Handle error correctly in `PyFloat_GetInfo`

[sobolevn]

Bug report

floatobject contains this code:

cpython/Objects/floatobject.c

Lines 101 to 123 in ba82a24

	#define SetIntFlag(flag) \
	PyStructSequence_SET_ITEM(floatinfo, pos++, PyLong_FromLong(flag))
	#define SetDblFlag(flag) \
	PyStructSequence_SET_ITEM(floatinfo, pos++, PyFloat_FromDouble(flag))
	SetDblFlag(DBL_MAX);
	SetIntFlag(DBL_MAX_EXP);
	SetIntFlag(DBL_MAX_10_EXP);
	SetDblFlag(DBL_MIN);
	SetIntFlag(DBL_MIN_EXP);
	SetIntFlag(DBL_MIN_10_EXP);
	SetIntFlag(DBL_DIG);
	SetIntFlag(DBL_MANT_DIG);
	SetDblFlag(DBL_EPSILON);
	SetIntFlag(FLT_RADIX);
	SetIntFlag(FLT_ROUNDS);
	#undef SetIntFlag
	#undef SetDblFlag
	if (PyErr_Occurred()) {
	Py_CLEAR(floatinfo);
	return NULL;
	}

It only shows the last error and swallows any others.

I propose to change the error handling with the smallest possible diff.

Linked PRs

• gh-116714: Handle errors correctly in PyFloat_GetInfo #116715
• [3.12] gh-116714: Handle errors correctly in PyFloat_GetInfo (GH-116715) #116722
• [3.11] gh-116714: Handle errors correctly in PyFloat_GetInfo (GH-116715) #116723