Re: [PATCH v15 05/13] hp-bioscfg: enum-attributes

From: Jorge Lopez
Date: Wed May 31 2023 - 12:00:47 EST

Next message: Jim Mattson: "Re: [PATCH] KVM: VMX: restore vmx_vmexit alignment"
Previous message: Jonathan Cameron: "Re: [PATCH v2 3/3] docs: perf: Add new description for HiSilicon UC PMU"
In reply to: Thomas Weißschuh: "Re: [PATCH v15 05/13] hp-bioscfg: enum-attributes"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Tue, May 30, 2023 at 4:42 PM Thomas Weißschuh <thomas@xxxxxxxx> wrote:
>
> On 2023-05-30 12:01:57-0500, Jorge Lopez wrote:
> > On Fri, May 26, 2023 at 10:35 AM Thomas Weißschuh <thomas@xxxxxxxx> wrote:
> > >
> > > On 2023-05-19 15:12:52-0500, Jorge Lopez wrote:
> > >
> > > <snip>
> > >
> > > > .../x86/hp/hp-bioscfg/enum-attributes.c | 465 ++++++++++++++++++
> > > > 1 file changed, 465 insertions(+)
> > > > create mode 100644 drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> > > >
> > > > diff --git a/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c b/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> > > > new file mode 100644
> > > > index 000000000000..80842835606d
> > > > --- /dev/null
> > > > +++ b/drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
> > > > @@ -0,0 +1,465 @@
> > > > +// SPDX-License-Identifier: GPL-2.0
> > > > +/*
> > > > + * Functions corresponding to enumeration type attributes under
> > > > + * BIOS Enumeration GUID for use with hp-bioscfg driver.
> > > > + *
> > > > + * Copyright (c) 2022 HP Development Company, L.P.
> > > > + */
> > > > +
> > > > +#include "bioscfg.h"
> > > > +
> > > > +GET_INSTANCE_ID(enumeration);
> > > > +
> > > > +static ssize_t current_value_show(struct kobject *kobj, struct kobj_attribute *attr, char *buf)
> > > > +{
> > > > + int instance_id = get_enumeration_instance_id(kobj);
> > > > +
> > > > + if (instance_id < 0)
> > > > + return -EIO;
> > > > +
> > > > + return sysfs_emit(buf, "%s\n",
> > > > + bioscfg_drv.enumeration_data[instance_id].current_value);
> > > > +}
> > > > +
> > > > +/**
> > > > + * validate_enumeration_input() -
> > > > + * Validate input of current_value against possible values
> > > > + *
> > > > + * @instance_id: The instance on which input is validated
> > > > + * @buf: Input value
> > > > + */
> >
> > <snip>
> >
> > > > +static int hp_populate_enumeration_elements_from_package(union acpi_object *enum_obj,
> > > > + int enum_obj_count,
> > > > + int instance_id)
> > > > +{
> > > > + char *str_value = NULL;
> > > > + int value_len;
> > > > + u32 size = 0;
> > > > + u32 int_value;
> > > > + int elem = 0;
> > > > + int reqs;
> > > > + int pos_values;
> > > > + int ret;
> > > > + int eloc;
> > > > + struct enumeration_data *enum_data = &bioscfg_drv.enumeration_data[instance_id];
> > > > +
> > > > + for (elem = 1, eloc = 1; elem < enum_obj_count; elem++, eloc++) {
> > > > + /* ONLY look at the first ENUM_ELEM_CNT elements */
> > > > + if (eloc == ENUM_ELEM_CNT)
> > > > + goto exit_enumeration_package;
> > > > +
> > > > + switch (enum_obj[elem].type) {
> > > > + case ACPI_TYPE_STRING:
> > > > + if (PREREQUISITES != elem && ENUM_POSSIBLE_VALUES != elem) {
> > > > + ret = hp_convert_hexstr_to_str(enum_obj[elem].string.pointer,
> > > > + enum_obj[elem].string.length,
> > > > + &str_value, &value_len);
> > > > + if (ret)
> > > > + return -EINVAL;
> > > > + }
> > > > + break;
> > > > + case ACPI_TYPE_INTEGER:
> > > > + int_value = (u32)enum_obj[elem].integer.value;
> > > > + break;
> > > > + default:
> > > > + pr_warn("Unsupported object type [%d]\n", enum_obj[elem].type);
> > > > + continue;
> > > > + }
> > > > +
> > > > + /* Check that both expected and read object type match */
> > > > + if (expected_enum_types[eloc] != enum_obj[elem].type) {
> > > > + pr_err("Error expected type %d for elem %d, but got type %d instead\n",
> > > > + expected_enum_types[eloc], elem, enum_obj[elem].type);
> > > > + return -EIO;
> > > > + }
> > > > +
> > > > + /* Assign appropriate element value to corresponding field */
> > > > + switch (eloc) {
> > > > + case NAME:
> > > > + case VALUE:
> > > > + break;
> > > > + case PATH:
> > > > + strscpy(enum_data->common.path, str_value,
> > > > + sizeof(enum_data->common.path));
> > > > + break;
> > > > + case IS_READONLY:
> > > > + enum_data->common.is_readonly = int_value;
> > > > + break;
> > > > + case DISPLAY_IN_UI:
> > > > + enum_data->common.display_in_ui = int_value;
> > > > + break;
> > > > + case REQUIRES_PHYSICAL_PRESENCE:
> > > > + enum_data->common.requires_physical_presence = int_value;
> > > > + break;
> > > > + case SEQUENCE:
> > > > + enum_data->common.sequence = int_value;
> > > > + break;
> > > > + case PREREQUISITES_SIZE:
> > > > + enum_data->common.prerequisites_size = int_value;
> > > > + if (int_value > MAX_PREREQUISITES_SIZE)
> > > > + pr_warn("Prerequisites size value exceeded the maximum number of elements supported or data may be malformed\n");
> > > > +
> > > > + /*
> > > > + * This HACK is needed to keep the expected
> > > > + * element list pointing to the right obj[elem].type
> > > > + * when the size is zero. PREREQUISITES
> > > > + * object is omitted by BIOS when the size is
> > > > + * zero.
> > > > + */
> > > > + if (int_value == 0)
> > > > + eloc++;
> > > > + break;
> > > > +
> > > > + case PREREQUISITES:
> > > > + size = min_t(u32, enum_data->common.prerequisites_size, MAX_PREREQUISITES_SIZE);
> > >
> > > We cannot blindly truncate this to a maximum value.
> > > The firmware reported an amount of elements it would return.
> > >
> > > If this value is to big than we can not just intpret the data as if it
> > > was something the firmware did not return.
> > >
> > > An error needs to be reported to userspace.
> > > A default value is not enough as userspace can not interpret this
> > > properly.
> > >
> >
> > It is ok to truncate prerequisite size to MAX_PREREQUISITES_SIZE.
> > MAX_PREREQUISITES_SIZE is a value predefined by BIOS when the
> > prerequisite values size is invalid ( > MAX_PREREQUISITES_SIZE) and/or
> > the prerequisite data is corrupted.
> > Neither PREREQUISITES nor PREREQUISITES_SIZE are reported to the
> > userspace so there is no need to report a failure on data that is not
> > exposed. One item that needs clarification is the fact that
> > regardless if PREREQUISITES or PREREQUISITES_SIZE are invalid, that
> > does not mean other values are invalid. It is for this reason, we
> > need to continue to read all remaining packages.
>
> It may be that prerequisites are not reported to userspace.
> But the following values are:
>
> security level, current value and possible values.
>
> And if prerequisites are garbage then those are now also garbage.

This statement is not correct. Each attribute value is isolated from
each other hence it only affects a single value and does not impact
the following values such security level, current value and possible
values
Here is a sample where the prerequisite_size is invalid but the
prerequisites, security, string min value and string max value are
correct.

8003.571287] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite_size 3145848
[ 8003.571287] String Prerequisites size value exceeded the maximum
number of elements supported or data may be malformed

[ 8003.571288] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[0]
/Pci(0x1F,0x6)/MAC(3
[ 8003.571288] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[1]
24D33F453F,0x0)/IPv6(000
[ 8003.571289] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[2]
:0000:0000:0000:0000:000
[ 8003.571290] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[3]
:0000:0000,0x0,Static,00
[ 8003.571290] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[4]
0:0000:0000:0000:0000:00
[ 8003.571291] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[5]
0:0000:0000,0x40,0000:00
[ 8003.571291] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[6]
0:0000:0000:0000:0000:00
[ 8003.571292] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[7]
0:0000)\r\n\tPciRoot(0x0)/P
[ 8003.571292] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[8]
i(0x0,0x0)/IPv4(0.0.0.0,0x0,DHCP,0.0.0.0,0.0.0.0,
[ 8003.571293] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[9]
.0.0.0)\r\n\tPciRoot(0x0)/P
[ 8003.571293] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[10]
i(0x0,0x0)/IPv6(0000:0000:0000:0000:0000:0000:000
[ 8003.571294] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[11]
:0000,0x0,Static,0000:00
[ 8003.571295] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[12]
0:0000:0000:0000:0000:00
[ 8003.571295] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[13]
0:0000,0x0,0000:0000:000
[ 8003.571296] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[14]
:0000:0000:0000:0000:000
[ 8003.571296] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[15] )
[ 8003.571297] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[16]
[ 8003.571297] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[17]
[ 8003.571298] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[18]
[ 8003.571298] Attribute: HP_Disk0MapForUefiBootOrder Prerequisite[19]

