Dotnet runtime master ingestion into Utf8String experiment #17

* [interp] Add intrinsic for Marvin.Block * [interp] Intrinsify ConvertAllAsciiCharsInUInt32ToUppercase * [interp] Intrinisfy UInt32OrdinalIgnoreCaseAscii * [interp] Intrinsify UInt64OrdinalIgnoreCaseAscii * [interp] Add intrinsic for UInt32ToDecStr * [interp] Add intrinsic opcode for WidenAsciiToUtf16 * [interp] Rename intrinsics to interp-intrins There seem to be some build failures on windows. Probably because of conflict with intrinsics source files from mini. * [interp] Add comments To prevent bcl and interp intrinsics going out of sync.

New property EnableMultipleHttp2Connections on WinHttpHandler enables multiple HTTP/2 connection to the same server. Contributes to #35088

Added EncodeString method that accounts for codepoint boundaries to Base64Stream and QEncodedStream. Used EncodeString instead of EncodeBytes where strings were encoded. Refactored EncodeBytes while its logic remained the same. Extracted encoding logic into ByteEncoder class. Fix dotnet#1485

…ng (#36604)

* [browser][wasm][tests] JavaScript Interop Marshal tests * Add primitive and string marshal tests * Add tests for object identity across marshaling calls for JS object and managed objects. - Tests to make sure the objects stay the same and are not created new. * Add tests and code cleanup * Add marshal of js function tests * Add delegate marshaling test * Fix License text * More tests * Fix an error where `mono_method_resolve` is called before BINDING object was initialized * Add more marshal tests * Add more tests * Address review comments

* enable GS Cookie on OSX * Actual fix

* Add always trimmed attributes list for browser config Example of size reduction on SPC with the most minimalistic Hello World sample. | System.Private.CoreLib | Size (kB) | Delta | |-|-|-| | Original | 1358 kB | - | | Trimmed | 1271 kB | - 87 kB | * Review feedback * Combine and embed the ILLink.LinkAttributes.xml file into the assembly. * Update linker command line to ignore attributes file. Add ExcludeFromCodeCoverageAttribute. * PR feedback. Add note about Obsolete. Only use assembly fullname="*" for compiler generated attributes. * Update ILLink.LinkAttributes.wasm.xml Co-authored-by: Eric Erhardt <[email protected]>

…38957) * DotProduct * DotProductBySelectedQuadruplet * MultiplyRoundedDoublingAndAddSaturateHigh * MultiplyRoundedDoublingAndAddSaturateHighScalar * MultiplyRoundedDoublingAndSubtractSaturateHigh * MultiplyRoundedDoublingAndSubtractSaturateHighScalar * MultiplyRoundedDoublingBySelectedScalarAndAddSaturateHigh * MultiplyRoundedDoublingBySelectedScalarAndSubtractSaturateHigh * MultiplyRoundedDoublingScalarBySelectedScalarAndAddSaturateHigh * MultiplyRoundedDoublingScalarBySelectedScalarAndSubtractSaturateHigh

Follow-up to dotnet/runtime#39130

Co-authored-by: Mitchell Hwang <[email protected]> Co-authored-by: Steve Pfister <[email protected]> Co-authored-by: Alexander Köplinger <[email protected]>

* Add documentation about libraries features switches * Review feedback * Casing tweaks

…te (#39111) * Inlined GC Polls for call to methods with SuppressGCTransitionAttribute. * Emit inlined GC Polls for methods with SuppressGCTransitionAttribute when possible and when optimizing. * Emit only one GC poll per basic block. * Move insertion of GC polls to a new phase `fgInsertGCPolls` that runs after most optimizations so that we don't insert unnecessary GC polls. * I plan to delete `fgCreateGCPolls` phase that was previously used to insert GC polls for platforms that don't support hijacking in a subsequent PR. We currently don't support such platforms. * Fix `fgCreateGCPoll` to be able to insert inlined GC polls for `BBJ_NONE` and `BBJ_THROW` basic blocks.

* Added field support to JSON serializer * Addressed issues * Added tests * Addressed issues

…(#39199) Once we release the semaphore, we no longer have ownership over _sendBuffer, so we have to release the latter before not after releasing the semaphore.

