This commit changes how stack extents are calculated for both the main
thread and other threads. Ruby uses the address of a local variable as
part of the calculation for machine stack extents:
* pthreads uses it as a lower-bound on the start of the stack, because
glibc (and maybe other libcs) can store its own data on the stack
before calling into user code on thread creation.
* win32 uses it as an argument to VirtualQuery, which gets the extent of
the memory mapping which contains the variable
However, the local being used for this is actually too low (too close to
the leaf function call) in both the main thread case and the new thread
case.
In the main thread case, we have the `INIT_STACK` macro, which is used
for pthreads to set the `native_main_thread->stack_start` value. This
value is correctly captured at the very top level of the program (in
main.c). However, this is _not_ what's used to set the execution context
machine stack (`th->ec->machine_stack.stack_start`); that gets set as
part of a call to `ruby_thread_init_stack` in `Init_BareVM`, using the
address of a local variable allocated _inside_ `Init_BareVM`. This is
too low; we need to use a local allocated closer to the top of the
program.
In the new thread case, the lolcal is allocated inside
`native_thread_init_stack`, which is, again, too low.
In both cases, this means that we might have VALUEs lying outside the
bounds of `th->ec->machine.stack_{start,end}`, which won't be marked
correctly by the GC machinery.
To fix this,
* In the main thread case: We already have `INIT_STACK` at the right
level, so just pass that local var to `ruby_thread_init_stack`.
* In the new thread case: Allocate the local one level above the call to
`native_thread_init_stack` in `call_thread_start_func2`.
[Bug #20001]
fix
nil is treated similarly to the empty hash in this case, passing
no keywords and not calling any conversion methods.
Fixes [Bug #20064]
Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
The implementation of `native_thread_init_stack` for the various
threading models can use the address of a local variable as part of the
calculation of the machine stack extents:
* pthreads uses it as a lower-bound on the start of the stack, because
glibc (and maybe other libcs) can store its own data on the stack
before calling into user code on thread creation.
* win32 uses it as an argument to VirtualQuery, which gets the extent of
the memory mapping which contains the variable
However, the local being used for this is actually allocated _inside_
the `native_thread_init_stack` frame; that means the caller might
allocate a VALUE on the stack that actually lies outside the bounds
stored in machine.stack_{start,end}.
A local variable from one level above the topmost frame that stores
VALUEs on the stack must be drilled down into the call to
`native_thread_init_stack` to be used in the calculation. This probably
doesn't _really_ matter for the win32 case (they'll be in the same
memory mapping so VirtualQuery should return the same thing), but
definitely could matter for the pthreads case.
[Bug #20001]
The environ is malloc'd, so it gets reported as a memory leak. This
commit adds ruby_free_proctitle which frees it during shutdown when
RUBY_FREE_AT_EXIT is set.
STACK OF 1 INSTANCE OF 'ROOT LEAK: <calloc in ruby_init_setproctitle>':
5 dyld 0x18b7090e0 start + 2360
4 ruby 0x10000e3a8 main + 100 main.c:58
3 ruby 0x1000b4dfc ruby_options + 180 eval.c:121
2 ruby 0x1001c5f70 ruby_process_options + 200 ruby.c:3014
1 ruby 0x10035c9fc ruby_init_setproctitle + 76 setproctitle.c:105
0 libsystem_malloc.dylib 0x18b8c7b78 _malloc_zone_calloc_instrumented_or_legacy + 100
When a forked process was started in a thread, this would result in a
double-free during the child process exit.
RUBY_FREE_AT_EXIT=1 ./miniruby -e 'Thread.new { fork { } }.join; Process.waitpid'
This is because the main thread in the forked process was not the
initial VM thread, and the new thread's stack was freed as part of
objectspace iteration.
This change also allows rb_threadptr_root_fiber_release to run without
EC being available.
Ractor's free iterates through its TLS keys so we need to keep this
memory available until after Ractors are freed.
Minimal reproduction:
RUBY_FREE_AT_EXIT=1 ./miniruby -e rand
Previously with RUBY_FREE_ON_EXIT, ractors where being xfree-ed which is incorrect since they are not xmalloced.
Instead we can free ractors with ractor free during shutdown. This change only effects main ractor freeing when RUBY_FREE_ON_EXIT is set.
Co-authored-by: John Hawthorn <john@hawthorn.email>
Our current implementation of rb_postponed_job_register suffers from
some safety issues that can lead to interpreter crashes (see bug #1991).
Essentially, the issue is that jobs can be called with the wrong
arguments.
