointer-to-a-struct but the struct itself. These can't be cast to a void * for printing as they can on linux, where the base type is a pointer. Let's fix all the usage of those to determine their own thread index in terms of the meaning to the program rather than as a tid.
2020-04-07 14:05:21 +01:00
# on unix, these are in the toolchain. On win32 you have to put them somewhere
# yourself and point to them here
set(LWS_EXT_PTHREAD_INCLUDE_DIR CACHE PATH "Path to an external pthreads include directory")
set(LWS_EXT_PTHREAD_LIBRARIES CACHE PATH "Path to an external pthreads library")
protocol generic sessions Signed-off-by: Andy Green <andy@warmcat.com>
2016-05-19 15:28:31 +08:00
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
http: compression methods Add generic http compression layer eanbled at cmake with LWS_WITH_HTTP_STREAM_COMPRESSION. This is wholly a feature of the HTTP role (used by h1 and h2 roles) and doesn't exist outside that context. Currently provides 'deflate' and 'br' compression methods for server side only. 'br' requires also -DLWS_WITH_HTTP_BROTLI=1 at cmake and the brotli libraries (available in your distro already) and dev package. Other compression methods can be added nicely using an ops struct. The built-in file serving stuff will use this is the client says he can handle it, and the mimetype of the file either starts with "text/" (html and css etc) or is the mimetype of Javascript. zlib allocates quite a bit while in use, it seems to be around 256KiB per stream. So this is only useful on relatively strong servers with lots of memory. However for some usecases where you are serving a lot of css and js assets, it's a nice help. The patch performs special treatment for http/1.1 pipelining, since the compression is performed on the fly the compressed content-length is not known until the end. So for h1 only, chunked transfer-encoding is automatically added so pipelining can continue of the connection. For h2 the chunking is neither supported nor required, so it "just works". User code can also request to add a compression transform before the reply headers were sent using the new api LWS_VISIBLE int lws_http_compression_apply(struct lws *wsi, const char *name, unsigned char **p, unsigned char *end, char decomp); ... this allows transparent compression of dynamically generated HTTP. The requested compression (eg, "deflate") is only applied if the client headers indicated it was supported, otherwise it's a NOP. Name may be NULL in which case the first compression method in the internal table at stream.c that is mentioned as acceptable by the client will be used. NOTE: the compression translation, same as h2 support, relies on the user code using LWS_WRITE_HTTP and then LWS_WRITE_HTTP_FINAL on the last part written. The internal lws fileserving code already does this.
2018-09-02 14:43:05 +08:00
if (LWS_WITH_HTTP_STREAM_COMPRESSION)
set(LWS_WITH_ZLIB 1)
endif()
CMake: convert all LWS_USE_... to LWS_WITH_... Almost all the CMake options begin with LWS_WITH_..., but many of the symbols passed to lws are LWS_USE_... , this causes neededless confusion, compounded by the fact that a few CMake options also begin with LWS_USE_. This patch globally converts all LWS_USE_... to LWS_WITH_..., so there is only one prefix to remember in both CMake and the code. The affected public CMake options are LWS_USE_BORINGSSL -> LWS_WITH_BORINGSSL LWS_USE_CYASSL -> LWS_WITH_CYASSL LWS_USE_WOLFSSL -> LWS_WITH_WOLFSSL LWS_USE_MBEDTLS -> LWS_WITH_MBEDTLS LWS_USE_BUNDLED_ZLIB -> LWS_WITH_BUNDLED_ZLIB
2017-09-28 11:29:03 +08:00
if (LWS_WITH_ZLIB AND NOT LWS_WITH_BUNDLED_ZLIB)
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
if ("${LWS_ZLIB_LIBRARIES}" STREQUAL "" OR "${LWS_ZLIB_INCLUDE_DIRS}" STREQUAL "")
else()
set(ZLIB_LIBRARIES ${LWS_ZLIB_LIBRARIES})
set(ZLIB_INCLUDE_DIRS ${LWS_ZLIB_INCLUDE_DIRS})
set(ZLIB_FOUND 1)
endif()
endif()
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
protocol generic sessions Signed-off-by: Andy Green <andy@warmcat.com>
2016-05-19 15:28:31 +08:00
if (LWS_WITH_SQLITE3)
if ("${LWS_SQLITE3_LIBRARIES}" STREQUAL "" OR "${LWS_SQLITE3_INCLUDE_DIRS}" STREQUAL "")
else()
set(SQLITE3_LIBRARIES ${LWS_SQLITE3_LIBRARIES})
set(SQLITE3_INCLUDE_DIRS ${LWS_SQLITE3_INCLUDE_DIRS})
set(SQLITE3_FOUND 1)
endif()
endif()
fix cmake quoting A few parameters (like "${PROJECT_BINARY_DIR}/lws_config.h" and "${PROJECT_SOURCE_DIR}/lib") are passed to CMake commands in your build script without enclosing them by quotation marks. I see that these places will result in build difficulties if the contents of the used variables will contain special characters like spaces. Signed-off-by: Markus Elfring <markus.elfring@web.de>
2013-10-26 20:03:43 +08:00
include_directories("${PROJECT_BINARY_DIR}")
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
Added check for inline keyword availability. Both to CMake project and Autoconf. inline will be defined to whatever inline replacement exists on the system, such as __inline or __inline__
2013-02-06 15:28:07 +09:00
# Check for different inline keyword versions.
foreach(KEYWORD "inline" "__inline__" "__inline")
set(CMAKE_REQUIRED_DEFINITIONS "-DKEYWORD=${KEYWORD}")
CHECK_C_SOURCE_COMPILES(
"
#include <stdio.h>
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
static KEYWORD void a() {}
Added check for inline keyword availability. Both to CMake project and Autoconf. inline will be defined to whatever inline replacement exists on the system, such as __inline or __inline__
2013-02-06 15:28:07 +09:00
int main(int argc, char **argv) { a(); return 0; }
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
" LWS_HAVE_${KEYWORD})
Added check for inline keyword availability. Both to CMake project and Autoconf. inline will be defined to whatever inline replacement exists on the system, such as __inline or __inline__
2013-02-06 15:28:07 +09:00
endforeach()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
if (NOT LWS_HAVE_inline)
if (LWS_HAVE___inline__)
Added check for inline keyword availability. Both to CMake project and Autoconf. inline will be defined to whatever inline replacement exists on the system, such as __inline or __inline__
2013-02-06 15:28:07 +09:00
set(inline __inline__)
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
elseif(LWS_HAVE___inline)
Added check for inline keyword availability. Both to CMake project and Autoconf. inline will be defined to whatever inline replacement exists on the system, such as __inline or __inline__
2013-02-06 15:28:07 +09:00
set(inline __inline)
endif()
endif()
Subject: Fix some typographical and grammatical errors.
2016-10-02 02:21:03 +03:00
# Put the libraries and binaries that get built into directories at the
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
# top of the build tree rather than in hard-to-find leaf directories.
fix cmake quoting A few parameters (like "${PROJECT_BINARY_DIR}/lws_config.h" and "${PROJECT_SOURCE_DIR}/lib") are passed to CMake commands in your build script without enclosing them by quotation marks. I see that these places will result in build difficulties if the contents of the used variables will contain special characters like spaces. Signed-off-by: Markus Elfring <markus.elfring@web.de>
2013-10-26 20:03:43 +08:00
SET(CMAKE_RUNTIME_OUTPUT_DIRECTORY "${PROJECT_BINARY_DIR}/bin")
SET(CMAKE_LIBRARY_OUTPUT_DIRECTORY "${PROJECT_BINARY_DIR}/lib")
SET(CMAKE_ARCHIVE_OUTPUT_DIRECTORY "${PROJECT_BINARY_DIR}/lib")
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
lejp-conf: substitute _lws_ddir_ with install dir
2016-07-06 10:48:41 +08:00
SET(LWS_INSTALL_PATH "${CMAKE_INSTALL_PREFIX}")
trac 27: use fullpath in OSX dynamic libs Signed-off-by: thinkski <chiszp@gmail.com>
