Hi,
On 11/9/22 21:52, Jorge Lopez wrote:
> On Wed, Nov 9, 2022 at 2:05 PM Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
>>
>> Hi,
>>
>> On 11/9/22 21:00, Jorge Lopez wrote:
>>> Hi Hans,
>>>
>>> On Wed, Nov 9, 2022 at 12:10 PM Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
>>>>
>>>> Hi,
>>>>
>>>> On 11/9/22 18:24, Jorge Lopez wrote:
>>>>> HI Hans,
>>>>>
>>>>> Please see questions and comments below.
>>>>>
>>>>> On Tue, Nov 8, 2022 at 8:51 AM Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
>>>>>>
>>>>>> Hi Jorge,
>>>>>>
>>>>>> Review comments inline.
>>>>>>
>>>>>> On 10/20/22 22:10, Jorge Lopez wrote:
>>>>>>> The purpose for this patch is submit HP BIOSCFG driver to be list of
>>>>>>> HP Linux kernel drivers. The driver include a total of 12 files
>>>>>>> broken in several patches. This is set 1 of 4.
>>>>>>>
>>>>>>> HP BIOS Configuration driver purpose is to provide a driver supporting
>>>>>>> the latest sysfs class firmware attributes framework allowing the user
>>>>>>> to change BIOS settings and security solutions on HP Inc.’s commercial
>>>>>>> notebooks.
>>>>>>>
>>>>>>> Many features of HP Commercial PC’s can be managed using Windows
>>>>>>> Management Instrumentation (WMI). WMI is an implementation of Web-Based
>>>>>>> Enterprise Management (WBEM) that provides a standards-based interface
>>>>>>> for changing and monitoring system settings. HP BISOCFG driver provides
>>>>>>> a native Linux solution and the exposed features facilitates the
>>>>>>> migration to Linux environments.
>>>>>>>
>>>>>>> The Linux security features to be provided in hp-bioscfg driver enables
>>>>>>> managing the BIOS settings and security solutions via sysfs, a virtual
>>>>>>> filesystem that can be used by user-mode applications. The new
>>>>>>> documentation cover features such Secure Platform Management, Sure
>>>>>>> Admin, and Sure Start. Each section provides security feature
>>>>>>> description and identifies sysfs directories and files exposed by
>>>>>>> the driver.
>>>>>>>
>>>>>>> Many HP Commercial PC’s include a feature called Secure Platform
>>>>>>> Management (SPM), which replaces older password-based BIOS settings
>>>>>>> management with public key cryptography. PC secure product management
>>>>>>> begins when a target system is provisioned with cryptographic keys
>>>>>>> that are used to ensure the integrity of communications between system
>>>>>>> management utilities and the BIOS.
>>>>>>>
>>>>>>> HP Commercial PC’s have several BIOS settings that control its behaviour
>>>>>>> and capabilities, many of which are related to security. To prevent
>>>>>>> unauthorized changes to these settings, the system can be configured
>>>>>>> to use a Sure Admin cryptographic signature-based authorization string
>>>>>>> that the BIOS will use to verify authorization to modify the setting.
>>>>>>>
>>>>>>> Signed-off-by: Jorge Lopez <jorge.lopez2@xxxxxx>
>>>>>>>
>>>>>>> ---
>>>>>>> Based on the latest platform-drivers-x86.git/for-next
>>>>>>> ---
>>>>>>> .../x86/hp/hp-bioscfg/biosattr-interface.c | 285 ++++++++
>>>>>>> drivers/platform/x86/hp/hp-bioscfg/bioscfg.h | 671 ++++++++++++++++++
>>>>>>> .../x86/hp/hp-bioscfg/enum-attributes.c | 521 ++++++++++++++
>>>>>>> 3 files changed, 1477 insertions(+)
>>>>>>> create mode 100644 drivers/platform/x86/hp/hp-bioscfg/biosattr-interface.c
>>>>>>> create mode 100644 drivers/platform/x86/hp/hp-bioscfg/bioscfg.h
>>>>>>> create mode 100644 drivers/platform/x86/hp/hp-bioscfg/enum-attributes.c
>>>>>>>
>>>>>>> diff --git a/drivers/platform/x86/hp/hp-bioscfg/biosattr-interface.c b/drivers/platform/x86/hp/hp-bioscfg/biosattr-interface.c
>>>>>>> new file mode 100644
>>>>>>> index 000000000000..f0c919bf3ab0
>>>>>>> --- /dev/null
>>>>>>> +++ b/drivers/platform/x86/hp/hp-bioscfg/biosattr-interface.c
>>>>>>> @@ -0,0 +1,285 @@
>>>>>>> +// SPDX-License-Identifier: GPL-2.0
>>>>>>> +/*
>>>>>>> + * Functions corresponding to methods under BIOS interface GUID
>>>>>>> + * for use with hp-bioscfg driver.
