When `CURLOPT_HAPPY_EYEBALLS_TIMEOUT_MS` expires, start the next ip
connect attempt, but keep all ongoing attempts alive.
Separate happy-eyeballs connection filter into own source files.
Closes#18105
To simplify dependencies, and sync tunits and units builds further.
`curlcheck.h` already depended on logic implemented within libtests:
it referenced a global variable (`unitfail`) defined in `first.c` and
declared in `test.h`.
Also:
- rename to `unitcheck.h` to indicate it's meant for unit tests.
- make `unitcheck.h` include `first.h` instead of `test.h`.
This brings header use closer to libtests. It also includes
`curlx/curlx.h` for all unit tests by default now.
- move `unitfail` declaration from `test.h` to `first.h`.
To match its definition in `first.c`.
- drop now redundant per-test curlx header includes.
Closes#17868
To make all src and test code refer to curlx headers the same way.
Also:
- src: move `curlx.h` include to `tool_setup.h`.
- src/tool_setup.h: drop stray `curlx/timeval.h`.
- servers: de-duplicate `curlx.h` and `curl_setup.h` includes.
- libtests, units: drop stray curlx sub-headers in favor of
`<curlx/curlx.h>`.
- tests: include `curlx.h` with `<>` instead of `""`. To match
other parts of the codebase.
Closes#17680
Connection filters had a method `get_host()` which had not really been
documented. Since then, the cf had the `query()` method added. Replace
the separate get_host with query.
Add `CF_QUERY_HOST_PORT` as query to connection filters to retrieve
which remote hostname and port the filter (or its sub-filter) is talking
to. The query is implemented by HTTP and SOCKS filters, all others pass
it through.
Add `Curl_conn_get_current_host()` to retrieve the remote host and port
for a connection. During connect, this will return the host the
connection is talking to right now. Before/After connect, this will
return `conn->host.name`.
This is used by SASL authentication.
Closes#17419
Make test bundles the default. Drop non-bundle build mode.
Also do all the optimizations and tidy-ups this allows, simpler builds,
less bundle exceptions, streamlined build mechanics.
Also rework the init/deinit macro magic for unit tests. The new method
allows using unique init/deinit function names, and calling them with
arguments. This is in turn makes it possible to reduce the use of global
variables.
Note this drop existing build options `-DCURL_TEST_BUNDLES=` from cmake
and `--enable-test-bundles` / `--disable-test-bundles` from autotools.
Also:
- rename test entry functions to have unique names: `test_<testname>`
This removes the last exception that was handled in the generator.
- fix `make dist` to not miss test sources with test bundles enabled.
- sync and merge `tests/mk-bundle.pl` into `scripts/mk-unity.pl`.
- mk-unity.pl: add `--embed` option and use it when `CURL_CLANG_TIDY=ON`
to ensure that `clang-tidy` does not miss external test C sources.
(because `clang-tidy` ignores code that's #included.)
- tests/unit: drop no-op setup/stop functions.
- tests: reduce symbol scopes, global macros, other fixes and tidy-ups.
- tool1621: fix to run, also fix it to pass.
- sockfilt: fix Windows compiler warning in certain unity include order,
by explicitly including `warnless.h`.
Follow-up to 6897aeb105#17468Closes#17590
Using a mixture of techniques to avoid symbols collisions:
- reduce scope.
- add `t*_` / `T*_` prefix.
- move shared functions to `testutil.c`.
(`suburl()`, `rlim2str()`)
- clone re-used lib*.c sources.
(lib587, lib645)
- include shared symbols just once in re-used `lib*.c` sources.
(using `LIB*_C` guards.)
- drop re-used `lib*.c` sources where they were identical or
unused.
- make macros global.
- #undef macros before use.
What remain is the entry functions `test`, and `unit_setup`,
`unit_stop` in unit tests.
Also:
- fix formatting and other minor things along the way.
- add `const` where possible.
- sync some symbol names between tests.
- drop `mk-bundle-hints.sh` that's no longer necessary.
Closes#17468
Move curlx_ functions into its own subdir.
The idea is to use the curlx_ prefix proper on these functions, and use
these same function names both in tool, lib and test suite source code.
Stop the previous special #define setup for curlx_ names.
The printf defines are now done for the library alone. Tests no longer
use the printf defines. The tool code sets its own defines. The printf
functions are not curlx, they are publicly available.