* add write queue to Http2Connection to optimize write handling * address feedback * remove AsyncMutex * remove AsyncMutex for real * use Channel<T> instead of custom queue * make channel options static Co-authored-by: Geoffrey Kizer <[email protected]>

- coreclr/build-test.sh now has a new subcommand: 'mono_aot', which builds a new target, named 'MonoAotCompileTests', added to coreclr/tests/src/runtest.proj. This target compiles the runtime tests using Mono LLVM AOT in a simple configuration where the host platform is identical to the target platform. Parallel compilation happens via a hack: actual compilation happens in mono/msbuild/aot-compile.proj, a single-target msbuild file, and runtest.proj invokes this single-purpose project and target using batching to create multiple parallelizable instances of this project. Future work: use the MonoAOTCompiler custom task currently used to build the iOS sample program. - Avoid using the runtimeVariant string when defining coreClrProductArtifactName in mono/templates/xplat-pipeline-job.yml. There are no "runtime variants" of CoreCLR configured with this parameter; instead, depend on a shared non-runtime-variant build of CoreCLR. - Mark function DISubprograms as local definitions--this is an LLVM 9 compatibility fix. - Use --tag=CXX when linking libmonosgen-2.0.so via libtool when LLVM is linked into Mono. This makes libtool use the C++ compiler driver when linking Mono--which uses whatever platform-specific flags are necessary to link against the C++ stdlib. Previously, libtool would use the C compiler driver, which didn't do this and would produce shared objects with no explicit dependency on libstdc++. This problem is normally masked because of the very lax dynamic linking semantics on ELF, but Mono on our CI setup is built in a CentOS 7 image (which does not contain a C++11 libstdc++) that has a GCC 7 compatibility package installed, along with a clang 9 installation that detects headers from the GCC 7 compatibility package. This compatibility package includes a libstdc++ linker script that links C++11 libstdc++ components statically into the target while dynamically linking against components present in pre-C++11 libstdc++. The end result of all of this is that Mono built with this configuration will dynamically depend on C++11 libstdc++ symbols that should have been statically linked into the library, and will outright fail to run on machines without a newer version of libstdc++ available. - Add tests that fail after LLVM AOT compilation to issues.targets.

Minimum viable support for portable pdb generation from ilasm. Supporting the following features of the portable pdb format - Portable PDB CodeView debug directory entry - Document table - MethodDebugInformation table - Local scope table - Local variable table This change extends the metadata writer to be able to write the portable pdb format. - All changes related to portable PDB metadata generation should be controlled with preprocessor directive FEATURE_METADATA_EMIT_PORT_PDB - New metadata code is not included in the runtime, as the runtime uses a managed parser when it works with portable pdbs Authored-by: Ivan Povazan <[email protected]>

This library was added to .NETStandard2.1 and thus can no-longer be changed.

They try to test various conditions that only work if you have multiple threads which is not the case on WebAssembly.

… globalization (#39228) Allows the test suite to pass on WebAssembly: `Tests run: 692, Errors: 0, Failures: 0, Skipped: 15. Time: 1.237011s`

* WASM: Fix System.Diagnostics.TraceSource tests One test was using `Assembly.GetEntryAssembly()` which returns null on WebAssembly. Others are testing that the output contains the stacktrace but it is empty on WASM right now. * PR feedback Co-authored-by: Stephen Toub <[email protected]> Co-authored-by: Stephen Toub <[email protected]>

We only support `Dns.GetHostName()` since that didn't throw in the older mono WebAssembly release and can be redirected to `Environment.MachineName`.

* [wasm] Enable System.Net.Http.Unit test suite * Remove System.Net.Http.Unit test project from the exclusion list Co-authored-by: Alexander Köplinger <[email protected]>

…9188) * [wasm] Modify OffsetOf_Decimal_ReturnsExpected for Browser * [wasm] Add ActiveIssues for GetDelegateForFunctionPointerTests and GetFunctionPointerForDelegateTests * [libraries] Remove System.Runtime.InteropServices project exclusion Co-authored-by: Mitchell Hwang <[email protected]> Co-authored-by: Alexander Köplinger <[email protected]>

* [wasm] Add ActiveIssue to LoadInDefaultContext for wasm * [libraries] Remove System.Runtime.Loader.DefaultContext project exclusion Co-authored-by: Mitchell Hwang <[email protected]> Co-authored-by: Alexander Köplinger <[email protected]>

* [wasm] Enable System.Runtime.Serialization.Json test suite * Remove System.Runtime.Serialization.Json.Tests project from the exclusion list * Remove System.Runtime.Serialization.Json.ReflectionOnly.Tests project from the exclusion list

…ialization.Xml.ReflectionOnly test suites (#39262)

… to create a function pointer to a native-to-managed wrapper. (#39260) The previous method of returning interp_no_native_to_managed doesn't work: * The runtime expects different function pointers for different methods, returning the same pointer leads to all kinds of weird errors. * On wasm, the native caller expects the function to have the right signature, so calling interp_no_native_managed () will leads a to signature mismatch error which is hard to debug.

…205)

Using `nameof(DebugView)` doesn't work because DebugView is a nested type, so this isn't the full name of the type. Fixing to use typeof instead.

…indows (#39266) Fixes #39258

Socket's connect start, stop and cancelled events are logged as activities via EventSource. Contributes to #37428

It got reactived but an unrelated PR introduced a test failure for it, disable it again.

