Alexander J. Semenuk
5d5e35fb9b
## Problem Windows builds break with a current local toolchain (Scoop LLVM 22.1.8, CMake 4.4.0, VS 2026), in two independent ways: 1. The build stops at curl's deliberate guard: `#error "no non-blocking method was found/used/set"` in `third-party/curl/lib/nonblock.c`. 2. From the second configure onward, `cmake --build` re-runs CMake in an endless loop (observed 42 consecutive reconfigure cycles in a single build). Likely the same mechanism behind the "endlessly building" VS 2026 note in `docs/setup/dev/vs.md`. ## Root cause 1. `third-party/curl/CMake/CurlTests.c` passes `int *` to `ioctlsocket()`, whose third parameter is `u_long *`. Clang 22 promotes `-Wincompatible-pointer-types` to a hard error in C, so the `HAVE_IOCTLSOCKET_FIONBIO` try_compile silently fails and `curl_config.h` never defines it. Upstream CI does not see this because the windows-2022 runner image ships an older LLVM. GCC 14 promotes the same warning to a hard error, which is very likely the `CurlTests.c.obj` failure reported from MSYS2 in open-goal/jak-project#3551. Upstream curl hit the identical problem with GCC 14 and fixed the probe in curl 8.8.0 (curl/curl#13578). 2. The root CMakeLists copies the build tree's `compile_commands.json` into `<src>/build/` for clangd using `configure_file()`, which registers its input as a configure dependency. CMake rewrites `compile_commands.json` late in every generation, after `CTestTestfile.cmake` and `cmake_install.cmake` (outputs of the same Ninja regen rule), so once the dependency is registered the rule is deterministically dirty and every `ninja` invocation re-runs CMake. A pristine first configure is safe (the file does not exist yet, so the `if(EXISTS ...)` guard skips the copy), which is why the loop looks machine- or IDE-specific. ## Fix 1. Per review, re-vendor `third-party/curl` at the `curl-8_21_0` tag (previously `curl-8_3_0`), which carries the upstream probe fix plus two years of upstream development; `vendor.yaml` updated to match. Adjustments the version jump forced: - curl 8.15 removed the native macOS Secure Transport backend (`CURL_USE_SECTRANSP`), so macOS now builds curl against OpenSSL like Linux. The two macOS workflows install Homebrew `openssl@3` and export `OPENSSL_ROOT_DIR` (keg-only), and the macOS setup docs gained the same two lines. - `CURL_BROTLI` / `CURL_ZSTD` switched to AUTO-detection in curl 8.10; pinned OFF to keep the previous no-compression behavior and avoid silently linking whatever the CI images happen to have. - curl's new top-level `BUILD_EXAMPLES` cache option (default ON) leaked into discord-rpc's identically named option and broke configure at a nonexistent `examples/send-presence` directory; pinned OFF ahead of the third-party subdirectories. The diff is dominated by the mechanical tag-tree swap under `third-party/curl` (linguist-vendored, collapsed in review). The hand-written changes are `CMakeLists.txt`, the two macOS workflows, `docs/setup/system/macos.md`, and `vendor.yaml`. 2. Swap `configure_file()` for `file(COPY ...)`: the same clangd copy with no configure dependency registered. (`file(COPY_FILE ... ONLY_IF_DIFFERENT)` would be cleaner still but requires CMake 3.21, above the declared `cmake_minimum_required(VERSION 3.10)`.) ## Test plan - [x] Fresh `cmake --preset Release-windows-clang` (LLVM 22, no cache seeding) completes and logs `Enabled SSL backends: Schannel`; the FIONBIO probe passes without the previous `#error` - [x] Full Windows Release build from scratch in the branch worktree (all 1422 targets) - [x] goalc-test suite: 1509 passed, 0 failed - [x] Second consecutive configure with `compile_commands.json` present: the regen rule in `build.ninja` has no `compile_commands.json` input; `<src>/build/compile_commands.json` is still refreshed for clangd - [x] Repeated `ninja` invocations after a full build no longer re-run CMake - [x] macOS Intel and ARM CI green (first exercise of the OpenSSL backend switch) --- I work off a self-hosted forge, so this GitHub account is quiet; the configure logs and ninja dirty-node traces from the investigation are available if anyone wants the raw data. (AI-assisted)
10 KiB
Vendored
Generated
c, SPDX-License-Identifier, Title, Section, Source, See-also, Protocol, Added-in
| c | SPDX-License-Identifier | Title | Section | Source | See-also | Protocol | Added-in | |||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al. | curl | libcurl | 3 | libcurl |
|
|
n/a |
NAME
libcurl - client-side URL transfers
DESCRIPTION
This is a short overview on how to use libcurl in your C programs. There are specific man pages for each function mentioned in here. See libcurl-easy(3), libcurl-multi(3), libcurl-share(3), libcurl-url(3), libcurl-ws(3) and libcurl-tutorial(3) for in-depth understanding on how to program with libcurl.
