Skip to content

bpo-46372: Try to mutate the LHS during fast float ops #30594

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

Closed
brandtbucher wants to merge 18 commits into from
Closed

bpo-46372: Try to mutate the LHS during fast float ops #30594

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
cf64e13
Perform specialized ops in-place when possible
brandtbucher Jan 11, 2022
5e49a13
Break out common logic
brandtbucher Jan 12, 2022
f7187d9
Clean things up a bit
brandtbucher Jan 12, 2022
dd19cec
Clean up macro
brandtbucher Jan 12, 2022
78cc5c4
Simplify the "small" int case
brandtbucher Jan 13, 2022
218b10b
blurb add
brandtbucher Jan 14, 2022
505fd8e
Catch up with main
brandtbucher Jan 18, 2022
6b4cc6d
Apply first round of code revew comments
brandtbucher Jan 18, 2022
f21a813
Fix 32-bit Windows errrors
brandtbucher Jan 19, 2022
d6a69d5
Perform the inplace checks when specializing
brandtbucher Jan 19, 2022
f11cf6d
Remove duplicate code
brandtbucher Jan 19, 2022
b6f041a
Simplify things further
brandtbucher Jan 19, 2022
eb586b1
Remove int and refcount == 2 optimizations
brandtbucher Jan 19, 2022
def9c64
Catch up with main
brandtbucher Jan 21, 2022
6e262ba
Revert int changes
brandtbucher Jan 24, 2022
37fab55
Catch up with main
brandtbucher Jan 28, 2022
04f186c
Add the "adaptive superinstructions" back
brandtbucher Feb 1, 2022
e0d0042
Catch up with main
brandtbucher Feb 1, 2022
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
Prev Previous commit
Next Next commit
Break out common logic
  • Loading branch information
@brandtbucher
brandtbucher committed Jan 12, 2022
commit 5e49a13266710b0a947286cd2ede26a33927ac39
201 changes: 63 additions & 138 deletions Python/ceval.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -863,6 +863,63 @@ static const binaryfunc binary_ops[] = {
[NB_INPLACE_XOR] = PyNumber_InPlaceXor,
};

#define BINARY_OP_INT_FAST(NAME, OP) \
do { \
PyObject *lhs = SECOND(); \
PyObject *rhs = TOP(); \
DEOPT_IF(!PyLong_CheckExact(lhs), BINARY_OP); \
DEOPT_IF(!PyLong_CheckExact(rhs), BINARY_OP); \
DEOPT_IF(1 < Py_ABS(Py_SIZE(lhs)), BINARY_OP); \
DEOPT_IF(1 < Py_ABS(Py_SIZE(rhs)), BINARY_OP); \
STAT_INC(BINARY_OP, hit); \
stwodigits l = Py_SIZE(lhs) * ((PyLongObject *)lhs)->ob_digit[0]; \
stwodigits r = Py_SIZE(rhs) * ((PyLongObject *)rhs)->ob_digit[0]; \
stwodigits i = l OP r; \
Py_DECREF(rhs); \
STACK_SHRINK(1); \
int op = GET_CACHE()->adaptive.original_oparg; \
if (Py_ABS(i) < PyLong_BASE && \
(_PY_NSMALLPOSINTS <= i || i < -_PY_NSMALLNEGINTS) && \
(_PY_NSMALLPOSINTS <= l || l < -_PY_NSMALLNEGINTS) && \
Py_REFCNT(lhs) == 1 + (op == NB_INPLACE_ ## NAME)) \
{ \
assert(i); \
((PyLongObject *)lhs)->ob_digit[0] = Py_ABS(i); \
Py_SET_SIZE(lhs, i < 0 ? -1 : 1); \
DISPATCH(); \
} \
Py_DECREF(lhs); \
PyObject *res = PyLong_FromLong(i); \
SET_TOP(res); \
if (res == NULL) { \
goto error; \
} \
DISPATCH(); \
} while (0)

