When HTTPS-RR is needed for the HTTP/3 handshake, delay the connect
until it arrives. Relevant only for TLS backends that support ECH, for
now.
Closes#21253
Make cf-https-connect work async correctly:
- only start first baller when at least one A/AAAA address
is available
- select first connect attempt after that with HTTPS-RR info
there or not.
- select second connect attempt only when HTTPS-RR is resolved
(may have resolved to "not known") and select possible ALPN
from things known by then. May not select any second attempt
when first already covers everything.
This means when the HTTPS-RR is known at/before the first address
is resolved, everything behaves as before. When the HTTPS-RR is
late, a first connection attempt will have been started. Any
ALPN preference from the HTTPS-RR that is not already ongoing will
then start the second attempt.
For HTTPS-RRs that recommend 2 or more ALPNs, the first will always
be attempted: either it is already ongong or it will be the ALPN
for the second attempt. The 2nd ALPN recommendation from HTTPS-RR
*may* be honored or not, depending on what is already selected.
The difference in behaviour between early/late HTTPS-RR resolve
cannot be helped - unless we do not perform any attempts before
it arrives. Trade offs.
Closes#21267
New connection filter `cf-dns` that manages DNS queries. If hands
out addresses and HTTPS-RR records to anyone interested. Used by
HTTPS and IP happy eyeballing.
Information may become available *before* the libcurl "dns entry"
is complete, e.g. all queries have been answered. The cf-ip-happy
filter uses this information to start connection attempts as soon
as the first address is available.
The multi MSTATE_RESOLVING was removed. A new connection always
goes to MSTATE_CONNECTING. The connectdata bit `dns_resolved`
indicates when DNS information is complete. This is used for
error reporting and starting the progress meter.
Removed dns entries `data->state.dns[i]`, as the `cf-dns` filter
now keeps the reference now.
Many minor tweaks for making this work and pass address information
around safely.
Closes#21027
Curl_1st_err() does not return the second error if the first result is
CURLE_AGAIN. This may cause errors to not become noticeable when they
should be.
Replace all use of Curl_1st_err() with Curl_1st_fatal(), which handles
CURLE_AGAIN as a not-a-real-error case.
Closes#20980
Also:
- support per-directory and per-upper-directory whitelist entries.
- convert badlist input grep tweak into the above format.
(except for 'And' which had just a few hits.)
- fix many code exceptions, but do not enforce.
(there also remain about 350 'will' uses in lib)
- fix badwords in example code, drop exceptions.
- badwords-all: convert to Perl.
To make it usable from CMake.
- FAQ: reword to not use 'will'. Drop exception.
Closes#20886
The 'id' struct field in 'struct h3_stream_ctx' is a uint64_t type so
should be output with PRIu64 - and it makes sense to be consistent.
Note that the field with the same name in the ngtcp2 version of this
struct is a *signed* 64-bit variable.
Reported by Codex Security
Closes#20849
Reported when running `HeaderFilterRegex: '.*'` in CI.
Also replace an underscored symbol with a regular one in macro
definition.
Cherry-picked from #20720Closes#20721
Backtrack on previous change that aimed to solve the wrong `share.h`
being included. It turns out it did not fix this issue. At the same time
it introduced relative header filenames and the need to include the same
headers differently depending on the source files' location, reducing
readability and editability.
Replace this method by re-adding curl's lib source directory to the
header path and addressing headers by the their full, relative name to
that base directory. Aligning with this method already used in src and
tests.
With these advantages:
- makes includes easier to read, recognize, grep, sort, write, and copy
between sources,
- syncs the way these headers are included across curl components,
- avoids the ambiguity between system `schannel.h`, `rustls.h` vs.
local headers using the same names in `lib/vtls`,
- silences clang-tidy `readability-duplicate-include` checker, which
detects the above issue,
Ref: https://clang.llvm.org/extra/clang-tidy/checks/readability/duplicate-include.html
- possibly silences TIOBE coding standard warnings:
`6.10.2.a: Don't use relative paths in #include statements.`
- long shot: it works well with concatenated test sources, for
clang-tidy-friendly custom unity builds. Ref: #20667
Slight downside: it's not enforced.
If there happens to be a collision between a local `lib/*.h` header and
a system one, the solution is to rename (possibly with its `.c`
counterpart) into the `curl_` namespace. This is also the method used by
curl in the past.
Also:
- curlx/inet_pton: reduce scope of an include.