[ 8003.571298] Attribute: HP_Disk0MapForUefiBootOrder Security 0
[ 8003.571299] Attribute: HP_Disk0MapForUefiBootOrder String Min Length 0
[ 8003.571299] Attribute: HP_Disk0MapForUefiBootOrder String Max Length 0

>
> hp_populate_enumeration_package_data() always returns "0".

Earlier in the code, failures reported for a particular attribute
value were ignored because each value is independent from each other.
It is up to the user space application to analyze the raw data read from BIOS.

>
> > In earlier reviews, it was agreed to report a warning that reads
> >
> > /* Report a message and limit prerequisite size to maximum value */
> > pr_warn("Enum Prerequisites size value exceeded the maximum number of
> > elements supported or data may be malformed\n");
> >
> > See lines 370-374
> >
> > > (Affects all attributes)
> > >
> > > > + for (reqs = 0; reqs < size; reqs++) {
> > > > + if (elem >= enum_obj_count) {
> > > > + pr_err("Error enum-objects package is too small\n");
> > > > + return -EINVAL;
> > > > + }
> > > > +
> > > > + ret = hp_convert_hexstr_to_str(enum_obj[elem + reqs].string.pointer,
> > > > + enum_obj[elem + reqs].string.length,
> > > > + &str_value, &value_len);
> > > > +
> > > > + if (ret)
> > > > + return -EINVAL;
> > > > +
> > > > + strscpy(enum_data->common.prerequisites[reqs],
> > > > + str_value,
> > > > + sizeof(enum_data->common.prerequisites[reqs]));
> > > > +
> > > > + kfree(str_value);
> > > > + }
> > > > + break;
> > > > +
> > > > + case SECURITY_LEVEL:
> > > > + enum_data->common.security_level = int_value;
> > > > + break;
> > > > +
> > > > + case ENUM_CURRENT_VALUE:
> > > > + strscpy(enum_data->current_value,
> > > > + str_value, sizeof(enum_data->current_value));
> > > > + break;
> > > > + case ENUM_SIZE:
> > > > + enum_data->possible_values_size = int_value;
> > > > + if (int_value > MAX_VALUES_SIZE)
> > > > + pr_warn("Possible number values size value exceeded the maximum number of elements supported or data may be malformed\n");
> > > > +
> > > > + /*
> > > > + * This HACK is needed to keep the expected
> > > > + * element list pointing to the right obj[elem].type
> > > > + * when the size is zero. POSSIBLE_VALUES
> > > > + * object is omitted by BIOS when the size is zero.
> > > > + */
> > > > + if (int_value == 0)
> > > > + eloc++;
> > > > + break;
> > > > +
> > > > + case ENUM_POSSIBLE_VALUES:
> > > > + size = enum_data->possible_values_size;
> > > > +
> > > > + for (pos_values = 0; pos_values < size && pos_values < MAX_VALUES_SIZE;
> > > > + pos_values++) {
> > > > + if (elem >= enum_obj_count) {
> > > > + pr_err("Error enum-objects package is too small\n");
> > > > + return -EINVAL;
> > > > + }
> > > > +
> > > > + ret = hp_convert_hexstr_to_str(enum_obj[elem + pos_values].string.pointer,
> > > > + enum_obj[elem + pos_values].string.length,
> > > > + &str_value, &value_len);
> > > > +
> > > > + if (ret)
> > > > + return -EINVAL;
> > > > +
> > > > + /*
> > > > + * ignore strings when possible values size
> > > > + * is greater than MAX_VALUES_SIZE
> > > > + */
> > > > + if (size < MAX_VALUES_SIZE)
> > > > + strscpy(enum_data->possible_values[pos_values],
> > > > + str_value,
> > > > + sizeof(enum_data->possible_values[pos_values]));
> > > > + }
> > > > + break;
> > > > + default:
> > > > + pr_warn("Invalid element: %d found in Enumeration attribute or data may be malformed\n", elem);
> > > > + break;
> > > > + }
> > > > +
> > > > + kfree(str_value);
> > > > + }
> > > > +
> > > > +exit_enumeration_package:
> > > > + kfree(str_value);
> > > > + return 0;
> > > > +}
> > > > +
> > > > +/**
> > > > + * hp_populate_enumeration_package_data() -
> > > > + * Populate all properties of an instance under enumeration attribute
> > > > + *
> > > > + * @enum_obj: ACPI object with enumeration data
> > > > + * @instance_id: The instance to enumerate
> > > > + * @attr_name_kobj: The parent kernel object
> > > > + */
> > > > +int hp_populate_enumeration_package_data(union acpi_object *enum_obj,
> > > > + int instance_id,
> > > > + struct kobject *attr_name_kobj)
> > > > +{
> > > > + struct enumeration_data *enum_data = &bioscfg_drv.enumeration_data[instance_id];
> > > > +
> > > > + enum_data->attr_name_kobj = attr_name_kobj;
> > > > +
> > > > + hp_populate_enumeration_elements_from_package(enum_obj,
> > > > + enum_obj->package.count,
> > > > + instance_id);
> > > > + hp_update_attribute_permissions(enum_data->common.is_readonly,
> > > > + &enumeration_current_val);
> > > > + /*
> > > > + * Several attributes have names such "MONDAY". Friendly
> > > > + * user nane is generated to make the name more descriptive
> > > > + */
> > > > + hp_friendly_user_name_update(enum_data->common.path,
> > > > + attr_name_kobj->name,
> > > > + enum_data->common.display_name,
> > > > + sizeof(enum_data->common.display_name));
> > > > + return sysfs_create_group(attr_name_kobj, &enumeration_attr_group);
> > > > +}
> > > > +
> > > > +static int hp_populate_enumeration_elements_from_buffer(u8 *buffer_ptr, u32 *buffer_size,
> > > > + int instance_id)
> > > > +{
> > > > + int reqs;
> > > > + int values;
> > > > + struct enumeration_data *enum_data = &bioscfg_drv.enumeration_data[instance_id];
> > > > +
> > > > + /*
> > > > + * In earlier implementation, reported errors were ignored
> > > > + * causing the data to remain uninitialized. It is for this
> > > > + * reason functions may return an error and no validation
> > > > + * takes place.
> > > > + */
> > >
> > > Where is this error returned?
> >
> > functions such hp_get_string_from_buffer, hp_get_integer_from_buffer
>
> But the errors returned from those functions are just thrown away, no?