#define BINARY_OP_FLOAT_FAST(NAME, OP) \
do { \
PyObject *lhs = SECOND(); \
PyObject *rhs = TOP(); \
DEOPT_IF(!PyFloat_CheckExact(lhs), BINARY_OP); \
DEOPT_IF(!PyFloat_CheckExact(rhs), BINARY_OP); \
STAT_INC(BINARY_OP, hit); \
double d = PyFloat_AS_DOUBLE(lhs) OP PyFloat_AS_DOUBLE(rhs); \
Py_DECREF(rhs); \
STACK_SHRINK(1); \
int op = GET_CACHE()->adaptive.original_oparg; \
if (Py_REFCNT(lhs) == 1 + (op == NB_INPLACE_ ## NAME)) { \
PyFloat_AS_DOUBLE(lhs) = d; \
DISPATCH(); \
} \
Py_DECREF(lhs); \
PyObject *res = PyFloat_FromDouble(d); \
SET_TOP(res); \
if (res == NULL) { \
goto error; \
} \
DISPATCH(); \
} while (0)

// PEP 634: Structural Pattern Matching

Expand Down Expand Up @@ -2054,107 +2111,19 @@ _PyEval_EvalFrameDefault(PyThreadState *tstate, InterpreterFrame *frame, int thr
}

TARGET(BINARY_OP_MULTIPLY_INT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyLong_CheckExact(left), BINARY_OP);
DEOPT_IF(!PyLong_CheckExact(right), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(left), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(right), BINARY_OP);
STAT_INC(BINARY_OP, hit);
stwodigits i = Py_SIZE(left) * ((PyLongObject *)left)->ob_digit[0];
i *= Py_SIZE(right) * ((PyLongObject *)right)->ob_digit[0];
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (_PY_NSMALLPOSINTS < i && i < PyLong_BASE &&
Py_REFCNT(left) == 1 + (op == NB_INPLACE_MULTIPLY))
{
((PyLongObject *)left)->ob_digit[0] = i;
Py_SET_SIZE(left, 1);
DISPATCH();
}
Py_DECREF(left);
PyObject *prod = PyLong_FromLong(i);
SET_TOP(prod);
if (prod == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_INT_FAST(MULTIPLY, *);
}

TARGET(BINARY_OP_MULTIPLY_FLOAT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyFloat_CheckExact(left), BINARY_OP);
DEOPT_IF(!PyFloat_CheckExact(right), BINARY_OP);
STAT_INC(BINARY_OP, hit);
double dprod = PyFloat_AS_DOUBLE(left) * PyFloat_AS_DOUBLE(right);
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (Py_REFCNT(left) == 1 + (op == NB_INPLACE_SUBTRACT)) {
PyFloat_AS_DOUBLE(left) = dprod;
DISPATCH();
}
Py_DECREF(left);
PyObject *sub = PyFloat_FromDouble(dprod);
SET_TOP(sub);
if (sub == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_FLOAT_FAST(MULTIPLY, *);
}

TARGET(BINARY_OP_SUBTRACT_INT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyLong_CheckExact(left), BINARY_OP);
DEOPT_IF(!PyLong_CheckExact(right), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(left), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(right), BINARY_OP);
STAT_INC(BINARY_OP, hit);
stwodigits i = Py_SIZE(left) * ((PyLongObject *)left)->ob_digit[0];
i -= Py_SIZE(right) * ((PyLongObject *)right)->ob_digit[0];
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (_PY_NSMALLPOSINTS < i && i < PyLong_BASE &&
Py_REFCNT(left) == 1 + (op == NB_INPLACE_SUBTRACT))
{
((PyLongObject *)left)->ob_digit[0] = i;
Py_SET_SIZE(left, 1);
DISPATCH();
}
Py_DECREF(left);
PyObject *sub = PyLong_FromLong(i);
SET_TOP(sub);
if (sub == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_INT_FAST(SUBTRACT, -);
}

TARGET(BINARY_OP_SUBTRACT_FLOAT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyFloat_CheckExact(left), BINARY_OP);
DEOPT_IF(!PyFloat_CheckExact(right), BINARY_OP);
STAT_INC(BINARY_OP, hit);
double dsub = PyFloat_AS_DOUBLE(left) - PyFloat_AS_DOUBLE(right);
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (Py_REFCNT(left) == 1 + (op == NB_INPLACE_SUBTRACT)) {
PyFloat_AS_DOUBLE(left) = dsub;
DISPATCH();
}
Py_DECREF(left);
PyObject *sub = PyFloat_FromDouble(dsub);
SET_TOP(sub);
if (sub == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_FLOAT_FAST(SUBTRACT, -);
}

TARGET(BINARY_OP_ADD_UNICODE) {
Expand Down Expand Up @@ -2203,55 +2172,11 @@ _PyEval_EvalFrameDefault(PyThreadState *tstate, InterpreterFrame *frame, int thr
}