The strcase defines are not curlx_ functions and should not be used by
tool or server code.
dynbuf, warnless, base64, strparse, timeval, timediff are now proper
curlx functions.
When libcurl is built statically, the functions from the library can be
used as-is. The key is then that the functions must work as-is, without
having to be recompiled for use in tool/tests. This avoids symbol
collisions - when libcurl is built statically, we use those functions
directly when building the tool/tests. When libcurl is shared, we
build/link them separately for the tool/tests.
Assisted-by: Jay Satiro
Closes#17253
Make it possible to build curl for Windows CE using the CeGCC toolchain.
With both CMake and autotools, including tests and examples, also in CI.
The build configuration is the default one with Schannel enabled. No
3rd-party dependencies have been tested.
Also revive old code to make Schannel build with Windows CE, including
certificate verification.
Builds have been throughougly tested. But, I've made no functional tests
for this PR. Some parts (esp. file operations, like truncate and seek)
are stubbed out and likely broken as a result. Test servers build, but
they do not work on Windows CE. This patch substitutes `fstat()` calls
with `stat()`, which operate on filenames, not file handles. This may or
may not work and/or may not be secure.
About CeGCC: I used the latest available macOS binary build v0.59.1
r1397 from 2009, in native `mingw32ce` build mode. CeGCC is in effect
MinGW + GCC 4.4.0 + old/classic-mingw Windows headers. It targets
Windows CE v3.0 according to its `_WIN32_WCE` value. It means this PR
restores portions of old/classic-mingw support. It makes the Windows CE
codepath compatible with GCC 4.4.0. It also adds workaround for CMake,
which cannot identify and configure this toolchain out of the box.
Notes:
- CMake doesn't recognize CeGCC/mingw32ce, necessitating tricks as seen
with Amiga and MS-DOS.
- CMake doesn't set `MINGW` for mingw32ce. Set it and `MINGW32CE`
manually as a helper variable, in addition to `WINCE` which CMake sets
based on `CMAKE_SYSTEM_NAME`.
- CMake fails to create an implib for `libcurl.dll`, due to not
recognizing the platform as a Windowsy one. This patch adds the
necessary workaround to make it work.
- headers shipping with CeGCC miss some things curl needs for Schannel
support. Fixed by restoring and renovating code previously deleted
old-mingw code.
- it's sometime non-trivial to figure out if a fallout is WinCE,
mingw32ce, old-mingw, or GCC version-specific.
- WinCE is always Unicode. With exceptions: no `wmain`,
`GetProcAddress()`.
- `_fileno()` is said to convert from `FILE *` to `void *` which is
a Win32 file `HANDLE`. (This patch doesn't use this, but with further
effort it probably could be.)
https://stackoverflow.com/questions/3989545/how-do-i-get-the-file-handle-from-the-fopen-file-structure
- WinCE has no signals, current directory, stdio/CRT file handles, no
`_get_osfhandle()`, no `errno`, no `errno.h`. Some of this stuff is
standard C89, yet missing from this platform. Microsoft expects
Windows CE apps to use Win32 file API and `FILE *` exclusively.
- revived CeGCC here (not tested for this PR):
https://building.enlyze.com/posts/a-new-windows-ce-x86-compiler-in-2024/
On `UNDER_CE` vs. `_WIN32_WCE`: (This patch settled on `UNDER_CE`)
- A custom VS2008 WinCE toolchain does not set any of these.
The compiler binaries don't contain these strings, and has no compiler
option for targeting WinCE, hinting that a vanilla toolchain isn't
setting any of them either.
- `UNDER_CE` is automatically defined by the CeGCC compiler.
https://cegcc.sourceforge.net/docs/details.html
- `UNDER_CE` is similar to `_WIN32`, except it's not set automatically
by all compilers. It's not supposed to have any value, like a version.
(Though e.g. OpenSSL sets it to a version)
- `_WIN32_WCE` is the CE counterpart of the non-CE `_WIN32_WINNT` macro.
That does return the targeted Windows CE version.
- `_WIN32_WCE` is not defined by compilers, and relies on a header
setting it to a default, or the build to set it to the desired target
version. This is also how `_WIN32_WINNT` works.
- `_WIN32_WCE` default is set by `windef.h` in CeGCC.
