monojenkins [Fri, 24 Apr 2020 10:52:31 +0000 (06:52 -0400)]
Prevent a possible assert in the debugger when a thread detaches (#34955)
When a new thread is attached to the VM, the debugger will add it to the
`thread_to_tls` hash table. When that thread detaches, it will be
removed. Once the thread is attached, if a client is debugging, the
client can use the `CMD_THREAD_GET_FRAME_INFO` command to ask for
details about that thread.
This is a possibility that the thread detaches before the
`CMD_THREAD_GET_FRAME_INFO` command is processed, so the thread may not
exist the the `thread_to_tls` hash table any more.
This is a race condition, but is a valid state, so instead of asserting,
the debugger agent should indicate to the client that the thread no
longer exists, using `ERR_UNLOADED`.
Kenneth Pouncey [Fri, 24 Apr 2020 09:37:37 +0000 (11:37 +0200)]
[wasm] Rename the browser targets to `TargetsBrowser` (#35390)
- The change also makes sure that the `TargetsBrowser` is used consistently.
- example TargetsWebAssembly and TargetsWASM
Jan Vorlicek [Fri, 24 Apr 2020 08:09:06 +0000 (10:09 +0200)]
Merge pull request #35363 from janvorli/fix-unloadability-tests
Fix failures in unloadability tests
monojenkins [Fri, 24 Apr 2020 07:47:09 +0000 (03:47 -0400)]
[wasm][debugger] Fix `evaluateOnCallFrame` for valuetype locals (#35361)
.. including generic, and async methods. This fixes `hover` for valuetype locals.
Co-authored-by: radical <radical@users.noreply.github.com>
Zoltan Varga [Fri, 24 Apr 2020 07:35:54 +0000 (03:35 -0400)]
Add build configuration to generate runtime packs for WebAssem… (#34936)
Jan Kotas [Fri, 24 Apr 2020 04:03:08 +0000 (21:03 -0700)]
Delete JIT_PATH host key/value pair (#35359)
This was originally introduced to deal with the JIT being in a separate NuGet package. No longer the case for a years.
Aaron Robinson [Fri, 24 Apr 2020 02:53:07 +0000 (19:53 -0700)]
Fix nethost.h header file when consuming as a static lib. (#35373)
Andrew Au [Fri, 24 Apr 2020 02:30:44 +0000 (19:30 -0700)]
Additional code cleanup (#35259)
* Additional code cleanup
* Code review feedback
* Remove associated clrconfigvalue.h entries
* More code review feedback
* Eliminate unnecessary entries in clrconfigvalues.h
Steve MacLean [Thu, 23 Apr 2020 23:15:05 +0000 (19:15 -0400)]
Revert "Build coreclr targetGeneric tests separately for CI (#34790)" (#35371)
This reverts commit
9188a5820ccc6dffc5763d22ee4a6cd040a4048c.
Sergey Andreenko [Thu, 23 Apr 2020 22:42:09 +0000 (15:42 -0700)]
Delete an extra new line when there is a child. (#35355)
Andy Ayers [Thu, 23 Apr 2020 21:48:16 +0000 (14:48 -0700)]
Jit: run throw helper merge phase before morph (#35255)
Now that we have pred lists before morph, we can move the throw helper
tail merge phase earlier in the phase list.
This has two benefits:
* we can now merge a few more cases, because morph can introduce unique
temps for otherwise identical calls;
* it saves some throughput, because we no longer need to morph duplicate
calls.
There is more opportunity here to reduce code size if we can find the right
heuristic in morph to decide if throw helpers should be called or tail-called,
though the overall benefit is small (~600 methods, ~2000k bytes). I left the
current heuristic in place as I couldn't come up with anything better.
Fixes #35135.
Tomáš Rylek [Thu, 23 Apr 2020 21:44:42 +0000 (23:44 +0200)]
Fix some overlooked template moves in the YML templates (#35311)
Jan Vorlicek [Thu, 23 Apr 2020 20:32:43 +0000 (22:32 +0200)]
Fix failures in unloadability tests
The assert in EEClass::Destroy is invalid after a change #1201 that has
removed ArrayTypeDesc. It was causing failures of a large portion of the
coreclr tests.
There were also three tests that were failing due to test
incompatibilities with unloadability, so I am marking them as such.
Jan Vorlicek [Thu, 23 Apr 2020 20:26:19 +0000 (22:26 +0200)]
Merge pull request #35332 from janvorli/fix-runincontext-path
Fix incorrect path of the runincontext script in Helix
Egor Bogatov [Thu, 23 Apr 2020 17:47:51 +0000 (20:47 +0300)]
[mono] Enable LLVM for iOS by default for AppleAppBuilder (#35338)
* Enable LLVM for iOS by default
* Update Makefile
Prashanth Govindarajan [Thu, 23 Apr 2020 17:29:24 +0000 (10:29 -0700)]
Changes triggered by the CA1834 analyzer (#35227)
* Cherry pick if needed
* Changes triggered by https://github.com/dotnet/roslyn-analyzers/pull/3481
* sq
* Revert "Cherry pick if needed"
This reverts commit
55a98c41f6d2143737ec707faf2a3cf213474c97.
