6.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Andy Shevchenko andriy.shevchenko@linux.intel.com
[ Upstream commit 046ece773cc77ef5d2a1431b188ac3d0840ed150 ]
In accordance with ACPI specificication and _DSD data buffer representation the data there is an array of bytes. Hence, accessing it with something longer will create a sparse data which is against of how device property APIs work in general and also not defined in the ACPI specification (see [1]). Fix the code to emit an error if non-byte accessor is used to retrieve _DSD buffer data.
Fixes: 369af6bf2c28 ("ACPI: property: Read buffer properties as integers") Link: https://uefi.org/specs/ACPI/6.5/19_ASL_Reference.html#buffer-declare-buffer-... # [1] Signed-off-by: Andy Shevchenko andriy.shevchenko@linux.intel.com [ rjw: Add missing braces ] Signed-off-by: Rafael J. Wysocki rafael.j.wysocki@intel.com Signed-off-by: Sasha Levin sashal@kernel.org --- drivers/acpi/property.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-)
diff --git a/drivers/acpi/property.c b/drivers/acpi/property.c index b8d9eb9a433ed..0565c18c2ee31 100644 --- a/drivers/acpi/property.c +++ b/drivers/acpi/property.c @@ -1114,25 +1114,26 @@ static int acpi_data_prop_read(const struct acpi_device_data *data, switch (proptype) { case DEV_PROP_STRING: break; - case DEV_PROP_U8 ... DEV_PROP_U64: + default: if (obj->type == ACPI_TYPE_BUFFER) { if (nval > obj->buffer.length) return -EOVERFLOW; - break; + } else { + if (nval > obj->package.count) + return -EOVERFLOW; } - fallthrough; - default: - if (nval > obj->package.count) - return -EOVERFLOW; break; } if (nval == 0) return -EINVAL;
- if (obj->type != ACPI_TYPE_BUFFER) - items = obj->package.elements; - else + if (obj->type == ACPI_TYPE_BUFFER) { + if (proptype != DEV_PROP_U8) + return -EPROTO; items = obj; + } else { + items = obj->package.elements; + }
switch (proptype) { case DEV_PROP_U8: