Commit graph

7185 commits

Author SHA1 Message Date
Viktor Szakats
74a2828279
runtests: log the required minimum number of tests in CI
For Test Clutch.

If set (via env or tflags), include the minimum number of tests required
in runtests' log output:
```
* Min tests: 1750
```

Follow-up to 3f1cd809ee #19942

Closes #19987
2025-12-16 13:43:31 +01:00
Stefan Eissing
3a485c2ea0
pytest: do not ignore server issues
When a test server is found or configured, do not silently ignore
errors to start and disable them when checking their version.

This forces pytest to fail when a server is not operating
as it should.

Closes #19996
2025-12-16 13:19:00 +01:00
Stefan Eissing
2de22a00c7
lib: keep timestamp in easy handle
Use `data->progress.now` as the timestamp of proecssing a transfer.
Update it on significant events and refrain from calling `curlx_now()`
in many places.

The problem this addresses is
a) calling curlx_now() has costs, depending on platform. Calling it
   every time results in 25% increase `./runtest` duration on macOS.
b) we used to pass a `struct curltime *` around to save on calls, but
   when some method directly use `curx_now()` and some use the passed
   pointer, the transfer experienes non-linear time. This results in
   timeline checks to report events in the wrong order.

By keeping a timestamp in the easy handle and updating it there, no
longer invoking `curlx_now()` in the "lower" methods, the transfer
can observer a steady clock progression.

Add documentation in docs/internals/TIME-KEEPING.md

Reported-by: Viktor Szakats
Fixes #19935
Closes #19961
2025-12-16 08:48:44 +01:00
Viktor Szakats
b10b7c1fce
tests/server: fix initialization on Windows Vista+
Make sure to call `curlx_now_init()` before the first call to
`curlx_now()`.

Before this patch the first `curlx_now()` used the non-Vista code path
calling `GetTickCount()` on Vista+. This is harmless, but the upcoming
PR #18009 is going to drop the non-Vista code path, causing a division
by zero at startup in test servers, without this fix.

Bug: https://github.com/curl/curl/pull/18009#issuecomment-3652154307

Closes #19973
2025-12-15 02:14:28 +01:00
Viktor Szakats
2399ec74c3
tidy-up: replace banned printf() with puts()
In `curlinfo` and CMake integration test app.

Closes #19972
2025-12-15 02:14:28 +01:00
Viktor Szakats
187e219616
runtests: improve XML prolog check, enable -w permanently, fix two tests
To really verify the presence of the XML prolog, also in CI.

- move the prolog check from `loadtest` to `checktest`.
  (load did a soft error, silently skipping the test instead of failing)
- runtests: enable `-w` functionality permanently for all test targets,
  drop the option. It has no measurable performance impact.
- test 798, 1665: add XML prolog.
  Follow-up to f0d277cb0e

Follow-up to b5ea0736bb #19946
Follow-up to 904e7ecb66 #19347

Closes #19970
2025-12-14 12:26:38 +01:00
Daniel Stenberg
f0d277cb0e
tests: verify new header undfolder
test 798 - incoming cookie header in a folded line

test 1665 - verify HTTP headers without final CRLF. Make sure all complete
headers are delivered even if the reponse is partial
2025-12-13 23:02:27 +01:00
Daniel Stenberg
67ae101666
http: unfold response headers earlier
Make the low-level HTTP header "builder" unfold headers so that
everything else can keep pretending folding does not exist.

This code no longer tries to reduce repeated leading whitespace (in the
continued folded header) to a single one. To avoid having to have a
special state for that.

Adjusted two test cases accordingly

Closes #19949
2025-12-13 23:02:18 +01:00
Dan Fandrich
e0a77ec90f tests: fix perl scalar warning
Fixes the warning "Scalar value @xml[0] better written as $xml[0]"

Follow-up to b5ea0736bb

Ref: #19946
2025-12-12 15:39:10 -08:00
Viktor Szakats
5e14d623cf
tests/data: delete stray comments
Closes #19950
2025-12-12 17:41:20 +01:00
Viktor Szakats
b5ea0736bb
tests/data: add XML prolog to test files
To formalize they are now XML-compliant (with some asterisks.)