…264)

Known issue: dotnet/runtime#39187

It was missing a few signatures in wasm_m2n_invoke.g.h that caused it to fail with `WASM-ERR: CANNOT HANDLE INTERP ICALL SIG IIILIIII`

…otnet/icu (#39158) * Update dependencies from https://github.com/mono/linker build 20200711.1 Microsoft.NET.ILLink.Tasks From Version 5.0.0-preview.3.20360.3 -> To Version 5.0.0-preview.3.20361.1 * Update dependencies from https://github.com/dotnet/runtime-assets build 20200710.1 System.ComponentModel.TypeConverter.TestData , System.Drawing.Common.TestData , System.IO.Compression.TestData , System.IO.Packaging.TestData , System.Net.TestData , System.Private.Runtime.UnicodeData , System.Security.Cryptography.X509Certificates.TestData , System.Windows.Extensions.TestData From Version 5.0.0-beta.20319.2 -> To Version 5.0.0-beta.20360.1 * Update dependencies from https://github.com/dotnet/icu build 20200709.7 Microsoft.NETCore.Runtime.ICU.Transport From Version 5.0.0-preview.8.20359.5 -> To Version 5.0.0-preview.8.20359.7 * Remove unsupported option * Bump SDK dependency to bring updated ILLink task * Update trimmer configuration property names * Keep debug symbols as they are needed for the tests * Disable trimming tests on Windows due to weird pdbs handling Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com> Co-authored-by: Marek Safar <[email protected]>

…arisons (#39209) * add OSPlatform.macOS, hide OSX * use OrdinalIgnoreCase for platform name comparisons * RuntimeInformation.IsOSPlatform(OSPlatform.macOS) must return true on OSX * add new test cases for IsOSPlatformEarlierThan and IsOSPlatformOrLater * cache the IsCurrent information * apply code review suggestion Co-authored-by: Jan Kotas <[email protected]> * apply code review suggestion Co-authored-by: Jan Kotas <[email protected]>

…OSArchitecture.ProcessArchitecture (#39196) * don't aquite lock when calling RuntimeInformation.OSArchitecture and OSArchitecture.ProcessArchitecture * make the arch fields volatile * proper fix of the Unix code * apply [SuppressGCTransition]

* WASM: Fix System.Diagnostics.TextWriterTraceListener tests It was using Process.GetCurrentProcess() which throws PNSE on WebAssembly. * PR feedback

* [Wasm] Disable System.Linq.Parallel test suite dotnet/runtime#38942 seems to be failing more frequently, so the suite will be disabled until we can provide a fix. * Adding it to the skipped test list instead of an attribute Co-authored-by: Steve Pfister <[email protected]>

* add ServerOptionsSelectionCallback to SslStream * remove System.IO * add missing file * more tests * feedback from review * feedback from review * skip SniSetVersion on win7 * fix IsNotWindows7 Co-authored-by: Tomas Weinfurt <[email protected]>

…9291)

* Add sleep before looking for dump * Enable helix only * typo

Co-authored-by: vargaz <[email protected]>

* StorePair * StorePairNonTemporal * StorePairScalar * StorePairScalarNonTemporal

* Preserve DebuggerTypeProxyAttribute classes Also, ensure the test app assembly is linked, and not copied during trimming tests, which was a bug in our test infrastructure. Fix #37307 * Add test for DebuggerTypeProxy string overload. * Update ref to match impl.

…re tests (#39276) * WASM: Fix System.IO.FileSystem.DriveInfo * WASM: Fix Microsoft.VisualBasic.Core tests Some tests need culture data. * Fix test that was accidentally reversed for non-Browser

* Fix static field names * Remove CachedEventArgs wrapper This was put in place initially when the Task-based APIs were new, and we wanted to make them pay-for-play such that other uses of Socket wouldn't incur the extra fields. But at this point, the Task-based APIs are the primary/recommended way to use Socket, and when they are used, the wrapper adds overhead (both in size and CPU). Deleting it. * Delete APM fallback from AcceptAsync * Delete APM fallbacks for Receive/SendAsync with tasks

* Remove some unnecessary tests * Handle more cases * Fix some bugs * Fix some comments * fix up docs some * add fallthrough * Formatting

…j is changed (#39239) * WASM: run tests on CI only in rolling and/or when libraries/tests.proj is changed * Fix build * Test change in tests.proj * Fix build * Flow variables from platform-matrix to jobTemplate * Fix installer base template variables propagation * Pass variables as array

* [wasm] Condition Finalizer_CollectibleWithNoAssemblyLoaded with precise gc supported * [wasm] Modify SatelliteAssemblies member data and condition on NonInvariantGlobalization * [libraries] Remove System.Runtime.Loader project from exclusions Co-authored-by: Mitchell Hwang <[email protected]>

…om single-file bundle (#39275) Changes made to PEFile to report empty string for path when it comes directly from the bundle. I experimented with changing the PEImage m_path to empty string, but way too many things break. It is valid for the m_path to be empty, but we only use it to memory-loaded assemblies where the assembly loaded via PEImage::LoadFlat which does things very differently. The bundle still relies on opening the assembly as a file (file handle and offset). The end result is that internally the assembly's PEImage has a non-existent m_path (it points to the base directory and the file name of the assembly, but no such file will normally exist). I don't think anything will actually use the path to access the file though. Also changed how we load CoreLib from bundle. We used to give it path "System.Private.CoreLib.dll" which causes asserts in the runtime in couple of places (the one I hit was when trying to load assembly from memory - it gets CoreLib as its creator and we call GetPath on the creator and then validate the path in debug builds). As far as I can tell this change should not make any observable effect, other than fixing the fact that we expect the m_path of a PEImage to be either empty or absolute path.

