[RyuJIT] Make casthelpers cold for sealed classes #49295
Conversation
dotnet-issue-labeler
bot
added
the
area-CodeGen-coreclr
|
I am wondering - could these cold blocks be safely marked as |
heh, I'm testing it now :-) |
|
I wonder if we should introduce a new throw helper for this? Among other things we might be able to common the calls (tail merge) which might mitigate some of the code size increase you are seeing. |
From my understanding, all of our corinfo throw helpers are paramterless - and this one requires arguments: so we can't merge same call but with different arguments. |
|
New diff is for other stuff the diff is small because the change introduces a very little size diff (0-few bytes). e.g. for this case: string Test(object o) => (string)o;old codegen: ; Method Pr:Test_sealed(System.Object):System.String:this
G_M52918_IG01:
sub rsp, 40
G_M52918_IG02:
mov rax, rdx
test rax, rax
je SHORT G_M52918_IG05
G_M52918_IG03:
mov rcx, 0xD1FFAB1E
cmp qword ptr [rax], rcx
je SHORT G_M52918_IG05
G_M52918_IG04:
call CORINFO_HELP_CHKCASTCLASS_SPECIAL
G_M52918_IG05:
nop
G_M52918_IG06:
add rsp, 40
ret
; Total bytes of code: 38new codegen: ; Method Pr:Test_sealed(System.Object):System.String:this
G_M52918_IG01:
sub rsp, 40
G_M52918_IG02:
mov rax, rdx
test rax, rax
je SHORT G_M52918_IG04
G_M52918_IG03:
mov rcx, 0xD1FFAB1E
cmp qword ptr [rax], rcx
jne SHORT G_M52918_IG05
G_M52918_IG04:
add rsp, 40
ret
G_M52918_IG05:
call CORINFO_HELP_CHKCASTCLASS_SPECIAL
int3
; Total bytes of code: 38
|
|
So the size regressions in say |
Here is the diff: https://www.diffchecker.com/7cVv3EsU I can take a look on why tails weren't merged or can file an issue for it and do it later if it already meets the bar to be merged as is now. |
|
Presumably it is this method: https://github.com/dotnet/roslyn/blob/6da1274c9d24c2f90a48290394a951b23617f2a3/src/Compilers/CSharp/Portable/BoundTree/BoundTreeVisitors.cs#L30-L42 where all the |
@AndyAyersMS no, it's a similar method but from VisualBasic 🙂 https://github.com/dotnet/roslyn/blob/6da1274c9d24c2f90a48290394a951b23617f2a3/src/Compilers/VisualBasic/Portable/BoundTree/BoundTreeVisitor.vb |
|
Regarding the |
|
Seems like the |
So the actual method from the diff is this one: https://github.com/dotnet/roslyn/blob/main/src/Compilers/VisualBasic/Portable/Generated/BoundNodes.xml.Generated.vb#L9290 it has a number of CType(variable, SomeRefClass) which are guarded by a custom enum (JIT doesn't know it relates to the underlying type) - so the cast helpers are valid. |
Got it. Thanks. |
* upstream/main: (83 commits) Fix a crash in llvm if the sreg of a setret is not set because the methods ends with a throw. (dotnet#49122) [macOS-arm64] Disable failing libraries tests (dotnet#49400) improve PriorityQueue documentation (dotnet#49392) [wasm] Fix debugger tests (dotnet#49206) [mono] Fix the emission of EnumEqualityComparer instances into the corlib AOT image. (dotnet#49402) jitutils M2M renaming reaction (dotnet#49430) WinHttpHandler: Read HTTP/2 trailing headers [RyuJIT] Make casthelpers cold for sealed classes (dotnet#49295) JIT: Non-void ThrowHelpers (dotnet#48589) Update package index for servicing (dotnet#49417) Remove unnecessary check on polymorphic serialization (dotnet#48464) Remove release build cron triggers from jitstress jobs (dotnet#49333) [main] Update dependencies from dotnet/arcade dotnet/llvm-project dotnet/runtime-assets (dotnet#49359) Implement AppleCryptoNative_X509GetRawData using SecCertificateCopyData [AndroidCrypto] Support a zero-length salt for HMACs. (dotnet#49384) Use managed implementation of pbkdf2 for Android's one-shot implementation. (dotnet#49314) Make 303 redirects do GET like Net Framework (dotnet#49095) Make sure event generation is incremental (dotnet#48903) Add amd and Surface arm64 perf runs (dotnet#49389) Enregister EH var that are single def (dotnet#47307) ...
ghost
locked as resolved and limited conversation to collaborators
Currently we optimize such casts in JIT to the following:
In case of "sealed" classes that
CORINFO_HELP_CHKCASTCLASS_SPECIALis only used to throwInvalidCastExceptionfor cases when the input object is of a wrong type, so we can mark the whole block as rarely-executed.Jit-diff:
Size improvement examples:
https://www.diffchecker.com/4hK014gu
https://www.diffchecker.com/tWgKrbyl (from my understanding - it merged cold tails)
but mostly it's a regression due to additional jumps (from cold blocks).
/cc @AndyAyersMS @VSadov