There are many bindings available that bring libcurl access to your favorite language. Look elsewhere for documentation on those.
TRANSFERS
To transfer files, you create an "easy handle" using curl_easy_init(3) for a single individual transfer (in either direction). You then set your desired set of options in that handle with curl_easy_setopt(3). Options you set with curl_easy_setopt(3) stick. They are then used for every repeated use of this handle until you either change the option, or you reset them all with curl_easy_reset(3).
To actually transfer data you have the option of using the "easy" interface, or the "multi" interface.
The easy interface is a synchronous interface with which you call curl_easy_perform(3) and let it perform the transfer. When it is completed, the function returns and you can continue. More details are found in the libcurl-easy(3) man page.
The multi interface on the other hand is an asynchronous interface, that you call and that performs only a little piece of the transfer on each invoke. It is perfect if you want to do things while the transfer is in progress, or similar. The multi interface allows you to select() on libcurl action, and even to easily download multiple files simultaneously using a single thread. See further details in the libcurl-multi(3) man page.
SUPPORT INTERFACES
There is also a series of other helpful functions and interface families to use, including these:
curl_version_info()
gets detailed libcurl (and other used libraries) version info. See curl_version_info(3)
curl_getdate()
converts a date string to time_t. See curl_getdate(3)
curl_easy_getinfo()
get information about a performed transfer. See curl_easy_getinfo(3)
curl_mime_addpart()
helps building an HTTP form POST. See curl_mime_addpart(3)
curl_slist_append()
builds a linked list. See curl_slist_append(3)
Sharing data between transfers
You can have multiple easy handles share certain data, even if they are used in different threads. Set that up using the share interface, as described in the libcurl-share(3) man page.
URL Parsing
URL parsing and manipulations. See libcurl-url(3)
WebSocket communication
See libcurl-ws(3)
LINKING WITH LIBCURL
On Unix-like machines, there is a tool named curl-config that gets installed with the rest of the curl stuff when 'make install' is performed.
curl-config is added to make it easier for applications to link with libcurl and developers to learn about libcurl and how to use it.
Run 'curl-config --libs' to get the (additional) linker options you need to link with the particular version of libcurl you have installed. See the curl-config(1) man page for further details.
Unix-like operating system that ship libcurl as part of their distributions often do not provide the curl-config tool, but only install the library and headers in the common path for this purpose.
Many Linux and similar systems use pkg-config to provide build and link options about libraries and libcurl supports that as well.
LIBCURL SYMBOL NAMES
All public functions in the libcurl interface are prefixed with 'curl_' (with a lowercase c). You can find other functions in the library source code, but other prefixes indicate that the functions are private and may change without further notice in the next release.
Only use documented functions and functionality.
PORTABILITY
libcurl works exactly the same, on any of the platforms it compiles and builds on.
THREADS
libcurl is thread-safe but there are a few exceptions. Refer to libcurl-thread(3) for more information.
PERSISTENT CONNECTIONS
Persistent connections means that libcurl can reuse the same connection for several transfers, if the conditions are right.
libcurl always attempts to use persistent connections. Whenever you use curl_easy_perform(3) or curl_multi_perform(3) etc, libcurl attempts to use an existing connection to do the transfer, and if none exists it opens a new one that is subject for reuse on a possible following call to curl_easy_perform(3) or curl_multi_perform(3).
To allow libcurl to take full advantage of persistent connections, you should do as many of your file transfers as possible using the same handle.