- `_WIN32_WCE` isn't set to a default by MSVC Windows CE headers (the
ones I checked at least).
- CMake sets `_WIN32_WCE=<ver>`, `UNDER_CE`, `WINCE` for MSVC WinCE.
- `_WIN32_WCE` seems more popular in other projects, including CeGCC
itself. `zlib` is a notable exception amongst curl dependencies,
which uses `UNDER_CE`.
- Since `_WIN32_WCE` needs "certain" headers to have it defined, it's
undefined depending on headers included beforehand.
- `curl/curl.h` re-uses `_WIN32_WCE`'s as a self-guard, relying on
its not-(necessarily)-defined-by-default property:
25b445e479/include/curl/curl.h (L77)
Toolchain downloads:
- Windows:
https://downloads.sourceforge.net/cegcc/cegcc/0.59.1/cegcc_mingw32ce_cygwin1.7_r1399.tar.bz2
- macOS Intel:
https://downloads.sourceforge.net/cegcc/cegcc/0.59.1/cegcc_mingw32ce_snowleopard_r1397.tar.bz2Closes#15975
Remove `blocking` argument from cfilter's connect method.
Implement blocking behaviour in Curl_conn_connect() instead for all
filter chains.
Update filters implementations. Several of which did never use the
paramter (QUIC for example). Simplifies connect handling in TLS filters
that no longer need to loop
Fixed a blocking connect call in FTP when waiting on a socket accept()
which only worked because the filter did not implement it.
Closes#16397
Always try ipv6 addresses first, ipv4 second after a delay.
If neither ipv4/6 are amongst the supplied addresses, start a happy
eyeballer for the first address family present. This is for AF_UNIX
connects.
Fixes#14761
Reported-by: janedenone on hackerone
Closes#14768
- disbable this test on WIN32 platforms. It uses the file describtor '1'
as valid socket without events. Not portable.
- reduce trace output somewhat on other runs
Fixes#14177
Reported-by: Viktor Szakats
Closes#14191
- add a DEBUGASSERT for when a transfer's pollset should not be empty.
- move write unpausing from transfer loop into curl_easy_pause. This
make sure that the url_updatesocket() finds the correct state when
updating socket events.
- fix HTTP/2 proxy during connect phase to set sockets correctly
- fix test2600 to simulate a socket set
- move write unpausing from transfer loop into curl_easy_pause. This
make sure that the url_updatesocket() finds the correct state when
updating socket events.
- waiting for the resolver to deliver might not involve any sockets to
wait for. Do not generate a warning.
Fixes#14047Closes#14074
This adds connection shutdown infrastructure and first use for FTP. FTP
data connections, when not encountering an error, are now shut down in a
blocking way with a 2sec timeout.
- add cfilter `Curl_cft_shutdown` callback
- keep a shutdown start timestamp and timeout at connectdata
- provide shutdown timeout default and member in
`data->set.shutdowntimeout`.
- provide methods for starting, interrogating and clearing
shutdown timers
- provide `Curl_conn_shutdown_blocking()` to shutdown the
`sockindex` filter chain in a blocking way. Use that in FTP.
- add `Curl_conn_cf_poll()` to wait for socket events during
shutdown of a connection filter chain.
This gets the monitoring sockets and events via the filters
"adjust_pollset()" methods. This gives correct behaviour when
shutting down a TLS connection through a HTTP/2 proxy.
- Implement shutdown for all socket filters
- for HTTP/2 and h2 proxying to send GOAWAY
- for TLS backends to the best of their capabilities
- for tcp socket filter to make a final, nonblocking
receive to avoid unwanted RST states
- add shutdown forwarding to happy eyeballers and
https connect ballers when applicable.
Closes#13904
Before this patch, two macros were used to guard IPv6 features in curl
sources: `ENABLE_IPV6` and `USE_IPV6`. This patch makes the source use
the latter for consistency with other similar switches.
`-DENABLE_IPV6` remains accepted for compatibility as a synonym for
`-DUSE_IPV6`, when passed to the compiler.
`ENABLE_IPV6` also remains the name of the CMake and `Makefile.vc`
options to control this feature.