>>>>>>> + *
>>>>>>> + * Copyright (c) 2022 Hewlett-Packard Inc.
>>>>>>> + */
>>>>>>> +
>>>>>>> +#include <linux/wmi.h>
>>>>>>> +#include "bioscfg.h"
>>>>>>> +
>>>>>>> +#define SET_DEFAULT_VALUES_METHOD_ID 0x02
>>>>>>> +#define SET_BIOS_DEFAULTS_METHOD_ID 0x03
>>>>>>> +#define SET_ATTRIBUTE_METHOD_ID 0x04
>>>>>>> +
>>>>>>> +/*
>>>>>>> + * set_attribute() - Update an attribute value
>>>>>>> + * @a_name: The attribute name
>>>>>>> + * @a_value: The attribute value
>>>>>>> + *
>>>>>>> + * Sets an attribute to new value
>>>>>>> + */
>>>>>>> +int hp_set_attribute(const char *a_name, const char *a_value)
>>>>>>> +{
>>>>>>> + size_t security_area_size;
>>>>>>> + size_t a_name_size, a_value_size;
>>>>>>> + u16 *buffer = NULL;
>>>>>>> + u16 *start = NULL;
>>>>>>> + int buffer_size;
>>>>>>
>>>>>> You have 2 spaces between int and buffer_size here, please drop
>>>>>> one.
>>>>>>
>>>>>>
>>>>>>> + int ret;
>>>>>>> + int instance;
>>>>>>> + char *auth_empty_value = " ";
>>>>>>> +
>>>>>>> + mutex_lock(&bioscfg_drv.mutex);
>>>>>>> + if (!bioscfg_drv.bios_attr_wdev) {
>>>>>>> + ret = -ENODEV;
>>>>>>> + goto out_set_attribute;
>>>>>>> + }
>>>>>>> +
>>>>>>> + instance = get_password_instance_for_type(SETUP_PASSWD);
>>>>>>> + if (instance < 0)
>>>>>>> + goto out_set_attribute;
>>>>>>> +
>>>>>>> + if (strlen(bioscfg_drv.password_data[instance].current_password) == 0)
>>>>>>> + strncpy(bioscfg_drv.password_data[instance].current_password,
>>>>>>> + auth_empty_value,
>>>>>>> + sizeof(bioscfg_drv.password_data[instance].current_password));
>>>>>>
>>>>>> strncpy does not guarantee 0 termination of the destination buffer,
>>>>>> please use strscpy.