Earlier in the code, failures reported for a particular attribute
value were ignored because each value is independent from each other.
It is for this reason, we are throwing away any errors returned from
those functions.

> > >
> > > > +
> > > > + // VALUE:
> > > > + hp_get_string_from_buffer(&buffer_ptr, buffer_size, enum_data->current_value,
> > > > + sizeof(enum_data->current_value));
> > > > +
> > > > + // PATH:
> > > > + hp_get_string_from_buffer(&buffer_ptr, buffer_size, enum_data->common.path,
> > > > + sizeof(enum_data->common.path));
> > > > +
> > > > + // IS_READONLY:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.is_readonly);
> > > > +
> > > > + //DISPLAY_IN_UI:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.display_in_ui);
> > > > +
> > > > + // REQUIRES_PHYSICAL_PRESENCE:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.requires_physical_presence);
> > > > +
> > > > + // SEQUENCE:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.sequence);
> > > > +
> > > > + // PREREQUISITES_SIZE:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.prerequisites_size);
> > > > +
> > > > + if (enum_data->common.prerequisites_size > MAX_PREREQUISITES_SIZE) {
> > > > + /* Report a message and limit prerequisite size to maximum value */
> > > > + pr_warn("Enum Prerequisites size value exceeded the maximum number of elements supported or data may be malformed\n");
> > > > + enum_data->common.prerequisites_size = MAX_PREREQUISITES_SIZE;
> > > > + }
> > > > +
> > > > + // PREREQUISITES:
> > > > + for (reqs = 0; reqs < enum_data->common.prerequisites_size; reqs++)
> > > > + hp_get_string_from_buffer(&buffer_ptr, buffer_size,
> > > > + enum_data->common.prerequisites[reqs],
> > > > + sizeof(enum_data->common.prerequisites[reqs]));
> > > > +
> > > > + // SECURITY_LEVEL:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->common.security_level);
> > >
> > > The reading of all the common elemtns can be extracted into a helper
> > > and reused from all the attributes.
> >
> > Is extracting all common elements into a helper routine absolutely
> > necessary now or can it be refactored after driver is accepted?
>
> It's not necessary.

Thank you
>
> > >
> > > > +
> > > > + // ENUM_CURRENT_VALUE:
> > > > + hp_get_string_from_buffer(&buffer_ptr, buffer_size,
> > > > + enum_data->current_value,
> > > > + sizeof(enum_data->current_value));
> > > > + // ENUM_SIZE:
> > > > + hp_get_integer_from_buffer(&buffer_ptr, buffer_size,
> > > > + &enum_data->possible_values_size);
> > > > +
> > > > + if (enum_data->possible_values_size > MAX_VALUES_SIZE) {
> > > > + /* Report a message and limit possible values size to maximum value */
> > > > + pr_warn("Enum Possible size value exceeded the maximum number of elements supported or data may be malformed\n");
> > > > + enum_data->possible_values_size = MAX_VALUES_SIZE;
> > > > + }
> > > > +
> > > > + // ENUM_POSSIBLE_VALUES:
> > > > +
> > > > + for (values = 0; values < enum_data->possible_values_size; values++)
> > > > + hp_get_string_from_buffer(&buffer_ptr, buffer_size,
> > > > + enum_data->possible_values[values],
> > > > + sizeof(enum_data->possible_values[values]));
> > > > +
> > > > + return 0;
> > > > +}
> > > > +
> > > > +/**
> > > > + * hp_populate_enumeration_buffer_data() -
> > > > + * Populate all properties of an instance under enumeration attribute
> > > > + *
> > > > + * @buffer_ptr: Buffer pointer
> > > > + * @buffer_size: Buffer size
> > > > + * @instance_id: The instance to enumerate
> > > > + * @attr_name_kobj: The parent kernel object
> > > > + */
> > > > +int hp_populate_enumeration_buffer_data(u8 *buffer_ptr, u32 *buffer_size,
> > > > + int instance_id,
> > > > + struct kobject *attr_name_kobj)
> > > > +{
> > > > + struct enumeration_data *enum_data = &bioscfg_drv.enumeration_data[instance_id];
> > > > +
> > > > + enum_data->attr_name_kobj = attr_name_kobj;
> > > > +
> > > > + /* Populate enumeration elements */
> > > > + hp_populate_enumeration_elements_from_buffer(buffer_ptr, buffer_size, instance_id);
> > > > +
> > > > + hp_update_attribute_permissions(enum_data->common.is_readonly,
> > > > + &enumeration_current_val);
> > > > + /*
> > > > + * Several attributes have names such "MONDAY". A Friendlier
> > > > + * user nane is generated to make the name more descriptive
> > > > + */
> > > > + hp_friendly_user_name_update(enum_data->common.path,
> > > > + attr_name_kobj->name,
> > > > + enum_data->common.display_name,
> > > > + sizeof(enum_data->common.display_name));
> > > > +
> > > > + return sysfs_create_group(attr_name_kobj, &enumeration_attr_group);
> > > > +}
> > > > +
> > > > +/**
> > > > + * hp_exit_enumeration_attributes() - Clear all attribute data
> > > > + *
> > > > + * Clears all data allocated for this group of attributes
> > > > + */
> > > > +void hp_exit_enumeration_attributes(void)
> > > > +{
> > > > + int instance_id;
> > > > +
> > > > + for (instance_id = 0; instance_id < bioscfg_drv.enumeration_instances_count;
> > > > + instance_id++) {
> > > > + struct enumeration_data *enum_data = &bioscfg_drv.enumeration_data[instance_id];
> > > > + struct kobject *attr_name_kobj = enum_data->attr_name_kobj;
> > > > +
> > > > + if (attr_name_kobj)
> > > > + sysfs_remove_group(attr_name_kobj, &enumeration_attr_group);
> > > > + }
> > > > + bioscfg_drv.enumeration_instances_count = 0;
> > > > +
> > > > + kfree(bioscfg_drv.enumeration_data);
> > > > + bioscfg_drv.enumeration_data = NULL;
> > > > +}
> > > > --
> > > > 2.34.1
> > > >

Next message: Jim Mattson: "Re: [PATCH] KVM: VMX: restore vmx_vmexit alignment"
Previous message: Jonathan Cameron: "Re: [PATCH v2 3/3] docs: perf: Add new description for HiSilicon UC PMU"
In reply to: Thomas Weißschuh: "Re: [PATCH v15 05/13] hp-bioscfg: enum-attributes"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]