Closes#13349
GCC 14 introduces a new -Walloc-size included in -Wextra which gives:
```
src/tool_operate.c: In function ‘add_per_transfer’:
src/tool_operate.c:213:5: warning: allocation of insufficient size ‘1’ for type ‘struct per_transfer’ with size ‘480’ [-Walloc-size]
213 | p = calloc(sizeof(struct per_transfer), 1);
| ^
src/var.c: In function ‘addvariable’:
src/var.c:361:5: warning: allocation of insufficient size ‘1’ for type ‘struct var’ with size ‘32’ [-Walloc-size]
361 | p = calloc(sizeof(struct var), 1);
| ^
```
The calloc prototype is:
```
void *calloc(size_t nmemb, size_t size);
```
So, just swap the number of members and size arguments to match the
prototype, as we're initialising 1 struct of size `sizeof(struct
...)`. GCC then sees we're not doing anything wrong.
Closes#12292
Connection filter had a `get_select_socks()` method, inspired by the
various `getsocks` functions involved during the lifetime of a
transfer. These, depending on transfer state (CONNECT/DO/DONE/ etc.),
return sockets to monitor and flag if this shall be done for POLLIN
and/or POLLOUT.
Due to this design, sockets and flags could only be added, not
removed. This led to problems in filters like HTTP/2 where flow control
prohibits the sending of data until the peer increases the flow
window. The general transfer loop wants to write, adds POLLOUT, the
socket is writeable but no data can be written.
This leads to cpu busy loops. To prevent that, HTTP/2 did set the
`SEND_HOLD` flag of such a blocked transfer, so the transfer loop cedes
further attempts. This works if only one such filter is involved. If a
HTTP/2 transfer goes through a HTTP/2 proxy, two filters are
setting/clearing this flag and may step on each other's toes.
Connection filters `get_select_socks()` is replaced by
`adjust_pollset()`. They get passed a `struct easy_pollset` that keeps
up to `MAX_SOCKSPEREASYHANDLE` sockets and their `POLLIN|POLLOUT`
flags. This struct is initialized in `multi_getsock()` by calling the
various `getsocks()` implementations based on transfer state, as before.
After protocol handlers/transfer loop have set the sockets and flags
they want, the `easy_pollset` is *always* passed to the filters. Filters
"higher" in the chain are called first, starting at the first
not-yet-connection one. Each filter may add sockets and/or change
flags. When all flags are removed, the socket itself is removed from the
pollset.
Example:
* transfer wants to send, adds POLLOUT
* http/2 filter has a flow control block, removes POLLOUT and adds
POLLIN (it is waiting on a WINDOW_UPDATE from the server)
* TLS filter is connected and changes nothing
* h2-proxy filter also has a flow control block on its tunnel stream,
removes POLLOUT and adds POLLIN also.
* socket filter is connected and changes nothing
* The resulting pollset is then mixed together with all other transfers
and their pollsets, just as before.
Use of `SEND_HOLD` is no longer necessary in the filters.
All filters are adapted for the changed method. The handling in
`multi.c` has been adjusted, but its state handling the the protocol
handlers' `getsocks` method are untouched.
The most affected filters are http/2, ngtcp2, quiche and h2-proxy. TLS
filters needed to be adjusted for the connecting handshake read/write
handling.
No noticeable difference in performance was detected in local scorecard
runs.
Closes#11833
Delete checks and guards for standard C89 headers and assume these are
available: `stdio.h`, `string.h`, `time.h`, `setjmp.h`, `stdlib.h`,
`stddef.h`, `signal.h`.
Some of these we already used unconditionally, some others we only used
for feature checks.
Follow-up to 9c7165e96a#11918 (for `stdio.h` in CMake)
Closes#11940
This was originally added to handle platforms that supported only 1
second granularity in connect timeouts, but after some recent changes
the test currently permafails on several Windows platforms.
The need for this special-case was removed in commit 8627416, which
increased the connect timeout in all cases to well above 1 second.
Fixes#11767Closes#11849
- refs #11355 where failures to to low cpu resources in CI
are reported
- vastly extend CURLOPT_CONNECTTIMEOUT_MS and max durations
to test cases
- trigger Curl_expire() in test filter to allow re-checks before
the usual 1second interval
Closes#11690
- connect timeout was used at half the configured value, if the
destination had 1 ip version 4 and other version 6 addresses
(or the other way around)
- extended test2600 to reproduce these cases
Reported-by: Michael Kaufmann
Fixes#10514Closes#10517
New cfilter HTTP-CONNECT for h3/h2/http1.1 eyeballing.
