Files
jak-project/third-party/curl/tests/libtest/lib1592.c
T
Alexander J. Semenuk 5d5e35fb9b fix: Windows toolchain compatibility (curl 8.21 re-vendor, endless reconfigure loop) (#4355)
## 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)
2026-07-27 19:19:18 -04:00

120 lines
4.6 KiB
C
Vendored
Generated

/***************************************************************************
* _ _ ____ _
* Project ___| | | | _ \| |
* / __| | | | |_) | |
* | (__| |_| | _ <| |___
* \___|\___/|_| \_\_____|
*
* Copyright (C) Daniel Stenberg, <daniel@haxx.se>, et al.
*
* This software is licensed as described in the file COPYING, which
* you should have received as part of this distribution. The terms
* are also available at https://curl.se/docs/copyright.html.
*
* You may opt to use, copy, modify, merge, publish, distribute and/or sell
* copies of the Software, and permit persons to whom the Software is
* furnished to do so, under the terms of the COPYING file.
*
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
* KIND, either express or implied.
*
* SPDX-License-Identifier: curl
*
***************************************************************************/
/*
* See https://github.com/curl/curl/issues/3371
*
* This test case checks whether curl_multi_remove_handle() cancels
* asynchronous DNS resolvers without blocking where possible. Obviously, it
* only tests whichever resolver curl is actually built with.
*/
/* We are waiting a generous two seconds for the removal. This is as low as we
can go while still easily supporting SIGALRM timing for the non-threaded
blocking resolver. It does not matter that much because when the test
passes, we never wait this long. We set it much higher via the default
TEST_HANG_TIMEOUT to avoid issues when running on overloaded CI machines. */
#include "first.h"
static CURLcode test_lib1592(const char *URL)
{
int stillRunning;
CURLM *multi = NULL;
CURL *curl = NULL;
CURLcode result = CURLE_OK;
CURLMcode mresult;
long timeout;
global_init(CURL_GLOBAL_ALL);
multi_init(multi);
easy_init(curl);
easy_setopt(curl, CURLOPT_VERBOSE, 1L);
easy_setopt(curl, CURLOPT_URL, URL);
/* Set a DNS server that hopefully does not respond when using c-ares. */
if(curl_easy_setopt(curl, CURLOPT_DNS_SERVERS, "0.0.0.0") == CURLE_OK)
/* Since we could set the DNS server, presume we are working with a
resolver that can be cancelled (i.e. c-ares). Thus,
curl_multi_remove_handle() should not block even when the resolver
request is outstanding. Thus, set a request timeout _longer_ than the
test hang timeout so we fail if the handle removal call incorrectly
blocks. */
timeout = TEST_HANG_TIMEOUT * 2;
else {
/* If we cannot set the DNS server, presume that we are configured to use
a resolver that cannot be cancelled (i.e. the threaded resolver or the
non-threaded blocking resolver). Thus, we test that the
curl_multi_remove_handle() call does finish well within our test
timeout.
But, it is unlikely that the resolver request takes any time at
all because we have not been able to configure the resolver to use an
non-responsive DNS server. At least we exercise the flow. */
curl_mfprintf(stderr,
"CURLOPT_DNS_SERVERS not supported; "
"assuming curl_multi_remove_handle() does block\n");
timeout = TEST_HANG_TIMEOUT / 2;
}
/* Setting a timeout on the request should ensure that even if we have to
wait for the resolver during curl_multi_remove_handle(), it does not take
longer than this, because the resolver request inherits its timeout from
this. */
easy_setopt(curl, CURLOPT_TIMEOUT_MS, timeout);
multi_add_handle(multi, curl);
/* This should move the handle from INIT => CONNECT => WAITRESOLVE. */
curl_mfprintf(stderr, "curl_multi_perform()...\n");
multi_perform(multi, &stillRunning);
curl_mfprintf(stderr, "curl_multi_perform() succeeded\n");
/* Start measuring how long it takes to remove the handle. */
curl_mfprintf(stderr, "curl_multi_remove_handle()...\n");
start_test_timing();
mresult = curl_multi_remove_handle(multi, curl);
if(mresult) {
curl_mfprintf(stderr, "curl_multi_remove_handle() failed, with code %d\n",
mresult);
result = TEST_ERR_MULTI;
goto test_cleanup;
}
curl_mfprintf(stderr, "curl_multi_remove_handle() succeeded\n");
/* Fail the test if it took too long to remove. This happens after the fact,
and says "it seems that it would have run forever", which is not true, but
it is close enough, and simple to do. */
abort_on_test_timeout();
test_cleanup:
curl_easy_cleanup(curl);
curl_multi_cleanup(multi);
curl_global_cleanup();
return result;
}