On macOS/BSD, trying to call sendto on a connected UDP socket fails
with a EISCONN error. Because the singleipconnect has already called
connect on the socket when we're trying to use it for QUIC transfers
we need to use plain send instead.
Fixes#4529
Closes https://github.com/curl/curl/pull/4533
The ngtcp2 QUIC backend was using the MSG_DONTWAIT flag for send/recv
in order to perform nonblocking operations. On Windows this flag does
not exist. Instead, the socket must be set to nonblocking mode via
ioctlsocket.
This change sets the nonblocking flag on UDP sockets used for QUIC on
all platforms so the use of MSG_DONTWAIT is not needed.
Fixes#4531Closes#4532
Appveyor provides 2 CPUs for each builder[1], make sure to use parallel
compilation, when running with CMake. CMake learned this new option in
version 3.12[2] and the version provided by appveyor is fresh enough.
Curl doesn't really take that long to build and it is using the slowest
builder available, msbuild, so expect only a moderate improvement in
build times.
[1] https://www.appveyor.com/docs/build-environment/
[2] https://cmake.org/cmake/help/v3.12/release/3.12.htmlCloses#4508
This way, we always have exactly one slash after the host name, making
the tests pass when curl is compiled with the MSYS GCC.
Closes https://github.com/curl/curl/pull/4512
The MSYS system on Windows can run the test suite for curl built with
any toolset. When built with the MSYS GCC, curl uses Unix line endings,
while it uses Windows line endings when built with the MinGW GCC, and
`^O` reports 'msys' in both cases. Use the curl executable itself to
determine the line endings instead, which reports 'x86_64-pc-msys' when
built with the MSYS GCC.
Closes https://github.com/curl/curl/pull/4506
To make sure that transfer is being dealt with. Streams without
Content-Length need a final read to notice the end-of-stream state.
Reported-by: Tom van der Woerdt
Fixes#4496
With the `isdigit` implementation that comes with MSYS2, the argument
is used as an array subscript, resulting in a -Wchar-subscripts
warning. `isdigit`'s behavior is undefined if the argument is negative
and not EOF [0]. As done in lib/curl_ctype.h, cast the `char` variable
to `unsigned char` to avoid that.
[0] https://en.cppreference.com/w/c/string/byte/isdigit
Closes https://github.com/curl/curl/pull/4503
In particular, this removes the case where configure would find an old
cyall installation rather than a wolfssl one if present. The library is
named wolfssl in modern days so there's no real need to keep support for
the former.
Reported-by: Jacob Barthelmeh
Closes#4502
Prior to this change some users did not understand that the "request"
starts when the handle is added to the multi handle, or probably they
did not understand that some of those transfers may be queued and that
time is included in timeout.
Reported-by: Jeroen Ooms
Fixes https://github.com/curl/curl/issues/4486
Closes https://github.com/curl/curl/pull/4489
- If server header Retry-After is being used for retry sleep time then
show that value to the user instead of the normal retry sleep time.
This is a follow-up to 640b973 (7.66.0) which changed curl tool so that
the value from Retry-After header overrides other retry timing options.
Closes https://github.com/curl/curl/pull/4498
The URL extracted with CURLINFO_EFFECTIVE_URL was returned as given as
input in most cases, which made it not get a scheme prefixed like before
if the URL was given without one, and it didn't remove dotdot sequences
etc.
Added test case 1907 to verify that this now works as intended and as
before 7.62.0.
Regression introduced in 7.62.0
Reported-by: Christophe Dervieux
Fixes#4491Closes#4493
With MinGW-w64, `curl_socket_t` is is a 32 or 64 bit unsigned integer,
while `read` expects a 32 bit signed integer.
Use `sread` instead of `read` to use the correct parameter type.
Closes https://github.com/curl/curl/pull/4483
Previosly all connect() failures would return CURLE_COULDNT_CONNECT, no
matter what errno said.
This makes for example --retry work on these transfer failures.
Reported-by: Nathaniel J. Smith
Fixes#4461
Clsoes #4462
If more commits are submitted to master between the time of triggering
the first Cirrus build and the time the final build gets started, the
desired commit is no longer at HEAD and the build will error out.
[skip ci]