2013-04-16 19:48:05 +08:00
# Put absolute path of dynamic libraries into the object code. Some
# architectures, notably Mac OS X, need this.
cmake and other fixes
2014-02-15 13:48:56 +08:00
SET(CMAKE_INSTALL_NAME_DIR "${CMAKE_INSTALL_PREFIX}/${LWS_INSTALL_LIB_DIR}${LIB_SUFFIX}")
trac 27: use fullpath in OSX dynamic libs Signed-off-by: thinkski <chiszp@gmail.com>
2013-04-16 19:48:05 +08:00
LWS_WITHOUT_BUILTIN_SHA1 Signed-off-by: Andy Green <andy.green@linaro.org>
2015-10-15 11:20:40 +08:00
if (LWS_WITHOUT_BUILTIN_SHA1)
set(LWS_SHA1_USE_OPENSSL_NAME 1)
endif()
glibc: if malloc_trim() exists, call it periodically
2019-03-16 08:10:47 +08:00
CHECK_C_SOURCE_COMPILES(
"#include <malloc.h>
int main(int argc, char **argv) { return malloc_trim(0); }
" LWS_HAVE_MALLOC_TRIM)
CHECK_C_SOURCE_COMPILES(
"#include <malloc.h>
int main(int argc, char **argv) { return (int)malloc_usable_size((void *)0); }
" LWS_HAVE_MALLOC_USABLE_SIZE)
bzero: improve detection of builtin
2018-10-09 15:14:12 +08:00
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
CHECK_FUNCTION_EXISTS(fork LWS_HAVE_FORK)
CHECK_FUNCTION_EXISTS(getenv LWS_HAVE_GETENV)
CHECK_FUNCTION_EXISTS(malloc LWS_HAVE_MALLOC)
CHECK_FUNCTION_EXISTS(memset LWS_HAVE_MEMSET)
CHECK_FUNCTION_EXISTS(realloc LWS_HAVE_REALLOC)
CHECK_FUNCTION_EXISTS(socket LWS_HAVE_SOCKET)
CHECK_FUNCTION_EXISTS(strerror LWS_HAVE_STRERROR)
CHECK_FUNCTION_EXISTS(vfork LWS_HAVE_VFORK)
cgi env Improve cgi support so it's capable of running cgit
2016-03-20 11:55:25 +08:00
CHECK_FUNCTION_EXISTS(execvpe LWS_HAVE_EXECVPE)
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
CHECK_FUNCTION_EXISTS(getifaddrs LWS_HAVE_GETIFADDRS)
windows deal with no snprintf Signed-off-by: Andy Green <andy.green@linaro.org>
2016-03-04 10:53:51 +08:00
CHECK_FUNCTION_EXISTS(snprintf LWS_HAVE_SNPRINTF)
CHECK_FUNCTION_EXISTS(_snprintf LWS_HAVE__SNPRINTF)
CHECK_FUNCTION_EXISTS(_vsnprintf LWS_HAVE__VSNPRINTF)
server status add getloadavg and lib version and more css Signed-off-by: Andy Green <andy@warmcat.com>
2016-04-26 07:45:45 +08:00
CHECK_FUNCTION_EXISTS(getloadavg LWS_HAVE_GETLOADAVG)
win32: enable 64-bit file lengths https://github.com/warmcat/libwebsockets/issues/916 AG added more appveyor warning cleaning and stitched in cmake tests
2017-06-09 20:20:42 +08:00
CHECK_FUNCTION_EXISTS(atoll LWS_HAVE_ATOLL)
CHECK_FUNCTION_EXISTS(_atoi64 LWS_HAVE__ATOI64)
CHECK_FUNCTION_EXISTS(_stat32i64 LWS_HAVE__STAT32I64)
unify us sorted waits There are quite a few linked-lists of things that want events after some period. This introduces a type binding an lws_dll2 for the list and a lws_usec_t for the duration. The wsi timeouts, the hrtimer and the sequencer timeouts are converted to use these, also in the common event wait calculation.
2019-08-08 20:05:03 +01:00
CHECK_FUNCTION_EXISTS(clock_gettime LWS_HAVE_CLOCK_GETTIME)
access-log: use localtime_r instead of localtime if possible Also replace the WIN32 check to global LWS_HAVE_LOCALTIME_R in logs.c
2021-03-25 09:57:49 +03:00
CHECK_FUNCTION_EXISTS(localtime_r LWS_HAVE_LOCALTIME_R)
use gmtime_r instead of gmtime if possible Also check tm result is not NULL before using it.
2021-03-25 10:51:01 +03:00
CHECK_FUNCTION_EXISTS(gmtime_r LWS_HAVE_GMTIME_R)
use ctime_r instead of ctime if possible
2021-03-25 10:33:05 +03:00
CHECK_FUNCTION_EXISTS(ctime_r LWS_HAVE_CTIME_R)
unix: use getpwnam_r and friends Figure out if the threadsafe versions are available on the platform and use them if so
2021-03-24 06:11:03 +00:00
CHECK_FUNCTION_EXISTS(getgrgid_r LWS_HAVE_GETGRGID_R)
CHECK_FUNCTION_EXISTS(getgrnam_r LWS_HAVE_GETGRNAM_R)
CHECK_FUNCTION_EXISTS(getpwuid_r LWS_HAVE_GETPWUID_R)
CHECK_FUNCTION_EXISTS(getpwnam_r LWS_HAVE_GETPWNAM_R)
http: date: use timegm
2021-08-01 08:55:34 +01:00
CHECK_FUNCTION_EXISTS(timegm LWS_HAVE_TIMEGM)
eventfd: restrict test to linux Seems otherwise it gets confused on iOS
2020-04-13 19:28:33 +01:00
OSX: Fixed can't find clock_gettime
2021-05-30 23:27:43 +08:00
if(CMAKE_SYSTEM_NAME MATCHES "Darwin")
if(CMAKE_OSX_DEPLOYMENT_TARGET LESS "10.12")
message("No clock_gettime found on macOS ${CMAKE_OSX_DEPLOYMENT_TARGET}. Disabling LWS_HAVE_CLOCK_GETTIME.")
set(LWS_HAVE_CLOCK_GETTIME 0)
endif()
endif()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
if (NOT LWS_HAVE_GETIFADDRS)
cmake and other fixes
2014-02-15 13:48:56 +08:00
if (LWS_WITHOUT_BUILTIN_GETIFADDRS)
message(FATAL_ERROR "No getifaddrs was found on the system. Turn off the LWS_WITHOUT_BUILTIN_GETIFADDRS compile option to use the supplied BSD version.")
Fixed linux compilation and added more compile options. - Tested and works on Linux now also, including SSL support. - Look for ZLIB not zlib. - Added CMake options for setting all LWS_ defines.
2013-02-06 15:27:27 +09:00
endif()
set(LWS_BUILTIN_GETIFADDRS 1)
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
endif()
windows: pthreads Add support for external pthreads lib on windows and some docs about how to do. It can build with LWS_WITH_THREADPOOL and LWS_WITH_MINIMAL_EXAMPLES including the pthreads-dependent ones without warnings or errors on windows platform as well with this. pthreads_t can be anything, including a struct - not a pointer-to-a-struct but the struct itself. These can't be cast to a void * for printing as they can on linux, where the base type is a pointer. Let's fix all the usage of those to determine their own thread index in terms of the meaning to the program rather than as a tid.
2020-04-07 14:05:21 +01:00
if (LWS_EXT_PTHREAD_INCLUDE_DIR)
set(CMAKE_REQUIRED_INCLUDES ${CMAKE_REQUIRED_INCLUDES};${LWS_EXT_PTHREAD_INCLUDE_DIR})
include_directories(${LWS_EXT_PTHREAD_INCLUDE_DIR})
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
list(APPEND LIB_LIST_AT_END ${LWS_EXT_PTHREAD_LIBRARIES})
windows: pthreads Add support for external pthreads lib on windows and some docs about how to do. It can build with LWS_WITH_THREADPOOL and LWS_WITH_MINIMAL_EXAMPLES including the pthreads-dependent ones without warnings or errors on windows platform as well with this. pthreads_t can be anything, including a struct - not a pointer-to-a-struct but the struct itself. These can't be cast to a void * for printing as they can on linux, where the base type is a pointer. Let's fix all the usage of those to determine their own thread index in terms of the meaning to the program rather than as a tid.
2020-04-07 14:05:21 +01:00
set(CMAKE_REQUIRED_FLAGS ${CMAKE_REQUIRED_FLAGS} " -DHAVE_STRUCT_TIMESPEC=1")
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
#
# add libs here that need to be at the end of the link order
#
if (LWS_EXT_PTHREAD_INCLUDE_DIR)
list(APPEND LIB_LIST_AT_END ${LWS_EXT_PTHREAD_LIBRARIES})
endif()
if (LWS_WITH_ZLIB AND NOT LWS_WITH_BUNDLED_ZLIB)
list(APPEND LIB_LIST_AT_END "${ZLIB_LIBRARIES}")
endif()
cmake: netbsd: no libdl
2020-12-06 19:16:48 +00:00
if (LWS_WITH_PLUGINS_API AND UNIX AND CMAKE_DL_LIBS AND NOT (${CMAKE_SYSTEM_NAME} MATCHES "QNX"))
list(APPEND LIB_LIST_AT_END ${CMAKE_DL_LIBS})
cmake: plugins: make sure about dl
2020-08-31 09:19:40 +01:00
endif()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
CHECK_INCLUDE_FILE(in6addr.h LWS_HAVE_IN6ADDR_H)
CHECK_INCLUDE_FILE(memory.h LWS_HAVE_MEMORY_H)
CHECK_INCLUDE_FILE(netinet/in.h LWS_HAVE_NETINET_IN_H)
CHECK_INCLUDE_FILE(stdint.h LWS_HAVE_STDINT_H)
CHECK_INCLUDE_FILE(stdlib.h LWS_HAVE_STDLIB_H)
CHECK_INCLUDE_FILE(strings.h LWS_HAVE_STRINGS_H)
CHECK_INCLUDE_FILE(string.h LWS_HAVE_STRING_H)
CHECK_INCLUDE_FILE(sys/prctl.h LWS_HAVE_SYS_PRCTL_H)
CHECK_INCLUDE_FILE(sys/socket.h LWS_HAVE_SYS_SOCKET_H)
Check for sys/sockio.h.