* sq
Anirudh Agnihotry [Thu, 23 Apr 2020 17:21:40 +0000 (10:21 -0700)]
correctly set runWorkingDirectory (#35322)
Ryan Lucia [Thu, 23 Apr 2020 15:49:30 +0000 (11:49 -0400)]
[meta] Add embedding API function to get the default ALC (#35191)
* Formatting fixes
* Add embedding API function get the default ALC
Steve MacLean [Thu, 23 Apr 2020 15:09:23 +0000 (11:09 -0400)]
Build coreclr targetGeneric tests separately for CI (#34790)
* Build CoreClr tests separately
* Build coreclr targetGeneric tests separately for CI
* Build coreclr targetGeneric tests separately in runtime.yml
* Distinguish generic build based on Libraries config
* Build OSX release libraries when CoreCLR changed
* Always use tar.gz compression for generic tests
* Mark all CoreCLR Interop/COM projects OsSpecific
Maira Wenzel [Thu, 23 Apr 2020 13:52:53 +0000 (06:52 -0700)]
Update docs links to 3.1 (#35309)
Stephen Toub [Thu, 23 Apr 2020 13:51:42 +0000 (09:51 -0400)]
Fix invalid cast in StringValues.this[int] (#35302)
vivekbm [Thu, 23 Apr 2020 13:50:34 +0000 (19:20 +0530)]
Updated assert statement to include a check for Timeout in FileSystem… (#35298)
* Updated assert statement to include a check for Timeout in FileSystem.Attributes.Windows
* fixes 28831
Added the check, the earlier commit had a syntax error.
Alexander Köplinger [Thu, 23 Apr 2020 12:56:38 +0000 (14:56 +0200)]
Update Android CI build image (#35333)
Includes newest Android NDK and defines the ANDROID_NDK_ROOT environment variable which is preferred to ANDROID_NDK_HOME.
Ryan Lucia [Thu, 23 Apr 2020 12:36:16 +0000 (08:36 -0400)]
[meta] Add GCHandle embedding API v2 with pointer-sized handles (#35197)
Jo Shields [Thu, 23 Apr 2020 12:22:49 +0000 (08:22 -0400)]
Build iOS AOT compiler against LLVM, always. (#35306)
Alexander Nikolaev [Thu, 23 Apr 2020 10:53:35 +0000 (12:53 +0200)]
Timeout increased to 10 miliseconds (#35289)
Timeout increased to 10 milliseconds in Timeout_TooShort_AllPendingOperationsCanceled
Fixes #32818
David Mason [Thu, 23 Apr 2020 10:43:06 +0000 (03:43 -0700)]
Merge pull request #35049 from k15tfu/fix-coreclr-profiler-path-arm-arm64
Fix handling CoreClr profiler path for 32/64 bit flavors on ARM/ARM64
Zoltan Varga [Thu, 23 Apr 2020 10:05:13 +0000 (06:05 -0400)]
Rename the 'WebAssembly' OS to 'Browser' and the RID to 'brows… (#35323)
Jan Jahoda [Thu, 23 Apr 2020 09:15:59 +0000 (11:15 +0200)]
enable Proxy_UseEnvironmentVariableToSetSystemProxy_RequestGoesThruProxy test (#34444)
* Disable test on windows and enable on other platforms
* enable test
* Add diag logs
* enhance logging
* Add body logging
* Add more logs
* Add logs to loopback server
* add trace
* Disallow run test on WinHttpHandler and remove logs
* Clean up
* rename variable
Andy Ayers [Thu, 23 Apr 2020 08:05:57 +0000 (01:05 -0700)]
JIT: revise how the jit tracks use of generics context (#34827)
The generics context is reported by the jit specially whenever it feeds into
runtime lookups, as expansion of those lookups can expose pointers into runtime
data structures, and we don't want those data structures to be collected if
jitted code is still using them.
Sometimes uses of the context are optimized away, and reporting costs
code size and GC space, so we don't want to report the context unless there
is an actual use.
This change revises how the jit keeps track of context use -- instead of trying
to incrementally ref count uses of the generics context, we now just leverage
existing passes which do local accounting.
Initial motivation for this came from #34641 where the context use was
over-reported, but investigation showed we also some times under-report as
the context var could be cloned without changing the ref count.
So this change fixes both under and over reporting.
Closes #34641.