If you use the easy interface, and you call curl_easy_cleanup(3), all the possibly open connections held by libcurl are closed and forgotten.
When you have created a multi handle and are using the multi interface, the connection pool is instead kept in the multi handle so closing and creating new easy handles to do transfers do not affect them. Instead all added easy handles can take advantage of the single shared pool.
GLOBAL CONSTANTS
There are a variety of constants that libcurl uses, mainly through its internal use of other libraries, which are too complicated for the library loader to set up. Therefore, a program must call a library function after the program is loaded and running to finish setting up the library code. For example, when libcurl is built for SSL capability via the GNU TLS library, there is an elaborate tree inside that library that describes the SSL protocol.
curl_global_init(3) is the function that you must call. This may allocate resources (e.g. the memory for the GNU TLS tree mentioned above), so the companion function curl_global_cleanup(3) releases them.
If libcurl was compiled with support for multiple SSL backends, the function curl_global_sslset(3) can be called before curl_global_init(3) to select the active SSL backend.
The global constant functions are thread-safe since libcurl 7.84.0 if curl_version_info(3) has the CURL_VERSION_THREADSAFE feature bit set (most platforms). Read libcurl-thread(3) for thread-safety guidelines.
If the global constant functions are not thread-safe, then you must not call them when any other thread in the program is running. It is not good enough that no other thread is using libcurl at the time, because these functions internally call similar functions of other libraries, and those functions are similarly thread-unsafe. You cannot generally know what these libraries are, or whether other threads are using them.
If the global constant functions are not thread-safe, then the basic rule for constructing a program that uses libcurl is this: Call curl_global_init(3), with a CURL_GLOBAL_ALL argument, immediately after the program starts, while it is still only one thread and before it uses libcurl at all. Call curl_global_cleanup(3) immediately before the program exits, when the program is again only one thread and after its last use of libcurl.
It is not actually required that the functions be called at the beginning and end of the program -- that is usually the easiest way to do it.
You can call both of these multiple times, as long as all calls meet these requirements and the number of calls to each is the same.
The global constant situation merits special consideration when the code you are writing to use libcurl is not the main program, but rather a modular piece of a program, e.g. another library. As a module, your code does not know about other parts of the program -- it does not know whether they use libcurl or not. Its code does not necessarily run at the start and end of the whole program.
A module like this must have global constant functions of its own, like curl_global_init(3) and curl_global_cleanup(3). The module thus has control at the beginning and end of the program and has a place to call the libcurl functions. If multiple modules in the program use libcurl, they all separately call the libcurl functions, and that is OK because only the first curl_global_init(3) and the last curl_global_cleanup(3) in a program change anything. (libcurl uses a reference count in static memory).
In a C++ module, it is common to deal with the global constant situation by defining a special class that represents the global constant environment of the module. A program always has exactly one object of the class, in static storage. That way, the program automatically calls the constructor of the object as the program starts up and the destructor as it terminates. As the author of this libcurl-using module, you can make the constructor call curl_global_init(3) and the destructor call curl_global_cleanup(3) and satisfy libcurl's requirements without your user having to think about it. (Caveat: If you are initializing libcurl from a Windows DLL you should not initialize it from DllMain or a static initializer because Windows holds the loader lock during that time and it could cause a deadlock.)
curl_global_init(3) has an argument that tells what particular parts of the global constant environment to set up. In order to successfully use any value except CURL_GLOBAL_ALL (which says to set up the whole thing), you must have specific knowledge of internal workings of libcurl and all other parts of the program of which it is part.
A special part of the global constant environment is the identity of the memory allocator. curl_global_init(3) selects the system default memory allocator, but you can use curl_global_init_mem(3) to supply one of your own. There is no way to use curl_global_init_mem(3) in a modular program -- all modules in the program that might use libcurl would have to agree on one allocator.
There is a failsafe in libcurl that makes it usable in simple situations without you having to worry about the global constant environment at all: curl_easy_init(3) sets up the environment itself if it has not been done yet. The resources it acquires to do so get released by the operating system automatically when the program exits.
This failsafe feature exists mainly for backward compatibility because there was a time when the global functions did not exist. Because it is sufficient only in the simplest of programs, it is not recommended for any program to rely on it.