- filter is installed when `--http3` in the tool is used (or
the equivalent CURLOPT_ done in the library)
- starts a QUIC/HTTP/3 connect right away. Should that not
succeed after 100ms (subject to change), a parallel attempt
is started for HTTP/2 and HTTP/1.1 via TCP
- both attempts are subject to IPv6/IPv4 eyeballing, same
as happens for other connections
- tie timeout to the ip-version HAPPY_EYEBALLS_TIMEOUT
- use a `soft` timeout at half the value. When the soft timeout
expires, the HTTPS-CONNECT filter checks if the QUIC filter
has received any data from the server. If not, it will start
the HTTP/2 attempt.
HTTP/3(ngtcp2) improvements.
- setting call_data in all cfilter calls similar to http/2 and vtls filters
for use in callback where no stream data is available.
- returning CURLE_PARTIAL_FILE for prematurely terminated transfers
- enabling pytest test_05 for h3
- shifting functionality to "connect" UDP sockets from ngtcp2
implementation into the udp socket cfilter. Because unconnected
UDP sockets are weird. For example they error when adding to a
pollset.
HTTP/3(quiche) improvements.
- fixed upload bug in quiche implementation, now passes 251 and pytest
- error codes on stream RESET
- improved debug logs
- handling of DRAIN during connect
- limiting pending event queue
HTTP/2 cfilter improvements.
- use LOG_CF macros for dynamic logging in debug build
- fix CURLcode on RST streams to be CURLE_PARTIAL_FILE
- enable pytest test_05 for h2
- fix upload pytests and improve parallel transfer performance.
GOAWAY handling for ngtcp2/quiche
- during connect, when the remote server refuses to accept new connections
and closes immediately (so the local conn goes into DRAIN phase), the
connection is torn down and a another attempt is made after a short grace
period.
This is the behaviour observed with nghttpx when we tell it to shut
down gracefully. Tested in pytest test_03_02.
TLS improvements
- ALPN selection for SSL/SSL-PROXY filters in one vtls set of functions, replaces
copy of logic in all tls backends.
- standardized the infof logging of offered ALPNs
- ALPN negotiated: have common function for all backends that sets alpn proprty
and connection related things based on the negotiated protocol (or lack thereof).
- new tests/tests-httpd/scorecard.py for testing h3/h2 protocol implementation.
Invoke:
python3 tests/tests-httpd/scorecard.py --help
for usage.
Improvements on gathering connect statistics and socket access.
- new CF_CTRL_CONN_REPORT_STATS cfilter control for having cfilters
report connection statistics. This is triggered when the connection
has completely connected.
- new void Curl_pgrsTimeWas(..) method to report a timer update with
a timestamp of when it happend. This allows for updating timers
"later", e.g. a connect statistic after full connectivity has been
reached.
- in case of HTTP eyeballing, the previous changes will update
statistics only from the filter chain that "won" the eyeballing.
- new cfilter query CF_QUERY_SOCKET for retrieving the socket used
by a filter chain.
Added methods Curl_conn_cf_get_socket() and Curl_conn_get_socket()
for convenient use of this query.
- Change VTLS backend to query their sub-filters for the socket when
checks during the handshake are made.
HTTP/3 documentation on how https eyeballing works.
TLS improvements
- ALPN selection for SSL/SSL-PROXY filters in one vtls set of functions, replaces
copy of logic in all tls backends.
- standardized the infof logging of offered ALPNs
- ALPN negotiated: have common function for all backends that sets alpn proprty
and connection related things based on the negotiated protocol (or lack thereof).
Scorecard with Caddy.
- configure can be run with `--with-test-caddy=path` to specify which caddy to use for testing
- tests/tests-httpd/scorecard.py now measures download speeds with caddy
pytest improvements
- adding Makfile to clean gen dir
- adding nghttpx rundir creation on start
- checking httpd version 2.4.55 for test_05 cases where it is needed. Skipping with message if too old.
- catch exception when checking for caddy existance on system.
Closes#10349
- add test2600 as a unit test that triggers various connect conditions
and monitors behaviour, available in a debug build only.
- this exposed edge cases in connect.c that have been fixed
Closes#10312