Santiago Fernandez Madero [Thu, 23 Apr 2020 07:35:04 +0000 (00:35 -0700)]
Re-enable Windows 7 and Windows8.1 helix queues (#35211)
Swaroop Sridhar [Thu, 23 Apr 2020 06:47:05 +0000 (23:47 -0700)]
Json-Parse: Parse with `RapidJson::kParseStopWhenDoneFlag` on all architectures (#35073)
RapidJson's `kParseStopWhenDoneFlag` indicates that parsing should stop once the root element is parsed, and should not throw an error for any further content.
This flag is useful when the input stream doesn't always have a null-terminator --
ex: an input `bytestream` (rather than a `stringstream`), files embeded within the single file bundle.
The limitation of using this flag is that any actual random text after the root element is silently ignored.
Currently, RapidJson is invoked with `kParseStopWhenDoneFlag` on Windows, but not on Unix.
This caused `kParseErrorDocumentRootNotSingular` failure on Unix when parsing json files from single-file bundles.
This change fixes this problem by passing `kParseStopWhenDoneFlag` on all platforms.
This also makes the host behavior consistent across platforms.
Tarek Mahmoud Sayed [Thu, 23 Apr 2020 03:57:13 +0000 (20:57 -0700)]
System.Diagnostics Tracing APIs (#35220)
* System.Diagnostics Tracing APIs
* address the feedback
AlFas [Thu, 23 Apr 2020 00:53:07 +0000 (03:53 +0300)]
Optimize VectorX.Clamp according to proposed change (#34896)
* Optimize Vector4.Clamp according to proposed change
* Apply clamp optimizations to all vector types
Maoni Stephens [Thu, 23 Apr 2020 00:42:51 +0000 (17:42 -0700)]
more code cleanup (#35276)
mostly for gc_low/gc_high usage.
there is one functional change which is there's an incorrect assert in plan_phase which is not true when mark_list is used
Egor Bogatov [Wed, 22 Apr 2020 21:07:00 +0000 (00:07 +0300)]
[mono] Bundle all tests to *.app for iOS as part of -subset Libs.Tests (#35283)
* Build *.app for iOS as part of Libs.Tests subset
* Add `-test` support
* Reduce *.app size (from 45 to 32 mb)
Mitchell Hwang [Wed, 22 Apr 2020 21:00:47 +0000 (17:00 -0400)]
Mdhwang/reduce mobile checks (#35105)
* [eng] Add TargetsMobile Property to Subsets.props
* Replace TargetOS mobile conditions with TargetsMobile Property
* [libraries] Remove leftover condition`
Co-authored-by: Mitchell Hwang <mihw@microsoft.com>
Nathan Ricci [Wed, 22 Apr 2020 20:06:26 +0000 (16:06 -0400)]
Implemented cmd version of excludemonofailures parameter. (#35292)
Carol Eidt [Wed, 22 Apr 2020 19:25:39 +0000 (12:25 -0700)]
Fix SuperPmi MCS tool (#35290)
It currently errors out if it finds a method without `CORINFO_FLG_STATIC`, `CORINFO_FLG_DONT_INLINE` or `CORINFO_FLG_CONSTRUCTOR`. Also, it only prints the first attribute it encounters.
It's unclear to me how this ever worked, but in any case it isn't useful to fail to generate a map or IL dump.
Brian Sullivan [Wed, 22 Apr 2020 18:42:22 +0000 (11:42 -0700)]
GCPOLL and Natural loop fixes (#34837)
* Fixes for GCPoll and COMPLUS_GCPollType=3
* This is a combination of 2 commits.
Fix for a rare case where fgReorderBlocks splits the blocks for a natural loop
* Address Code review feedback from Aaron
Egor Chesakov [Wed, 22 Apr 2020 18:01:23 +0000 (11:01 -0700)]
[Arm64] Implement ASIMD Extract Insert ExtractVector64 ExtractVector128 (#35030)
* Implements Extract, Insert, ExtractVector64 and ExtractVector128 intrinsics.
* Implements a way to generate a fallback mechanism for intrinsics accepting an immediate operand when the operand is not constant.
* Renames NoContainment flag to SupportsContainment on Arm64 (presumably, there should be fewer intrinsics supporting containment analysis so it makes more sense to have NoContainment as default)
* Removes ival column from hwintrinsiclistarm64.h table and the corresponding field in HWIntrinsicInfo struct.
Tanner Gooding [Wed, 22 Apr 2020 17:21:52 +0000 (10:21 -0700)]
Remove the explicit IEquatable implementation from the IntPtr/UIntPtr reference API (#35256)
Sergey Andreenko [Wed, 22 Apr 2020 16:56:27 +0000 (09:56 -0700)]
Fix store/load float LCL_FIELD with unaligned offsets on arm32. (#34990)
* Fix arm32 access to unaligned float fields.
* Add an additional test for the issue.
* Reenable the test.
* We don't need to require 8 bytes alignment for double, 4 is enough.