…9321) * Fix 2 cases of wrong prinf formatting. old typos. * Change `fgIsIndirOfAddrOfLocal` return type to the precise.

…(#39153) * Comment dispose * Fix service tests * typo

Issue: dotnet/runtime#38892

ComputeAutoFieldLayout() function ignores calculated fieldSizeAndAlignment.Alignment and uses context.Target.LayoutPointerSize instead. It produces a problems if fieldSizeAndAlignment.Alignment > context.Target.LayoutPointerSize on arm for example. I suggest to use maximum of that values to calculate offset. Fixes Issue #38822 Signed-off-by: Timur <[email protected]>

…9048) * Add TryReadTo(out ReadOnlySpan<T> span, ReadOnlySpan<T> delimiter * Tweaks

- Unify the handling of promoted args when assigning offsets; there's no need to handle independent vs. dependent promotion separately. - Use `varTypeUsesFloatReg` where appropriate. - Some minor refactoring of initial reg assignment. - Add a 'Clear' method to `StructPromotionHelper` for use after inlining (not yet in use, as it produces diffs). - Wrap a SIMD field node in `GT_OBJ` only if needed. These changes are preparatory to improved struct promotion for args.

Recognize `get_AllBitsSet` instead of `get_AllOnes`; this internal intrinsic was renamed in dotnet/runtime#36579 (1c66ad1). Implement `Vector<T>.One` as an intrinsic.

* Porting new netfx configurations for packages * Add Netfx configurations to Microsoft.Extensions and missing System packages * Making sure net461 configurations build against targeting pack assembly versions * Undo changes to add netfx configurations on reference projects that don't ship in their package * Fix issue with M.E.FP.P package by removing NetStandard.Library metapackage dependency * Fix issues with depending on higher version than the one available in net461 * Disable UTF8String verifyClosure on package * Undo changes to Microsoft.XmlSerializer.Generator * Update src/libraries/Microsoft.Extensions.FileProviders.Physical/src/Microsoft.Extensions.FileProviders.Physical.csproj Co-authored-by: Eric StJohn <[email protected]> Co-authored-by: Eric StJohn <[email protected]>

… (#39087)

Issue: dotnet/runtime#39309

PAL_VirtualUnwindOutOfProc for MacOS Add MacOS compact unwind section parsing (__unwind_info) and stepping: SearchCompactEncodingSection, SearchDwarfSection, GetProcInfo, StepWithCompactEncoding*. Add back some (from a previous commit) of the dwarf unwind info parsing functions to use for the dwarf unwind info (__eh_frame section): ReadValue*, ReadULEB128, ReadSLEB128, ReadEncodedPointer, ParseCie, ExtractFde, ExtractProcInfoFromFde. These functions were used and fairly well tested in 3.x. Build libunwind8 source on MacOS for those dwarf unwind info cases. The methoddesc's enum memory region code needed add more of the code version manager's data to the coredump.

…9316) The previous code would convert exceptions into errors which would be rethrown later, clearing the resume state. Hopefully fixes dotnet/runtime#38337. Co-authored-by: vargaz <[email protected]>

…39324)

Fixes CVE-2020-1147 https://portal.msrc.microsoft.com/en-us/security-guidance/advisory/CVE-2020-1147 See also https://go.microsoft.com/fwlink/?linkid=2132227.

One test failed with an issue: dotnet/runtime#39311 Also noticed that StringTests.cs is included from the CommonPath, this is a historical artifact from when System.Memory built for netfx/netstandard so we could run tests on those configs. We no longer need it since StringTests.cs is already ran in System.Runtime.Tests.

* Update dependencies from https://github.com/mono/linker build 20200713.5 Microsoft.NET.ILLink.Tasks From Version 5.0.0-preview.3.20361.1 -> To Version 5.0.0-preview.3.20363.5 * Update dependencies from https://github.com/dotnet/icu build 20200714.1 Microsoft.NETCore.Runtime.ICU.Transport From Version 5.0.0-preview.8.20359.7 -> To Version 5.0.0-preview.8.20364.1 Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>

* IDL Change * Store generated code for ICorDebugProcess11 * Implement ICorDebugProcess11 * PR Feedback * Fix clang builds

…#39280) This change ensures correct unwinding of stack when stepping through these methods under a debugger.