2016-09-09 06:48:24 +08:00
CHECK_INCLUDE_FILE(sys/sockio.h LWS_HAVE_SYS_SOCKIO_H)
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
CHECK_INCLUDE_FILE(sys/stat.h LWS_HAVE_SYS_STAT_H)
CHECK_INCLUDE_FILE(sys/types.h LWS_HAVE_SYS_TYPES_H)
CHECK_INCLUDE_FILE(unistd.h LWS_HAVE_UNISTD_H)
CHECK_INCLUDE_FILE(vfork.h LWS_HAVE_VFORK_H)
capabilities support
2017-06-07 06:10:02 +08:00
CHECK_INCLUDE_FILE(sys/capability.h LWS_HAVE_SYS_CAPABILITY_H)
mbedtls wrapper: check for malloc.h before using https://github.com/warmcat/libwebsockets/issues/1163
2018-01-26 07:22:35 +08:00
CHECK_INCLUDE_FILE(malloc.h LWS_HAVE_MALLOC_H)
pthreads: build test server based on pthreads.h availability ... and fix signed / unsigned comparison
2018-03-02 09:01:22 +08:00
CHECK_INCLUDE_FILE(pthread.h LWS_HAVE_PTHREAD_H)
solve compilation problems on vs2012
2019-07-31 16:31:22 +08:00
CHECK_INCLUDE_FILE(inttypes.h LWS_HAVE_INTTYPES_H)
context: properly check for sys/resource.h Instead of hardcoding where sys/resource.h is available, check for it using cmake.
2020-11-01 11:24:31 +01:00
CHECK_INCLUDE_FILE(sys/resource.h LWS_HAVE_SYS_RESOURCE_H)
capabilities support
2017-06-07 06:10:02 +08:00
mingw: update cross toolchain and add 32 and 64 default builds to sai
2020-06-03 09:29:28 +01:00
if (WIN32 OR MSVC)
CHECK_C_SOURCE_COMPILES("#include <winsock2.h>
#include <afunix.h>
int main() { return 0; }" LWS_HAVE_WIN32_AFUNIX_H)
if (LWS_UNIX_SOCK AND NOT LWS_HAVE_WIN32_AFUNIX_H)
cmake: mingw: fix build with default options afunix.h doesn't exist on mingw.
2020-09-17 14:30:23 +03:00
message("No afunix.h found. Disabling LWS_UNIX_SOCK.")
set(LWS_WITH_UNIX_SOCK OFF)
mingw: update cross toolchain and add 32 and 64 default builds to sai
2020-06-03 09:29:28 +01:00
endif()
endif()
role: dbus This adds support for the integrating libdbus into the lws event loop. Unlike the other roles, lws doesn't completely adopt the fd and libdbus insists to retain control over the fd lifecycle. However libdbus provides apis for foreign code (lws) to provide event loop services to libdbus for the fd. Accordingly, unlike the other roles rx and writeable are not subsumed into lws callback messages and the events remain the property of libdbus. A context struct wrapper is provided that is available in the libdbus callbacks to bridge between the lws and dbus worlds, along with a minimal example dbus client and server.
2018-10-02 10:50:24 +08:00
CHECK_LIBRARY_EXISTS(cap cap_set_flag "" LWS_HAVE_LIBCAP)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (LWS_WITH_ZLIB AND NOT LWS_WITH_BUNDLED_ZLIB)
if (LWS_WITH_MINIZ)
CHECK_INCLUDE_FILE(miniz.h LWS_HAVE_ZLIB_H)
else()
CHECK_INCLUDE_FILE(zlib.h LWS_HAVE_ZLIB_H)
role: dbus This adds support for the integrating libdbus into the lws event loop. Unlike the other roles, lws doesn't completely adopt the fd and libdbus insists to retain control over the fd lifecycle. However libdbus provides apis for foreign code (lws) to provide event loop services to libdbus for the fd. Accordingly, unlike the other roles rx and writeable are not subsumed into lws callback messages and the events remain the property of libdbus. A context struct wrapper is provided that is available in the libdbus callbacks to bridge between the lws and dbus worlds, along with a minimal example dbus client and server.
2018-10-02 10:50:24 +08:00
endif()
endif()
mbed3 build support Signed-off-by: Andy Green <andy.green@linaro.org>
2015-11-02 08:21:08 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
CHECK_INCLUDE_FILES("stdlib.h;stdarg.h;string.h" STDC_HEADERS)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
if (NOT CMAKE_REQUIRED_FLAGS)
set(CMAKE_REQUIRED_FLAGS "")
endif()
if (NOT CMAKE_REQUIRED_INCLUDES)
set(CMAKE_REQUIRED_INCLUDES "")
endif()
if (NOT CMAKE_REQUIRED_LIBRARIES)
set(CMAKE_REQUIRED_LIBRARIES "")
endif()
lws_intptr_t
2017-07-07 08:32:04 +08:00
CHECK_C_SOURCE_COMPILES("#include <stdint.h>
int main(void) {
intptr_t test = 1;
return 0;
}" LWS_HAS_INTPTR_T)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if ((CMAKE_C_COMPILER_ID MATCHES "Clang") OR
(CMAKE_CXX_COMPILER_ID MATCHES "Clang"))
set(COMPILER_IS_CLANG ON)
endif()
cmake: old osx clang: suppress check failures sue to -pthread only recognized at compile time
2020-01-24 13:55:21 +00:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (LWS_HAVE_PTHREAD_H AND NOT LWS_PLAT_FREERTOS)
cmake: check if _GNU_SOURCE is already defined
2021-03-25 11:09:13 +03:00
CHECK_C_SOURCE_COMPILES("
#ifndef _GNU_SOURCE
#define _GNU_SOURCE
#endif
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
#include <pthread.h>
int main(void) {
windows: work well with vcpkg pthreads
2021-04-04 18:50:05 +01:00
#ifdef __PTW32_H
pthread_t th = {0,0};
#else
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
pthread_t th = 0;
windows: work well with vcpkg pthreads
2021-04-04 18:50:05 +01:00
#endif
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
pthread_setname_np(th, NULL);
return 0;
}" LWS_HAS_PTHREAD_SETNAME_NP)
endif()
Fix type checks in CMake Use CHECK_TYPE_SIZE to check existence of types.
2014-02-28 00:42:08 +01:00
x509-warning-fixes
2019-01-11 18:19:21 +08:00
CHECK_C_SOURCE_COMPILES("#include <stddef.h>
openssl: detect apis correctly when pthreads not available Since version 3.1.0 and commit aa4143aebdb59f4f21094781fb3b6ab4649643a6, -pthread is unconditionally added to CMAKE_REQUIRED_FLAGS even if pthread.h is not found, this will result in a build failure with openssl if the toolchain doesn't support threads: [ 5%] Building C object CMakeFiles/websockets_shared.dir/lib/core/lws_dll2.c.o In file included from /home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/include/libwebsockets.h:570, from /home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/lib/core/private.h:130, from /home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/lib/core/lws_dll2.c:22: /home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/include/libwebsockets/lws-genhash.h:79:18: error: field 'ctx' has incomplete type HMAC_CTX ctx; ^~~ This build failure is raised because openssl functions are not correcly detected: Determining if the function SSL_CTX_set1_param exists failed with the following output: Change Dir: /home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/CMakeFiles/CMakeTmp Run Build Command(s):/usr/bin/make cmTC_06946/fast && make[1]: Entering directory '/home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/CMakeFiles/CMakeTmp' /usr/bin/make -f CMakeFiles/cmTC_06946.dir/build.make CMakeFiles/cmTC_06946.dir/build make[2]: Entering directory '/home/buildroot/autobuild/instance-2/output-1/build/libwebsockets-3.2.0/CMakeFiles/CMakeTmp' Building C object CMakeFiles/cmTC_06946.dir/CheckFunctionExists.c.o /home/buildroot/autobuild/instance-2/output-1/host/bin/arm-linux-gcc --sysroot=/home/buildroot/autobuild/instance-2/output-1/host/arm-buildroot-linux-uclibcgnueabihf/sysroot -DKEYWORD=__inline -Wall -Wsign-compare -Wuninitialized -Werror -Wundef -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Os -DCHECK_FUNCTION_EXISTS=SSL_CTX_set1_param -pthread -DNDEBUG -o CMakeFiles/cmTC_06946.dir/CheckFunctionExists.c.o -c /home/buildroot/autobuild/instance-2/output-1/host/share/cmake-3.15/Modules/CheckFunctionExists.c Linking C executable cmTC_06946 /home/buildroot/autobuild/instance-2/output-1/host/bin/cmake -E cmake_link_script CMakeFiles/cmTC_06946.dir/link.txt --verbose=1 /home/buildroot/autobuild/instance-2/output-1/host/bin/arm-linux-gcc --sysroot=/home/buildroot/autobuild/instance-2/output-1/host/arm-buildroot-linux-uclibcgnueabihf/sysroot -Wall -Wsign-compare -Wuninitialized -Werror -Wundef -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -Os -DCHECK_FUNCTION_EXISTS=SSL_CTX_set1_param -pthread -DNDEBUG CMakeFiles/cmTC_06946.dir/CheckFunctionExists.c.o -o cmTC_06946 /home/buildroot/autobuild/instance-2/output-1/host/arm-buildroot-linux-uclibcgnueabihf/sysroot/usr/lib/libssl.so /home/buildroot/autobuild/instance-2/output-1/host/arm-buildroot-linux-uclibcgnueabihf/sysroot/usr/lib/libcrypto.so -lssl -lcrypto -lm -lcap /home/buildroot/autobuild/instance-2/output-1/host/opt/ext-toolchain/bin/../lib/gcc/arm-buildroot-linux-uclibcgnueabihf/8.3.0/../../../../arm-buildroot-linux-uclibcgnueabihf/bin/ld: cannot find -lpthread collect2: error: ld returned 1 exit status CMakeFiles/cmTC_06946.dir/build.make:88: recipe for target 'cmTC_06946' failed Fixes: - http://autobuild.buildroot.org/results/6186b4718db285edadf7203d00ed72f8d76a31e4 Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
2020-01-03 14:02:06 +00:00
#include <getopt.h>
int main(void) {
qnx: qnx6.5 compatibility
2019-01-11 16:48:53 +08:00
void *p = (void *)getopt_long;
return p != NULL;
}" LWS_HAS_GETOPT_LONG)
netlink: check for RTA_PREF before use Old kernels like v3.10 don't have it [#2079](https://github.com/warmcat/libwebsockets/issues/2079)
2020-10-09 17:48:39 +08:00
CHECK_C_SOURCE_COMPILES("#include <linux/rtnetlink.h>
int main(void) {
int test = RTA_PREF;
return 0;
}" LWS_HAVE_RTA_PREF)
qnx: qnx6.5 compatibility
2019-01-11 16:48:53 +08:00
type comparisons: fixes This is a huge patch that should be a global NOP. For unix type platforms it enables -Wconversion to issue warnings (-> error) for all automatic casts that seem less than ideal but are normally concealed by the toolchain. This is things like passing an int to a size_t argument. Once enabled, I went through all args on my default build (which build most things) and tried to make the removed default cast explicit. With that approach it neither change nor bloat the code, since it compiles to whatever it was doing before, just with the casts made explicit... in a few cases I changed some length args from int to size_t but largely left the causes alone. From now on, new code that is relying on less than ideal casting will complain and nudge me to improve it by warnings.
