Merge branch 'pm-cpuidle'

 - Avoid using deep idle states during initialization in the intel_idle
   driver to work around device handling issues (Rafael Wysocki)

 - Fix and refactor the ACPI processor driver code related to ACPI _LPI
   support and add ACPI _LPI support to intel_idle based on that ACPI
   processor driver update (Rafael Wysocki)

* pm-cpuidle:
  intel_idle: Avoid using deep idle states during initialization
  intel_idle: Update documentation after adding ACPI _LPI support
  intel_idle: Add ACPI _LPI support
  intel_idle: Prepare for adding ACPI _LPI support
  ACPI: processor: idle: Add switch for strict _LPI processing
  ACPI: processor: idle: Relocate acpi_processor_extract_lpi_info()
  ACPI: processor: idle: Introduce acpi_processor_extract_lpi_info()
  ACPI: processor: idle: Introduce too_many_states() for _LPI
  ACPI: processor: idle: Rework flatten_lpi_states()
  ACPI: processor: idle: Rearrange loop in acpi_processor_get_lpi_info()
  ACPI: processor: idle: Drop redundant _LPI presence checks
  ACPI: processor: idle: Rework first-level _LPI states processing
  ACPI: processor: idle: Rearrange acpi_processor_get_lpi_info()
  ACPI: processor: idle: Introduce lpi_state_debug()
  ACPI: processor: idle: Split acpi_processor_evaluate_lpi()
  ACPI: processor: idle: Rearrange acpi_processor_evaluate_lpi()
  ACPI: processor: idle: Unify debug in acpi_processor_evaluate_lpi()
  ACPI: processor: idle: Ignore _LPI states with SYSTEMIO entry method
  ACPI: processor: idle: Expand _LPI package sanity checks

5 files changed, 583 insertions, 325 deletions

diff --git a/Documentation/admin-guide/pm/intel_idle.rst b/Documentation/admin-guide/pm/intel_idle.rst
index 188d52cd26e8..996e5bd83c42 100644
--- a/Documentation/admin-guide/pm/intel_idle.rst
+++ b/Documentation/admin-guide/pm/intel_idle.rst
@@ -87,17 +87,22 @@ tables with any processor model recognized by it; see
 `below <intel-idle-parameters_>`_.]
 
 If the ACPI tables are going to be used for building the list of available idle
-states, ``intel_idle`` first looks for a ``_CST`` object under one of the ACPI
-objects corresponding to the CPUs in the system (refer to the ACPI specification
-[2]_ for the description of ``_CST`` and its output package).  Because the
-``CPUIdle`` subsystem expects that the list of idle states supplied by the
-driver will be suitable for all of the CPUs handled by it and ``intel_idle`` is
-registered as the ``CPUIdle`` driver for all of the CPUs in the system, the
-driver looks for the first ``_CST`` object returning at least one valid idle
-state description and such that all of the idle states included in its return
-package are of the FFH (Functional Fixed Hardware) type, which means that the
-``MWAIT`` instruction is expected to be used to tell the processor that it can
-enter one of them.  The return package of that ``_CST`` is then assumed to be
+states, ``intel_idle`` will be looking for ``_LPI`` or ``_CST`` objects in them
+(refer to the ACPI specification [2]_ for the definitions of the ``_LPI`` and
+``_CST`` objects).  If ``_LPI`` is present under at least one of the ACPI
+objects representing the CPUs in the system and ``_LPI`` processing produces a
+non-empty list of valid idle states, it will be used.  Otherwise, ``_CST`` will
+be used so long as it is present under at least one of the ACPI objects
+representing the CPUs in the system and it returns a non-empty list of valid
+idle states.  In either case, since the ``CPUIdle`` subsystem expects that the
+list of idle states supplied by the driver will be suitable for all of the CPUs
+handled by it and ``intel_idle`` is registered as the ``CPUIdle`` driver for all
+of the CPUs in the system, ``intel_idle`` looks for the first CPU where the
+ACPI-supplied list of idle states (coming from either ``_LPI`` or ``_CST``)
+is not empty.  Moreover, all of the states in that list need to be of the FFH
+(Functional Fixed Hardware) type, which means that the ``MWAIT`` instruction is
+expected to be used to tell the processor that the given idle state may be
+entered.  If that expectation is met, the list of idle states is assumed to be
 applicable to all of the other CPUs in the system and the idle state
 descriptions extracted from it are stored in a preliminary list of idle states
 coming from the ACPI tables.  [This step is skipped if ``intel_idle`` is
@@ -129,18 +134,21 @@ If the given processor model is not recognized by ``intel_idle``, but it
 supports ``MWAIT``, the preliminary list of idle states coming from the ACPI
 tables is used for building the final list that will be supplied to the
 ``CPUIdle`` core during driver registration.  For each idle state in that list,
-the description, ``MWAIT`` hint and exit latency are copied to the corresponding
-entry in the final list of idle states.  The name of the idle state represented
-by it (to be returned by the ``name`` idle state attribute in ``sysfs``) is
-"CX_ACPI", where X is the index of that idle state in the final list (note that
-the minimum value of X is 1, because 0 is reserved for the "polling" state), and
-its target residency is based on the exit latency value.  Specifically, for
-C1-type idle states the exit latency value is also used as the target residency
-(for compatibility with the majority of the "internal" tables of idle states for
-various processor models recognized by ``intel_idle``) and for the other idle
-state types (C2 and C3) the target residency value is 3 times the exit latency
-(again, that is because it reflects the target residency to exit latency ratio
-in the majority of cases for the processor models recognized by ``intel_idle``).
+the description, ``MWAIT`` hint and exit (wake) latency are copied to the
+corresponding entry in the final list of idle states.  If the preliminary list
+of idle states has been obtained through ``_LPI`` processing, the minimum
+residency parameter of the given idle state is taken as its target residency.
+Otherwise, for C1-type idle states, the exit latency value is also used as the
+target residency (for compatibility with the majority of the "internal" tables
+of idle states for various processor models recognized by ``intel_idle``), and
+for the other idle state types (C2 and C3) the target residency value is 3 times
+the exit latency (again, that is because it reflects the target residency to
+exit latency ratio in the majority of cases for the processor models recognized
+by ``intel_idle``).  The name of the idle state (to be returned by the ``name``
+idle state attribute in ``sysfs``) is either "Cx_LPI" (if it comes from ``_LPI``
+processing) or "Cx_ACPI", where x is the index of that idle state in the final
+list (note that the minimum value of x is 1, because 0 is reserved for the
+"polling" state), and its target residency is based on the exit latency value.
 All of the idle states in the final list are enabled by default in this case.
 
 
diff --git a/drivers/acpi/acpi_processor.c b/drivers/acpi/acpi_processor.c
index 00775b91bd41..5fbab54171b7 100644
--- a/drivers/acpi/acpi_processor.c
+++ b/drivers/acpi/acpi_processor.c
@@ -994,3 +994,357 @@ end:
 }
 EXPORT_SYMBOL_NS_GPL(acpi_processor_evaluate_cst, "ACPI_PROCESSOR_IDLE");
 #endif /* CONFIG_ACPI_PROCESSOR_CSTATE */
+
+#ifdef CONFIG_ACPI_PROCESSOR_IDLE
+struct acpi_lpi_states_array {
+	unsigned int size;
+	unsigned int composite_states_size;
+	struct acpi_lpi_state *entries;
+	struct acpi_lpi_state *composite_states[ACPI_PROCESSOR_MAX_POWER];
+};
+
+static int obj_get_integer(union acpi_object *obj, u32 *value)
+{
+	if (obj->type != ACPI_TYPE_INTEGER)
+		return -EINVAL;
+
+	*value = obj->integer.value;
+	return 0;
+}
+
+#define lpi_state_debug(handle, message, state_idx)	\
+	acpi_handle_debug(handle, message " for _LPI state %u\n", state_idx)
+
+static void process_lpi_state_package(union acpi_object *lpi_pkg,
+				      struct acpi_lpi_state *lpi_state,
+				      acpi_handle handle,
+				      unsigned int state_idx, bool strict)
+{
+	union acpi_object *lpi_pkg_elem, *obj;
+
+	if (lpi_pkg->type != ACPI_TYPE_PACKAGE || lpi_pkg->package.count < 7)
+		return;
+
+	lpi_pkg_elem = lpi_pkg->package.elements;
+
+	/* Get the entry method first and skip the state if that fails. */
+	obj = &lpi_pkg_elem[6];
+	if (obj->type == ACPI_TYPE_BUFFER) {
+		struct acpi_power_register *reg;
+
+		if (obj->buffer.length < sizeof(*reg)) {
+			lpi_state_debug(handle, "Invalid register data", state_idx);
+			return;
+		}
+
+		reg = (struct acpi_power_register *)obj->buffer.pointer;
+		if (reg->space_id != ACPI_ADR_SPACE_FIXED_HARDWARE) {
+			lpi_state_debug(handle, "Unsupported entry method", state_idx);
+			return;
+		}
+
+		lpi_state->entry_method = ACPI_CSTATE_FFH;
+		lpi_state->address = reg->address;
+	} else if (obj->type == ACPI_TYPE_INTEGER) {
+		lpi_state->entry_method = ACPI_CSTATE_INTEGER;
+		lpi_state->address = obj->integer.value;
+	} else {
+		lpi_state_debug(handle, "Invalid entry method", state_idx);
+		return;
+	}
+
+	if (obj_get_integer(&lpi_pkg_elem[0], &lpi_state->min_residency)) {
+		if (strict) {
+			lpi_state_debug(handle, "No min. residency", state_idx);
+			return;
+		}
+
+		lpi_state_debug(handle, "Assuming 10 us min. residency", state_idx);
+		lpi_state->min_residency = 10;
+	}
+
+	if (obj_get_integer(&lpi_pkg_elem[1], &lpi_state->wake_latency)) {
+		if (strict) {
+			lpi_state_debug(handle, "No wake latency", state_idx);
+			return;
+		}
+
+		lpi_state_debug(handle, "Assuming 10 us wake latency", state_idx);
+		lpi_state->wake_latency = 10;
+	}
+
+	if (obj_get_integer(&lpi_pkg_elem[2], &lpi_state->flags))
+		lpi_state->flags = 0;
+
+	if (obj_get_integer(&lpi_pkg_elem[3], &lpi_state->arch_flags))
+		lpi_state->arch_flags = 0;
+
+	if (obj_get_integer(&lpi_pkg_elem[4], &lpi_state->res_cnt_freq))
+		lpi_state->res_cnt_freq = 1;
+
+	if (obj_get_integer(&lpi_pkg_elem[5], &lpi_state->enable_parent_state))
+		lpi_state->enable_parent_state = 0;
+
+	/* Skip elements [7-8] i.e. Residency/Usage counters. */
+
+	/*
+	 * Avoid out-of-bounds access if the size of the package is less than
+	 * expected.
+	 */
+	if (lpi_pkg->package.count < 10)
+		return;
+
+	obj = &lpi_pkg_elem[9];
+	if (obj->type == ACPI_TYPE_STRING)
+		strscpy(lpi_state->desc, obj->string.pointer, ACPI_CX_DESC_LEN);
+}
+
+static int acpi_processor_evaluate_lpi(acpi_handle handle,
+				       struct acpi_lpi_states_array *info,
+				       bool strict)
+{
+	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
+	union acpi_object *lpi_data, *lpi_pkg;
+	unsigned int lpi_pkg_count, state_idx;
+	struct acpi_lpi_state *lpi_state;
+	acpi_status status;
+	int ret = 0;
+
+	status = acpi_evaluate_object(handle, "_LPI", NULL, &buffer);
+	if (ACPI_FAILURE(status)) {
+		acpi_handle_debug(handle, "No _LPI, giving up\n");
+		return -ENODEV;
+	}
+
+	lpi_data = buffer.pointer;
+
+	/* There must be at least 4 elements = 3 elements + 1 package */
+	if (!lpi_data || lpi_data->type != ACPI_TYPE_PACKAGE ||
+	    lpi_data->package.count < 4) {
+		acpi_handle_debug(handle, "Not enough elements in _LPI\n");
+		ret = -ENODATA;
+		goto end;
+	}
+
+	lpi_pkg_count = lpi_data->package.elements[2].integer.value;
+
+	/* Validate number of power states. */
+	if (!lpi_pkg_count || lpi_pkg_count != lpi_data->package.count - 3) {
+		acpi_handle_debug(handle, "Invalid _LPI state count\n");
+		ret = -ENODATA;
+		goto end;
+	}
+
+	lpi_state = kzalloc_objs(*lpi_state, lpi_pkg_count);
+	if (!lpi_state) {
+		ret = -ENOMEM;
+		goto end;
+	}
+
+	info->size = lpi_pkg_count;
+	info->entries = lpi_state;
+
+	/* _LPI State packages start at index 3. */
+	lpi_pkg = &lpi_data->package.elements[3];
+
+	for (state_idx = 1; state_idx <= lpi_pkg_count; state_idx++) {
+		lpi_state->index = state_idx;
+		process_lpi_state_package(lpi_pkg++, lpi_state++, handle,
+					  state_idx, strict);
+	}
+
+	acpi_handle_debug(handle, "Found %u power states\n", lpi_pkg_count);
+end:
+	kfree(buffer.pointer);
+	return ret;
+}
+
+/**
+ * combine_lpi_states - combine local and parent LPI states to form a composite LPI state
+ *
+ * @local: local LPI state
+ * @parent: parent LPI state
+ * @result: composite LPI state
+ */
+static bool combine_lpi_states(struct acpi_lpi_state *local,
+			       struct acpi_lpi_state *parent,
+			       struct acpi_lpi_state *result)
+{
+	if (parent->entry_method == ACPI_CSTATE_INTEGER) {
+		if (!parent->address) /* 0 means autopromotable */
+			return false;
+		result->address = local->address + parent->address;
+	} else {
+		result->address = parent->address;
+	}
+
+	result->min_residency = max(local->min_residency, parent->min_residency);
+	result->wake_latency = local->wake_latency + parent->wake_latency;
+	result->enable_parent_state = parent->enable_parent_state;
+	result->entry_method = local->entry_method;
+
+	result->flags = parent->flags;
+	result->arch_flags = parent->arch_flags;
+	result->index = parent->index;
+
+	scnprintf(result->desc, ACPI_CX_DESC_LEN, "%s+%s", local->desc, parent->desc);
+	return true;
+}
+
+#define ACPI_LPI_STATE_FLAGS_ENABLED			BIT(0)
+
+static void stash_composite_state(struct acpi_lpi_states_array *curr_level,
+				  struct acpi_lpi_state *t)
+{
+	curr_level->composite_states[curr_level->composite_states_size++] = t;
+}
+
+static bool too_many_states(acpi_handle handle, unsigned int state_count)
+{
+	if (state_count < ACPI_PROCESSOR_MAX_POWER)
+		return false;
+
+	acpi_handle_info(handle, "No space for more _LPI states than %d\n",
+			 ACPI_PROCESSOR_MAX_POWER);
+	return true;
+}
+
+static unsigned int flatten_lpi_states(acpi_handle handle,
+				       struct acpi_lpi_state *lpi_states,
+				       unsigned int state_count,
+				       struct acpi_lpi_states_array *curr,
+				       struct acpi_lpi_states_array *prev)
+{
+	struct acpi_lpi_state *parent_lpi = curr->entries;
+	unsigned int j;
+
+	/*
+	 * Combine each of the "raw" _LPI states from the current (processor
+	 * container) level with all of the composite _LPI states from the
+	 * previous (processor or processor container) level.
+	 */
+	for (j = 0; j < curr->size; j++, parent_lpi++) {
+		struct acpi_lpi_state *flpi;
+		int i;
+
+		if (!(parent_lpi->flags & ACPI_LPI_STATE_FLAGS_ENABLED))
+			continue;
+
+		if (too_many_states(handle, state_count))
+			break;
+
+		flpi = &lpi_states[state_count];
+
+		for (i = 0; i < prev->composite_states_size; i++) {
+			struct acpi_lpi_state *local_lpi = prev->composite_states[i];
+
+			if (parent_lpi->index > local_lpi->enable_parent_state)
+				continue;
+
+			if (!combine_lpi_states(local_lpi, parent_lpi, flpi))
+				continue;
+
+			stash_composite_state(curr, flpi);
+			state_count++;
+			flpi++;
+
+			if (state_count >= ACPI_PROCESSOR_MAX_POWER)
+				break;
+		}
+	}
+
+	return state_count;
+}
+
+int acpi_processor_extract_lpi_info(acpi_handle pr_handle,
+				    struct acpi_processor_power *pr_power,
+				    bool strict)
+{
+	struct acpi_lpi_states_array info[2], *prev, *curr;
+	acpi_handle handle = pr_handle;
+	unsigned int state_count = 0;
+	unsigned int i;
+	int ret;
+
+	if (!osc_pc_lpi_support_confirmed)
+		return -EOPNOTSUPP;
+
+	curr = &info[0];
+	curr->composite_states_size = 0;
+
+	ret = acpi_processor_evaluate_lpi(handle, curr, strict);
+	if (ret)
+		return ret;
+
+	/* Copy all of the usable first-level states to power.lpi_states[]. */
+	for (i = 0; i < curr->size; i++) {
+		struct acpi_lpi_state *lpi = &curr->entries[i];
+		struct acpi_lpi_state *flpi;
+
+		/*
+		 * Skip states that are not enabled or have an inadequate entry
+		 * method for this level.
+		 */
+		if (!(lpi->flags & ACPI_LPI_STATE_FLAGS_ENABLED) ||
+		    lpi->entry_method == ACPI_CSTATE_INTEGER)
+			continue;
+
+		if (too_many_states(pr_handle, state_count))
+			break;
+
+		flpi = &pr_power->lpi_states[state_count++];
+		memcpy(flpi, lpi, sizeof(*lpi));
+		stash_composite_state(curr, flpi);
+	}
+
+	kfree(curr->entries);
+
+	/*
+	 * If there are no _LPI states at the first level, there are no _LPI
+	 * states at all.
+	 */
+	if (!state_count)
+		return -ENODATA;
+
+	prev = curr;
+	curr = &info[1];
+
+	for (;;) {
+		struct acpi_lpi_states_array *tmp;
+		struct acpi_device *d;
+
+		if (ACPI_FAILURE(acpi_get_parent(handle, &handle)))
+			break;
+
+		d = acpi_fetch_acpi_dev(handle);
+		if (!d)
+			break;
+
+		if (strcmp(acpi_device_hid(d), ACPI_PROCESSOR_CONTAINER_HID))
+			break;
+
+		curr->composite_states_size = 0;
+
+		ret = acpi_processor_evaluate_lpi(handle, curr, strict);
+		if (ret)
+			break;
+
+		/* flatten all the LPI states in this level of hierarchy */
+		state_count = flatten_lpi_states(pr_handle, pr_power->lpi_states,
+						 state_count, curr, prev);
+
+		kfree(curr->entries);
+
+		tmp = prev, prev = curr, curr = tmp;
+	}
+
+	/* reset the index after flattening */
+	for (i = 0; i < state_count; i++)
+		pr_power->lpi_states[i].index = i;
+
+	pr_power->count = state_count;
+
+	return 0;
+}
+EXPORT_SYMBOL_NS_GPL(acpi_processor_extract_lpi_info, "ACPI_PROCESSOR_IDLE");
+#endif /* CONFIG_ACPI_PROCESSOR_IDLE */
diff --git a/drivers/acpi/processor_idle.c b/drivers/acpi/processor_idle.c
index 4482cf28f56a..e113bcbbb882 100644
--- a/drivers/acpi/processor_idle.c
+++ b/drivers/acpi/processor_idle.c
@@ -853,223 +853,6 @@ static int acpi_processor_setup_cstates(struct acpi_processor *pr)
 
