| Age | Commit message (Collapse) | Author |
|---|
|
Notes:
Merged-By: k0kubun <[email protected]>
|
|
Co-authored-by: Alan Wu <[email protected]>
Co-authored-by: Alan Wu <[email protected]>
Notes:
Merged-By: k0kubun <[email protected]>
|
|
When RUBY_DEBUG is enabled, shape ids are 16 bits. I would like to do
16 bit comparisons, so I need to load halfwords sometimes. This commit
adds LDURH so that I can load halfwords.
https://developer.arm.com/documentation/ddi0596/2021-12/Base-Instructions/LDURH--Load-Register-Halfword--unscaled--?lang=en
I verified the bytes using clang:
```
$ cat asmthing.s
.global _start
.align 2
_start:
ldurh w10, [x1]
ldurh w10, [x1, #123]
$ as asmthing.s -o asmthing.o && objdump --disassemble asmthing.o
asmthing.o: file format mach-o arm64
Disassembly of section __TEXT,__text:
0000000000000000 <ltmp0>:
0: 2a 00 40 78 ldurh w10, [x1]
4: 2a b0 47 78 ldurh w10, [x1, #123]
```
Notes:
Merged: https://github.com/ruby/ruby/pull/6729
|
|
We don't need this constant to be exposed anymore, so remove it
Notes:
Merged: https://github.com/ruby/ruby/pull/6728
|
|
* YJIT: Instrument global allocations on stats build
* Just use GLOVAL_ALLOCATOR.stats()
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
Previously, there is no instruction boundary patch point after
the call to a non-leaf C function we generate for
OPTIMIZED_METHOD_TYPE_CALL. This meant that if code GC is triggered
while inside the C function, we would keep running invalidated code when
we return from the C function. This had the effect of running
stale branch stubs, jumping to bad code, etc.
Use jit_prepare_routine_call() to make sure we exit from the invalidated
region as soon as possible after the C call in case of invalidation.
Notes:
Merged: https://github.com/ruby/ruby/pull/6711
|
|
* Enable --yjit-stats for release builds
In order for people in the real world to report information about how their application runs with YJIT, we want to expose stats without requiring rebuilding ruby. We can do this without overhead, with the exception of count ratio in yjit, since this relies on the interpreter also counting instructions.
This change exposes those stats, while not showing ratio in yjit if we are not in a stats build.
* Update yjit.rb
Co-authored-by: Takashi Kokubun <[email protected]>
Co-authored-by: Maxime Chevalier-Boisvert <[email protected]>
Notes:
Merged-By: maximecb <[email protected]>
|
|
Since object shapes store the capacity of an object, we no longer
need the numiv field on RObjects. This gives us one extra slot which
we can use to give embedded objects one more instance variable (for a
total of 3 ivs). This commit removes the concept of numiv from RObject.
Notes:
Merged: https://github.com/ruby/ruby/pull/6699
|
|
This commit adds a `capacity` field to shapes, and adds shape
transitions whenever an object's capacity changes. Objects which are
allocated out of a bigger size pool will also make a transition from the
root shape to the shape with the correct capacity for their size pool
when they are allocated.
This commit will allow us to remove numiv from objects completely, and
will also mean we can guarantee that if two objects share shapes, their
IVs are in the same positions (an embedded and extended object cannot
share shapes). This will enable us to implement ivar sets in YJIT using
object shapes.
Co-Authored-By: Aaron Patterson <[email protected]>
Notes:
Merged: https://github.com/ruby/ruby/pull/6699
|
|
We switch to a new page when we detect dropped_bytes flipping from false
to true. Previously, when we patch code for invalidation during code gc,
we start with the flag being set to true, so we failed to apply patches
that straddle pages. We would write out jumps half way and then stop,
which left the code corrupted.
Reset the flag before patching so we patch across pages properly.
Notes:
Merged: https://github.com/ruby/ruby/pull/6686
|
|
This dispatches to a c func for doing the dynamic lookup. I experimented with chain on the proc but wasn't able to detect which call sites would be monomorphic vs polymorphic. There is definitely room for optimization here, but it does reduce exits.
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
* YJIT: Fix a wrong type reference
* YJIT: Just remove CapturedSelfOpnd for now
Notes:
Merged-By: maximecb <[email protected]>
|
|
Co-Authored-By: Alan Wu <[email protected]>
Co-authored-by: Alan Wu <[email protected]>
Notes:
Merged-By: maximecb <[email protected]>
|
|
* YJIT: Support invokeblock
* Update yjit/src/backend/arm64/mod.rs
* Update yjit/src/codegen.rs
Co-authored-by: Maxime Chevalier-Boisvert <[email protected]>
Notes:
Merged-By: maximecb <[email protected]>
|
|
Co-Authored-By: Alan Wu <[email protected]>
Co-Authored-By: Maxime Chevalier-Boisvert <[email protected]>
Co-authored-by: Alan Wu <[email protected]>
Co-authored-by: Maxime Chevalier-Boisvert <[email protected]>
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
While experimenting I found that it's easy to change Context and forget
to also change the copying operation in limit_block_versions(). Add an
assert to make sure we substitute a compatible generic context when
limiting the number of versions.