…tead, since the test still fails. (#39348)

With this change: * Building up hashes, piecemeal: IncrementalHash * One-shot hashes, polymorphically: HashAlgorithm.ComputeHash * One-shot hashes, statically: [Algorithm].HashData

…aining thunks (#39063) * [Crossgen2] Add DelayLoadMethodCallThunks table, fix stack walks containing thunks GC stress runs with Crossgen2 binaries frequently see AVs with stack-walk failures. This occurs when a delay load thunk is being executed as a stack-walk is performed. The runtime resolves any thunk address to the code info for the last managed method in the runtime functions table. Use the `DelayLoadMethodCallThunks` header entry to avoid finding an R2R method when the actual IP address is for a stub. The issue is that the binary search algorithm in `NativeUnwindInfoLookupTable::LookupUnwindInfoForMethod` which doesn't correctly return -1 if the code address being searched for is larger than any present. That issue cannot easily be fixed inside of that routine as identifying the end of a function via the `RUNTIME_FUNCTION` structure is slow on some platforms and impossible on others (Windows X86). This old bug was not hit because Crossgen emits stubs earlier in the image than managed code, whereas the two are reversed with Crossgen2. Fix generation of the RuntimeFunctionsTable so that the terminating sentinel node is not treated as part of the table in the R2R header entry for the table. This was causing the sentinel node to be treated as its own runtime function by the runtime and could lead to reading beyond the table bounds in `NativeUnwindInfoLookupTable::LookupUnwindInfoForMethod`. The bug in the binary search of the table hid this bug until now. Add the `DelayLoadMethodCallThunks` header entry. `DelayLoadMethodCallThunkNodeRange` is a placeholder node (not an `ObjectNode`) to provide a symbol to relocate against from the header. It represents the range of import thunks. Add support to the object writer for defining a range between two symbols in the same section. The relocation phase can then resolve the difference between the two symbols to determine the range size. Fixes dotnet/runtime#38482 Co-authored-by: David Wrighton <[email protected]>

…ockets (#39346)

* use arm64 intrinsics in FindFirstCharacterToEncodeUtf8 * remove whitespace * shim intrinsics apis for netstandard & netcoreapp targets * use shimmed APIs in DefaultJavascripEncoder.cs * implement MoveMask over AdvSimd * Fix build and remove x86 shims * remove stub goto labels * implement optimizations for FindFirstCharacterToEncodeUtf8 * optimize FindFirstCharacterToEncode * fix bug * implement optimizations for FindFirstCharacterToEncodeUtf8 * remove goto labels * optimize FindFirstCharacterToEncode * Optimiize FindFirstCharacterToEncodeUtf8 * bug fix * add missing shim methods * fix bug * minor cleanups * cleanup and add comments * address feedback * address feedback and add checks for endianness * address feedback * Use shims from CoreLib * Remove AdvSimdHelper.MoveMask and factor non-ascii byte locator logic into architecture-specific helper methods * address feedback * reinstate goto labels * revert removed goto labels * revert more changes * address feedback * address feedback * further optimize GetIndexOfFirstNonAsciiByte * add TODO comment

…237) * Remove debug only attributes when Debugger.IsSupported is false. Contributes to dotnet/linker#1093 * Move debug only attributes to shared CoreLib.

* [interp] Fix definition of instruction * [interp] Intrinsify span ctor It is heavily used in bcl for static read only spans. In the future we should do copy propagation through its fields so we avoid creating the span in the first place and remove other unecessary opcodes. * [interp] Allow method inlining if constructor is intrinsified * Fix case where has_references is not yet initialized

Co-authored-by: Alexander Köplinger <[email protected]>

Now that we have a new SDK with the latest linker settings, we need to change how we invoke the linker during the trimming tests.

Abort jitting with BADCODE if there's a type the jit can't handle, rather than allowing an AV in the jit. Closes #38956.

… morph. (#39137) * Fix the issue. * add an IL repo. * reenable the optimization. * Don't do the transformation when optimizations are disabled.

Co-authored-by: Geoffrey Kizer <[email protected]>

* [mono] Improve RuntimeConstructorInfo.ToString Use CoreCLR's code. Fixes dotnet/runtime#36688 * Use CoreCLR's code for RuntimeMethodInfo too

`fgOptimizeBranch` can now run during fgInsertGCPolls() so it needs additional processing for newly-created statements. Fixes #39308.

