ruby.git/vm_insnhelper.c, branch v4.0.3 The Ruby Programming Language Store ractor_id directly on EC 2025-12-18T21:43:45+00:00 John Hawthorn john@hawthorn.email 2025-12-17T20:12:17+00:00 63b082cf0e87942dcea28cbdeb1c8a9e616e903a This is easier to access as ec->ractor_id instead of pointer-chasing through ec->thread->ractor->ractor_id Co-authored-by: Luke Gruber <luke.gru@gmail.com> 
This is easier to access as ec->ractor_id instead of pointer-chasing through
ec->thread->ractor->ractor_id

Co-authored-by: Luke Gruber <luke.gru@gmail.com>
Respect encoding of ID in exception messages 2025-12-17T01:15:59+00:00 Daisuke Aritomo osyoyu@osyoyu.com 2025-12-16T06:15:28+00:00 0fe111caa6885407960541bdb6de7ba3cd6aad73 






Make tracepoints with set_trace_func or TracePoint.new ractor local (#15468)
2025-12-16T19:06:55+00:00

Luke Gruber
luke.gruber@shopify.com

2025-12-16T19:06:55+00:00

4fb537b1ee28bb37dbe551ac65c279d436c756bc

Before this change, GC'ing any Ractor object caused you to lose all
enabled tracepoints across all ractors (even main). Now tracepoints are
ractor-local and this doesn't happen. Internal events are still global.

Fixes [Bug #19112]




Before this change, GC'ing any Ractor object caused you to lose all
enabled tracepoints across all ractors (even main). Now tracepoints are
ractor-local and this doesn't happen. Internal events are still global.

Fixes [Bug #19112]






Change bmethod defined_ractor to use id instead
2025-12-04T21:25:45+00:00

John Hawthorn
john@hawthorn.email

2025-10-30T19:05:25+00:00

1d3fe2c382fd543bf12b2e12c2633162ea4b3a0a

When defining a bmethod, we recorded the current Ractor's object in the
method. However that was never marked and so could be GC'd and reused by
a future Ractor. Instead we can use the Ractor's id, which we expect to
be unique forever.

Co-authored-by: Luke Gruber <luke.gru@gmail.com>





When defining a bmethod, we recorded the current Ractor's object in the
method. However that was never marked and so could be GC'd and reused by
a future Ractor. Instead we can use the Ractor's id, which we expect to
be unique forever.

Co-authored-by: Luke Gruber <luke.gru@gmail.com>







Use ALWAYS_INLINE for vm_getinstancevariable
2025-11-28T17:50:38+00:00

John Hawthorn
john@hawthorn.email

2025-11-28T06:57:48+00:00

8c70def42c4596a02d45464d0d694d0c234fb1d7

Recently rb_vm_getinstancevariable was introduced exposing this method
to ZJIT. On clang specifically this ended up causing the compiler not to
inline into vm_exec_core and cause a significant performance regression
in optcarrot for the interpreter.

Co-authored-by: Luke Gruber <luke.gru@gmail.com>





Recently rb_vm_getinstancevariable was introduced exposing this method
to ZJIT. On clang specifically this ended up causing the compiler not to
inline into vm_exec_core and cause a significant performance regression
in optcarrot for the interpreter.

Co-authored-by: Luke Gruber <luke.gru@gmail.com>







vm_cc_new: don't assume `cme` is present.
2025-11-25T15:40:31+00:00

Jean Boussier
jean.boussier@gmail.com

2025-11-17T19:01:12+00:00

a36ebb18a6d4c4726915b6d7c16cfdbf4e5d417b

[Bug #21694]

`vm_search_super_method` explictly calls `vm_cc_new` with `cme=NULL`
when there is no super class.





[Bug #21694]

`vm_search_super_method` explictly calls `vm_cc_new` with `cme=NULL`
when there is no super class.







ZJIT: Rename array length reference to make the code easier to follow
2025-11-20T22:35:09+00:00

Kevin Menard
kevin@nirvdrum.com

2025-11-13T16:56:23+00:00

604fc059618b8f1f94b19efa51d468d827a766d1












ZJIT: Put optional interpreter cache on both GetIvar and SetIvar
2025-11-20T16:32:17+00:00

Max Bernstein
ruby@bernsteinbear.com

2025-11-20T15:28:23+00:00

f8cb9f320ceddbfe6feffb6410bb6212ed27673c












ZJIT: Fix assertion failure when profiling VM_BLOCK_HANDLER_NONE
2025-11-19T23:14:01+00:00

Alan Wu
XrXr@users.noreply.github.com

2025-11-18T22:46:52+00:00

2cd792a1cfefeaee948b321bbc14cb86acc2d456

As can be seen in vm_block_handler_verify(), VM_BLOCK_HANDLER_NONE is
not a valid argument for vm_block_handler(). Store nil in the profiler
when seen instead of crashing.





As can be seen in vm_block_handler_verify(), VM_BLOCK_HANDLER_NONE is
not a valid argument for vm_block_handler(). Store nil in the profiler
when seen instead of crashing.







Win32: Drop support for older than MSVC 8.0/_MSC_VER 1400
2025-11-19T02:03:42+00:00

Nobuyoshi Nakada
nobu@ruby-lang.org

2024-12-09T05:56:58+00:00

cdb9893c552f67a6065dcb165b2040d35c57aee3

Visual C++ 2005 (8.0):
- _MSC_VER: 1400
- MSVCRT_VERSION: 80





Visual C++ 2005 (8.0):
- _MSC_VER: 1400
- MSVCRT_VERSION: 80