- toolx/tool_time: apply this to an include, and update VS project
files accordingly. Also dropping unnecessary lib/curlx header path.
- clang-tidy: enable `readability-duplicate-include`.
Follow-up to 3887069c66#19676
Follow-up to 625f2c1644#16991#16949Closes#20623
- fix internal macro `AN_APPLE_OS` reused between sources without
resetting it. It may potentially have left the system sha256
function unused.
- fix to define `WOLFSSL_OPTIONS_IGNORE_SYS` so that it always applies
to wolfSSL headers, also during feature detection.
- md4, md5, sha256: simplify fallback logic.
- delete 20+ unused macros.
- scope or move macros to avoid `-Wunused-macros` warnings.
- examples: delete unused code.
The warning detects macros defined but not used within the same C
source. It does not warn for macros defined in headers. It also works
with unity builds, but to a lesser extent.
Closes#20593
Report UDP packets with SOCKEMSGSIZE as being "sent" to progress
the send buffer properly on PMTUD probes.
Reported-by: Daniil Gentili
Fixes#20440Closes#20448
For HTTP/2, add error code description to close failures.
For HTTP/3, add special handling like in HTTP/2 when streams
have been rejected or an error comes during the response body
and we are not interested in the body.
Closes#20207
When the compiler supports C99.
- map logging functions to macro stubs when verbose logging is disabled
and the compiler is C99. Make sure these stubs silence unused variable
warnings for non-variadic arguments.
Before this patch they mapped to function stubs, the same codepath
used for C89 compiler in this configuration.
- introduce new macros to tell the compiler which code to include
when verbose code is active, or inactive:
- `CURLVERBOSE`: defined when verbose code is active.
To enclose blocks of code only used for verbose logging.
- `VERBOSE(statement);`:
compile statement when verbose code is active.
To mark code lines only used for verbose logging.
- `NOVERBOSE(statement);`:
compile statement when verbose code is inactive.
To suppress warnings for arguments passed to logging functions via
printf masks, e.g. `NOVERBOSE((void)ipaddress);`, yet keeping
the warning in verbose builds.
Note these macros are not the same as `CURL_DISABLE_VERBOSE_STRINGS`.
Verbose code is always active in C89 mode (without variadic macro
support).
- drop existing uses of `CURL_DISABLE_VERBOSE_STRINGS` where redundant,
or replace with the above macros. Ending up reducing the number of
`#ifdef`s, and also the number of lines.
Assisted-by: Daniel Stenberg
Assisted-by: Jay Satiro
Reported-by: Dan Fandrich
Fixes#20341
Refs: #12105#12167Closes#20353
Use non-deprecated CRT function variants on Windows.
- introduce `curlx_fdopen()`, `curlx_close()` and use them. Map them to
non-deprecated, underscored, CRT functions on Windows.
- replace `close()` uses with either `sclose()` (for sockets) or
`curlx_close()` (for files).
- map `fileno`, `unlink`, `isatty` to their non-deprecated, underscored,
versions on Windows.
- tool_dirhie: map `mkdir` to `_mkdir` on Windows.
- easy: use `_strdup()` on Windows, regardless of how `HAVE_STRDUP` is
set.
- cmake: assume `HAVE_STRDUP` on Windows. To allow dropping a detection
hack using `_CRT_NONSTDC_NO_DEPRECATE` with MSVC. Windows always has
`_strdup()` which the code uses, but also needs `HAVE_STRDUP` defined
to disable curl's own `strdup()` implementation.
- curl_setup.h: drop `_CRT_NONSTDC_NO_DEPRECATE` as no longer necessary.
Closes#20212
This allows builds know about all schemes - but only have the protocol
implementations for those actually built-in.
It further allows multiple protocols to reuse the same protocol setup
and functions for both TLS and non-TLS implementations instead of
needing two (or more) structs.
The scheme information is now in 'struct Curl_scheme' and all the
function pointers for each scheme/protocol implementation are in struct
Curl_protocol.
The URL API now always work with all known protocols.
Closes#20351
- It is slower and uses more memory than the alternatives and is only
experimental in curl.
- We disable a few tests for OpenSSL-QUIC because of flakiness
- It gets little attention from OpenSSL and we have no expectation of the
major flaws getting corrected anytime soon.
- No one has spoken up for keeping it
- curl users building with vanilla OpenSSL can still use QUIC through the
means of ngtcp2
Closes#20226
Add casts to `bool`, or use `bit` type in local variables, where
neccessary to avoid MSVC compiler warnings C4242.
Note: There may remain places needing the above updates, where not
tested in CI, and missed in manual review.
Also:
- urldata: convert struct field `connect_only` to bitfield to match its
counterpart in another struct.
- rename curl-specific `bit` type to `curl_bit`.
Closes#20142
When a download size is known and rate limiting is in effect, adjust the
duration of each measurement step and its rate for maximum precision.
Since it is unpredictable how long the last bytes of a download will
take, download speed can be thrown off if the "last bytes" are a
significant amount of the total download. Make the "last bytes" small in
comparision to the rest and "stretch" the rate limit intervals to
accommodate the difference.
Fix ngtcp2 receive data acknowldgements to be based on a local window
size tracking. This allows window updates controlled by rate limits.
Fix ratelimit wait time calculation to accomodate negative tokens.
h3 rate limit, update timeers
Make download rate limits work correctly in ngtcp2. Fix multi handling
of rate limits to set a timer for when limits will update again.
Without running the transfer on limit updates, protocols like h2/h3 may
stall if the server does not send due to stream windows being too small.
scorecard: measure download speedlimits
When running scorecard with --limit-rate=n, show the reported download
speed averages plus percentage deviation from the limit.
Closes#20228
When receiving on a stream that already failed or has already been closed,
return the matching error code without touching the connection. In case
the connection shows errors, e.g. the server closed, those errors should
not have impact on an already failed/closed stream.
This might mitigate flakiness in pytest 07_13 where unexpected errors
occur after a successful upload.
Closes#20220
We recently allowed a larger send buffer in ngtcp2 streams. This allowed
curl to send more early data then previously when the server was slow in
performing the handshake. This led to flaky test failures when the
amount of early data was larger than expected.
Change test expectations to allow for varying amount of early data.
Ref: #20112Closes#20161
The special handling for draining server connections during a connect
attempt was only done on CURLE_RECV_ERROR. But it may also happen when
ngtcp2 errors on writing data. Check for CURLE_SEND_ERROR also.
Ref: #20112Closes#20162
- asyn-thrdd.c: scope an include.
- apply more clang-format suggestions.
- tidy-up PP guard comments.
- delete empty line from the top of headers.
- add empty line after `curl_setup.h` include where missing.
- fix indent.
- CODE_STYLE.md: add `strcpy`.
Follow-up to 8636ad55df#20088
- lib1901.c: drop unnecessary line.
Follow-up to 436e67f65b#20076Closes#20070
- they rarely catch any problems
- we have other ways to test different send/recv problems
- the number of such calls vary much more per invoke than others, making
memdebugging harder
- reducing the total number of fallible functions per test is good
- they were not used as intended anyway
Closes#20097
This function REQUIRES the size of the target buffer as well as the
length of the source string. Meant to make it harder to do a bad
strcpy().
Removes 23 calls to strcpy().
Closes#20067
- replace `sendf.h` with `curl_trc.h` where it was included just for it.
- drop unused `curl_trc.h` includes.
- easy: delete obsolete comment about `send.h` include reason.
Also:
- move out `curl_trc.h` include from `sendf.h` and include it directly
in users, where not done already. To flatten the include tree and
to less rely on indirect includes.
- stop including `sendf.h` from other headers, replace it with forward
declaration of `Curl_easy`, as done already elsewhere.
Verified with an all non-unity CI run.
Closes#20061
To make it available for all files. Drop includes from individual
sources. This header was already included from most sources and not
specific to any internal subsystem.
Also to ensure that two system symbol redefines on Windows (`read()` and
`write()`) get applied to all sources. Move them to `curl_setup.h`.
Closes#20056
Also:
- vquic-tls.h: do not include unused headers for non-H3 builds.
- autotools: stop looking for `openssl/x509.h` header.
- cmp-config.pl: delete exception for `openssl/x509.h`.
- examples: format/comment sync between the two touched files.
- openssl: drop unused `curlx/wait.h` include.
Closes#20049
With 24b36fd stream flow control window sizes have been set too
restrictive, crippling transfer rates when no rate limit is in effect.
Disable ngtcp3 stream window auto-tuning and extend the stream window
from the small initial size to the effective rate limit. If no rate
limit is configured, extend stream window to maximum value right away.
This cannot shrink the stream window later, however. But growing the
limit or removing it, will work mid download.
Fixes#20030
Reported-by: koujaz on github
Closes#20033
