2024-05-22 21:37:10

by Armin Wolf

[permalink] [raw]
Subject: [PATCH 1/2] ACPI: EC: Abort address space access upon error

Then a multi-byte address space access is requested, acpi_ec_read()/
acpi_ec_write() is being called multiple times.

Abort such operations if a single call to acpi_ec_read()/
acpi_ec_write() fails, as the data read from/written to the EC
might be incomplete.

Signed-off-by: Armin Wolf <[email protected]>
---
drivers/acpi/ec.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
index e7793ee9e649..a68dce2147a4 100644
--- a/drivers/acpi/ec.c
+++ b/drivers/acpi/ec.c
@@ -1333,10 +1333,13 @@ acpi_ec_space_handler(u32 function, acpi_physical_address address,
if (ec->busy_polling || bits > 8)
acpi_ec_burst_enable(ec);

- for (i = 0; i < bytes; ++i, ++address, ++value)
+ for (i = 0; i < bytes; ++i, ++address, ++value) {
result = (function == ACPI_READ) ?
acpi_ec_read(ec, address, value) :
acpi_ec_write(ec, address, *value);
+ if (result < 0)
+ break;
+ }

if (ec->busy_polling || bits > 8)
acpi_ec_burst_disable(ec);
--
2.39.2



2024-05-22 21:37:21

by Armin Wolf

[permalink] [raw]
Subject: [PATCH 2/2] ACPI: EC: Avoid returning AE_OK upon unknown error

If an error code other than EINVAL, ENODEV or ETIME is returned
by acpi_ec_read()/acpi_ec_write(), then AE_OK is wrongly returned.

Fix this by only returning AE_OK if the return code is 0, and
return AE_ERROR otherwise.

Signed-off-by: Armin Wolf <[email protected]>
---
drivers/acpi/ec.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
index a68dce2147a4..68dd17f96f63 100644
--- a/drivers/acpi/ec.c
+++ b/drivers/acpi/ec.c
@@ -1351,8 +1351,10 @@ acpi_ec_space_handler(u32 function, acpi_physical_address address,
return AE_NOT_FOUND;
case -ETIME:
return AE_TIME;
- default:
+ case 0:
return AE_OK;
+ default:
+ return AE_ERROR;
}
}

--
2.39.2


2024-05-27 10:45:45

by Rafael J. Wysocki

[permalink] [raw]
Subject: Re: [PATCH 2/2] ACPI: EC: Avoid returning AE_OK upon unknown error

On Wed, May 22, 2024 at 11:37 PM Armin Wolf <[email protected]> wrote:
>
> If an error code other than EINVAL, ENODEV or ETIME is returned
> by acpi_ec_read()/acpi_ec_write(), then AE_OK is wrongly returned.
>
> Fix this by only returning AE_OK if the return code is 0, and
> return AE_ERROR otherwise.
>
> Signed-off-by: Armin Wolf <[email protected]>
> ---
> drivers/acpi/ec.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/acpi/ec.c b/drivers/acpi/ec.c
> index a68dce2147a4..68dd17f96f63 100644
> --- a/drivers/acpi/ec.c
> +++ b/drivers/acpi/ec.c
> @@ -1351,8 +1351,10 @@ acpi_ec_space_handler(u32 function, acpi_physical_address address,
> return AE_NOT_FOUND;
> case -ETIME:
> return AE_TIME;
> - default:
> + case 0:
> return AE_OK;
> + default:
> + return AE_ERROR;
> }
> }
>
> --

Applied (with some edits in the subject and changelog) along with the
[1/2] as 6.10-rc material, thanks!