2020-12-12 06:21:40 +00:00
CHECK_C_SOURCE_COMPILES("#include <sys/types.h>
int main(void) {
suseconds_t x = 0;
return (int)x;
}" LWS_HAVE_SUSECONDS_T)
Fix type checks in CMake Use CHECK_TYPE_SIZE to check existence of types.
2014-02-28 00:42:08 +01:00
if (NOT PID_T_SIZE)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
set(pid_t int)
Fix type checks in CMake Use CHECK_TYPE_SIZE to check existence of types.
2014-02-28 00:42:08 +01:00
endif()
if (NOT SIZE_T_SIZE)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
set(size_t "unsigned int")
endif()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
if (NOT LWS_HAVE_MALLOC)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
set(malloc rpl_malloc)
endif()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
if (NOT LWS_HAVE_REALLOC)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
set(realloc rpl_realloc)
endif()
cmake add gcc compile options Signed-off-by: Andy Green <andy.green@linaro.org>
2013-03-16 12:32:59 +08:00
cmake: clang: Fix LWS_WITH_ASAN Building with LWS_WITH_ASAN was broken when using Clang, due to some flags being given that Clang doesn't recognize. Fixed this by adding those flags only when the compiler isn't Clang.
2019-06-20 16:16:03 -07:00
if (CMAKE_COMPILER_IS_GNUCC OR CMAKE_COMPILER_IS_GNUCXX OR COMPILER_IS_CLANG)
Add compiler checking for visibility flag Signed-off-by: Steven Lai <steven.lai@liricco.com>
2015-01-27 19:46:26 +08:00
include (CheckCCompilerFlag)
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
CHECK_C_COMPILER_FLAG(-fvisibility=hidden LWS_HAVE_VISIBILITY)
gcc: support -fanalyzer static analysis Starting with gcc 10 (in fedora 32) there's a new static analyzer built into gcc you can enable with -fanalyzer. It doesn't slow compilation much, but it's a bit hit and miss, it only analyzes each compilation unit standalone so it reports issues that can never happen. Enable it if the option LWS_WITH_FANALYZER is enabled and cmake can see the actual compiler supports it.
2020-03-27 18:35:23 +00:00
if (LWS_WITH_FANALYZER)
CHECK_C_COMPILER_FLAG(-fanalyzer LWS_HAVE_FANALYZER)
endif()
Use LWS_HAVE_ instead of just HAVE_ Since we include lws_config.h in the public headers, at least our HAVE_ macros should be kind of unique, so that we don't get redefinitions when used with other libraries using config files as well.
2015-06-24 16:46:02 +02:00
if (LWS_HAVE_VISIBILITY)
Add compiler checking for visibility flag Signed-off-by: Steven Lai <steven.lai@liricco.com>
2015-01-27 19:46:26 +08:00
set(VISIBILITY_FLAG -fvisibility=hidden)
endif()
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
if (LWS_WITH_GCOV)
cmake: override build system release optimization policy The cmake config on the build system actually decides the release build optimization policy. On Fedora, it's -O2. On Ubuntu, it's -O3. Anything given in CMakeLists.txt is overridden by the build system policy since it goes at the end of the compiler commandline. When you are building cross, the build system's opinion of your cross binary optimization level is irrelevant, and at worst destructive. Some versions of gcc contain broken optimizations that are applied only at -O3. This patch removes any doomed attempt to set -O in CMakeLists.txt, which has no effect since the build system policy is still added at the end, but removes confusion; and adds code to all the cross build files to forcibly override release optimization level to -O2, removing the build system's opinion of how your cross build should look.
2019-02-28 07:05:12 +08:00
set (GCOV_FLAGS "-fprofile-arcs -ftest-coverage ")
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
else()
set(GCOV_FLAGS "")
gcov: add in cmake and adapt defaults 1) Introduce LWS_WITH_GCOV to build with gcc / clang coverage instrumentation. $ cd build $ make clean && rm -f `find . -name "*.gcno" -o -name "*.gcda"` && make -j16 && sudo make install && sudo /usr/local/bin/libwebsockets-test-server -s ... $ gcov `find . -name *.c.gcno | grep -v test-apps` -b | sed "/\.h.\$/,/^$/d" The above are available in two helper scripts - scripts/build-gcov.sh - scripts/gcov.sh 2) CMake defaults changed: - LWS_WITH_ZIP_FOPS: OFF - LWS_WITH_RANGES: OFF - LWS_WITHOUT_EXTENSIONS: ON - LWS_WITH_ZLIB: OFF New CMake controls that default-OFF: - LWS_WITH_GENHASH - LWS_WITH_GENRSA these are implied by LWS_WITH_JWS (which is implied by LWS_WITH_ACME) 3) rename ./lib/tls/XXX/server.c and client.c to XXX-server.c / XXX-client.c. This is because gcov dumps its results using the .c filename part only, the copies overwrite each other if there are different .c files in the tree with the same filename part. 4) Add onetime test-client mode and test to ./test-apps/attack.sh 5) Add gcov howto in READMEs/README.build.md using attack.sh
2017-11-14 11:25:54 +08:00
endif()
autobahn fixes This replaces the old test-app for echo with separate client and server minimal versions. The autobahn test script is made more autonomous and tests both client and server.
2018-04-20 10:33:23 +08:00
LWS_WITH_ASAN This lets you build using the runtime Address Sanitizer in gcc. LWS is heavily tested with valgrind routinely during development. But ASAN did find some theoretical-only issues with shifting, strictly ~(1 << 31) is a signed int, it should be ~(1u << 31). Gcc does the same for both, but it's good to have the ability to find these.
2018-11-03 14:47:48 +08:00
if (LWS_WITH_ASAN)
cmake: clang: Fix LWS_WITH_ASAN Building with LWS_WITH_ASAN was broken when using Clang, due to some flags being given that Clang doesn't recognize. Fixed this by adding those flags only when the compiler isn't Clang.
2019-06-20 16:16:03 -07:00
set (ASAN_FLAGS "-fsanitize=address -fsanitize=undefined -fsanitize-address-use-after-scope -fsanitize-undefined-trap-on-error")
if (NOT COMPILER_IS_CLANG)
set (ASAN_FLAGS "${ASAN_FLAGS} -fsanitize=pointer-compare -fsanitize=pointer-subtract -fsanitize=leak")
endif()
LWS_WITH_ASAN This lets you build using the runtime Address Sanitizer in gcc. LWS is heavily tested with valgrind routinely during development. But ASAN did find some theoretical-only issues with shifting, strictly ~(1 << 31) is a signed int, it should be ~(1u << 31). Gcc does the same for both, but it's good to have the ability to find these.
2018-11-03 14:47:48 +08:00
message("Enabling ASAN")
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
else()
set(ASAN_FLAGS "")
LWS_WITH_ASAN This lets you build using the runtime Address Sanitizer in gcc. LWS is heavily tested with valgrind routinely during development. But ASAN did find some theoretical-only issues with shifting, strictly ~(1 << 31) is a signed int, it should be ~(1u << 31). Gcc does the same for both, but it's good to have the ability to find these.
2018-11-03 14:47:48 +08:00
endif()
autobahn fixes This replaces the old test-app for echo with separate client and server minimal versions. The autobahn test script is made more autonomous and tests both client and server.
2018-04-20 10:33:23 +08:00
cmake: probe gcc flags capability old gcc doesn't have some of the extended warning flags we want to apply... use cmake to confirm the gcc has them before trying to use them.
2019-05-15 07:27:18 +01:00
check_c_compiler_flag("-Wignored-qualifiers" LWS_GCC_HAS_IGNORED_QUALIFIERS)
check_c_compiler_flag("-Wtype-limits" LWS_GCC_HAS_TYPE_LIMITS)
opensslv3: handle deprecation of EC apis openssl v3-alpha11 has marked EC_KEY pieces as deprecated... we use it in LWS_WITH_GENCRYPTO but the related RSA etc pieces were already deprecated for that. We use EC_KEY pieces in vhost init... The apis are not removed but deprecated, we should have a way to keep trucking, but as it is the deprecation warning is promoted to an error. Let's add LWS_SUPPRESS_DEPRECATED_API_WARNINGS option off by default. If enabled at cmake, external deprecated api warnings are suppressed. This gives a general workaround for now for opensslv3. In addition, even if you don't do that, let's notice we are on openssl v3 and don't build the EC curve selection stuff, I don't think anyone is actually using it anyway.