>>>>>>
>>>>>>> +
>>>>>>> + a_name_size = calculate_string_buffer(a_name);
>>>>>>> + a_value_size = calculate_string_buffer(a_value);
>>>>>>> + security_area_size = calculate_security_buffer(bioscfg_drv.password_data[instance].current_password);
>>>>>>> + buffer_size = a_name_size + a_value_size + security_area_size;
>>>>>>> +
>>>>>>> + buffer = kzalloc(buffer_size, GFP_KERNEL);
>>>>>>> + if (!buffer) {
>>>>>>> + ret = -ENOMEM;
>>>>>>> + goto out_set_attribute;
>>>>>>> + }
>>>>>>> +
>>>>>>> + /* build variables to set */
>>>>>>> + start = buffer;
>>>>>>> + start = ascii_to_utf16_unicode(start, a_name);
>>>>>>> + if (!start)
>>>>>>> + goto out_set_attribute;
>>>>>>> +
>>>>>>> + start = ascii_to_utf16_unicode(start, a_value);
>>>>>>> + if (!start)
>>>>>>> + goto out_set_attribute;
>>>>>>> +
>>>>>>> + populate_security_buffer(start, bioscfg_drv.password_data[instance].current_password);
>>>>>>> + ret = hp_wmi_set_bios_setting(buffer, buffer_size);
>>>>>>> +
>>>>>>> +
>>>>>>> +out_set_attribute:
>>>>>>> + kfree(buffer);
>>>>>>> + mutex_unlock(&bioscfg_drv.mutex);
>>>>>>> + return ret;
>>>>>>> +}
>>>>>>> +
>>>>>>> +/*
>>>>>>> + * hp_wmi_perform_query
>>>>>>> + *
>>>>>>> + * query: The commandtype (enum hp_wmi_commandtype)
>>>>>>> + * write: The command (enum hp_wmi_command)
>>>>>>> + * buffer: Buffer used as input and/or output
>>>>>>> + * insize: Size of input buffer
>>>>>>> + * outsize: Size of output buffer
>>>>>>> + *
>>>>>>> + * returns zero on success
>>>>>>> + * an HP WMI query specific error code (which is positive)
>>>>>>> + * -EINVAL if the query was not successful at all
>>>>>>> + * -EINVAL if the output buffer size exceeds buffersize
>>>>>>> + *
>>>>>>> + * Note: The buffersize must at least be the maximum of the input and output
>>>>>>> + * size. E.g. Battery info query is defined to have 1 byte input
>>>>>>> + * and 128 byte output. The caller would do:
>>>>>>> + * buffer = kzalloc(128, GFP_KERNEL);
>>>>>>> + * ret = hp_wmi_perform_query(HPWMI_BATTERY_QUERY, HPWMI_READ, buffer, 1, 128)
>>>>>>> + */
>>>>>>> +int hp_wmi_perform_query(int query, enum hp_wmi_command command, void *buffer, int insize, int outsize)
>>>>>>> +{
>>>>>>> + struct acpi_buffer input, output = { ACPI_ALLOCATE_BUFFER, NULL };
>>>>>>> + struct bios_return *bios_return;
>>>>>>> + union acpi_object *obj = NULL;
>>>>>>> + struct bios_args *args = NULL;
>>>>>>> + int mid, actual_insize, actual_outsize;
>>>>>>> + size_t bios_args_size;
>>>>>>> + int ret;
>>>>>>> +
>>>>>>> + mid = encode_outsize_for_pvsz(outsize);
>>>>>>> + if (WARN_ON(mid < 0))
>>>>>>> + return mid;
>>>>>>> +
>>>>>>> + actual_insize = insize;
>>>>>>> + bios_args_size = struct_size(args, data, insize);
>>>>>>> + args = kmalloc(bios_args_size, GFP_KERNEL);
>>>>>>> + if (!args)
>>>>>>> + return -ENOMEM;
>>>>>>> +
>>>>>>> + input.length = bios_args_size;
>>>>>>> + input.pointer = args;
>>>>>>> +
>>>>>>> + args->signature = 0x55434553;
>>>>>>> + args->command = command;
>>>>>>> + args->commandtype = query;
>>>>>>> + args->datasize = insize;
>>>>>>> + memcpy(args->data, buffer, flex_array_size(args, data, insize));
>>>>>>> +
>>>>>>> + ret = wmi_evaluate_method(HP_WMI_BIOS_GUID, 0, mid, &input, &output);
>>>>>>> + bioscfg_drv.last_wmi_status = ret;
>>>>>>> + if (ret)
>>>>>>> + goto out_free;
>>>>>>> +
>>>>>>> + obj = output.pointer;
>>>>>>> + if (!obj) {
>>>>>>> + ret = -EINVAL;
>>>>>>> + goto out_free;
>>>>>>> + }
>>>>>>> +
>>>>>>
>>>>>> You need to check the type of obj here before dereferencing
>>>>>> obj as if it is a buffer.