* Add comments.
* Move `buildInternalRegisterUses` after we define all of them.
* Extract `IsOffsetMisaligned`.
* Fix for the `IND float(ADD byref(LCL_VAR byref, CNT_INT 1))`.
Alexander Nikolaev [Wed, 22 Apr 2020 16:07:42 +0000 (18:07 +0200)]
Ping timeout in SendPingWithIPAddress test increased to 10 seconds (#35286)
Ping timeout in SendPingWithIPAddress test increased to 10 seconds to prevent failures.
Viktor Hofer [Wed, 22 Apr 2020 12:43:31 +0000 (14:43 +0200)]
Stop passing in auto-discovered xunit traits (#35237)
* Stop passing in auto-discovered xunit traits
With https://github.com/dotnet/arcade/pull/5314, framework and platform
traits don't need to be passed in manually anymore. Also removing the
obsolete vstest.runner.json as it isn't needed for VSTest.
Kenneth Pouncey [Wed, 22 Apr 2020 10:32:46 +0000 (12:32 +0200)]
Merge pull request #35275 from kjpou1/doc/requirements/mono
[docs] Helpful links when building the first time.
dotnet-maestro[bot] [Wed, 22 Apr 2020 10:19:33 +0000 (10:19 +0000)]
[master] Update dependencies from Microsoft/vstest dotnet/xharness (#35198)
* Update dependencies from https://github.com/microsoft/vstest build
20200420-01
- Microsoft.NET.Test.Sdk: 16.7.0-preview-
20200416-02 -> 16.7.0-preview-
20200420-01
* Update dependencies from https://github.com/microsoft/vstest build
20200421-02
- Microsoft.NET.Test.Sdk: 16.7.0-preview-
20200416-02 -> 16.7.0-preview-
20200421-02
* Update dependencies from https://github.com/dotnet/xharness build
20200421.1
- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20220.2 -> 1.0.0-prerelease.20221.1
* Update dependencies from https://github.com/dotnet/xharness build
20200421.3
- Microsoft.DotNet.XHarness.Tests.Runners: 1.0.0-prerelease.20220.2 -> 1.0.0-prerelease.20221.3
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Co-authored-by: Alexander Köplinger <alex.koeplinger@outlook.com>
monojenkins [Wed, 22 Apr 2020 10:03:54 +0000 (06:03 -0400)]
Fix a warning. (#35272)
<!--
Thank you for your Pull Request!
If you are new to contributing to Mono, please try to do your best at conforming to our coding guidelines http://www.mono-project.com/community/contributing/coding-guidelines/ but don't worry if you get something wrong. One of the project members will help you to get things landed.
Does your pull request fix any of the existing issues? Please use the following format: Fixes #issue-number
-->
Co-authored-by: vargaz <vargaz@users.noreply.github.com>
monojenkins [Wed, 22 Apr 2020 09:09:26 +0000 (05:09 -0400)]
[wasm][debugger] Fix enumerating for `this` on valuetypes (#35267)
- This also fixes the same thing for `evaluateOnCallFrame`
- And, showing type arguments in generic method names
Co-authored-by: radical <radical@users.noreply.github.com>
Vlad Brezae [Wed, 22 Apr 2020 08:49:25 +0000 (11:49 +0300)]
[amd64] Fix tailcall with arguments passed on stack (#34814)
On all platforms, OP_TAILCALL copies the arguments (that are passed on the stack) from the param area in the current frame, to the param area in the caller frame, meaning that it expects the arguments to be passed normally on the stack. However, on amd64, mono_arch_emit_call was storing these arguments directly in the param area of the caller (EMIT_NEW_ARGSTORE), instead of the param area of the current frame. OP_TAILCALL would then override the already set stack parameters with random data from the uninitialized param area of the current frame.
Jan Vorlicek [Wed, 22 Apr 2020 08:38:06 +0000 (10:38 +0200)]
Merge pull request #35173 from am11/feature/solaris/coreclr-port-without-libunwind-changes
Port CoreCLR to SunOS
Andy Ayers [Wed, 22 Apr 2020 07:47:31 +0000 (00:47 -0700)]
JIT: add suppressed zero inits if we transform tail recursive call to… (#35148)
Track whether the jit has suppressed in-body zero initialization for any locals
because at the time it appeared prolog initialization would be sufficient. If
the jit subsequently decides to change a tail recursive call into a loop, use
this information to add back the missing zero initializations.
Added a test case.
Addresses #33529.
Kenneth Pouncey [Wed, 22 Apr 2020 05:23:46 +0000 (07:23 +0200)]
[docs] Helpful links when building the first time.
- Add requirements link to the mono build README.md
- Add a link to the requirements document when missing openssl.
Nathan Ricci [Wed, 22 Apr 2020 04:18:31 +0000 (00:18 -0400)]
Enable CI to run CoreCLR tests on Mono (#34054)
Run the runtime tests on mono in ci.