Notes:
Merged: https://github.com/ruby/ruby/pull/6656
|
|
Context::new() is the same as Context::default() and
Context::new_with_stack_size() was only used in tests.
Notes:
Merged: https://github.com/ruby/ruby/pull/6656
|
|
* YJIT: Add RubyVM::YJIT.code_gc
* Rename compiled_page_count to live_page_count
Notes:
Merged-By: maximecb <[email protected]>
|
|
Reduce default --yjit-exec-mem-size to 128MiB instead of 256
Notes:
Merged-By: maximecb <[email protected]>
|
|
This code used to roll its own heap object check before we made a better
version in guard_known_class(). The improved version uses one fewer
comparison, so let's use that.
Notes:
Merged-By: maximecb <[email protected]>
|
|
Co-authored-by: John Hawthorn <[email protected]>
Co-authored-by: John Hawthorn <[email protected]>
Notes:
Merged-By: maximecb <[email protected]>
|
|
* YJIT: Invalidate i-cache for the other cb on next_page
* YJIT: Invalidate only what's written by jmp_ptr
* YJIT: Move the code to the arm64 backend
Notes:
Merged-By: maximecb <[email protected]>
|
|
when it fails to allocate a new page.
Co-authored-by: Alan Wu <[email protected]>
Notes:
Merged-By: k0kubun <[email protected]>
|
|
`iv_count` is a misleading name because when IVs are unset, the new
shape doesn't decrement this value. `next_iv_count` is an accurate, and
more descriptive name.
Notes:
Merged: https://github.com/ruby/ruby/pull/6608
|
|
Previously, we found the current page by rounding the current pointer to
the closest smaller page size. This is incorrect because pages are
relative to the start of the address we reserve. For example, if the
starting address is 12KiB modulo the 16KiB page size, once we have more
than 4KiB of code, calculating with the address would incorrectly give
us page 1 when we're actually still on page 0.
Previously, I can reproduce crashes with:
make btest RUN_OPTS=--yjit-code-page-size=32
on ARM64 macOS, where system page sizes are 16KiB.
Notes:
Merged: https://github.com/ruby/ruby/pull/6607
Merged-By: XrXr
|
|
This patch makes sure that we're not accidentally reading rb_num_t
instruction arguments as VALUE and accidentally baking them into
code and marking them. Some of these are simply moving the cast earlier,
but some of these avoid potential problems for flag and ID arguments.
Follow-up for 39f7eddec4c55711d56f05b085992a83bf23159e.
Notes:
Merged: https://github.com/ruby/ruby/pull/6606
|
|
The expandarray instruction interpreters its arguments as rb_num_t.
YJIT was treating the num argument as a VALUE previously and when
it has a certain bit pattern, it can look like a GC pointer. The
argument is not a pointer, so YJIT crashed when trying to mark those
pointers.
This bug existed previously, but our test suite didn't expose it until
f55212bce939f736559709a8cd16c409772389c8. TestArgf#test_to_io has a
line like:
a1, a2, a3, a4, a5, a6, a7, a8 = array
Which maps to an expandarray with an argument of 8. Qnil happened to
be defined as 8, which masked the issue.
Fix it by not using the argument as a VALUE.
Notes:
Merged: https://github.com/ruby/ruby/pull/6603
|
|
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6600
|
|
Notes:
Merged: https://github.com/ruby/ruby/pull/6599
|
|
YJIT: Respect writable_addrs on --yjit-dump-iseq-disasm
as well
Notes:
Merged-By: maximecb <[email protected]>
|
|
YJIT: Skip dumping code for the other cb
on --yjit-dump-disasm
Notes:
Merged-By: maximecb <[email protected]>
|
|
|
|
Previously, enabling only "disasm" didn't actually build. Since these
two features are closely related and we don't really use one without the
other, let's simplify and merge the two features together.
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: k0kubun <[email protected]>
|
|
* YJIT: Allow --yjit-dump-disasm to dump into a file
* YJIT: Move IO implementation to disasm.rs
* YJIT: More consistent naming
Notes:
Merged-By: k0kubun <[email protected]>
|
|
Co-authored-by: Alan Wu <[email protected]>
Co-authored-by: Maxime Chevalier-Boisvert <[email protected]>
Notes:
Merged-By: k0kubun <[email protected]>
|
|
* Make mjit_cont sharable with YJIT
* Update dependencies
* Update YJIT binding
Notes:
Merged-By: k0kubun <[email protected]>
|
|
The commented out instance of free_block() is left over from the port.
The addition in gen_single_block() was a place we missed. The new block
is allocated in the same function and could have invariants associated
with it even though there is no space to hold all the code.
Notes:
Merged: https://github.com/ruby/ruby/pull/6551
|
|
We should make this function static and remove it from YJIT bindings.
Notes:
Merged: https://github.com/ruby/ruby/pull/6553
|
|
* YJIT: Count freed ISEQs
* YJIT: Avoid creating payloads for non-JITed ISEQs
Notes:
Merged-By: k0kubun <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
|
* fixes more clippy warnings
* Fix x86 c_callable to have doc_strings
Notes:
Merged-By: maximecb <[email protected]>
|
|
Notes:
Merged-By: maximecb <[email protected]>
|
