| Age | Commit message (Collapse) | Author |
|---|
|
HttpClient.SendAsync() should not attempt to read response body on a HEAD request.
|
|
PUT request. (#38129)
* HttpClient.SendAsync() should not attempt to read response body on a HEAD or PUT request.
* Update HttpMethod.cs
* Address feedback.
* Cleanup.
* Update HttpRequestMessageTest.cs
(cherry picked from commit 73e610f847f9c5100a0cec4a80692969836ff35d)
|
|
|
|
|
|
|
|
|
|
|
|
They cause issues on iOS where all p/invokes need to be resolved.
|
|
|
|
Should match dotnet/corefx#36710 with slight differences due to
divergences in the files.
|
|
* Update SecurityProtocolType to support TLS 1.3
* Added Tls13 to sslprotocols
|
|
Add more `GlobalizationMode.Invariant` conditionals.
|
|
|
|
and `IsDynamicCodeCompiled`).
|
|
ThreadPoolTests.QueueRegisterPositiveAndFlowTest
Only observed a couple of time in CoreFX, corlib-xunit Windows x64 Mono CI,
https://jenkins.mono-project.com/view/Mono/job/z/label=w64/8715/testReport/junit/Test%20collection%20for%20System.Threading.ThreadPools.Tests/ThreadPoolTests/System_Threading_ThreadPools_Tests_ThreadPoolTests_QueueRegisterPositiveAndFlowTest/
Test is using a local variable to set state between threads and very infrequent,
the reading thread seems to pick the old value.
Looks like coordination between reader/writer thread works as expected and since
this is an infrequent failure, doesn't look like a coordination issues or problem
with the ThreadPool (would have expected failures in many tests) implementation.
The test is however using a fragile construction using a local variable (backgroundAsyncLocalValue)
to share state between the writer/reader thread without any synchronization involved.
That could lead to infrequent failures due to memory read/write optimizations.
Since this is a shared local variable captured and shared between main thread and thread pool
it needs to have proper acquire/release semantics, or it could cause failures like these due
to memory optimization.
|
|
|
|
|
|
|
|
|
|
[master] Use fallbacks when link() fails
|
|
System.IO.Filesystem's MoveFile() and ReplaceFile() use the link() syscall under some conditions to avoid a copy. This fixes two problems with that:
* MoveFile()'s fall-back does not recognize a possible Android error as non-critical, and bails out on the fall-back with an exception. From logs we know that 'strerror_r(android_error)' = 'Permission denied' so the Android error should be EACCES (we already check for EPERM)
* While MoveFile() was using a fall-back, ReplaceFile() wasn't.
|
|
|
|
It was a Theory instead of a Fact and the method parameter wasn't being used.
|
|
available (#273)
These tests failed on Xamarin.iOS since the linker strips out most Debugger* attributes in release builds.
We can gracefully handle this case and simply skip the tests instead.
Port of https://github.com/dotnet/corefx/pull/36785 to the mono corefx fork.
|
|
|
|
Cherrypick dotnet/corefx@991edc840b0240aeb9b245673eb78dc3b80f3a3b
to initialize `header->msg_flags`
Addresses mono/mono#13859
Coverity: 1443505
|
|
Add `GlobalizationMode.Invariant` conditionals to some calendar code.
This allows the Japanese, Taiwan and Hebrew calendars to be disabled.
|
|
Cherry-picked from https://github.com/dotnet/corefx/pull/36134
|
|
They are not supported by Bitcode/Xmarin.WatchOS.
|
|
|
|
Yes, AIX has different values.
Merged upstream at dotnet#35971
|
|
They're moved to System.Drawing.Common.dll in Mono 2019-02.
|
|
|
|
|
|
Add new static `HttpClient.CreateDefaultHandler()` and use it in the
default constructor.
Mono will provide a custom version in a parial class part, to allow
us to replace the default handler in our mobile profiles.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Move BackendSupportsAlpn check to CoreFx.Private.TestUtilities (#31592)
* add BackendSupportsAlpn to comon
* address feedback
* address feedback
(cherry picked from commit e4c68f518455c921b5d425df845dc12eba8081e4)
* client side ALPN for OSX (#31905)
* ALPN client side for OSX
* rework code to use new API introduced in 10.13.4 + feedback from review
* reviews feedback round 2
* feedback from review
(cherry picked from commit 219ea1bea0f12660398a75160b1d68c2885696fd)
* Fix ServerAsyncAuthenticateTest. (#35170)
In ServerAsyncSslHelper, pass a certificate validation callback to both
client and server.
(cherry picked from commit a46d64a21591dd36292b0f6f0946fc55babced11)
* Use `PlatformDetection.IsMono`; going to close the corefx PR.
* Ignore some tests for Mono.
|
|
|
|
|
|
dotnet/darc-master-be4ecac4-471a-4e11-961e-78e7628e8c2c
Update dependencies from dotnet/coreclr
(cherry picked from commit 09986997bfa24b73fcffbbdabe363cea60d53eff)
|
|
|
|
* Index and Range updates
* Address @mikedn feedback
* Address Feedback
* more feedback
* Use Deconstruct in Range.GetOffsetAndLength
* Rename GetArrayRange to GetSubArray
* Temporary disable the old Corefx Range tests
* Return back the TimeSpan test disabling
* Fix Range jit test
* Exclude the jit test
* revert the changes in the jit Range test
* Address Suggested Feedback
Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
(cherry picked from commit b0b16179288058f42d95fe89faa76bf86496bbb9)
|
|
* Expose Index and Range with adding tests
* More tests and more exposed APIs
* More changes
* address feedback
* Fix ApiCompat break
* Fix UAPAOT app compat
(cherry picked from commit 75a6b034b4a478db5bb50d6cd618cecabbf3ecf1)
|
|
* Reduce allocation in Index/Range.ToString
* Address PR feedback
Signed-off-by: dotnet-bot <dotnet-bot@microsoft.com>
(cherry picked from commit ceee48b5894c8f128792d7ad2696cd8418a62878)
|