Co-authored-by: Joseph Shields <directhex@apebox.org>
Levi Broderick [Wed, 22 Apr 2020 03:44:47 +0000 (20:44 -0700)]
Expose BinaryReader / BinaryWriter 7-bit encoding methods publicly (#35170)
Simon Nattress [Wed, 22 Apr 2020 03:31:23 +0000 (20:31 -0700)]
Add Checked configuration to crossgen2 project (#35266)
The runtime is often built Checked during development for faster code that's still fairly debuggable. Add a Checked build option to Crossgen2's projects which lets runtime developers build / debug in VS and consume an existing Checked JIT.
Andrew Au [Wed, 22 Apr 2020 02:43:08 +0000 (19:43 -0700)]
More clrconfig cleanup (#35261)
* More clrconfig cleanup
* Fixed a build error
Aaron Robinson [Wed, 22 Apr 2020 01:56:13 +0000 (18:56 -0700)]
Stop passing GC types in IL tests. (#35207)
Steve Pfister [Tue, 21 Apr 2020 19:35:58 +0000 (15:35 -0400)]
Enable the installer for mono Linux_x64 (#35127)
Since mono distributes only runtime packs, this change removes steps that assume the
shared framework exists along with installers.
Fixes https://github.com/dotnet/runtime/issues/35082
Mitchell Hwang [Tue, 21 Apr 2020 19:34:35 +0000 (15:34 -0400)]
Consolidate Microsoft.NETCore.App property usage (#35116)
* [eng] Add SharedFrameworkName Microsoft.NETCore.App property
* [src] Remove extra SharedFrameworkName declarations
Co-authored-by: Mitchell Hwang <mihw@microsoft.com>
Steve Harter [Tue, 21 Apr 2020 18:44:48 +0000 (13:44 -0500)]
Remove DisallowNull annotation in JsonConverter.Write() (#35022)
Lee Coward [Tue, 21 Apr 2020 18:42:24 +0000 (11:42 -0700)]
Merge pull request #35253 from dotnet/leecow-patch
Update 5.0 snap yaml
Lee Coward [Tue, 21 Apr 2020 18:32:19 +0000 (11:32 -0700)]
add 5.0 snap yamls
Lee Coward [Tue, 21 Apr 2020 18:30:02 +0000 (11:30 -0700)]
Merge branch 'master' of https://github.com/dotnet/runtime into leecow-patch
Juan Hoyos [Tue, 21 Apr 2020 17:46:15 +0000 (10:46 -0700)]
Fix dactablegen's search for MSDia and remove DIADll from the tree (#35086)
* Copy msdia140 to binary path to ensure correct bitness
* Remove DIALib binary from the tree
Simon Nattress [Tue, 21 Apr 2020 17:42:16 +0000 (10:42 -0700)]
Emit type layout check fixups (#34927)
* Emit type layout check fixups
When a ready-to-run compiled method uses value types from other version bubbles, emit a precode fixup to check at runtime that the type is still compatible (same size, alignment, GC layout).
Bring over `GCPointerMap*.cs` from CoreRT repo to get the compute the GC layout bytes that match what the runtime expects in the layout blob.
Simplify the behavior of the Crossgen test scripting. If tests opt out of crossgen by setting `<CrossGenTest>false</CrossGenTest>`, skip emitting the crossgen commands into the test run script. Also, stop setting the `RunCrossGen` and `RunCrossGen2` environment variables in the test script based on compile-time settings. These get set by runtest.py anyway. This set of tweaks allow for crossgen tests which we don't want to run the default crossgen commands on, and allow choosing either crossgen or crossgen2 as part of test execution.
Vladimir Sadov [Tue, 21 Apr 2020 17:13:52 +0000 (10:13 -0700)]
[GC] Fix the bounds of desired size equalization loop . (#35226)
* Fix the bounds of desired size equalization loop .
* Use <= in the right place.
Anirudh Agnihotry [Tue, 21 Apr 2020 16:44:01 +0000 (09:44 -0700)]
update branding to preview5 (#35215)
Kevin Jones [Tue, 21 Apr 2020 16:27:30 +0000 (12:27 -0400)]
Add API docs for HashAlgorithmName.{Try}FromOid.
John [Tue, 21 Apr 2020 16:23:33 +0000 (17:23 +0100)]
Add new IntPtr/UIntPtr API surface (#307)
* Add new IntPtr surface
* Add new UIntPtr API surface
* Add sequential layout to match [U]Int32
* Add interfaces and sequential layout
* Add interfaces
* Add namespaces
* Add namespaces
* Update UIntPtr.cs
* Update IntPtr.cs
* Change style
* make non versionable, elide copy
* fix style, elide copy, make non versionable
* Fix syntax error
* Fix style issues
* Fix style issues
* Update IntPtr.cs
* Update UIntPtr.cs
* Update ref assembly
* Allow comparison of intptr/uintptr in Array
* Fixed ELEMENT_TYPE cases, added tests based on Int32/UInt32 tests
* Fixes
* Update Array.cs
* Update Array.cs
* Update Array.cs
* Update ArrayTests.cs
* Update ArrayTests.cs
* Update UIntPtrTests.cs
* Update ArrayTests.cs
* fix instance methods??