 #endif /* CONFIG_ACPI_PROCESSOR_CSTATE */
 
-struct acpi_lpi_states_array {
-	unsigned int size;
-	unsigned int composite_states_size;
-	struct acpi_lpi_state *entries;
-	struct acpi_lpi_state *composite_states[ACPI_PROCESSOR_MAX_POWER];
-};
-
-static int obj_get_integer(union acpi_object *obj, u32 *value)
-{
-	if (obj->type != ACPI_TYPE_INTEGER)
-		return -EINVAL;
-
-	*value = obj->integer.value;
-	return 0;
-}
-
-static int acpi_processor_evaluate_lpi(acpi_handle handle,
-				       struct acpi_lpi_states_array *info)
-{
-	acpi_status status;
-	int ret = 0;
-	int pkg_count, state_idx = 1, loop;
-	struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL };
-	union acpi_object *lpi_data;
-	struct acpi_lpi_state *lpi_state;
-
-	status = acpi_evaluate_object(handle, "_LPI", NULL, &buffer);
-	if (ACPI_FAILURE(status)) {
-		acpi_handle_debug(handle, "No _LPI, giving up\n");
-		return -ENODEV;
-	}
-
-	lpi_data = buffer.pointer;
-
-	/* There must be at least 4 elements = 3 elements + 1 package */
-	if (!lpi_data || lpi_data->type != ACPI_TYPE_PACKAGE ||
-	    lpi_data->package.count < 4) {
-		pr_debug("not enough elements in _LPI\n");
-		ret = -ENODATA;
-		goto end;
-	}
-
-	pkg_count = lpi_data->package.elements[2].integer.value;
-
-	/* Validate number of power states. */
-	if (pkg_count < 1 || pkg_count != lpi_data->package.count - 3) {
-		pr_debug("count given by _LPI is not valid\n");
-		ret = -ENODATA;
-		goto end;
-	}
-
-	lpi_state = kzalloc_objs(*lpi_state, pkg_count);
-	if (!lpi_state) {
-		ret = -ENOMEM;
-		goto end;
-	}
-
-	info->size = pkg_count;
-	info->entries = lpi_state;
-
-	/* LPI States start at index 3 */
-	for (loop = 3; state_idx <= pkg_count; loop++, state_idx++, lpi_state++) {
-		union acpi_object *element, *pkg_elem, *obj;
-
-		element = &lpi_data->package.elements[loop];
-		if (element->type != ACPI_TYPE_PACKAGE || element->package.count < 7)
-			continue;
-
-		pkg_elem = element->package.elements;
-
-		obj = pkg_elem + 6;
-		if (obj->type == ACPI_TYPE_BUFFER) {
-			struct acpi_power_register *reg;
-
-			reg = (struct acpi_power_register *)obj->buffer.pointer;
-			if (reg->space_id != ACPI_ADR_SPACE_SYSTEM_IO &&
-			    reg->space_id != ACPI_ADR_SPACE_FIXED_HARDWARE)
-				continue;
-
-			lpi_state->address = reg->address;
-			lpi_state->entry_method =
-				reg->space_id == ACPI_ADR_SPACE_FIXED_HARDWARE ?
-				ACPI_CSTATE_FFH : ACPI_CSTATE_SYSTEMIO;
-		} else if (obj->type == ACPI_TYPE_INTEGER) {
-			lpi_state->entry_method = ACPI_CSTATE_INTEGER;
-			lpi_state->address = obj->integer.value;
-		} else {
-			pr_debug("Entry method of state-%d is invalid, disable it.\n",
-				 state_idx);
-			continue;
-		}
-
-		/* elements[7,8] skipped for now i.e. Residency/Usage counter*/
-
-		obj = pkg_elem + 9;
-		if (obj->type == ACPI_TYPE_STRING)
-			strscpy(lpi_state->desc, obj->string.pointer,
-				ACPI_CX_DESC_LEN);
-
-		lpi_state->index = state_idx;
-		if (obj_get_integer(pkg_elem + 0, &lpi_state->min_residency)) {
-			pr_debug("No min. residency found, assuming 10 us\n");
-			lpi_state->min_residency = 10;
-		}
-
-		if (obj_get_integer(pkg_elem + 1, &lpi_state->wake_latency)) {
-			pr_debug("No wakeup residency found, assuming 10 us\n");
-			lpi_state->wake_latency = 10;
-		}
-
-		if (obj_get_integer(pkg_elem + 2, &lpi_state->flags))