Also to help syntax highlighters work on them to make their content more
readable.

Also:
- Delete empty comment decorations.
- GHA/checksrc: simplify XML check.
- runtests: fail to load test data with XML prolog missing.

Follow-up to bfe6eb1c06 #19927
Follow-up to 87ba80a6df

Closes #19946
2025-12-12 17:17:24 +01:00
Daniel Stenberg
7d79ef926b
test567: add a header separating CRLF
To make it a valid response.

Closes #19948
2025-12-12 16:58:49 +01:00
Viktor Szakats
77eb6697f5
test568: fix codespell, catch it next time early in CI
Also:
- GHA/checksrc: do not exclude `tests/data/*` changes.

Follow-up to 407d2f3d57 #19944

Closes #19945
2025-12-12 13:33:51 +01:00
Viktor Szakats
d248d17cf3
cmake: update a comment 2025-12-12 13:09:52 +01:00
Viktor Szakats
3f1cd809ee
runtests: add options to set minimum number of tests, use them
To detect mistakes made in the runtests framework that reduce
the number of test runs. Before this patch it could go undetected with
a green CI.

The minimum thresholds will need light maintenance going forward (either
bumping them periodically, or adjust if some may fell below minimums for
justified reasons). We may also make minimums tighter or looser, or more
job-specific.

Latest number of test runs for each job can be seen at Test Clutch:
https://testclutch.curl.se/static/reports/feature-matrix.html

Also:
- GHA: set minimums.

Assisted-by: Dan Fandrich

Follow-up to f2a75a14dd
Follow-up to bb1391f943 #19510

Closes #19942
2025-12-12 12:20:22 +01:00
Stefan Eissing
eb39fee40b
pytest: add tests using sshd
With either /usr/sbin/sshd found or configured via --with-test-sshd=path
add tests for SCP down- and uploads, insecure, with known hosts or not,
with authorized user key or unauthorized one.

Working now with libssh and libssh2, using a hashed known_hosts file.

Closes #19934
2025-12-12 11:58:57 +01:00
Daniel Stenberg
407d2f3d57
test568: remove what looks like an email and a URL
Closes #19944
2025-12-12 11:50:29 +01:00
Viktor Szakats
fe8393d7db
tidy-up: miscellaneous
- drop stray duplicate empty lines in docs, scripts, test data, include,
  examples, tests.
- drop duplicate PP parenthesis.
- curl-functions.m4: move literals to the right side in if expressions,
  to match rest of the source code.
- FAQ.md: delete language designator from an URL.
- packages: apply clang-format (OS400, VMS).
- scripts/schemetable.c: apply clang-format.
- data320: delete duplicate empty line that doesn't change the outcome.
- spacecheck: extend to check for duplicate empty lines
  (with exceptions.)
- fix whitespace nits

Closes #19936
2025-12-12 04:18:48 +01:00
Viktor Szakats
421f931e7a
test1165: drop reference to deleted CURL_DISABLE_TESTS
Follow-up to bf823397ba #16134

Closes #19929
2025-12-11 11:42:28 +01:00
Daniel Stenberg
f72c377914
tests: verify setting bearer and doing redirects
Test 778, 779 and 795
2025-12-11 09:42:54 +01:00
Daniel Stenberg
bb134bba29
test3214: allow a larger struct Curl_easy
In my local build it is now 5840 bytes. Add a 10 byte margin.

Closes #19932
2025-12-11 09:41:45 +01:00
Viktor Szakats
319298484f
runtests: add support for single-quoted attributes, use it
With this, all test data files are XML-compliant.

Also:
- test1158, test1186: use single quotes for the test filename attribute
  containing a double quote. For XML-compliance.
- drop support for unquoted attributes. For XML-compliance.

Closes #19926
2025-12-11 00:08:16 +01:00
Viktor Szakats
2397be8349
tests/data: replace <, > with %LT, %GT
For XML-compliance.

Closes #19925
2025-12-11 00:08:16 +01:00
Viktor Szakats
7109b427fb
tests/data: replace < with %LT
For XML-compliance.

Closes #19924
2025-12-11 00:08:16 +01:00
Viktor Szakats
73b732e3e8
tests/data: replace & with %AMP
For XML-compliance.

Closes #19923
2025-12-11 00:08:15 +01:00
Stefan Eissing
a73040ac8a
quiche: fix version for skip due to flakiness
0.24.6 is the quiche version without the fix for proper handling fo
RESET streams. Require a verion higher than that to run test_05_02.

Follow-up to 14478429e7 #19916

Closes #19921
2025-12-10 18:07:51 +01:00
Stefan Eissing
14478429e7
pytest: quiche flakiness
Let nghttpx only use http/1.1 to backend. This reproduces the bug in
quiche with higher frequency. Allow test_14_05 to now return a 400 in
addition to the 431 we get from a h2 backend to nghttpx.

Skip test_05_02 in h3 on quiche not newer than version 0.24.4 in which
its bug is fixed: https://github.com/cloudflare/quiche/pull/2278

Ref: https://github.com/cloudflare/quiche/issues/2277
Closes #19770 (original Issue)
Closes #19916
2025-12-10 15:26:27 +01:00
Daniel Stenberg
b11b67c96f
test318: tweak the name a little
to make it properly differ from test 317
2025-12-10 09:29:43 +01:00
Stefan Eissing
d371288de7
test: increase altsvc test reliability
Move new tests from test_12 to test_06 (eyeballing) where they better
fit. Increase reliability by check Alt-Svc redirects from h3 to a lower
version for a port where no h3 is available.

Closes #19903
2025-12-10 00:23:33 +01:00
Stefan Eissing
44e64919cf
pytest: socksd startup delay
Add a small delay after the startup of the danted socks daemon to
give it more time to become responsive.

Closes #19895
2025-12-09 16:03:38 +01:00
Stefan Eissing
5ed7b5b01b
alt-svc: more flexibility on same destination
When the Alt-Svc points to the same host and port, add the destination
ALPN to the `wanted` versions and set it also as the `preferred` version
in negotiations.

This allows Alt-Svc for h3 to point to h2 and have it tried first. Also,
this allows Alt-Svc to say http/1.1 is preferred and changes the ALPN
protocol ordering for the TLS handshake.

Add tests in various combination to verify this works.

Reported-by: yushicheng7788 on github
Fixes #19740
Closes #19874
2025-12-09 15:59:09 +01:00
Yedaya Katsman
e76080fb73
test: add test 2090 for SSLKEYLOGFILE structure
Only the TLS 1.2 structure for now since it's simpler, and only has a
single label type. This has the bonus of also testing libressl that only
supports logging keys in TLS 1.2

Closes #19816
2025-12-09 14:14:50 +01:00
Viktor Szakats
d14bf19fda
test1025, 1221: fold long command-lines
Closes #19893
2025-12-09 12:40:21 +01:00
Viktor Szakats
1dd758b36d
test1464: mark XML-compliant
Closes #19892
2025-12-09 12:40:21 +01:00
Viktor Szakats
70d71e8761
tests/data: move section data to external files
To make the test files XML-compliant, and the expected results
possibly easier to manage by keeping them in `.md`, `.html`, `.1`
and `.txt` files.

Non-XML-compliant files are down to 31 (1.6%) after this patch.

Closes #19882
2025-12-09 10:12:49 +01:00
Viktor Szakats
cce660693c
tests/data: use more %TESTNUMBER macro for previous patch
Follow-up to d75716e4e5 #19799
Cherry-picked from #19882
Closes #19885
2025-12-09 09:20:10 +01:00
Viktor Szakats
d75716e4e5
tests/data: move --libcurl output to external data files
To make the test files XML-compliant, and the expected results
possibly easier to manage by keeping them in `.c` files.

Non-XML-compliant files are down to 36 after this patch.

Also:
- make all macro expansions apply to `%includetext` contents.

Closes #19799
2025-12-09 00:19:10 +01:00
dependabot[bot]
cc853ddc3d
GHA: bump pip-dependencies ruff, psutil, pytest
- update `ruff` from 0.14.5 to 0.14.8
- update `psutil` from 7.1.2 to 7.1.3
- update `pytest` from 8.4.2 to 9.0.1

Closes #19876
2025-12-08 15:37:43 +01:00
Fabian Keil
d360ddb1b2
test1475: consistently use %CR in headers
Gets the test working when using Privoxy as proxy.

Closes #19870
2025-12-08 10:44:00 +01:00
Daniel Stenberg
2535c4298f
hostcheck: fail wildcard match if host starts with a dot
A hostname cannot start with a dot when DNS is used, but there are other
ways.

Amend unit test 1397

Closes #19869
2025-12-08 10:42:49 +01:00
Daniel Stenberg
a78a07d3a9
cookie: cleanups and improvements
- Stricter cookie validation with earlier rejection of empty/invalid
  cookie names

- secure and httponly attributes no longer accept = with empty values
  (only bare keywords)

- Validation checks (length, TAB, prefixes) moved into the first
  name/value pair block for better code organization

- Deferred time(NULL) calls for better performance when expires/max-age
  aren't used

- Simplified loop control flow by removing done flag

- The cookie size restriction now only applies to name + value, not other
  parts of the header line.

- Fixed a gcc 4.8.1 quirk

Closes #19868
2025-12-08 09:52:58 +01:00
Daniel Stenberg
a093c93994
cookie: only keep and use the canonical cleaned up path
Instead of keeping both versions around.

Closes #19864
2025-12-08 08:03:57 +01:00
Viktor Szakats
f08417c425
runner.pm: run memanalyzer as a Perl module
To improve performance of torture tests.

Also on Windows, where this patch may make those viable for CI.

Linux   !FTP    4m47 ->  4m24 (-shallow=25)
Linux    FTP    2m30 ->  2m23 (-shallow=25)
macOS   !FTP   14m30 -> 13m07 (-shallow=25)
macOS    FTP    3m57 ->  3m59 (-shallow=25)
Windows !FTP  >25m   ->  4m47 to 14m45 (-shallow=5 to 25) (not in CI)

Linux
Before: https://github.com/curl/curl/actions/runs/20006771767/job/57370205514
After: https://github.com/curl/curl/actions/runs/20006783210/job/57370236911?pr=19863

macOS:
Before: https://github.com/curl/curl/actions/runs/20006771786/job/57370205769
After: https://github.com/curl/curl/actions/runs/20006783177/job/57370236995?pr=19863

Windows:
Before: https://github.com/curl/curl/actions/runs/19667198537/job/56326962912?pr=19675
After: https://github.com/curl/curl/actions/runs/20007175773/job/57371768734?pr=19863
After shallow=25: https://github.com/curl/curl/actions/runs/20008523913/job/57374427449?pr=19865

Ref: #19675
Follow-up to 472bc90323 #19821
Closes #19863
2025-12-07 19:51:45 +01:00
Stefan Eissing
891566c72d
ftp: make EPRT connections non-blocking
On platforms where neither accept4 nor fcntl was available, an
EPRT connection did not send the accepted socket as non-blocking.

This became apparent when TLS was in use and the test receive
on shutdown did simply hang.

Reported-by: Denis Goleshchikhin
Fixes #19753
Closes #19851
2025-12-06 14:48:04 +01:00
Viktor Szakats
cca815ccfd
test1498: disable 'HTTP PUT from stdin' test on Windows
Test became flaky with memanalyze errors after merging #19845,
in a TrackMemory Windows Unicode c-ares openssl-quic build:
GHA/windows: mingw, AM x86_64 c-ares U.

Disable it until further investigation.

This test uses the Windows-specific multi-threaded stdin code
that caused issues in the past. It's also using `TerminateThread()`,
that apps aren't supposed to.

Examples:
https://github.com/curl/curl/pull/19845#issuecomment-3614921298
https://github.com/curl/curl/actions/runs/19948992659/job/57205061260?pr=19845#step:13:3028
https://github.com/curl/curl/actions/runs/19966429786/job/57259325027?pr=19852#step:13:3030

Also seen to fail earlier while testing torture tests on Windows:
https://github.com/curl/curl/pull/19675#issuecomment-3573154110

Ref: 4e051ff550 #19845

Closes #19855
2025-12-05 17:45:16 +01:00
Viktor Szakats
af5def0738
tidy-up: avoid (()), clang-format fixes and more
- drop redundant parentheses from macro definitions.
- apply clang-format in some places missed earlier.
- wolfssl: fix a macro guard comment.
- curl_setup.h: drop empty lines
- FAQ: fix C formatting.

Closes #19854
2025-12-05 16:07:03 +01:00
Viktor Szakats
4e051ff550
curlx: limit use of system allocators to the minimum possible
Clone a multibye conversion function into curlx/fopen, and use that
local copy from curlx/fopen functions. Adjust allocators in curlx/fopen
to use curl's in normal builds, and system allocators in TrackMemory
builds to avoid recursion.

This allows to switch curlx/multibyte functions to curl allocators in
all configurations, as they are no longer called by curlx/fopen, and
a recursive call can no longer happen.

After this patch the system allocator is only used in TrackMemory
Windows builds, within curlx `fopen`, `freopen`, `stat` and `open`
functions.

Also:
- test 1, 440, 767: raise allocation limitsto fit the extra allocations
  in Windows Unicode builds.
- replace all uses of `curlx_unicodefree()` macro with `curlx_free()`
  across the codebase.
- curlx/multibyte: delete `curlx_unicodefree()`.
- ldap: join Windows and non-Windows codepaths that became
  identical after moving from `curlx_unicodefree()` to `curlx_free()`.
- vauth: drop a strdup from standard to curl allocator since
  the original allocation is now already done by curl's.
- tool_doswin: drop now superfluous strdup from `FindWin32CACert()`.
- memanalyzer.pm: sync weirdo `calloc` log message with `malloc`'s.

Fixes #19748
Closes #19845
2025-12-05 15:32:59 +01:00
Stefan Eissing
8d9aa6d6e1
tests/servers: put unix-domain-path inside LOGDIR
Change Unix domain socket paths from `/tmp/curl-socksd-<random>` to
`/drive/path/to/LOGDIR/PIDDIR/*-uds` to avoid having to create and
delete them before use. Also to use a path which remains an absolute one
while passed from MSYS2 Perl to native Windows curl tool and test server
via the command-line, and keep pointing to the same location, fixing:
```
=== Start of file commands.log
../src/curl.exe -q --output log/3/curl1468.out --include --trace-ascii log/3/trace1468
  --trace-time http://this.is.a.host.name:64405/1468
  --proxy socks5h://localhost/tmp/curl-socksd-YnbvRo98 [...]
=== End of file commands.log
=== Start of file socks2_server.log
[...]
14:11:54.597968 Listening on Unix socket D:/a/_temp/msys64/tmp/curl-socksd-YnbvRo98
```
Ref: https://github.com/curl/curl/actions/runs/19896583933/job/57028545111?pr=19812

The curl tool is pending #19825 to fix accepting an absolute unix domain
socket path on Windows.

Assisted-by: Viktor Szakats
Closes #19810
2025-12-05 13:00:19 +01:00
Viktor Szakats
f97e62ff14
sws: fix binding to unix socket on Windows
Windows 10.17063+ (having unix socket support) fails to set for unix
sockets the `SO_REUSEADDR` option, with error 10045 (`WSAEOPNOTSUPP`),
and also fails to set `SO_KEEPALIVE` with error 10042 (`WSAENOPROTOOPT`).

Fix by not enabling these socket options on Windows for unix sockets.

Also:
- fixing test 1435, 1436 to run in CI.
- fixing the `socksd` test server for test 1467, 1468, 1470. But, also
  disable these for now due to another Windows issue: #19825

Ref: https://stackoverflow.com/questions/68791319/unix-domain-socket-bind-failed-in-windows/68794755#68794755
Ref: #19810
Closes #19812
2025-12-04 22:49:03 +01:00
Viktor Szakats
0476e4fc65
tidy-up: one more round of formatting nits
Closes #19835
2025-12-04 19:30:59 +01:00