- 03 11月, 2022 5 次提交
-
-
由 dotnet-maestro[bot] 提交于
* Update dependencies from https://github.com/dotnet/linker build 20221027.4 Microsoft.NET.ILLink.Tasks From Version 7.0.100-1.22525.3 -> To Version 7.0.100-1.22527.4 * Update dependencies from https://github.com/dotnet/linker build 20221031.1 Microsoft.NET.ILLink.Tasks From Version 7.0.100-1.22525.3 -> To Version 7.0.100-1.22531.1 * Update dependencies from https://github.com/dotnet/linker build 20221031.2 Microsoft.NET.ILLink.Tasks From Version 7.0.100-1.22525.3 -> To Version 7.0.100-1.22531.2 * Update dependencies from https://github.com/dotnet/linker build 20221101.1 Microsoft.NET.ILLink.Tasks From Version 7.0.100-1.22525.3 -> To Version 7.0.100-1.22551.1 Co-authored-by: Ndotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
-
由 Adeel Mujahid 提交于
-
由 Brian Bohe 提交于
* Removing empty variable live ranges The debugger is not using empty variable live ranges. We are reporting them because they can get extended later if the variable becomes alive in the immediately next emitted instruction. If an empty live range is not getting extended, which we can realize after emitting all the code or creating a new live range for the same variable, we can remove it. * Extending variable live ranges in more cases When the emitter moved to the next group but has not emitted any instruction, and the variable died and becomes alive again, we would like to extend its range. * Avoiding creating a new debug range when previous is empty * Updating check for empty debug ranges * Updating print * Avoiding printing twice variable live range * Avoiding reporting empty variable ranges to the vm * Revert "Avoiding printing twice variable live range" This reverts commit 4e1cf47dd6cdf9d45ce4a51eaa05b3ec6e4b3b41. * Revert "Updating print" This reverts commit 7b79b0d955daa4dc9770b604d6780591b51c9ee1. * Revert "Updating check for empty debug ranges" This reverts commit e8b102d489d79028750068decf4f427ad8e5f69f. * Revert "Avoiding creating a new debug range when previous is empty" This reverts commit a11fd5d0ffaa98631d731bec1f2619f1d108d33a. * Revert "Extending variable live ranges in more cases" This reverts commit 609605a1ca7cf1c0c843dbaf353432ec9ed846e2. * Revert "Removing empty variable live ranges" This reverts commit 66d18e031f83c8efe864981b35c0548af49e0714. * Freeing vm memory when there is no debug info * Persisting JIT-EE contract on empty debug info * Update src/coreclr/jit/ee_il_dll.cpp Co-authored-by: NJakob Botsch Nielsen <Jakob.botsch.nielsen@gmail.com> Co-authored-by: NBrian Bohe <brianbohe@microsoft.com> Co-authored-by: NJakob Botsch Nielsen <Jakob.botsch.nielsen@gmail.com>
-
由 Tanner Gooding 提交于
* Mark and expose additional Vector functions as Intrinsic * Applying formatting patch * Ensure Vector64.CreateScalarUnsafe for long/ulong/double is handled * Ensure we use the right Count * Ensure small types don't sign extend up to int * Ensure reflection invocation is correct for Vector<T> bitwise ops * Adding tests and fixing a couple small typing issues * Fixing the name of the One property * Assert GT_RSH for simd isn't TYP_LONG on xarch * Ensure the correct operation is picked for vector ShiftRightArithmetic * Fixing StoreAlignedNonTemporal on Arm64
-
由 Aleksey Kliger (λgeek) 提交于
* Add new test ReflectionAddNewMethod * FIXME: get_param_names, get_marshal_info and custom_attrs need work * WIP - add a method param reverse lookup * look up params from added methods * Remove FIXMEs and unused field * remove writelines from test * fix test on coreclr * why does coreclr have 2 attributes here?? * There should be 2 attributes on the 4th param * one more place that looks at params * A couple more places where we look at the Params table * Check default values on params on added method * fix lookup if table is empty * add a gratuitious typeof assert otherwise the CancellationToken type is trimmed on wasm * Add a single mono_metadata_get_method_params function remove duplicated code
-
- 02 11月, 2022 21 次提交
-
-
由 Jakob Botsch Nielsen 提交于
Split up "we have a profile" and "we have profile weights" questions.