2021-02-14 06:05:16 +00:00
check_c_compiler_flag("-Wno-deprecated-declarations" LWS_GCC_HAS_NO_DEPRECATED_DECLARATIONS)
cmake: probe gcc flags capability old gcc doesn't have some of the extended warning flags we want to apply... use cmake to confirm the gcc has them before trying to use them.
2019-05-15 07:27:18 +01:00
if (LWS_GCC_HAS_IGNORED_QUALIFIERS)
set(CMAKE_C_FLAGS "-Wignored-qualifiers ${CMAKE_C_FLAGS}" )
endif()
if (LWS_GCC_HAS_TYPE_LIMITS)
set(CMAKE_C_FLAGS "-Wtype-limits ${CMAKE_C_FLAGS}" )
endif()
gcc: support -fanalyzer static analysis Starting with gcc 10 (in fedora 32) there's a new static analyzer built into gcc you can enable with -fanalyzer. It doesn't slow compilation much, but it's a bit hit and miss, it only analyzes each compilation unit standalone so it reports issues that can never happen. Enable it if the option LWS_WITH_FANALYZER is enabled and cmake can see the actual compiler supports it.
2020-03-27 18:35:23 +00:00
if (LWS_WITH_FANALYZER AND LWS_HAVE_FANALYZER)
set(CMAKE_C_FLAGS "-fanalyzer ${CMAKE_C_FLAGS}" )
endif()
cmake: use -Wuninitialized only for GCC >= 4.4 It doesn't work with 4.3.3
2020-09-15 15:00:09 +03:00
if (CMAKE_COMPILER_IS_CLANG OR CMAKE_C_COMPILER_VERSION VERSION_GREATER 4.4)
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
set(CMAKE_C_FLAGS "-Wuninitialized ${CMAKE_C_FLAGS}")
endif()
CMakeLists: add the option to disable Werror https://github.com/warmcat/libwebsockets/pull/1926 Signed-off-by: Jakov Smolic <jakov.smolic@sartura.hr>
2020-05-26 16:35:16 +02:00
# always warn all and generate debug info
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
if (UNIX AND NOT LWS_PLAT_FREERTOS)
type comparisons: fixes This is a huge patch that should be a global NOP. For unix type platforms it enables -Wconversion to issue warnings (-> error) for all automatic casts that seem less than ideal but are normally concealed by the toolchain. This is things like passing an int to a size_t argument. Once enabled, I went through all args on my default build (which build most things) and tried to make the removed default cast explicit. With that approach it neither change nor bloat the code, since it compiles to whatever it was doing before, just with the casts made explicit... in a few cases I changed some length args from int to size_t but largely left the causes alone. From now on, new code that is relying on less than ideal casting will complain and nudge me to improve it by warnings.
2020-12-12 06:21:40 +00:00
set(CMAKE_C_FLAGS "-Wall -Wconversion -Wsign-compare -Wstrict-aliasing ${VISIBILITY_FLAG} -Wundef ${GCOV_FLAGS} ${CMAKE_C_FLAGS} ${ASAN_FLAGS}" )
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
else()
cmake: remove fixed -g
2020-09-29 16:58:05 +01:00
set(CMAKE_C_FLAGS "-Wall -Wsign-compare ${VISIBILITY_FLAG} ${GCOV_FLAGS} ${CMAKE_C_FLAGS}" )
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
endif()
cmake add gcc compile options Signed-off-by: Andy Green <andy.green@linaro.org>
2013-03-16 12:32:59 +08:00
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
if ("${DISABLE_WERROR}" STREQUAL "OFF")
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Werror")
endif()
opensslv3: handle deprecation of EC apis openssl v3-alpha11 has marked EC_KEY pieces as deprecated... we use it in LWS_WITH_GENCRYPTO but the related RSA etc pieces were already deprecated for that. We use EC_KEY pieces in vhost init... The apis are not removed but deprecated, we should have a way to keep trucking, but as it is the deprecation warning is promoted to an error. Let's add LWS_SUPPRESS_DEPRECATED_API_WARNINGS option off by default. If enabled at cmake, external deprecated api warnings are suppressed. This gives a general workaround for now for opensslv3. In addition, even if you don't do that, let's notice we are on openssl v3 and don't build the EC curve selection stuff, I don't think anyone is actually using it anyway.
2021-02-14 06:05:16 +00:00
if (LWS_SUPPRESS_DEPRECATED_API_WARNINGS)
set(CMAKE_C_FLAGS "-Wno-deprecated ${CMAKE_C_FLAGS}")
if (LWS_GCC_HAS_NO_DEPRECATED_DECLARATIONS)
set(CMAKE_C_FLAGS "-Wno-deprecated-declarations ${CMAKE_C_FLAGS}")
endif()
endif()
cmake: gcc: allow build with GCC < 4.4 Prior to 4.4, -Wuninitialized could only be used with -O.
2020-08-30 09:28:35 +03:00
endif ()
optee: remove build system
2019-01-15 06:59:48 +08:00
pthreads only as needed This just lets you build lws 1.6 without pthreads if your OS / toolchain makes that possible, in the case you don't build the test apps (libwebsockets-test-server-pthreads needs it) Signed-off-by: Andy Green <andy.green@linaro.org>
2016-01-25 21:39:07 +08:00
if ((CMAKE_COMPILER_IS_GNUCC OR CMAKE_COMPILER_IS_GNUCXX) AND NOT LWS_WITHOUT_TESTAPPS)
cmake: qnx adaptations https://github.com/warmcat/libwebsockets/issues/2034
2020-09-03 08:20:16 +01:00
if (UNIX AND LWS_HAVE_PTHREAD_H AND NOT (${CMAKE_SYSTEM_NAME} MATCHES "QNX"))
osx clang blows up if pthreads flag at link time Signed-off-by: Andy Green <andy.green@linaro.org>
2015-11-29 19:41:13 +08:00
# jeez clang understands -pthread but dies if he sees it at link time!
# http://stackoverflow.com/questions/2391194/what-is-gs-pthread-equiv-in-clang
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -pthread" )
wip: ss c++ classes C++ APIs wrapping SS client These are intended to provide an experimental protocol-independent c++ api even more abstracted than secure streams, along the lines of "wget -Omyfile https://example.com/thing" WIP
2020-09-28 10:13:39 +01:00
list(APPEND LIB_LIST_AT_END -lpthread)
osx clang blows up if pthreads flag at link time Signed-off-by: Andy Green <andy.green@linaro.org>
2015-11-29 19:41:13 +08:00
endif()
endif()
cmake: clang: Fix LWS_WITH_ASAN Building with LWS_WITH_ASAN was broken when using Clang, due to some flags being given that Clang doesn't recognize. Fixed this by adding those flags only when the compiler isn't Clang.
2019-06-20 16:16:03 -07:00
if (COMPILER_IS_CLANG)
freebsd: clang requires -pthread In the now default case that THREADPOOL is enabled... not sure what OSX will make of that as previously it blew up with -pthread on clang cmdline at link time
2018-11-05 10:25:36 +08:00
osx clang quench deprecated api errors after approach by redhat https://bugzilla.redhat.com/show_bug.cgi?id=1155181 Signed-off-by: Andy Green <andy.green@linaro.org>
2015-11-29 19:19:09 +08:00
# otherwise osx blows a bunch of openssl deprecated api errors
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -Wno-deprecated-declarations" )
freebsd: clang requires -pthread In the now default case that THREADPOOL is enabled... not sure what OSX will make of that as previously it blew up with -pthread on clang cmdline at link time
2018-11-05 10:25:36 +08:00
if (UNIX AND LWS_HAVE_PTHREAD_H)
cmake: old osx clang: suppress check failures sue to -pthread only recognized at compile time
2020-01-24 13:55:21 +00:00
set(CMAKE_C_FLAGS "${CMAKE_C_FLAGS} -pthread -Wno-error=unused-command-line-argument" )
freebsd: clang requires -pthread In the now default case that THREADPOOL is enabled... not sure what OSX will make of that as previously it blew up with -pthread on clang cmdline at link time
2018-11-05 10:25:36 +08:00
endif()
osx clang quench deprecated api errors after approach by redhat https://bugzilla.redhat.com/show_bug.cgi?id=1155181 Signed-off-by: Andy Green <andy.green@linaro.org>
2015-11-29 19:19:09 +08:00
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (WINCE)
list(APPEND LIB_LIST_AT_END ws2.lib)
elseif (WIN32)
openssl: windows: also link with crypt32 in case used by openssl
2020-08-05 12:01:14 +01:00
list(APPEND LIB_LIST_AT_END ws2_32.lib userenv.lib psapi.lib iphlpapi.lib crypt32.lib)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
endif()
if (MSVC)
# Turn off pointless microsoft security warnings.
add_definitions(-D_CRT_SECURE_NO_DEPRECATE -D_CRT_NONSTDC_NO_DEPRECATE)
# Fail the build if any warnings
add_compile_options(/W3 /WX)
endif(MSVC)
if (MINGW)
set(LWS_MINGW_SUPPORT 1)
set(CMAKE_C_FLAGS "-D__USE_MINGW_ANSI_STDIO ${CMAKE_C_FLAGS}")
add_definitions(-DWINVER=0x0601 -D_WIN32_WINNT=0x0601)
endif()
windows: for data exports use explicit extern LWS_EXTERN needs to be empty for windows when declaring functions in the headers. But for data, it needs the explicit extern otherwise on windows or mingw based builds, it thinks we are redeclaring the data each time.
