Skip to content

gh-149238: avoid bool objects when branching on int/str in JIT #149418

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
NekoAsakura wants to merge 3 commits into
base: main
Choose a base branch
from
Open

gh-149238: avoid bool objects when branching on int/str in JIT #149418

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f5a3a71
gh-149238: avoid bool objects when branching on int/str in JIT
NekoAsakura May 5, 2026
362f6cf
Merge branch 'main' into gh-149238/bit-to-bool
NekoAsakura May 5, 2026
810d782
address feedback
NekoAsakura May 6, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 4 additions & 4 deletions Include/internal/pycore_opcode_metadata.h
View file
Open in desktop

Some generated files are not rendered by default. Learn more about how customized files appear on GitHub.

33 changes: 32 additions & 1 deletion Include/internal/pycore_stackref.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -73,10 +73,28 @@ static const _PyStackRef PyStackRef_ERROR = { .index = (1 << Py_TAGGED_SHIFT) };
#define PyStackRef_None ((_PyStackRef){ .index = (2 << Py_TAGGED_SHIFT) } )
#define _Py_STACKREF_FALSE_INDEX (3 << Py_TAGGED_SHIFT)
#define _Py_STACKREF_TRUE_INDEX (4 << Py_TAGGED_SHIFT)
/* Tier-2 transient bit values (gh-149238). */
#define _Py_STACKREF_BIT_0_INDEX (5 << Py_TAGGED_SHIFT)
Copy link
Copy Markdown
Member

@markshannon markshannon May 6, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Rather than expose the internal representation, can you add new PyStackRef_WrapBit PyStackRef_UnwrapBit functions to present a nicer API.

How there are stored doesn't need to leak into the rest of the code.

#define _Py_STACKREF_BIT_1_INDEX (6 << Py_TAGGED_SHIFT)
#define PyStackRef_False ((_PyStackRef){ .index = _Py_STACKREF_FALSE_INDEX })
#define PyStackRef_True ((_PyStackRef){ .index = _Py_STACKREF_TRUE_INDEX })

#define INITIAL_STACKREF_INDEX (5 << Py_TAGGED_SHIFT)
static inline _PyStackRef
PyStackRef_WrapBit(int truthy)
{
return (_PyStackRef){ .index = truthy ? _Py_STACKREF_BIT_1_INDEX
: _Py_STACKREF_BIT_0_INDEX };
}

static inline int
PyStackRef_UnwrapBit(_PyStackRef bit)
{
assert(bit.index == _Py_STACKREF_BIT_0_INDEX ||
bit.index == _Py_STACKREF_BIT_1_INDEX);
return (bit.index == _Py_STACKREF_BIT_1_INDEX) ? 1 : 0;
}

#define INITIAL_STACKREF_INDEX (7 << Py_TAGGED_SHIFT)

#define PyStackRef_ZERO_BITS PyStackRef_NULL

Expand Down Expand Up @@ -472,6 +490,19 @@ static const _PyStackRef PyStackRef_NULL = { .bits = PyStackRef_NULL_BITS };
#define PyStackRef_False ((_PyStackRef){.bits = ((uintptr_t)&_Py_FalseStruct) | Py_TAG_REFCNT })
#define PyStackRef_None ((_PyStackRef){.bits = ((uintptr_t)&_Py_NoneStruct) | Py_TAG_REFCNT })

static inline _PyStackRef
PyStackRef_WrapBit(int truthy)
{
return (_PyStackRef){ .bits = (uintptr_t)(truthy ? 1 : 0) };
}

static inline int
PyStackRef_UnwrapBit(_PyStackRef bit)
{
assert(bit.bits == 0 || bit.bits == 1);
return (int)bit.bits;
}

#ifdef Py_GIL_DISABLED
// Checks that mask out the deferred bit in the free threading build.
#define PyStackRef_IsNone(REF) (((REF).bits & ~Py_TAG_REFCNT) == (uintptr_t)&_Py_NoneStruct)
Expand Down
Loading
Loading