-			lpi_state->flags = 0;
-
-		if (obj_get_integer(pkg_elem + 3, &lpi_state->arch_flags))
-			lpi_state->arch_flags = 0;
-
-		if (obj_get_integer(pkg_elem + 4, &lpi_state->res_cnt_freq))
-			lpi_state->res_cnt_freq = 1;
-
-		if (obj_get_integer(pkg_elem + 5, &lpi_state->enable_parent_state))
-			lpi_state->enable_parent_state = 0;
-	}
-
-	acpi_handle_debug(handle, "Found %d power states\n", state_idx);
-end:
-	kfree(buffer.pointer);
-	return ret;
-}
-
-/**
- * combine_lpi_states - combine local and parent LPI states to form a composite LPI state
- *
- * @local: local LPI state
- * @parent: parent LPI state
- * @result: composite LPI state
- */
-static bool combine_lpi_states(struct acpi_lpi_state *local,
-			       struct acpi_lpi_state *parent,
-			       struct acpi_lpi_state *result)
-{
-	if (parent->entry_method == ACPI_CSTATE_INTEGER) {
-		if (!parent->address) /* 0 means autopromotable */
-			return false;
-		result->address = local->address + parent->address;
-	} else {
-		result->address = parent->address;
-	}
-
-	result->min_residency = max(local->min_residency, parent->min_residency);
-	result->wake_latency = local->wake_latency + parent->wake_latency;
-	result->enable_parent_state = parent->enable_parent_state;
-	result->entry_method = local->entry_method;
-
-	result->flags = parent->flags;
-	result->arch_flags = parent->arch_flags;
-	result->index = parent->index;
-
-	scnprintf(result->desc, ACPI_CX_DESC_LEN, "%s+%s", local->desc, parent->desc);
-	return true;
-}
-
-#define ACPI_LPI_STATE_FLAGS_ENABLED			BIT(0)
-
-static void stash_composite_state(struct acpi_lpi_states_array *curr_level,
-				  struct acpi_lpi_state *t)
-{
-	curr_level->composite_states[curr_level->composite_states_size++] = t;
-}
-
-static unsigned int flatten_lpi_states(struct acpi_processor *pr,
-				       unsigned int flat_state_cnt,
-				       struct acpi_lpi_states_array *curr_level,
-				       struct acpi_lpi_states_array *prev_level)
-{
-	int i, j, state_count = curr_level->size;
-	struct acpi_lpi_state *p, *t = curr_level->entries;
-
-	curr_level->composite_states_size = 0;
-	for (j = 0; j < state_count; j++, t++) {
-		struct acpi_lpi_state *flpi;
-
-		if (!(t->flags & ACPI_LPI_STATE_FLAGS_ENABLED))
-			continue;
-
-		if (flat_state_cnt >= ACPI_PROCESSOR_MAX_POWER) {
-			pr_warn("Limiting number of LPI states to max (%d)\n",
-				ACPI_PROCESSOR_MAX_POWER);
-			pr_warn("Please increase ACPI_PROCESSOR_MAX_POWER if needed.\n");
-			break;
-		}
-
-		flpi = &pr->power.lpi_states[flat_state_cnt];
-
-		if (!prev_level) { /* leaf/processor node */
-			memcpy(flpi, t, sizeof(*t));
-			stash_composite_state(curr_level, flpi);
-			flat_state_cnt++;
-			continue;
-		}
-
-		for (i = 0; i < prev_level->composite_states_size; i++) {
-			p = prev_level->composite_states[i];
-			if (t->index <= p->enable_parent_state &&
-			    combine_lpi_states(p, t, flpi)) {
-				stash_composite_state(curr_level, flpi);
-				flat_state_cnt++;
-				flpi++;
-				if (flat_state_cnt >= ACPI_PROCESSOR_MAX_POWER)
-					break;
-			}
-		}
-	}
-
-	kfree(curr_level->entries);
-	return flat_state_cnt;
-}
-
 int __weak acpi_processor_ffh_lpi_probe(unsigned int cpu)
 {
 	return -EOPNOTSUPP;
@@ -1077,64 +860,16 @@ int __weak acpi_processor_ffh_lpi_probe(unsigned int cpu)
 