* fixes
* Fixwa
* fix tests
* Add non versionables
* fix compare methods
* Fix comparison error
* fix boundary
* fix compares
* fix maxvals
* remove xunit buggy data
* silly var name error
* Update src/libraries/System.Private.CoreLib/src/Resources/Strings.resx
Co-Authored-By: Tanner Gooding <tagoo@outlook.com>
* Fix ToString
Co-authored-by: Tanner Gooding <tagoo@outlook.com>
Kevin Jones [Tue, 21 Apr 2020 16:07:09 +0000 (12:07 -0400)]
Add API docs for HashAlgorithmName.{Try}FromOid.
Adeel Mujahid [Tue, 21 Apr 2020 15:00:20 +0000 (18:00 +0300)]
Merge branch 'master' into feature/solaris/coreclr-port-without-libunwind-changes
Egor Bogatov [Tue, 21 Apr 2020 14:29:25 +0000 (17:29 +0300)]
[mono] Fix iOS sample (#35241)
Stephen Toub [Tue, 21 Apr 2020 13:37:38 +0000 (09:37 -0400)]
Avoid unnecessary ToLower calls in RegexCompiler generated code (#35185)
* Avoid unnecessary ToLower calls in RegexCompiler generated code
We can skip calling ToLower{Invariant} if the only character that could possibly lower-case to the character we're comparing against is that character itself. This then also lets us employ optimizations like using IndexOf when searching for `\n` as part of expressions like `.*`.
* Fix a comment
* Add test to validate UnicodeCategory assumptions
Jan Vorlicek [Tue, 21 Apr 2020 12:14:44 +0000 (14:14 +0200)]
Merge pull request #34997 from janvorli/implement-native-getmoduleindex
Implement native GetModuleIndex
Egor Bogatov [Tue, 21 Apr 2020 10:07:28 +0000 (13:07 +0300)]
[mono] Add LLVM support for iOS test runner; use runtime packs (#35106)
Eirik Tsarpalis [Tue, 21 Apr 2020 09:06:34 +0000 (10:06 +0100)]
[CBOR] Implement support for tagged types (#34767)
* [CBOR] Implement support for dateTime types
* fix test errors
* refine RFC3339 dateTime formatting logic
* add invalid date string tests
* Implement BigInteger support
* initial decimal support implementation
* use stack allocated buffers instead of thread local buffers
* add test cases for negative unix time dates
* add WriteNegativeIntegerEncoding tests
* implement checkpointing logic and apply to SkipValue() method
* implement checkpointing for tagged type readers
* store reader checkpoints in the stack
* add tests for rollback logic in tagged type readers
* add tested that utilizes nested reader state rollbacks
* use culture invariant decimal parsing in tests
* add WriteUnixTimeSeconds(double) method
* split unix time and string time readers into separate methods
* add WriteUnixTestSeconds invalid input test
* add support and tests for indefinite-length buffers in tagged types
* add PeekTag() tests
* add Read/WriteInt32() convenience methods
* add pervasive buffer postcondition checks on negative reader tests; fix related bug.
* add comment on DateTime parsing
* Rename CborReader.Peek() to CborReader.PeekState()
* Use correct overload for int32 handling
* add caching to CborReader.PeekState()
* Use "simple" instead of "special" for cbor major type
* add Write/ReadUInt32() convenience methods
* add uint32 negative tests
Stephen Toub [Tue, 21 Apr 2020 01:09:00 +0000 (21:09 -0400)]
Improve RegexCompiler's ToLower perf (#35203)
* Pool LocalBuilders in RegexCompiler
Currently GenerateGo/FindFirstChar create a few _tempLocal1, _tempLocal2, etc. LocalBuilders that various parts of the methods may use. The developer then need to keep track of which of these locals is currently being used such that we maximize the re-use of them while not accidentally overlapping their use. And the names end up being non-descriptive, so in some cases additional names are added for convenience.
The primary purpose of the PR is cleanup. It adds a pool of LocalBuilders that lets them be rented just where they're needed, such that we only create locals when necessary, can give them appropriate names, etc. However, it also makes it a lot easier for helpers to create temporary locals (e.g. as was done in EmitMatchCharacterClass) which then makes it easier to add additional optimizations.