-
由 Matous Kozak 提交于
* Remove Vector128 checks for amd64 APIs * refactor emit_sum_vector (amd64)
-
由 Jay Cho 提交于
-
由 Zoltan Varga 提交于
* [mono][wasm] Handle delegates decorated with [UnmanagedFunctionPointer] in the interp-to-native generator. Fixes https://github.com/dotnet/runtime/issues/76930. * Update src/tasks/WasmAppBuilder/PInvokeTableGenerator.cs Co-authored-by: NAnkit Jain <radical@gmail.com> * Update src/tasks/WasmAppBuilder/PInvokeTableGenerator.cs Co-authored-by: NAnkit Jain <radical@gmail.com> * Update src/tasks/WasmAppBuilder/PInvokeTableGenerator.cs Co-authored-by: NAnkit Jain <radical@gmail.com> Co-authored-by: NAnkit Jain <radical@gmail.com>
-
由 mfkl 提交于
* Add option for truncated stream detection fix https://github.com/dotnet/runtime/issues/47563 * Use RemoteExecutor move the test to concrete classes as abstracted classes are not supported by RemoteExecutor * review feedback * use same error text message * cache appcontext getswitch * fix failing test * slice byte array for assertion * renaming * add missing RemoteExecutor.IsSupported * fast check first
-
由 Ankit Jain 提交于
This reverts commit 0e24ea7c. It broke all the perf pipeline wasm builds. It happens with v8 `10.9.130`: ``` Error: Invalid 'startMark' argument: No numeric 'startTime' field at pe (/home/helixbot/work/B2E7090F/w/A520093B/e/performance/artifacts/bin/for-running/MicroBenchmarks/659adab0-5c86-4e74-8913-672bd56fd58e/bin/net7.0/browser-wasm/AppBundle/dotnet.js:3:12994) at /home/helixbot/work/B2E7090F/w/A520093B/e/performance/artifacts/bin/for-running/MicroBenchmarks/659adab0-5c86-4e74-8913-672bd56fd58e/bin/net7.0/browser-wasm/AppBundle/dotnet.js:5:95874 exiting due to exception: Error: Invalid 'startMark' argument: No numeric 'startTime' field ``` - It works fine with older versions like `10.7.193` (current stable). - It wasn't caught in the CI checks because those are using an older version of v8.
-
由 Andy Ayers 提交于
Allow up to 1/50 variance between the expected and computed values. Fixes #77450.
-
由 Tarek Mahmoud Sayed 提交于
-
由 Tarek Mahmoud Sayed 提交于
-
由 Michelle McDaniel 提交于
* Move docker tags to -latest As part of https://github.com/dotnet/arcade/issues/10123, we have added a floating -latest tag to all currently in-support docker container images. This change moves all container reference to the -latest version so runtime can get all of the latest updates to the containers. * Change all of the centos-8-rpmpkg images to centos-7-rpmpkg CentOS 8 was EOL and has been removed as a supported docker image in dotnet-buildtools-prereqs-docker * Replace -latest tags with new tag schema * Move tests off eol docker containers * Update the images in the new infra
-
由 Parker Bibus 提交于
-
由 MSDN.WhiteKnight 提交于
-
由 Brian Bohe 提交于
Co-authored-by: NBrian Bohe <brianbohe@microsoft.com>
-
由 Aaron Robinson 提交于
* Expand ComWrappers mask for vtable lookup on 64-bit
-
由 Pavel Savara 提交于
* wip * wip * cleanup * feedback * feedback
-
由 Aaron Robinson 提交于
* Update SuperPMI artifact logging
-
由 Ankit Jain 提交于
- Add `[assembly:System.Runtime.Versioning.SupportedOSPlatform("browser")]` to the browser, and console templates. This would allow the analyzers, if enabled, to treat the assembly as one that will run only on browser. - Populate `@(SupportedPlatform)` with only `browser`, for *wasm* projects, similar to https://github.com/dotnet/sdk/blob/fef8cedfb6b4ac85a7e135f3e4f155e29cdcbdf1/src/BlazorWasmSdk/Targets/Microsoft.NET.Sdk.BlazorWebAssembly.5_0.targets#L75-L79 - This also fixes the firefox CI build which regressed recently
-
由 Egor Bogatov 提交于
-
由 Kevin Jones 提交于
* Recommend using Visual Studio's ninja
-
由 Jakob Botsch Nielsen 提交于
* Update references to Ngen variants of JIT environment variables * Update crossgen to crossgen2 * SPC instead of mscorlib
-
https://github.com/dotnet/arcade由 dotnet-maestro[bot] 提交于
Microsoft.DotNet.ApiCompat , Microsoft.DotNet.Arcade.Sdk , Microsoft.DotNet.Build.Tasks.Archives , Microsoft.DotNet.Build.Tasks.Feed , Microsoft.DotNet.Build.Tasks.Installers , Microsoft.DotNet.Build.Tasks.Packaging , Microsoft.DotNet.Build.Tasks.TargetFramework , Microsoft.DotNet.Build.Tasks.Templating , Microsoft.DotNet.Build.Tasks.Workloads , Microsoft.DotNet.CodeAnalysis , Microsoft.DotNet.GenAPI , Microsoft.DotNet.GenFacades , Microsoft.DotNet.Helix.Sdk , Microsoft.DotNet.PackageTesting , Microsoft.DotNet.RemoteExecutor , Microsoft.DotNet.SharedFramework.Sdk , Microsoft.DotNet.VersionTools.Tasks , Microsoft.DotNet.XUnitConsoleRunner , Microsoft.DotNet.XUnitExtensions From Version 8.0.0-beta.22524.5 -> To Version 8.0.0-beta.22528.1 Co-authored-by: Ndotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
-
- 01 11月, 2022 14 次提交
-
-
由 Steve Dunn 提交于
-
由 Andrew Au 提交于
-
由 Jakob Botsch Nielsen 提交于
Fix some component governance warnings.