 static int acpi_processor_get_lpi_info(struct acpi_processor *pr)
 {
-	int ret, i;
-	acpi_status status;
-	acpi_handle handle = pr->handle, pr_ahandle;
-	struct acpi_device *d = NULL;
-	struct acpi_lpi_states_array info[2], *tmp, *prev, *curr;
-	unsigned int state_count;
+	int ret;
 
 	/* make sure our architecture has support */
 	ret = acpi_processor_ffh_lpi_probe(pr->id);
 	if (ret == -EOPNOTSUPP)
 		return ret;
 
-	if (!osc_pc_lpi_support_confirmed)
-		return -EOPNOTSUPP;
-
-	if (!acpi_has_method(handle, "_LPI"))
-		return -EINVAL;
-
-	prev = &info[0];
-	curr = &info[1];
-	handle = pr->handle;
-	ret = acpi_processor_evaluate_lpi(handle, prev);
+	ret = acpi_processor_extract_lpi_info(pr->handle, &pr->power, false);
 	if (ret)
 		return ret;
-	state_count = flatten_lpi_states(pr, 0, prev, NULL);
-
-	status = acpi_get_parent(handle, &pr_ahandle);
-	while (ACPI_SUCCESS(status)) {
-		d = acpi_fetch_acpi_dev(pr_ahandle);
-		if (!d)
-			break;
-
-		handle = pr_ahandle;
-
-		if (strcmp(acpi_device_hid(d), ACPI_PROCESSOR_CONTAINER_HID))
-			break;
-
-		/* can be optional ? */
-		if (!acpi_has_method(handle, "_LPI"))
-			break;
-
-		ret = acpi_processor_evaluate_lpi(handle, curr);
-		if (ret)
-			break;
-
-		/* flatten all the LPI states in this level of hierarchy */
-		state_count = flatten_lpi_states(pr, state_count, curr, prev);
-
-		tmp = prev, prev = curr, curr = tmp;
-
-		status = acpi_get_parent(handle, &pr_ahandle);
-	}
-
-	/* reset the index after flattening */
-	for (i = 0; i < state_count; i++)
-		pr->power.lpi_states[i].index = i;
-
-	pr->power.count = state_count;
 
 	/* Tell driver that _LPI is supported. */
 	pr->flags.has_lpi = 1;
diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
index d74b478db280..651408df9c24 100644
--- a/drivers/idle/intel_idle.c
+++ b/drivers/idle/intel_idle.c
@@ -53,6 +53,7 @@
 #include <linux/notifier.h>
 #include <linux/cpu.h>
 #include <linux/moduleparam.h>
+#include <linux/pm_qos.h>
 #include <linux/sysfs.h>
 #include <asm/cpuid/api.h>
 #include <asm/cpu_device_id.h>
@@ -1779,6 +1780,7 @@ module_param_named(no_native, no_native, bool, 0444);
 MODULE_PARM_DESC(no_native, "Ignore cpu specific (native) idle states in lieu of ACPI idle states");
 
 static struct acpi_processor_power acpi_state_table __initdata;
+static bool acpi_lpi_available __initdata;
 
 /**
  * intel_idle_cst_usable - Check if the _CST information can be used.
@@ -1803,18 +1805,37 @@ static bool __init intel_idle_cst_usable(void)
 	return true;
 }
 
-static bool __init intel_idle_acpi_cst_extract(void)
+static bool __init intel_idle_acpi_extract_lpi_cstates(void)
 {
 	unsigned int cpu;
 
-	if (no_acpi) {
-		pr_debug("Not allowed to use ACPI _CST\n");
-		return false;
+	for_each_possible_cpu(cpu) {
+		struct acpi_processor *pr;
+
+		pr = per_cpu(processors, cpu);
+		if (!pr)
+			continue;
+
+		if (acpi_processor_extract_lpi_info(pr->handle,
+						    &acpi_state_table, true))
+			continue;
+
+		acpi_lpi_available = true;
+		return true;
 	}
 
+	pr_debug("No ACPI _LPI idle states\n");
+	return false;
+}
+
+static bool __init intel_idle_acpi_extract_cst_cstates(void)
+{
+	unsigned int cpu;
+
 	for_each_possible_cpu(cpu) {
-		struct acpi_processor *pr = per_cpu(processors, cpu);
+		struct acpi_processor *pr;
 
+		pr = per_cpu(processors, cpu);
 		if (!pr)
 			continue;
 
@@ -1826,18 +1847,96 @@ static bool __init intel_idle_acpi_cst_extract(void)
 		if (!intel_idle_cst_usable())
 			continue;
 
-		if (!acpi_processor_claim_cst_control())
-			break;
+		return true;
+	}
+
+	pr_debug("ACPI _CST not found or not usable\n");
+	return false;
+}
 
+static bool __init intel_idle_acpi_extract_cstates(void)
+{
+	if (intel_idle_acpi_extract_lpi_cstates())
 		return true;
+
+	if (intel_idle_acpi_extract_cst_cstates())
+		return true;
+
+	return false;
+}
+
+static bool __init intel_idle_acpi_probe(void)
+{
+	if (no_acpi) {
+		pr_debug("Not allowed to use ACPI for C-states extraction\n");
+		return false;
 	}
 