…#37607) * Forward CancellationToken to methods that can take it * Address suggestions * address suggestions * Revert using * Make rule 2016 a warning in runtime * Extra cases found * Revert changes that should not pass a token, change rule to info to verify build and test CI results * Pass CancellationToken.None to base.CreateContentReadStreamAsync, as in its default overload * Address stack overflow in MultipartContent * HostingAbstractionsHostExtensions address using * Update nuget package version to latest, Info->Warning again, fix one more case. * System.Diagnostics.ProcessWaitState.Unix token * Remove two unnecessary tokens

* Initial snapshort of vxsort implementation incorporating source code from Dan Shechter. * Bug fix from Dan. * Use bigger mark list for experiments. * Give up if the mark list size is bigger than a reasonable fraction of the ephemeral space. * Latest version from Dan, disable for ARM64. * Fixes for Linux compile, ended up disabling vxsort for Linux for now. * Experimenting with 32-bit sort - the variation that gathers mark list entries pertaining to the local heap by reading the mark lists from all the heaps appears to be too slow and scales very badly with increasing number of heaps. * 32-bit sort - preserve failing case. * Do the pointer compression/decompression in place, to improve performance, optionally write mark lists and associated information to binary files for further analysis. * Introduce runtime check whether CPU supports AVX2 instruction set. * Implement mark list growth. * Integrate new version including AVX512 code path. * Implement runtime test for AVX512 support. * Move the files for the vectorized sort to their own directory, add stubs to call AVX2 or AVX512 flavor of the sort. * Get rid of unneeded #include statement in two files. * Address codereview feedback to specifically say AVX512F instead of just AVX512 as there are multiple subsets. * Fix CMakeLists.tx files for non-x64 non-Windows targets, introduce separate max mark list sizes for WKS, remove dead code from grow_mark_list, add #ifdef to AVX512 detection to make the other architectures build. * Instead of modifying the tool-generated header file corinfoinstructionset.h, modify InstructionSetDesc.txt that it is generated from, and run the tools that generates all the files from it. * Move AVX2/AVX512 instruction set detection to GC side. * Use vectorized packer, switch packed range from uint32_t to int32_t, because that makes the sorting a bit more efficient. * Add GCConfig setting to turn vectorized sorting off, streamline ISA detection (but require initialization), rename to IsSupportedInstructionSet. * Several small improvements: - Don't waste time sorting the mark list if background GC is running as we are not going to use it. - Use smaller max mark list size if we cannot use AVX2/AVX512 instruction sets - Fix mark list overflow detection for server GC. * Address codereview feedback - add constants for the thresholds above which we use AVX/AVX512F instruction sets. Add space before parameter lists as per GC codebase coding conventions. Improve some comments. * Add license headers and entry in THIRD-PARTY_NOTICES.TXT for Dan's vectorized sorting code. * Update license headers * Address code review feedback: - fix typo in comment in InitSupportedInstructionSet - move test for full GC to beginning of sort_mark_list - in WKS GC, we can use the tighter range shigh - slow for the surviving objects instead of the full ephemeral range. - make the description for the new config setting GCEnabledInstructionSets more explicit by enumerating the legal values and their meanings. * Snapshot for Linux changes * Add more definitions to immintrinh.h * Fix cmake warnings about mismatched endif clauses. * Disable Linux support for now due to multiple compile & link errors. * Address code review feedback: - add instructions to bitonic_gen.py - centralize range and instruction set checks in do_vxsort - add parentheses around expressions. - removed some printfs, converted others to dprintf - strengthened assert

Crossgen2 currently prints many unnecessary details to the console using strings that are regarded as temporary - Update the type system exceptions to use descriptive strings instead of [TEMPORARY ERROR MESSAGE]XXX - Do not print any error when a method isn't compiled due to lack of IL code - This is common, and not a failure of any form - Tie logging of non-compiled methods to the --verbose switch - We may wish to provide variable logging levels, but that can easily be done in the future

* don't mark stack indir as 'X'. * Fold OBJ(ADDR(x)) under ASG when possible. * FIx issues found by Carol

* use new API on new windows to get TLS13 * use unmanaged ref to avoid copy * feedback from review * feedback from review * feedback from review * kick the build

Create manifest attributes that are consistent with arcade's usage.

Fixes #13591

…ative path (#39331) * Get absolute path

* Add JsonIgnoreCondition.WhenWritingNull * Address review feedback

Applications are given the option of disabling BinaryFormatter.Serialize and BinaryFormatter.Deserialize, meaning they'll throw NotSupportedException if anybody tries calling them. WASM projects always disable these APIs via a hardcoded mechanism; there is no re-enablement mechanism for those projects. This commit only introduces the switch. It does not disable it in any project types by default. (Exception: wasm, as mentioned above.) The related PR dotnet/sdk#12434 causes aspnet 5.0+ applications to disable BinaryFormatter functionality by default, but apps can re-enable if needed.

Fixes issue #dotnet/runtime#2014 Adds back ELF unwind info parsing code that has been around since 3.0 so it is fairly well tested. It is conditional on CLR_CMAKE_USE_SYSTEM_LIBUNWIND, otherwise, PAL_VirtualUnwindOutOfProc uses the new libunwind8 _OOP_find_proc_info function.