TARGET(BINARY_OP_ADD_FLOAT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyFloat_CheckExact(left), BINARY_OP);
DEOPT_IF(Py_TYPE(right) != Py_TYPE(left), BINARY_OP);
STAT_INC(BINARY_OP, hit);
double dsum = PyFloat_AS_DOUBLE(left) + PyFloat_AS_DOUBLE(right);
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (Py_REFCNT(left) == 1 + (op == NB_INPLACE_SUBTRACT)) {
PyFloat_AS_DOUBLE(left) = dsum;
DISPATCH();
}
Py_DECREF(left);
PyObject *sub = PyFloat_FromDouble(dsum);
SET_TOP(sub);
if (sub == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_FLOAT_FAST(ADD, +);
}

TARGET(BINARY_OP_ADD_INT) {
PyObject *left = SECOND();
PyObject *right = TOP();
DEOPT_IF(!PyLong_CheckExact(left), BINARY_OP);
DEOPT_IF(!PyLong_CheckExact(right), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(left), BINARY_OP);
DEOPT_IF(1 < (size_t)Py_SIZE(right), BINARY_OP);
STAT_INC(BINARY_OP, hit);
stwodigits i = Py_SIZE(left) * ((PyLongObject *)left)->ob_digit[0];
i += Py_SIZE(right) * ((PyLongObject *)right)->ob_digit[0];
Py_DECREF(right);
STACK_SHRINK(1);
int op = GET_CACHE()->adaptive.original_oparg;
if (_PY_NSMALLPOSINTS < i && i < PyLong_BASE &&
Py_REFCNT(left) == 1 + (op == NB_INPLACE_ADD))
{
((PyLongObject *)left)->ob_digit[0] = i;
Py_SET_SIZE(left, 1);
DISPATCH();
}
Py_DECREF(left);
PyObject *sum = PyLong_FromLong(i);
SET_TOP(sum);
if (sum == NULL) {
goto error;
}
DISPATCH();
BINARY_OP_INT_FAST(ADD, +);
}

TARGET(BINARY_SUBSCR) {
Expand Down
12 changes: 6 additions & 6 deletions Python/specialize.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1621,8 +1621,8 @@ _Py_Specialize_BinaryOp(PyObject *lhs, PyObject *rhs, _Py_CODEUNIT *instr,
_Py_OPARG(*instr));
goto success;
}
if (PyLong_CheckExact(lhs) &&
(size_t)Py_SIZE(lhs) < 1 && (size_t)Py_SIZE(rhs) < 1)
if (PyLong_CheckExact(lhs) &&
Py_ABS(Py_SIZE(lhs)) < 2 && Py_ABS(Py_SIZE(rhs)) < 2)
{
*instr = _Py_MAKECODEUNIT(BINARY_OP_ADD_INT, _Py_OPARG(*instr));
goto success;
Expand All @@ -1639,8 +1639,8 @@ _Py_Specialize_BinaryOp(PyObject *lhs, PyObject *rhs, _Py_CODEUNIT *instr,
SPECIALIZATION_FAIL(BINARY_OP, SPEC_FAIL_DIFFERENT_TYPES);
goto failure;
}
if (PyLong_CheckExact(lhs) &&
(size_t)Py_SIZE(lhs) < 1 && (size_t)Py_SIZE(rhs) < 1)
if (PyLong_CheckExact(lhs) &&
Py_ABS(Py_SIZE(lhs)) < 2 && Py_ABS(Py_SIZE(rhs)) < 2)
{
*instr = _Py_MAKECODEUNIT(BINARY_OP_MULTIPLY_INT,
_Py_OPARG(*instr));
Expand All @@ -1658,8 +1658,8 @@ _Py_Specialize_BinaryOp(PyObject *lhs, PyObject *rhs, _Py_CODEUNIT *instr,
SPECIALIZATION_FAIL(BINARY_OP, SPEC_FAIL_DIFFERENT_TYPES);
goto failure;
}
if (PyLong_CheckExact(lhs) &&
(size_t)Py_SIZE(lhs) < 1 && (size_t)Py_SIZE(rhs) < 1)
if (PyLong_CheckExact(lhs) &&
Py_ABS(Py_SIZE(lhs)) < 2 && Py_ABS(Py_SIZE(rhs)) < 2)
{
*instr = _Py_MAKECODEUNIT(BINARY_OP_SUBTRACT_INT,
_Py_OPARG(*instr));
Expand Down