We made two attempts to fix this whilst keeping the promised semantics,
but:
* The first one involved masking/unmasking when flushing jobs, which
was believed to be too expensive
* The second one involved a lock-free, multi-producer, single-consumer
ringbuffer, which was too complex
The critical insight behind this third solution is that essentially the
only user of these APIs are a) internal, or b) profiling gems.
For a), none of the usages actually require variable data; they will
work just fine with the preregistration interface.
For b), generally profiling gems only call a single callback with a
single piece of data (which is actually usually just zero) for the life
of the program. The ringbuffer is complex because it needs to support
multi-word inserts of job & data (which can't be atomic); but nobody
actually even needs that functionality, really.
So, this comit:
* Introduces a pre-registration API for jobs, with a GVL-requiring
rb_postponed_job_prereigster, which returns a handle which can be
used with an async-signal-safe rb_postponed_job_trigger.
* Deprecates rb_postponed_job_register (and re-implements it on top of
the preregister function for compatability)
* Moves all the internal usages of postponed job register
pre-registration
This patch introduces thread specific storage APIs
for tools which use `rb_internal_thread_event_hook` APIs.
* `rb_internal_thread_specific_key_create()` to create a tool specific
thread local storage key and allocate the storage if not available.
* `rb_internal_thread_specific_set()` sets a data to thread and tool
specific storage.
* `rb_internal_thread_specific_get()` gets a data in thread and tool
specific storage.
Note that `rb_internal_thread_specific_get|set(thread_val, key)`
can be called without GVL and safe for async signal and safe for
multi-threading (native threads). So you can call it in any internal
thread event hooks. Further more you can call it from other native
threads. Of course `thread_val` should be living while accessing the
data from this function.
Note that you should not forget to clean up the set data.
when the RUBY_FREE_ON_SHUTDOWN environment variable is set, manually free memory at shutdown.
Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
Co-authored-by: Peter Zhu <peter@peterzhu.ca>
`src_ep[VM_ENV_DATA_INDEX_ME_CREF]` was read out and held without
marking across the allocation in vm_env_new(). In case vm_env_new() ran
compaction, an invalid reference could have been written into
`copied_env`.
It might've been hard to actually produce a crash with this issue due to
the pinning marking of the field in rb_execution_context_mark().
Previously, the following crashed with
`vm_assert_env:imemo_type_p(obj, imemo_env)` due to missing a missing
WB:
o = Object.new
def o.foo(n)
freeze
GC.stress = 1
# inflate block nesting get an imemo_env for each level
n.tap do |i|
i.tap do |local|
return Ractor.make_shareable(-> do
local + i + n
end)
end
end
ensure
GC.stress = false
GC.verify_internal_consistency
end
p o.foo(1)[]
By the time the recursive env_copy() call returns, `copied_env` could
have aged or have turned greyed, so we need a WB for the
`ep[VM_ENV_DATA_INDEX_SPECVAL]` assignment which adds an edge.
Fix: 674eb7df7f
The original order of events is:
1. Allocate env_body.
2. Fill env_body using elements in src_env, and it performs operations
that can trigger a GC.
3. Create the copied_env using vm_env_new.
However, if GC compaction runs during step 2, then copied_env would not
have yet been created and objects on env_body could move but it would
not be reference updated.
This commit changes the the order to be (1), (3), (2).
Previously, vm_make_env_each() did:
1. ALLOC env_body
2. Copy locals into env_body
3. Allocate imemo_env
4. Set up imemo_env with env_body
If compaction runs during (3), locals copied to env_body could be
moved and the imemo_env could end up with invalid references.
Move (2) down so it reads references after potential movement.
`rb_jmpbuf_t` type is considerably large due to inline-allocated
Asyncify buffer, and it leads to stack overflow even with small number
of C-method call frames. This commit allocates the Asyncify buffer used
by `rb_wasm_setjmp` in heap to mitigate the issue.
This patch introduces a new type `rb_vm_tag_jmpbuf_t` to abstract the
representation of a jump buffer, and init/deinit hook points to manage
lifetime of the buffer. These changes are effectively NFC for non-wasm
platforms.
This is an experimental commit that uses a functional red-black tree to
create an index of the ancestor shapes. It uses an Okasaki style
functional red black tree:
https://www.cs.tufts.edu/comp/150FP/archive/chris-okasaki/redblack99.pdf
This tree is advantageous because:
* It offers O(n log n) insertions and O(n log n) lookups.
* It shares memory with previous "versions" of the tree
When we insert a node in the tree, only the parts of the tree that need
to be rebalanced are newly allocated. Parts of the tree that don't need
to be rebalanced are not reallocated, so "new trees" are able to share
memory with old trees. This is in contrast to a sorted set where we
would have to duplicate the set, and also resort the set on each
insertion.