2020-06-07 07:53:36 +01:00
if (HDR_PRIVATE)
source_group("Headers Private" FILES ${HDR_PRIVATE})
endif()
if (HDR_PUBLIC)
source_group("Headers Public" FILES ${HDR_PUBLIC})
endif()
if (SOURCES)
source_group("Sources" FILES ${SOURCES})
endif()
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
if (RESOURCES)
source_group("Resources" FILES ${RESOURCES})
endif()
Some more Cmake stuff.
2013-02-06 15:25:26 +09:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
Some more Cmake stuff.
2013-02-06 15:25:26 +09:00
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# ZLIB (needed for deflate extension and if LWS_WITH_HTTP_STREAM_COMPRESSION)
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
#
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
if (LWS_WITH_ZLIB)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (NOT ZLIB_FOUND)
cmake: miniz: add as option
2019-07-03 19:46:23 +01:00
if (LWS_WITH_MINIZ)
find_package(Miniz REQUIRED)
set(ZLIB_INCLUDE_DIRS ${MINIZ_INCLUDE_DIRS})
set(ZLIB_LIBRARIES ${MINIZ_LIBRARIES})
else()
find_package(ZLIB REQUIRED)
endif()
Added "make install" support to the CMake project. - Does everything as the autoconf script does (I think). - Generates SSL certs for the test-server - Installs all test apps. - Defaults to /usr/local/ - "DESTDIR=/bla make install" works - Append "libwebsockets" to start of the test-app names like the autoconf stuff does. - Only tested on OSX so far...
2013-02-19 10:14:37 +08:00
endif()
cmake: miniz: add as option
2019-07-03 19:46:23 +01:00
message("zlib/miniz include dirs: ${ZLIB_INCLUDE_DIRS}")
message("zlib/miniz libraries: ${ZLIB_LIBRARIES}")
Added "make install" support to the CMake project. - Does everything as the autoconf script does (I think). - Generates SSL certs for the test-server - Installs all test apps. - Defaults to /usr/local/ - "DESTDIR=/bla make install" works - Append "libwebsockets" to start of the test-app names like the autoconf stuff does. - Only tested on OSX so far...
2013-02-19 10:14:37 +08:00
include_directories(${ZLIB_INCLUDE_DIRS})
cmake: make sure libz is at end of link if enabled https://github.com/warmcat/libwebsockets/issues/1840
2020-02-06 06:51:19 +00:00
# done later at end of link list
# list(APPEND LIB_LIST ${ZLIB_LIBRARIES})
windows: fix VS cmake build error #1929
2020-05-28 21:55:08 +03:00
set(CMAKE_REQUIRED_LIBRARIES ${CMAKE_REQUIRED_LIBRARIES} ${ZLIB_LIBRARIES})
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
list(APPEND LIB_LIST_AT_END ${ZLIB_LIBRARIES})
Subject: [PATCH] Multiple changes in the build process * Default CMAKE_BUILD_TYPE to Release * Use CMAKE_BUILD_TYPE to define _DEBUG (NDEBUG is more standard though, but cmake defines it) * Drop LWS_WITHOUT_DEBUG (use CMAKE_BUILD_TYPE for that) * Drop LWS_NO_EXTERNAL_POLL (was not used) * Drop CMAKE_BUILD (CMake is the only build system now) * Add LWS_WITH_STATIC and LWS_WITH_SHARED to choose what version(s) to build * Add LWS_XXX_LIBRARIES and LWS_XXX_INCLUDE_DIRS for each library dependency (zlib, openssl, libev, cyassl) * Support setting of XXX_LIBRARIES, XXX_LIBRARIES and XXX_INCLUDE_DIRS by parent project (when included with add_subdirectory()) * Rename LWS_USE_EXTERNAL_ZLIB to LWS_USE_BUNDLED_ZLIB and default it to NO (since it's Windows only) * Default LWS_WITHOUT_DAEMONIZE to NO (since network lib shouldn't know how to do it anyway) * Rename config.h.cmake to lws_config.h.in * Rename shared library to websockets_shared so linker will not be confused * Fix inline keyword detection in clang * Explicitly set MACOSX_RPATH to YES on MacOS
2015-04-22 10:52:13 -07:00
endif()
Started redoing CMake support based on the up to date repos
2013-02-01 14:52:53 +01:00
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
lws_fsmount: overlayfs apis
2020-02-15 10:09:50 +00:00
if (LWS_WITH_FSMOUNT AND ${CMAKE_SYSTEM_NAME} STREQUAL "Linux")
cmakelists: clean out pthreads and normalize libmount args
2020-06-03 17:14:03 +01:00
if (NOT LWS_LIBMOUNT_INCLUDE_DIRS STREQUAL "")
include_directories(${LWS_LIBMOUNT_INCLUDE_DIRS})
message("libmount include dir: ${LWS_LIBMOUNT_INCLUDE_DIRS}")
endif()
if (NOT LWS_LIBMOUNT_LIBRARIES STREQUAL "")
message("libmount libraries: ${LWS_LIBMOUNT_LIBRARIES}")
list(APPEND LIB_LIST ${LWS_LIBMOUNT_LIBRARIES})
else()
list(APPEND LIB_LIST mount)
endif()
lws_fsmount: overlayfs apis
2020-02-15 10:09:50 +00:00
endif()
event-lib-glib Adds support for glib event library.
2020-02-07 11:39:32 +00:00
protocol generic sessions Signed-off-by: Andy Green <andy@warmcat.com>
2016-05-19 15:28:31 +08:00
if (LWS_WITH_SQLITE3)
if (NOT SQLITE3_FOUND)
find_path(SQLITE3_INCLUDE_DIRS NAMES sqlite3.h)
find_library(SQLITE3_LIBRARIES NAMES sqlite3)
if(SQLITE3_INCLUDE_DIRS AND SQLITE3_LIBRARIES)
set(SQLITE3_FOUND 1)
endif()
endif()
message("sqlite3 include dir: ${SQLITE3_INCLUDE_DIRS}")
message("sqlite3 libraries: ${SQLITE3_LIBRARIES}")
include_directories("${SQLITE3_INCLUDE_DIRS}")
list(APPEND LIB_LIST ${SQLITE3_LIBRARIES})
endif()
http proxy: client: unix socket support This allows the client stuff to understand that addresses beginning with '+' represent unix sockets. If the first character after the '+' is '@', it understands that the '@' should be read as '\0', in order to use Linux "abstract namespace" sockets. Further the lws_parse_uri() helper is extended to understand the convention that an address starting with + is a unix socket, and treats the socket path as delimited by ':', eg http://+/var/run/mysocket:/my/path HTTP Proxy is updated to allow mounts to these unix socket paths. Proxy connections go out on h1, but are dynamically translated to h1 or h2 on the incoming side. Proxy usage of libhubbub is separated out... LWS_WITH_HTTP_PROXY is on by default, and LWS_WITH_HUBBUB is off by default.
2018-09-04 08:06:46 +08:00
if (LWS_WITH_HUBBUB)
Correct the library name of LIBHUBBUB_LIBRARIES It should be `hubbub` in `find_library` to make the function work.
2016-10-10 20:10:34 +08:00
find_library(LIBHUBBUB_LIBRARIES NAMES hubbub)
proxy rewrite If you enable -DLWS_WITH_HTTP_PROXY=1 at cmake, the test server has a new URI path http://localhost:7681/proxytest If you visit here, a client connection to http://example.com:80 is spawned, and the results piped on to your original connection. Also with LWS_WITH_HTTP_PROXY enabled at cmake, lws wants to link to an additional library, "libhubbub". This allows lws to do html rewriting on the fly, adjusting proxied urls in a lightweight and fast way.
2016-03-20 11:59:53 +08:00
list(APPEND LIB_LIST ${LIBHUBBUB_LIBRARIES} )
endif()
cmake: Link with cap if it's found and used
2020-07-19 16:29:07 +02:00
if (LWS_HAVE_LIBCAP)
find_library(LIBCAP_LIBRARIES NAMES cap)
list(APPEND LIB_LIST ${LIBCAP_LIBRARIES} )
endif()
plugins: add option LWS_WITH_PLUGINS_BUILTIN Add the ability to just build plugins into the main library. They are already designed to have a pinhole export for when they are used as dynamic lib plugins so their namespace does not conflict.
2021-01-31 05:50:12 +00:00
if (LWS_WITH_PLUGINS_BUILTIN)
add_subdirectory(plugins)
endif()
cmakelists: clean out pthreads and normalize libmount args
2020-06-03 17:14:03 +01:00
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# Append the "at end" pieces to the lib list
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
list(APPEND LIB_LIST ${LIB_LIST_AT_END})
qnx: add support https://libwebsockets.org/pipermail/libwebsockets/2018-January/003600.html
2018-01-04 10:37:25 +08:00
cmake: make sure libz is at end of link if enabled https://github.com/warmcat/libwebsockets/issues/1840
2020-02-06 06:51:19 +00:00
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# Second-level CMakeLists
cmake: make sure libz is at end of link if enabled https://github.com/warmcat/libwebsockets/issues/1840
2020-02-06 06:51:19 +00:00
#
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
include_directories("${PROJECT_SOURCE_DIR}/lib")
Subject: Haiku: initial platform support https://github.com/warmcat/libwebsockets/pull/1033
2017-09-24 15:30:42 +00:00
plugins: add option LWS_WITH_PLUGINS_BUILTIN Add the ability to just build plugins into the main library. They are already designed to have a pinhole export for when they are used as dynamic lib plugins so their namespace does not conflict.