+	if (intel_idle_acpi_extract_cstates() &&
+	    acpi_processor_claim_cst_control())
+		return true;
+
 	acpi_state_table.count = 0;
-	pr_debug("ACPI _CST not found or not usable\n");
 	return false;
 }
 
-static void __init intel_idle_init_cstates_acpi(struct cpuidle_driver *drv)
+static void __init intel_idle_complete_state_init(struct cpuidle_state *state)
+{
+	if (intel_idle_state_needs_timer_stop(state))
+		state->flags |= CPUIDLE_FLAG_TIMER_STOP;
+
+	state->enter = intel_idle;
+	state->enter_dead = intel_idle_enter_dead;
+	state->enter_s2idle = intel_idle_s2idle;
+}
+
+static void __init intel_idle_init_cstates_acpi_lpi(struct cpuidle_driver *drv)
+{
+	int index;
+
+	for (index = 0; index < acpi_state_table.count; index++) {
+		struct acpi_lpi_state *lpi_state;
+		struct cpuidle_state *state;
+
+		if (intel_idle_max_cstate_reached(index))
+			break;
+
+		lpi_state = &acpi_state_table.lpi_states[index];
+
+		state = &drv->states[drv->state_count++];
+
+		scnprintf(state->name, CPUIDLE_NAME_LEN, "C%d_LPI", index + 1);
+		strscpy(state->desc, lpi_state->desc, CPUIDLE_DESC_LEN);
+		state->exit_latency = lpi_state->wake_latency;
+		state->target_residency = lpi_state->min_residency;
+		state->flags = MWAIT2flg(lpi_state->address);
+		/*
+		 * Assume that entering any of the idle states extracted from
+		 * _LPI except for the first two will cause the TLB to be
+		 * flushed and let the core call leave_mm() for them upfront
+		 * to avoid unnecessary wakeups due to TLB shootdowns.
+		 */
+		if (index > 1)
+			state->flags |= CPUIDLE_FLAG_TLB_FLUSHED;
+
+		if (disabled_states_mask & BIT(index + 1))
+			state->flags |= CPUIDLE_FLAG_OFF;
+
+		intel_idle_complete_state_init(state);
+
+		pr_info("%s: MWAIT hint 0x%x\n", state->name, flg2MWAIT(state->flags));
+	}
+
+	/*
+	 * Assume the first idle state in the table to be C1 and if any deeper
+	 * idle states are exposed while X86_FEATURE_NONSTOP_TSC is unset, mark
+	 * the TSC as unstable.
+	 */
+	if (index > 1 && !boot_cpu_has(X86_FEATURE_NONSTOP_TSC))
+		mark_tsc_unstable("TSC halts in idle");
+}
+
+static void __init intel_idle_init_cstates_acpi_cst(struct cpuidle_driver *drv)
 {
 	int cstate, limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count);
 
@@ -1879,49 +1978,76 @@ static void __init intel_idle_init_cstates_acpi(struct cpuidle_driver *drv)
 		if (disabled_states_mask & BIT(cstate))
 			state->flags |= CPUIDLE_FLAG_OFF;
 
-		if (intel_idle_state_needs_timer_stop(state))
-			state->flags |= CPUIDLE_FLAG_TIMER_STOP;
-
 		if (cx->type > ACPI_STATE_C1 && !boot_cpu_has(X86_FEATURE_NONSTOP_TSC))
 			mark_tsc_unstable("TSC halts in idle");
 
-		state->enter = intel_idle;
-		state->enter_dead = intel_idle_enter_dead;
-		state->enter_s2idle = intel_idle_s2idle;
+		intel_idle_complete_state_init(state);
 	}
 }
 
-static bool __init intel_idle_off_by_default(unsigned int flags, u32 mwait_hint)
+static void __init intel_idle_init_cstates_acpi(struct cpuidle_driver *drv)
 {
-	int cstate, limit;
+	if (acpi_lpi_available)
+		intel_idle_init_cstates_acpi_lpi(drv);
+	else
+		intel_idle_init_cstates_acpi_cst(drv);
+}
 
-	/*
-	 * If there are no _CST C-states, do not disable any C-states by
-	 * default.
-	 */
-	if (!acpi_state_table.count)
-		return false;
+static bool __init intel_idle_acpi_hint_match(unsigned int flags, u32 acpi_hint,
+					      u32 table_hint)
+{
+	if (flags & CPUIDLE_FLAG_PARTIAL_HINT_MATCH) {
+		acpi_hint &= ~MWAIT_SUBSTATE_MASK;
+		table_hint &= ~MWAIT_SUBSTATE_MASK;
+	}
+	return acpi_hint == table_hint;
+}
+
+static bool __init intel_idle_off_by_default_lpi(unsigned int flags, u32 mwait_hint)
+{
+	int index;
+
+	for (index = 0; index < acpi_state_table.count; index++) {
+		u32 acpi_hint = acpi_state_table.lpi_states[index].address;
+
+		if (intel_idle_acpi_hint_match(flags, acpi_hint, mwait_hint))
+			return false;
+	}
+	return true;
+}
+
+static bool __init intel_idle_off_by_default_cst(unsigned int flags, u32 mwait_hint)
+{
+	int cstate, limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count);
 