-
由 Alan Hayward 提交于
* Arm64: Add If conversion pass * Minor review fixups * Return a PhaseStatus * Fix formatting * Check for side effects on NOPs * Add function block comments for the phase * Remove creation of AND chains from if conversion pass * Update middleBlock flow * Check for order side effects * Remove COLON_COND check * Remove flag toggling * Move the conditional assignment to the JTRUE block * Fix formatting * Allow conditions with side effects * Fix formatting * Correct all moved SSA statements * Add size costing check * Only move middle block ssa defs * Fix formatting * Fewer SSA assumptions * Use implicit func for value numbering * Update header for gtFoldExprConditional * Cost based on speed * Add Stress mode for inner loops * Rework costings * Check for invalid VNs * Ignore compares against zero * Ensure float compares are contained * Allow if conversion of test compares * Do not contain test compares within compare chains * Add float versions of the JIT/opt/Compares tests * Fix formatting * Compare chains use CmpCompares, selects use Compares * Fix flow checking for empty blocks * Fix to contexts setting JitStdOutFile * Fix attr and reg producing in select generation
-
由 SingleAccretion 提交于
Mostly I was after the FP morphing, but removed some other things along the way as well. No diffs except a nice TP boost as expected.
-
由 Egor Bogatov 提交于
Co-authored-by: NJan Kotas <jkotas@microsoft.com> Co-authored-by: NSingleAccretion <62474226+SingleAccretion@users.noreply.github.com>
-
由 Vladimir Sadov 提交于
* Delete managed CoreClr WR * all but com works * COM tests are passing * HandleTagBits const for NativeAot * Exclusive Set * fix * Use sign bit * Platforms not supporting COM can mask only one bit. * new approach * fix mono build * check for FEATURE_COMWRAPPERS too * stub NativeAOT support (NYI). * current * moved handle tags on the managed side to one location * Getter optimizations * Optimizations for Setter * accessibility of some members * ensure identity of the rehydrated RCW * make ComWeakRefToObject a QCall * delete unused pWeakReferenceOfTCanonMT and pWeakReferenceMT * byte-aligned * cleanup unreachable code * renamed WeakReferenceObject::m_Handle -> WeakReferenceObject::m_taggedHandle * Apply suggestions from code review Co-authored-by: NAaron Robinson <arobins@microsoft.com> * some PR feedback * GetWeakHandle no longer cares about inlining. * turn ObjectToComWeakRef into a QCall * revert changes under coreclr\gc * added a note to eventually remove HNDTYPE_WEAK_NATIVE_COM * Update src/coreclr/gc/gcinterface.h Co-authored-by: NMaoni Stephens <Maoni0@users.noreply.github.com> Co-authored-by: NAaron Robinson <arobins@microsoft.com> Co-authored-by: NMaoni Stephens <Maoni0@users.noreply.github.com>
-
由 Elinor Fung 提交于
-
由 Neale Ferguson 提交于
-
由 Filip Navara 提交于
-
由 Will Smith 提交于
-
由 Andy Ayers 提交于
Record some information about each SSA def and try and keep it conservatively correct through the first few optimization phases. We note: * total number of uses * whether all uses are in the same block as the def * whether there are any phi uses Subsequent phases that introduce new uses must now call `optRecordSsaUses` on the new trees they create to update these accounts. This information is cross-checked versus the IR in post phase checking. Because we don't have a well-defined mechanism to track when nodes are deleted the recorded counts may end up being overestimates in subsequent phases. This is ok, but underestimates are flagged as errors.
-
由 Aaron Robinson 提交于
This was mechanism for IBC to track metadata access. It is no longer used.
-
由 Roman Marusyk 提交于
* Trim trailing SP and HTAB for header value on HTTP/1.1 * Fix condition Co-authored-by: NMiha Zupan <mihazupan.zupan1@gmail.com> * Add tests * Update src/libraries/System.Net.Http/tests/FunctionalTests/SocketsHttpHandlerTest.cs Co-authored-by: NMiha Zupan <mihazupan.zupan1@gmail.com> Co-authored-by: NMiha Zupan <mihazupan.zupan1@gmail.com>
-