From: "alanwu (Alan Wu) via ruby-core" <ruby-core@...>
Date: 2025-01-29T00:57:43+00:00
Subject: [ruby-core:120815] [Ruby master Bug#21021] "try to mark T_NONE object" with 3.4.1

Issue #21021 has been updated by alanwu (Alan Wu).


I have a patch that should address the issue. It definitely fixes *a* bug, seems to fix the workload in #21087. If you'd like, try it on for size and share if it works!

```patch
diff --git a/yjit/src/codegen.rs b/yjit/src/codegen.rs
index 37ddbce0bb..e8153c17b0 100644
--- a/yjit/src/codegen.rs
+++ b/yjit/src/codegen.rs
@@ -8068,7 +8068,6 @@ fn gen_send_iseq(
         }
     }
 
-    // Don't nil fill forwarding iseqs
     if !forwarding {
         // Nil-initialize missing optional parameters
         nil_fill(
@@ -8103,9 +8102,13 @@ fn gen_send_iseq(
         assert_eq!(1, num_params);
         // Write the CI in to the stack and ensure that it actually gets
         // flushed to memory
+        asm_comment!(asm, "put call info for forwarding");
         let ci_opnd = asm.stack_opnd(-1);
         asm.ctx.dealloc_reg(ci_opnd.reg_opnd());
         asm.mov(ci_opnd, VALUE(ci as usize).into());
+
+        // Nil-initialize other locals which are above the CI
+        nil_fill("nil-initialize locals", 1..num_locals, asm);
     }
 
     // Points to the receiver operand on the stack unless a captured environment is used

```
(This is https://github.com/ruby/ruby/pull/12660 modulo tests so it applies onto 3.4.1 cleanly.)

----------------------------------------
Bug #21021: "try to mark T_NONE object" with 3.4.1
https://bugs.ruby-lang.org/issues/21021#change-111684

* Author: Benoit_Tigeot (Benoit Tigeot)
* Status: Open
* ruby -v: ruby 3.4.1 (2024-12-25 revision 48d4efcb85) +YJIT +PRISM [x86_64-linux]                                                                                                                                  ���
* Backport: 3.1: UNKNOWN, 3.2: UNKNOWN, 3.3: UNKNOWN, 3.4: UNKNOWN
----------------------------------------
Hello

We upgraded to 3.4.1 yesterday but we are seeing crash since then.

```
/bundle/ruby/3.4.0/gems/activejob-7.2.2.1/lib/active_job/enqueuing.rb:93: [BUG] try to mark T_NONE object
```

I saw the other issue related to ffi gem https://bugs.ruby-lang.org/issues/20694

But in our case the `C level backtrace information` looks different.

https://gist.github.com/benoittgt/13507c2000281aa7740bc782adab68c5


We migrated this part of the code to parallel->concurrent-ruby and we do not see the error yet again but I am a little bit worried we could see the issue again. 



-- 
https://bugs.ruby-lang.org/
 ______________________________________________
 ruby-core mailing list -- ruby-core@ml.ruby-lang.org
 To unsubscribe send an email to ruby-core-leave@ml.ruby-lang.org
 ruby-core info -- https://ml.ruby-lang.org/mailman3/lists/ruby-core.ml.ruby-lang.org/