2021-01-31 05:50:12 +00:00
add_subdirectory(lib)
capabilities support
2017-06-07 06:10:02 +08:00
openssl: make sure tests have -ldl v2 https://github.com/warmcat/libwebsockets/issues/1662
2020-01-17 12:33:28 +00:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if(WIN32 AND NOT CYGWIN)
set(DEF_INSTALL_CMAKE_DIR cmake)
else()
set(DEF_INSTALL_CMAKE_DIR lib${LIB_SUFFIX}/cmake/libwebsockets)
cmake: make sure libz is at end of link if enabled https://github.com/warmcat/libwebsockets/issues/1840
2020-02-06 06:51:19 +00:00
endif()
cmake: windows: try to auto include LwsCheckRequirements
2020-11-25 10:58:09 +00:00
configure_file(${PROJECT_SOURCE_DIR}/cmake/LwsCheckRequirements.cmake
${PROJECT_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/LwsCheckRequirements.cmake
@ONLY)
configure_file(${PROJECT_SOURCE_DIR}/cmake/LwsCheckRequirements.cmake
${PROJECT_BINARY_DIR}/LwsCheckRequirements.cmake
@ONLY)
windows: CHECK_FUNCTION_EXISTS workaround
2017-11-08 15:53:43 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# Generate version info for both build-tree and install-tree.
configure_file(${PROJECT_SOURCE_DIR}/cmake/libwebsockets-config-version.cmake.in
${PROJECT_BINARY_DIR}/libwebsockets-config-version.cmake
@ONLY)
esp32: shift to support latest esp-idf Esp-idf has an improved but still kind of abused cmake- based build system now. If we see ESP_PLATFORM coming as a cmake var, we can know we are being built from inside the esp-idf config system. Leave the existing esp32 arrangements alone but triggered off ESP_PLATFORM, adapt to use the cross toolchain file and various quirks automatically. In this way you can build lws a part of your project in a much cleaner way. Prepare a minimal esp32 test app for use in Sai Adapt .sai.json to build for esp32
2020-05-14 21:28:48 +01:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# Generate the config file for the build-tree.
set(LWS__INCLUDE_DIRS
"${PROJECT_SOURCE_DIR}/lib"
"${PROJECT_BINARY_DIR}")
set(LIBWEBSOCKETS_INCLUDE_DIRS ${LWS__INCLUDE_DIRS} CACHE PATH "Libwebsockets include directories")
configure_file(${PROJECT_SOURCE_DIR}/cmake/libwebsockets-config.cmake.in
${PROJECT_BINARY_DIR}/libwebsockets-config.cmake
@ONLY)
set(LWS_INSTALL_CMAKE_DIR ${DEF_INSTALL_CMAKE_DIR} CACHE PATH "Installation directory for CMake files")
mbedtls: finer-grained enable checks and OP-TEE
2019-01-11 13:14:32 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
# Export targets (This is used for other CMake projects to easily find the libraries and include files).
if (LWS_WITH_EXPORT_LWSTARGETS)
export(TARGETS ${LWS_LIBRARIES}
FILE "${PROJECT_BINARY_DIR}/LibwebsocketsTargets.cmake")
esp32: workaround cant find TLS_client_method
2017-04-07 11:29:57 +08:00
endif()
pipe2 where possible
2018-04-12 21:37:59 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
set(libwebsockets_DIR ${PROJECT_BINARY_DIR})
set(CMAKE_MODULE_PATH "${PROJECT_SOURCE_DIR}/cmake")
message("DIR ${libwebsockets_DIR} CMP ${CMAKE_MODULE_PATH}")
mbedtls: support older versions without net_sockets.h
2020-03-19 10:14:25 +00:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (LWS_WITH_MINIMAL_EXAMPLES)
add_subdirectory(minimal-examples)
endif()
mbedtls: support older versions without net_sockets.h
2020-03-19 10:14:25 +00:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (NOT LWS_WITHOUT_TESTAPPS)
add_subdirectory(test-apps)
endif()
tcp keepalive: user TCP_USER_TIMEOUT on linux if extant https://github.com/warmcat/libwebsockets/issues/1223
2018-04-18 18:45:27 +08:00
plugins: add option LWS_WITH_PLUGINS_BUILTIN Add the ability to just build plugins into the main library. They are already designed to have a pinhole export for when they are used as dynamic lib plugins so their namespace does not conflict.
2021-01-31 05:50:12 +00:00
if (NOT LWS_WITH_PLUGINS_BUILTIN)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
add_subdirectory(plugins)
plugins: add option LWS_WITH_PLUGINS_BUILTIN Add the ability to just build plugins into the main library. They are already designed to have a pinhole export for when they are used as dynamic lib plugins so their namespace does not conflict.
2021-01-31 05:50:12 +00:00
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
add_subdirectory(lwsws)
tcp keepalive: user TCP_USER_TIMEOUT on linux if extant https://github.com/warmcat/libwebsockets/issues/1223
2018-04-18 18:45:27 +08:00
Add testapps support for client certs and CRLs AG: plumb into cmake to avoid travis mac blowing up
2016-04-13 22:17:05 +02:00
# Generate the lws_config.h that includes all the public compilation settings.
configure_file(
clean up top level of project
2017-09-22 09:46:58 +08:00
"${PROJECT_SOURCE_DIR}/cmake/lws_config.h.in"
Add testapps support for client certs and CRLs AG: plumb into cmake to avoid travis mac blowing up
2016-04-13 22:17:05 +02:00
"${PROJECT_BINARY_DIR}/lws_config.h")
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
add_custom_command(
OUTPUT ${PROJECT_BINARY_DIR}/include/lws_config.h
${PROJECT_BINARY_DIR}/include/libwebsockets
${PROJECT_BINARY_DIR}/include/libwebsockets.h
COMMENT "Creating build include dir"
COMMAND ${CMAKE_COMMAND} -E copy ${CMAKE_CURRENT_SOURCE_DIR}/include/libwebsockets.h
${CMAKE_CURRENT_BINARY_DIR}/include/libwebsockets.h
COMMAND ${CMAKE_COMMAND} -E copy_directory ${CMAKE_CURRENT_SOURCE_DIR}/include/libwebsockets/
${CMAKE_CURRENT_BINARY_DIR}/include/libwebsockets
COMMAND ${CMAKE_COMMAND} -E copy ${PROJECT_BINARY_DIR}/lws_config.h
${CMAKE_CURRENT_BINARY_DIR}/include/lws_config.h
MAIN_DEPENDENCY ${PROJECT_BINARY_DIR}/lws_config.h
)
Add testapps support for client certs and CRLs AG: plumb into cmake to avoid travis mac blowing up
2016-04-13 22:17:05 +02:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
add_custom_target(GENHDR DEPENDS ${PROJECT_BINARY_DIR}/include/lws_config.h)
esp32: align fops member name defines
2017-03-08 11:11:41 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
file(GLOB HDR_PUBLIC1 RELATIVE ${CMAKE_CURRENT_SOURCE_DIR} include/libwebsockets/*.h)
file(GLOB HDR_PUBLIC2 RELATIVE ${CMAKE_CURRENT_SOURCE_DIR} include/libwebsockets.h)
file(GLOB HDR_PUBLIC3 RELATIVE ${CMAKE_CURRENT_SOURCE_DIR} ${CMAKE_CURRENT_BINARY_DIR}/include/lws_config.h)
list(APPEND HDR_PUBLIC ${HDR_PUBLIC1} ${HDR_PUBLIC2} ${HDR_PUBLIC3})
esp32: align fops member name defines
2017-03-08 11:11:41 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
set_source_files_properties(${HDR_PUBLIC} PROPERTIES GENERATED 1)
esp32: align fops member name defines
2017-03-08 11:11:41 +08:00
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (LWS_WITH_STATIC)
add_dependencies(websockets GENHDR)
esp32: align fops member name defines
2017-03-08 11:11:41 +08:00
endif()
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
if (LWS_WITH_SHARED)
add_dependencies(websockets_shared GENHDR)
esp32: align fops member name defines
2017-03-08 11:11:41 +08:00
endif()
Add testapps support for client certs and CRLs AG: plumb into cmake to avoid travis mac blowing up
2016-04-13 22:17:05 +02:00
CMake support + fixed windows build. - Finalized CMake support (tested on windows only so far). - Uses a generated lws_config.h that is included in private-libwebsocket to pass defines, only used if CMAKE_BUILD is set. - Support for SSL on Windows. - Initial support for CyaSSL replacement of OpenSSL (This has been added to my older CMake-fork but haven't been tested on this version yet). - Fixed windows build (see below for details). - Fixed at least the 32-bit Debug build for the existing Visual Studio Project. (Not to keen fixing all the others when we have CMake support anyway (which can generate much better project files)...) - BUGFIXES: - handshake.c - used C99 definition of handshake_0405 function - libwebsocket.c - syslog not available on windows, put in ifdefs. - Fixed previous known crash bug on Windows where WSAPoll in Ws2_32.dll would not be present, causing the poll function pointer being set to NULL. - Uninitialized variable context->listen_service_extraseen would result in stack overflow because of infinite recursion. Fixed by initializing in libwebsocket_create_context - SO_REUSADDR means something different on Windows compared to Unix. - Setting a socket to nonblocking is done differently on Windows. (This should probably broken out into a helper function instead) - lwsl_emit_syslog -> lwsl_emit_stderr on Windows. - private-libwebsocket.h - PATH_MAX is not available on Windows, define as MAX_PATH - Always define LWS_NO_DAEMONIZE on windows. - Don't define lws_latency as inline that does nothing. inline is not support by the Microsoft compiler, replaced with an empty define instead. (It's __inline in MSVC) - server.c - Fixed nonblock call on windows - test-ping.c - Don't use C99 features (Microsoft compiler does not support it). - Move non-win32 headers into ifdefs. - Skip use of sighandler on Windows. - test-server.c - ifdef syslog parts on Windows.