* update branding to rc1 * update branding to rc1 Co-authored-by: Jan Jahoda <[email protected]>

…714.3 (#39392) Microsoft.DotNet.XUnitExtensions , Microsoft.DotNet.GenFacades , Microsoft.DotNet.Build.Tasks.Feed , Microsoft.DotNet.Build.Tasks.Packaging , Microsoft.DotNet.Build.Tasks.SharedFramework.Sdk , Microsoft.DotNet.Build.Tasks.TargetFramework.Sdk , Microsoft.DotNet.CodeAnalysis , Microsoft.DotNet.GenAPI , Microsoft.DotNet.XUnitConsoleRunner , Microsoft.DotNet.Helix.Sdk , Microsoft.DotNet.RemoteExecutor , Microsoft.DotNet.VersionTools.Tasks , Microsoft.DotNet.Arcade.Sdk , Microsoft.DotNet.ApiCompat From Version 5.0.0-beta.20330.3 -> To Version 5.0.0-beta.20364.3 Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>

…s.CreateDefaultBuilder_RegistersEventSourceLogger (#39411) * ifdefing SuppressGCTransition out of Logging.Console Fixes #39404 Co-authored-by: Adam Sitnik <[email protected]>

…ment (#39427)

* Sprinkle SuppressGCTransition across Unix shim P/Invokes Did a quick audit and added it to all the trivial functions that just return a constant or return some static or map one value to another or such things. * Address PR feedback * Update src/libraries/Native/Unix/System.Native/pal_io.c Co-authored-by: Aaron Robinson <[email protected]> Co-authored-by: Aaron Robinson <[email protected]>

* Allow building libs tests for all configurations Building libraries tests and passing in the -allconfigurations switch results in the libraries packages being built and tested. Conditioning the packages build/test on the /p:TestPackages switch.

…39389) * Make ILBlittablePtrMarshaler participate in marshaller inheritance * Add test for class layout with Out attribute scenario.

…#39408)

* Remove and sort usings * Remove AnsiSystemConsole * Remove and sort usings on tests * flip ifdefs * Add missing xml docs * PR feedback

* Mmap R2R images at and non-page-boundary Assemblies embedded within a single-file are aligned at 16 byte boundaries. However, currently MappedImageLayout assumes that assemblies are mapped at 4K offsets. This results in map failure and subsequent failure to load the image in a mapped layout. Subsequently, the image was loaded as IL as a fallback. The change fixes this problem by: mmap() pages containing sections/headers using appropriate ROUND_DOWN and ROUND_UP operations Load sections at unaligned offsets in MAPMapPEFile. Fixes: #38061 * Relaxing map alignment check to 16 bytes * do not round up header size * asserts * reject sections where data and VA hint have different page alignment * PR feedback * revert init of loadedHeaderBase and add a comment Co-authored-by: Swaroop Sridhar <[email protected]>

* Enable logging for CI runs * mistaken variable name * realized I was fixing the wrong thing. This should just set logging params for CI jobs. * fix indent * mistakenly moved variable groups to the wrong block. * remove newline * remove newline

System.Resources.Extensions writes a type-reference into the .resources file it writes but RAR will never see this when considering conflicts for bindingRedirects. Force a bindingRedirect whenever this package is used.

…#39151)

…n after metadata properties (#39254)

* Fix issue with Trimming tests on Windows * Update the IsSymbolFile metadata calculation

* Add validation to ConfigureOptions * Improve comments

* WASM: fix System.Runtime.Tests * Enable System.Runtime.Tests * Disable test and don't strip link attributes on all product assemblies * PR feedback

* Unit test * Fix * Typo * CR feedback

Add documentation to contructor in RuntimeLibrary.

* Move libraries test build to product build * Publish artifacts needed by coreclr tests * Don't build corelib anymore since it is a P2P from libraries now

…tInvalidChar (#39050) * AdvSimd support for System.Text.Unicode.Utf16Utility.GetPointerToFirstInvalidChar * Move using directive outside #if. Improve Arm64MoveMask. * Change overloads * UIn64 in Arm64MoveMask * Build error implicit conversion fix * Rename method and use simpler version * Use ShiftRightArithmetic instead of CompareEqual + And. * Remove unnecessary comment * Add missing shims causing Linux build to fail

This PR overhauls runtime startup/asset loading and adds support for ICU integration. The mono-config.js format is reworked and simplified, with new functionality added: Individual assets can be loaded from one or more remote sources with configurable fallback behavior In addition to the existing support for loading assemblies, you can now pre-load arbitrary files into the native heap or into emscripten's virtual file system. VFS support previously only existed in runtime-test.js but now is available to any consumer of dotnet.js. Assets can have a virtual path set so that their application-facing path does not necessarily have to match their path on the server. One or more ICU data archives can be added to the assets list and will be automatically loaded and used to enable ICU-based globalization support. Many configuration knobs that previously required API calls can now be set declaratively in the configuration file (environment variables, etc.) WasmAppBuilder is updated to add ICUDataFiles and RemoteSources parameters that can be used to add the associated information to the config file declaratively from a msbuild project. Various adjustments are made to existing tests and test cases so that they will pass with the addition of ICU integration. Co-authored-by: EgorBo <[email protected]> Co-authored-by: Alexander Köplinger <[email protected]> Co-authored-by: Larry Ewing <[email protected]>