* Avoid CultureInfo.TextInfo virtual calls in RegexCompiler
To support RegexOptions.IgnoreCase, RegexCompiler ends up generating a call to `char.ToLower(c, _culture)` for every character compared. That `ToLower` call in turn needs to access `_culture.TextInfo`, and then call its `ToLower(char)` virtual method. We can instead get the `TextInfo` once and call its `ToLower` method directly, saving a virtual call on every character being compared.
Note that we were already doing this (for .NET 5) in RegexInterpreter, but hadn't done it in RegexCompiler because of complications around locals management. The previous commit fixed that and made this possible.
* Update src/libraries/System.Text.RegularExpressions/src/System/Text/RegularExpressions/RegexCompiler.cs
Dan Moseley [Mon, 20 Apr 2020 23:03:09 +0000 (16:03 -0700)]
Add microsoft-extensions to area owners (#35186)
* Add extensions
* Add missing data rows
* Add area-CodeGen-LLVM-mono
Co-authored-by: Marek Safar <marek.safar@gmail.com>
Juan Hoyos [Mon, 20 Apr 2020 22:47:45 +0000 (15:47 -0700)]
Guard against chaching of dangling global parent class of module (#35150)
Levi Broderick [Mon, 20 Apr 2020 22:41:24 +0000 (15:41 -0700)]
Address some edge cases in EncoderNLS.GetBytes draining logic (#35182)
Aaron Robinson [Mon, 20 Apr 2020 22:31:06 +0000 (15:31 -0700)]
Add test for native COM client consuming managed COM server with Events. (#35151)
Vladimir Sadov [Mon, 20 Apr 2020 19:14:44 +0000 (12:14 -0700)]
Use pinned arrays in Sockets (#34175)
* Use pinned arrays in SocketAsyncEventArgs
* Missed one case of defensive pinning.
Levi Broderick [Mon, 20 Apr 2020 19:06:01 +0000 (12:06 -0700)]
Improve char.ToUpperInvariant / ToLowerInvariant perf (#35194)
Adeel [Mon, 20 Apr 2020 17:55:16 +0000 (17:55 +0000)]
Address PR feedback
Fan Yang [Mon, 20 Apr 2020 17:49:40 +0000 (13:49 -0400)]
Remove passing tests from exclude list (#35139)
monojenkins [Mon, 20 Apr 2020 17:44:30 +0000 (13:44 -0400)]
[tests] Disable sgen-bridge.exe on Windows x64 FullAOT+LLVM (#35201)
See https://github.com/mono/mono/issues/19603
Co-authored-by: akoeplinger <akoeplinger@users.noreply.github.com>
Marcus Turewicz [Mon, 20 Apr 2020 17:03:29 +0000 (03:03 +1000)]
Override IsByRefLike on S.R.E.*Builder classes (#34846)
* Override IsByRefLike on S.R.E.*Builder classes
Fixes #8828
* Adds tests
* Update System.Reflection.Emit.Tests.csproj
* Bumping CI
* Update mono as well
monojenkins [Mon, 20 Apr 2020 16:55:49 +0000 (12:55 -0400)]
[wasm][debugger] Implement `Runtime.callFunctionOn`, and support array expansion (#35156)
- This was prompted by vscode/vs requiring this for array expansion.
- It runs a given js function string, on a given object id.
- We handle the case for dotnet object ids.
Implementation:
- We build a proxy object that reflects the details of the real dotnet object/array
- And run the js function on that proxy object
- Then if `returnByValue` was requested, the result is returned as-is
- Else we cache that object, with a new special object id (`dotnet:cfo_res:..`),
and return that id in the response.
- Subsequently, we would get `Runtime.getProperties` request on this new object id
- And a `Runtime.releaseObject` request, to free that function result.
- All this is handled currently
Tests:
- A few tests were added which invoke `callFunctionOn`, on JS objects, and checks the result
- And the same checks are performed on equivalent dotnet objects, with the same js function
- This should help stay in sync with what we think `callFunctionOn` should work as
- These tests have an additional parameter - `roundtrip`, which calls a simple function
`function () { return this; }`
.. on a dotnet object, and gets the result object id. If `roundtrip == true`, then it
runs that same function on the result object, and gets a new result object id. And the
tests run their checks on that.
- this helps to check that outside of the proxy, the result object behaves same as any
other dotnet object.
- Also, most of the tests were modified to support an additional mode (`use_cfo`).
- With this, whenever a test wants to run `Runtime.getProperties on an object, we pass that
object through the earlier "identity" function, and then run `getProperties` on that
resultant object.
- This helps to take advantage most of the existing tests, and use them for `callFunctionOn`
testing too, by ensuring that the results of CFO returns results same as any other object.