2013-02-06 15:26:58 +09:00
#
cmake and other fixes
2014-02-15 13:48:56 +08:00
#
# Installation preparations.
#
export(PACKAGE libwebsockets)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
install(DIRECTORY include/libwebsockets
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
DESTINATION "${LWS_INSTALL_INCLUDE_DIR}" COMPONENT dev)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
install(FILES ${PROJECT_BINARY_DIR}/include/libwebsockets.h ${PROJECT_BINARY_DIR}/include/lws_config.h
event libs: default to building as dynamically loaded plugins Event lib support as it has been isn't scaling well, at the low level libevent and libev headers have a namespace conflict so they can't both be built into the same image, and at the distro level, binding all the event libs to libwebsockets.so makes a bloaty situation for packaging, lws will drag in all the event libs every time. This patch implements the plan discussed here https://github.com/warmcat/libwebsockets/issues/1980 and refactors the event lib support so they are built into isolated plugins and bound at runtime according to what the application says it wants to use. The event lib plugins can be packaged individually so that only the needed sets of support are installed (perhaps none of them if the user code is OK with the default poll() loop). And dependent user code can mark the specific event loop plugin package as required so pieces are added as needed. The eventlib-foreign example is also refactored to build the selected lib support isolated. A readme is added detailing the changes and how to use them. https://libwebsockets.org/git/libwebsockets/tree/READMEs/README.event-libs.md
2020-08-27 15:37:14 +01:00
DESTINATION "${LWS_INSTALL_INCLUDE_DIR}" COMPONENT dev)
cmake and other fixes
2014-02-15 13:48:56 +08:00
# Generate the config file for the installation tree.
Fix installation into relative installation directory This is required to make the other CPack generators like "ZIP" work.
2014-02-28 16:29:28 +01:00
get_filename_component(LWS_ABSOLUTE_INSTALL_CMAKE_DIR ${LWS_INSTALL_CMAKE_DIR} ABSOLUTE)
get_filename_component(LWS_ABSOLUTE_INSTALL_INCLUDE_DIR ${LWS_INSTALL_INCLUDE_DIR} ABSOLUTE)
cmake and other fixes
2014-02-15 13:48:56 +08:00
file(RELATIVE_PATH
REL_INCLUDE_DIR
Fix installation into relative installation directory This is required to make the other CPack generators like "ZIP" work.
2014-02-28 16:29:28 +01:00
"${LWS_ABSOLUTE_INSTALL_CMAKE_DIR}"
"${LWS_ABSOLUTE_INSTALL_INCLUDE_DIR}") # Calculate the relative directory from the cmake dir.
cmake and other fixes
2014-02-15 13:48:56 +08:00
cmake: move LWS__INCLUDE_DIRS def to after REL_INCLIDE_DIR def Generate the config file for the installation tree fix wrong LWS__INCLUDE_DIRS definition
2020-12-24 16:18:46 +03:00
if (DEFINED REL_INCLUDE_DIR)
set(LWS__INCLUDE_DIRS "\${LWS_CMAKE_DIR}/${REL_INCLUDE_DIR}")
endif()
configure_file(${PROJECT_SOURCE_DIR}/cmake/libwebsockets-config.cmake.in
${PROJECT_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/libwebsockets-config.cmake
@ONLY)
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
set_target_properties(${LWS_LIBRARIES}
PROPERTIES PUBLIC_HEADER "${HDR_PUBLIC}")
plugins This adds support for dynamically loaded plugins at runtime, which can expose their own protocols or extensions transparently. With these changes lwsws defaults to OFF in cmake, and if enabled it automatically enables plugins and libuv support. Signed-off-by: Andy Green <andy@warmcat.com>
2016-04-06 16:15:40 +08:00
cmake additionally install cmake config
2015-12-08 10:26:57 +08:00
# Install the LibwebsocketsConfig.cmake and LibwebsocketsConfigVersion.cmake
install(FILES
cmakelist: Augean Stables refactor Establish a new distributed CMake architecture with CMake code related to a source directory moving to be in the subdir in its own CMakeLists.txt. In particular, there's now one in ./lib which calls through to ones further down the directory tree like ./lib/plat/xxx, ./lib/roles/xxx etc. This cuts the main CMakelists.txt from 98KB -> 33KB, about a 66% reduction, and it's much easier to maintain sub-CMakeLists.txt that are in the same directory as the sources they manage, and conceal all the details that that level. Child CMakelists.txt become responsible for: - include_directories() definition (this is not supported by CMake directly, it passes it back up via PARENT_SCOPE vars in helper macros) - Addition child CMakeLists.txt inclusion, for example toplevel -> role -> role subdir - Source file addition to the build - Dependent library path resolution... this is now a private thing in the child CMakeLists.txt, it just passes back any adaptations to include_directories() and the LIB_LIST without filling the parent namespace with the details
2020-05-22 16:47:40 +01:00
"${PROJECT_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/libwebsockets-config.cmake"
"${PROJECT_BINARY_DIR}/libwebsockets-config-version.cmake"
cmake: windows: try to auto include LwsCheckRequirements
2020-11-25 10:58:09 +00:00
"${PROJECT_BINARY_DIR}${CMAKE_FILES_DIRECTORY}/LwsCheckRequirements.cmake"
cmake additionally install cmake config
2015-12-08 10:26:57 +08:00
DESTINATION "${LWS_INSTALL_CMAKE_DIR}" COMPONENT dev)
revert cmake remove targets from install path Signed-off-by: Andy Green <andy@warmcat.com>
2016-03-29 18:27:32 +08:00
# Install exports for the install-tree.
Add a flag to disable CMake exports Add LWS_EXPORT_LWSTARGETS (defaults to ON) that can be used to disable CMake export targets as well as install export for libwebsockets. AG: add _WITH_
2017-11-20 12:41:29 +01:00
if (LWS_WITH_EXPORT_LWSTARGETS)
install(EXPORT LibwebsocketsTargets
DESTINATION "${LWS_INSTALL_CMAKE_DIR}" COMPONENT dev)
endif()
revert cmake remove targets from install path Signed-off-by: Andy Green <andy@warmcat.com>
2016-03-29 18:27:32 +08:00
cmake ignore non source things for make dist Signed-off-by: Andy Green <andy.green@linaro.org>
2013-03-10 08:11:48 +08:00
# build subdir is not part of sources
cmake: CPACK_SOURCE_IGNORE_FILES is a list of regular expressions
2018-05-26 09:19:54 +08:00
set(CPACK_SOURCE_IGNORE_FILES $(CPACK_SOURCE_IGNORE_FILES) "/.git/" "/build/" "\\\\.tgz$" "\\\\.tar\\\\.gz$")
cmake ignore non source things for make dist Signed-off-by: Andy Green <andy.green@linaro.org>
2013-03-10 08:11:48 +08:00
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
# Most people are more used to "make dist" compared to "make package_source"
fix cmake quoting A few parameters (like "${PROJECT_BINARY_DIR}/lws_config.h" and "${PROJECT_SOURCE_DIR}/lib") are passed to CMake commands in your build script without enclosing them by quotation marks. I see that these places will result in build difficulties if the contents of the used variables will contain special characters like spaces. Signed-off-by: Markus Elfring <markus.elfring@web.de>
2013-10-26 20:03:43 +08:00
add_custom_target(dist COMMAND "${CMAKE_MAKE_PROGRAM}" package_source)
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
cmake and other fixes
2014-02-15 13:48:56 +08:00
LWS_WITH_MINIMAL_EXAMPLES: default-off build all minimal samples after lws for QA This adds an lws cmake option that builds all the minimal examples as part of lws, it's useful for QA. It adds a macro to examples that depend on a particular lws configuration to understand they should just null out their project definition in builds where the lws configuration requirement is not met, and we are building as part of lws. It also adapts all the example library additions to select the just-built-but-not-yet-installed library in the case it is built as part of lws. If built standalone, it now uses the cmake platform-abstracted way to add the library requirement too.
2018-03-17 06:43:12 +08:00
Added CPack support + some more. - "make dist" will now produce a tar.gz file. This includes everything (not really what we want but a start). - Got rid of a bunch of defines and variables that aren't used. - Added the option LINK_TESTAPPS_DYNAMIC that enables linking to the shared library version of the lib for the test apps, instead of doing it statically like the default is. - Fixed proper support for the --prefix stuff (-DCMAKE_INSTALL_PREFIX for cmake). (Don't specify /usr/local explicitly, that is the default anyway and will break other platforms). - Note: I noticed a problem with the "INSTALL_DATADIR" define used by the test-apps. Since we hard code the path to the certs using this, doing "DESTDIR=/bla make install" will result in not being able to use the SSL mode for the test-apps since they won't find the certs. (This also applies to the autoconf project). Fixed this by setting "../share" as the default location instead of using the prefix.
2013-02-22 09:28:08 +08:00
# This must always be last!
include(CPack)
Reference in a new issue Copy permalink