>>>>>>
>>>>>>> + bios_return = (struct bios_return *)obj->buffer.pointer;
>>>>>>> + ret = bios_return->return_code;
>>>>>>> + bioscfg_drv.last_wmi_status = ret;
>>>>>>> + if (ret) {
>>>>>>> + if (ret != HPWMI_RET_UNKNOWN_COMMAND &&
>>>>>>> + ret != HPWMI_RET_UNKNOWN_CMDTYPE)
>>>>>>> + pr_warn("query 0x%x returned error 0x%x\n", query, ret);
>>>>>>> + goto out_free;
>>>>>>> + }
>>>>>>> +
>>>>>>> + /* Ignore output data of zero size */
>>>>>>> + if (!outsize)
>>>>>>> + goto out_free;
>>>>>>> +
>>>>>>> + actual_outsize = min(outsize, (int)(obj->buffer.length - sizeof(*bios_return)));
>>>>>>> + memcpy(buffer, obj->buffer.pointer + sizeof(*bios_return), actual_outsize);
>>>>>>> + memset(buffer + actual_outsize, 0, outsize - actual_outsize);
>>>>>>> +
>>>>>>> +out_free:
>>>>>>> + kfree(obj);
>>>>>>> + kfree(args);
>>>>>>> + return ret;
>>>>>>> +}
>>>>>>> +
>>>>>>> +/*
>>>>>>> + * ascii_to_utf16_unicode - Convert ascii string to UTF-16 unicode
>>>>>>> + *
>>>>>>> + * @p: Unicode buffer address
>>>>>>> + * @str: string to convert to unicode
>>>>>>> + *
>>>>>>> + * Returns a void pointer to the buffer containing unicode string
>>>>>>> + */
>>>>>>> +void *ascii_to_utf16_unicode(u16 *p, const u8 *str)
>>>>>>> +{
>>>>>>> + int len = strlen(str);
>>>>>>> + int ret;
>>>>>>> +
>>>>>>> + /*
>>>>>>> + * Add null character when reading an empty string
>>>>>>> + */
>>>>>>> + if (len == 0) {
>>>>>>> + *p++ = 2;
>>>>>>> + *p++ = (u8)0x00;
>>>>>>> + return p;
>>>>>>
>>>>>> This does not match with calculate_string_buffer() which will
>>>>>> return 2 for a 0 length string while you are using 4 bytes here.
>>>>>>
>>>>>> I guess this may also be why you need to use " " for
>>>>>> auth_empty_value above, so as to avoid this bug.
>>>>>>
>>>>> HP BIOS expects 2 characters when an empty string is being converted
>>>>> to u16 hence the reason for returning 2 instead of zero. This is an
>>>>> intended behavior and needed when allocating a buffer and writing to
>>>>> BIOS.
>>>>
>>>> Right I understand that, it wants 2 characters for the 16 bit length
>>>> word, but why not write 0 to that 16 bit length word. Why actually
>>>> say the string-buffer length is 2 bytes long / and then write
>>>> a 16-bit word with value 0?
>>>>
>>>> What you are doing now creates a 4 byte buffer like this:
>>>>
>>>> u8 buf[4] = { 0x02, 0x00, 0x00, 0x00 }
>>>>
>>>> Why not just create a 2 byte buffer like this:
>>>>
>>>> u8 buf[2] = { 0x00, 0x00 }
>>>>
>>>> ?
>>>>
>>>>
>>>> Also I'm wondering why the empty auth string is " " and
>>>> not "" ?
>>>>
>>>>
>>>
>>> The string returned for an empty string is 4 bytes. The returned
>>> string includes two bytes for the string size in bytes and the
>>> remaining bytes are the string.
>>> Size = 0x02, 0x00
>>> String = 0x00, 0x00
>>>
>>> All strings return include the string size in bytes followed by the u16 string
>>
>> Right I understand that, but why is the "String = 0x00, 0x00"
>> there ? All the non-0-length strings are not 0 terminated,
>> why does the zero length string needs to be specified as length 2
>> (1 u16) and then have that u16 be a 0 terminator ?
>>
> It is a specific format required by BIOS for any zero-length strings.
> I don't know the reason why BIOS wants that format.
>
>
>> Have you tried just using Size[2] = 0x00, 0x00 and String[0] for
>> an empty string?
>>
>
> I tried during the development process and each time the data is
> rejected by BIOS.
Ok, well in that case you are going to need your own
ascii_to_utf16_unicode() to handle the weird case for the
0 sized string, so no need to move the dell-wmi-sysman versions
to the shared wmi code.
But please do start with a copy of the Dell function and then
add the special case for the 0 len string, since the original
version above does not properly handle errors.
Also this means you need to update the hp calculate_string_buffer()
copy to properly return 4 rather then 2 for the bufsize for
a 0 length string.
Regards,
Hans