Fixes mono/mono#19229, mono/mono#19531
Co-authored-by: radical <radical@users.noreply.github.com>
John Salem [Mon, 20 Apr 2020 16:12:21 +0000 (09:12 -0700)]
Add Reverse Diagnostics Server (#33307)
* Add Advertise IPC Command
* Update diagnostics server to use both reverse and traditional modes
* Change DOTNET_DiagnosticsServerAddress to DOTNET_DiagnosticsMonitorAddress and only use for reverse connection
* Add IpcStreamFactory abstraction
* IpcStreamFactory::Poll is now more similar to the poll API from Linux
* IpcPollHandle struct is used to abstract listening for client and server connections
* use overlapped io for all io on windows
* Add ConnectionState abstraction
* Implement timeout read/write
Swaroop Sridhar [Mon, 20 Apr 2020 15:29:38 +0000 (08:29 -0700)]
Single-File Bundler: Add a FileSize test (#35149)
Add a bundler consistency tests that verifies that the size of each embedded file
(recorded in the bundle manifest) matches its original size of the file on disk.
monojenkins [Mon, 20 Apr 2020 14:39:32 +0000 (10:39 -0400)]
Add ability to turn on just the portable executable verifier. (#35102)
Helps fix issue: https://issuetracker.unity3d.com/issues/unity-crashes-when-opening-the-project
We were hitting a crash were a corrupted dll would get loaded and then hit an assert later on when something tried to use it. The verifier suite would correctly identify the dll as corrupt but it was impossible to turn on just the PE verification. I've added a new PE only mode above "off" to allow for just the PE verifier to be used instead of the larger verification suite.
Co-authored-by: UnityAlex <UnityAlex@users.noreply.github.com>
Jan Vorlicek [Mon, 20 Apr 2020 13:52:38 +0000 (15:52 +0200)]
Fix incorrect path of the runincontext script in Helix
The source path of the runincontext.cmd/sh script in
helixpublishwitharcade.proj is incorrectly referring to the parent
directory instead of the current one as the base for the scripts folder.
Alexander Nikolaev [Mon, 20 Apr 2020 13:32:54 +0000 (15:32 +0200)]
Http2LoopbackConnection properly handles stream termination (#35100)
PR adds an optional stream termination handling to `Http2LoopbackConnection.ReadBodyAsync` and fixes `HttpClient` construction missing the `SocketHttpHandler` created in the test.
Fixes #31220
Jan Vorlicek [Thu, 16 Apr 2020 08:07:09 +0000 (01:07 -0700)]
Remove MicrosoftFileFormatsVersion from eng/Versions.props
Jan Vorlicek [Wed, 15 Apr 2020 21:16:58 +0000 (23:16 +0200)]
Remove notions of the buildtools and the GetModuleIndex
Ryan Lucia [Mon, 20 Apr 2020 11:55:53 +0000 (07:55 -0400)]
Add Mono-specific check for uninitialized default ALC on Resolving event (#35192)
Unlike the other resolvers, this one *can* exist on the default ALC, so we have to add the null check like in GetAssemblyLoadContext.
Move the now-unused ResolveUsingResolvingEvent method back to the CoreCLR-specific file.
Vladimir Sadov [Mon, 20 Apr 2020 02:49:14 +0000 (19:49 -0700)]
CastTable perf tweaks. (#34427)
* No null check in Get.
* bypass AuxData
monojenkins [Sun, 19 Apr 2020 23:34:08 +0000 (19:34 -0400)]
[mono] Fix multidimensional array construction when using programmer-specified lower bounds. (#35091)
* [mono] Fix multidimensional array construction when using programmer-specified lower bounds.
mono_array_full_new_checked and mono_array_full_new (which is marked as
a MONO_API function) both take two pointers to buffers containing
lengths and lower bounds. mono_array_new_n_icall can split the incoming
parameter list in two before forwarding the results to
mono_array_full_new_checked, so mono_array_full_new_checked was
receiving two buffers containing interleaved lower bounds and lengths.
ECMA-335 states that array constructors that specify both lower bounds
and lengths interleave these values. Deinterleave these
in method_to_ir.
Add some tests to iltests.il that verify that multidimensional arrays
with custom lower bounds work.
Fixes https://github.com/dotnet/runtime/issues/34377.
Fixes https://github.com/dotnet/runtime/issues/34378.
Fixes https://github.com/dotnet/runtime/issues/34381.
* Deinterleave lower bounds and array lengths in the interpreter.
Also make the condition in which the deinterleaving code path is taken
more precise. Perhaps this branch should be marked as unlikely.
* Fix a typo.
Co-authored-by: imhameed <imhameed@users.noreply.github.com>
nietras [Sun, 19 Apr 2020 21:23:26 +0000 (23:23 +0200)]
Update shared-generics.md to use `Method<>` instead of `Func<>` (#35181)
To avoid confusion with `Func<>` delegate and object. At least to me it create a bit of unnecessary confusion. Whether this should be `Method` or just `F` can be discussed. However, `Method` makes it clear the discussion is about a generic **method** instantiation.