Merged
Conversation
This PR removes the `SORT_MARK_STACK` feature and related `rqsort1` function from the GC. These aren't used and if we need them, we can always bring them back from history. I don't know the full history here but I assume it's not enabled because the performance benefits weren't worth the sorting cost. Plus we have prefetch in the mark phase now which should help hide memory latency.
Co-authored-by: Milos Kotlar <kotlarmilos@gmail.com>
…stack arg sizes through our data structures where possible (#123016) A quick analysis of the current use of PInvokeArgIterator leads me to believe that we would use it to calculate a value we never actually used. This PR changes the implementation of the runtime to remove those uses which were tied to construction of function pointers from delegates where the delegate parameters were a structure which required marshalling. Should be a small code size win for Unix X64 platforms along with a negligible perf boost on all non-windows platforms. We still need this thing for COM scenarios on Windows so I haven't touched it there. This work is not removing the PInvokeArgIterator itself, as it appears to be necessary for the interpreter call stub generator (although it also needs to be fixed to be correct in those cases as well.)
…custom VARIANT types in libraries (#123060) Co-authored-by: Aaron R Robinson <arobins@microsoft.com>
…rtion (#121527) When RangeCheck inspects, say, "i + cns" tree, it tries to get the "assertion-based" range for `i` via its block's `bbAssertionIn`. It's too conservative and doesn't take into account assertions created before that `i` in the block (inter-block assertions). Let's see if we can cheaply just accumulate those by hands (@AndyAyersMS's idea). A simplified version of the repro is the following (thanks @BoyBaykiller for the repro): ```cs void Test(int[] arr, int i) { arr[i] = 0; // creates 'i >= 0 && i < arr.Length' assertion i++; // same block as ^ if (i < arr.Length) arr[i] = 0; } ``` Codegen diff: ```diff ; Method Benchmarks:Test(int[],int):this (FullOpts) G_M59621_IG01: sub rsp, 40 G_M59621_IG02: mov eax, dword ptr [rdx+0x08] cmp r8d, eax jae SHORT G_M59621_IG05 mov ecx, r8d xor r10d, r10d mov dword ptr [rdx+4*rcx+0x10], r10d inc r8d cmp eax, r8d jle SHORT G_M59621_IG04 G_M59621_IG03: - cmp r8d, eax - jae SHORT G_M59621_IG05 mov eax, r8d xor ecx, ecx mov dword ptr [rdx+4*rax+0x10], ecx G_M59621_IG04: add rsp, 40 ret G_M59621_IG05: call CORINFO_HELP_RNGCHKFAIL int3 -; Total bytes of code: 56 +; Total bytes of code: 51 ```
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
6 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
See Commits and Changes for more details.
Created by
pull[bot] (v2.0.0-alpha.4)
Can you help keep this open source service alive? 💖 Please sponsor : )