I've added a new stat to RubyVM.stat so we can understand how the red
black tree increases.
* Port call threshold logic from Rust to C for performance
* Prefix global/field names with yjit_
* Fix linker error
* Fix preprocessor condition for rb_yjit_threshold_hit
* Fix third linker issue
* Exclude yjit_calls_at_interv from RJIT bindgen
---------
Co-authored-by: Takashi Kokubun <takashikkbn@gmail.com>
This patch introduce M:N thread scheduler for Ractor system.
In general, M:N thread scheduler employs N native threads (OS threads)
to manage M user-level threads (Ruby threads in this case).
On the Ruby interpreter, 1 native thread is provided for 1 Ractor
and all Ruby threads are managed by the native thread.
From Ruby 1.9, the interpreter uses 1:1 thread scheduler which means
1 Ruby thread has 1 native thread. M:N scheduler change this strategy.
Because of compatibility issue (and stableness issue of the implementation)
main Ractor doesn't use M:N scheduler on default. On the other words,
threads on the main Ractor will be managed with 1:1 thread scheduler.
There are additional settings by environment variables:
`RUBY_MN_THREADS=1` enables M:N thread scheduler on the main ractor.
Note that non-main ractors use the M:N scheduler without this
configuration. With this configuration, single ractor applications
run threads on M:1 thread scheduler (green threads, user-level threads).
`RUBY_MAX_CPU=n` specifies maximum number of native threads for
M:N scheduler (default: 8).
This patch will be reverted soon if non-easy issues are found.
[Bug #19842]
This commit moves IO#readline to Ruby. In order to call C functions,
keyword arguments must be converted to hashes. Prior to this commit,
code like `io.readline(chomp: true)` would allocate a hash. This
commits moves the keyword "denaturing" to Ruby, allowing us to send
positional arguments to the C API and avoiding the hash allocation.
Here is an allocation benchmark for the method:
```
x = GC.stat(:total_allocated_objects)
File.open("/usr/share/dict/words") do |f|
f.readline(chomp: true) until f.eof?
end
p ALLOCATIONS: GC.stat(:total_allocated_objects) - x
```
Before this commit, the output was this:
```
$ make run
./miniruby -I./lib -I. -I.ext/common -r./arm64-darwin22-fake ./test.rb
{:ALLOCATIONS=>707939}
```
Now it is this:
```
$ make run
./miniruby -I./lib -I. -I.ext/common -r./arm64-darwin22-fake ./test.rb
{:ALLOCATIONS=>471962}
```
[Bug #19890] [ruby-core:114803]
All kind of AST nodes use same struct RNode, which has u1, u2, u3 union members
for holding different kind of data.
This has two problems.
1. Low flexibility of data structure
Some nodes, for example NODE_TRUE, don’t use u1, u2, u3. On the other hand,
NODE_OP_ASGN2 needs more than three union members. However they use same
structure definition, need to allocate three union members for NODE_TRUE and
need to separate NODE_OP_ASGN2 into another node.
This change removes the restriction so make it possible to
change data structure by each node type.
2. No compile time check for union member access
It’s developer’s responsibility for using correct member for each node type when it’s union.
This change clarifies which node has which type of fields and enables compile time check.
This commit also changes node_buffer_elem_struct buf management to handle
different size data with alignment.
This is an internal only function not exposed to the C extension API.
It's only use so far is from rb_vm_mark, where it's used to mark the
values in the vm->trap_list.cmd array.
There shouldn't be any reason why these cannot move.
This commit allows them to move by updating their references during the
reference updating step of compaction.
To do this we've introduced another internal function
rb_gc_update_values as a partner to rb_gc_mark_values.
This allows us to refactor rb_gc_mark_values to not pin
Revert "Revert "Skip calling jit_exec on Wasm""
This reverts commit 2e94610f70.
It's not about whether it's optimized away or not. I just don't want to
leave and maintain the callsite (e.g. signature) in the path where
YJIT is never built.
This reverts commit e80752f9bb.
RJIT and YJIT are never enabled on Wasm. When both are disabled,
`jit_exec` is defined to return `Qundef` constantly, and is optimized
away.
We often break Wasm build when we modify how jit_exec works. I'm
planning to modify it again soon.
We actually don't support running Ruby JIT on Wasm, so it doesn't seem
worth the maintenance effort.
Nobuyoshi Nakada
Peter Zhu
KJ Tsanaktsidis
Jeremy Evans
John Hawthorn
HParker
Koichi Sasada
Adam Hess
Alan Wu
Yuta Saito
Aaron Patterson
Takashi Kokubun
Maxime Chevalier-Boisvert
yui-knk
Matt Valentine-House
Jean Boussier