* Removing unused Resources from System.Threading.Overlapped * Removing unused code from S.T.O.csproj and add Resources check to test

* native libs location * pass artifacts locations as CMake variables. * Link native libs * use LibrariesConfiguration to discover libraries artifacts * redirect to current exe * remove bundled binaries from the netcoreapp/pkg * delete exports file * no bundling of native libs on OSX * only link native libs on Linux * Support linking of coreclr on Linux only * PR feedback * move the shim to dllimport.cpp * Use library names vs. file names * removed ShouldRedirectToCurrentLibrary. NULL library name in PAL_LoadLibrary will be always considered self-referring.

…ther Read (#39336)

… (#39441) Crossgen2 would try to override input files if they are on a different drive or a share with respect to _compositeRootPath.

… (#39481)

… (#39400)

…(#39486) This reverts commit 211bc88.

* use static method calls * remove unnecessary helpers calls

… (#39511) Co-authored-by: Jan Jahoda <[email protected]>

* move all OSPlatformAttributes to a single file so they can be easily referenced in < .NET 5 libraries * Adds System.Runtime.Versioning*Platform* annotation attributes to < .NET 5 builds * introduce MinimiumSupportedWindowsPlatform * introduce IsWindowsSpecific setting that adds MinimumOSPlatform attribute for Windows Specific libraries * mark all Windows-specific libraries as such * annotate Windows-specific System.Console methods and properites * annotate Windows-specific DpapiProtectedConfigurationProvider type * annotate throwing Windows-specific APIs from System.Diagnostics.Process namespace * annotate throwing Windows-specific APIs from System.IO.MemoryMappedFiles namespace * annotate Windows-specific APIs from System.IO.Pipes namespace * the new attributes should support enums * annotate Windows-specific APIs from HttpListenerTimeoutManager class * annotate Windows-specific APIs from System.Net.Sockets namespace * annotate Windows-specific APIs from System.Runtime.InteropServices namespace * annotate Windows-specific APIs from System.Security.Cryptography.Csp namespace * annotate Windows-specific APIs from System.Security.Cryptography.X509Certificates namespace * annotate Windows-specific APIs from System.Threading namespace * address code review feedback: dont introduce new constants, reuse existing SYSTEM_PRIVATE_CORELIB * code review: import the namespace, simplify the attribute name * add missing Socket.DuplicateAndClose and Socket(SocketInformation) * fix the test name (it does not throw) * add missing DriveInfo.set_VolumeLabel * add System.IO.FileSystem File Encrypt and Descrypt methods * simplify the < .NET 5 check * enable nullable in explicit way as it's not enabled in all the projects where this file is included * include platform attributes in projects that ask for it in explicit way * set IncludePlatformAttributes to true for projects that include files from other projects * use suggestions from Viktor and Santi Co-authored-by: Viktor Hofer <[email protected]>

…lying type. (#39207) runtime\src\tests\tracing\common\TraceConfiguration.cs tries to call a constructor of using type System.Diagnostics.Tracing.EventPipeSerializationFormat with a different (but compatbile) enum type using reflection. Both enums are of underlying type, Int32, but the call fails on Mono since it fails to recognize that the two enums are compatible in CheckValue. Fix will make sure call to IsConvertibleToPrimitiveType uses underlying type as target type when checking if current value is convertible to a compatible value.

This change adds the set of signatures required for SkiaSharp to work untouched using the signatures already provided for the existing platform. Co-authored-by: mattleibow <[email protected]>

…642) Fixes interpreter test lane not running on interpreter

…ResourceManager.Tests test suites (#39513)

…on wasm. (#38991) * Add a --enable-minimal=eventpipe configure option. Disable eventpipe on wasm. * Address review comments. * Disable FeaturePerfTracing for wasm. * Disable calls to ep_init()/ep_cleanup() as well.

* Small performance improvment * Update Matrix3x2 as well

…ss (#39399) Issue: dotnet/runtime#39398

* Add telemetry to System.Net.NameResolution * Rename hostName to hostNameOrAddress * Use constants for event IDs * Address PR feedback * Reduce overhead when only the counter is needed * Make NameResolution Telemetry an activity * Remove redundant check * Fix indentation * Check IsEnabled before AfterResolution to help linker * Add Telemetry tests * Throws => ThrowsAny

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Dotnet runtime master ingestion into Utf8String experiment #17

Dotnet runtime master ingestion into Utf8String experiment #17

Commits on Jul 13, 2020

Commits on Jul 14, 2020

Commits on Jul 15, 2020

Commits on Jul 16, 2020

Commits on Jul 17, 2020