-	limit = min_t(int, CPUIDLE_STATE_MAX, acpi_state_table.count);
 	/*
 	 * If limit > 0, intel_idle_cst_usable() has returned 'true', so all of
 	 * the interesting states are ACPI_CSTATE_FFH.
 	 */
 	for (cstate = 1; cstate < limit; cstate++) {
 		u32 acpi_hint = acpi_state_table.states[cstate].address;
-		u32 table_hint = mwait_hint;
-
-		if (flags & CPUIDLE_FLAG_PARTIAL_HINT_MATCH) {
-			acpi_hint &= ~MWAIT_SUBSTATE_MASK;
-			table_hint &= ~MWAIT_SUBSTATE_MASK;
-		}
 
-		if (acpi_hint == table_hint)
+		if (intel_idle_acpi_hint_match(flags, acpi_hint, mwait_hint))
 			return false;
 	}
 	return true;
 }
 
+static bool __init intel_idle_off_by_default(unsigned int flags, u32 mwait_hint)
+{
+	/*
+	 * If there is no C-states information in the ACPI tables, do not
+	 * disable any C-states by default.
+	 */
+	if (!acpi_state_table.count)
+		return false;
+
+	if (acpi_lpi_available)
+		return intel_idle_off_by_default_lpi(flags, mwait_hint);
+
+	return intel_idle_off_by_default_cst(flags, mwait_hint);
+}
+
 static inline bool ignore_native(void)
 {
 	return no_native && !no_acpi;
@@ -1929,7 +2055,7 @@ static inline bool ignore_native(void)
 #else /* !CONFIG_ACPI_PROCESSOR_CSTATE */
 #define force_use_acpi	(false)
 
-static inline bool intel_idle_acpi_cst_extract(void) { return false; }
+static inline bool intel_idle_acpi_probe(void) { return false; }
 static inline void intel_idle_init_cstates_acpi(struct cpuidle_driver *drv) { }
 static inline bool intel_idle_off_by_default(unsigned int flags, u32 mwait_hint)
 {
@@ -2697,6 +2823,9 @@ error:
 	pr_info("Failed to adjust C-states with data from 'intel_idle.table'\n");
 }
 
+#define INTEL_IDLE_INIT_QOS	20
+static struct pm_qos_request qos_req __initdata;
+
 static int __init intel_idle_init(void)
 {
 	const struct x86_cpu_id *id;
@@ -2741,7 +2870,7 @@ static int __init intel_idle_init(void)
 	if (icpu) {
 		if (icpu->state_table)
 			cpuidle_state_table = icpu->state_table;
-		else if (!intel_idle_acpi_cst_extract())
+		else if (!intel_idle_acpi_probe())
 			return -ENODEV;
 
 		auto_demotion_disable_flags = icpu->auto_demotion_disable_flags;
@@ -2750,8 +2879,8 @@ static int __init intel_idle_init(void)
 		if (icpu->c1_demotion_supported)
 			c1_demotion_supported = true;
 		if (icpu->use_acpi || force_use_acpi)
-			intel_idle_acpi_cst_extract();
-	} else if (!intel_idle_acpi_cst_extract()) {
+			intel_idle_acpi_probe();
+	} else if (!intel_idle_acpi_probe()) {
 		return -ENODEV;
 	}
 
@@ -2766,6 +2895,13 @@ static int __init intel_idle_init(void)
 	if (retval)
 		pr_warn("failed to initialized sysfs");
 
+	/*
+	 * Some platforms, in particular the Intel S1200BTL motherboard, have a
+	 * problem with using package idle states too early, so prevent that
+	 * from taking place until the device_initcall() phase is over.
+	 */
+	cpu_latency_qos_add_request(&qos_req, INTEL_IDLE_INIT_QOS);
+
 	retval = cpuidle_register_driver(&intel_idle_driver);
 	if (retval) {
 		struct cpuidle_driver *drv = cpuidle_get_driver();
@@ -2790,6 +2926,9 @@ hp_setup_fail:
 	intel_idle_cpuidle_devices_uninit();
 	cpuidle_unregister_driver(&intel_idle_driver);
 init_driver_fail:
+	if (cpu_latency_qos_request_active((&qos_req)))
+		cpu_latency_qos_remove_request(&qos_req);
+
 	intel_idle_sysfs_uninit();
 	free_percpu(intel_idle_cpuidle_devices);
 	return retval;
@@ -2797,6 +2936,15 @@ init_driver_fail:
 }
 subsys_initcall_sync(intel_idle_init);
 
+static int __init intel_idle_init_complete(void)
+{
+	if (cpu_latency_qos_request_active((&qos_req)))
+		cpu_latency_qos_remove_request(&qos_req);
+
+	return 0;
+}
+device_initcall_sync(intel_idle_init_complete);
+
 /*
  * We are not really modular, but we used to support that.  Meaning we also
  * support "intel_idle.max_cstate=..." at boot and also a read-only export of
diff --git a/include/linux/acpi.h b/include/linux/acpi.h
index 60ab50cb8930..9e418b23373c 100644
--- a/include/linux/acpi.h
+++ b/include/linux/acpi.h
@@ -315,6 +315,19 @@ static inline int acpi_processor_evaluate_cst(acpi_handle handle, u32 cpu,
 }
 #endif
 
+#ifdef CONFIG_ACPI_PROCESSOR_IDLE
+int acpi_processor_extract_lpi_info(acpi_handle pr_handle,
+				    struct acpi_processor_power *pr_power,
+				    bool strict);
+#else
+static inline int acpi_processor_extract_lpi_info(acpi_handle pr_handle,
+				    struct acpi_processor_power *pr_power,
+				    bool strict)
+{
+	return -ENODEV;
+}
+#endif
+
 #ifdef CONFIG_ACPI_HOTPLUG_CPU
 /* Arch dependent functions for cpu hotplug support */
 int acpi_map_cpu(acpi_handle handle, phys_cpuid_t physid, u32 acpi_id,