Getting a report (e.g. feature report) from a device requires us to send
a request indicating which report we want to retrieve and then waiting
for the corresponding response containing that report. We already
provide the response structure to the request call, but the request
isn't marked as a request that expects a response. Thus the request
returns before we receive the response and the response buffer indicates
a zero length response due to that.
This essentially means that the get-report calls are broken and will
always indicate that a report of length zero has been read.
Fix this by appropriately marking the request.
Fixes: b05ff1002a5c ("HID: Add support for Surface Aggregator Module HID transport")
Signed-off-by: Maximilian Luz <[email protected]>
---
drivers/hid/surface-hid/surface_hid.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/hid/surface-hid/surface_hid.c b/drivers/hid/surface-hid/surface_hid.c
index 3477b31611ae..a3a70e4f3f6c 100644
--- a/drivers/hid/surface-hid/surface_hid.c
+++ b/drivers/hid/surface-hid/surface_hid.c
@@ -143,7 +143,7 @@ static int ssam_hid_get_raw_report(struct surface_hid_device *shid, u8 rprt_id,
rqst.target_id = shid->uid.target;
rqst.instance_id = shid->uid.instance;
rqst.command_id = SURFACE_HID_CID_GET_FEATURE_REPORT;
- rqst.flags = 0;
+ rqst.flags = SSAM_REQUEST_HAS_RESPONSE;
rqst.length = sizeof(rprt_id);
rqst.payload = &rprt_id;
--
2.31.1
On Tue, 8 Jun 2021, Maximilian Luz wrote:
> Getting a report (e.g. feature report) from a device requires us to send
> a request indicating which report we want to retrieve and then waiting
> for the corresponding response containing that report. We already
> provide the response structure to the request call, but the request
> isn't marked as a request that expects a response. Thus the request
> returns before we receive the response and the response buffer indicates
> a zero length response due to that.
>
> This essentially means that the get-report calls are broken and will
> always indicate that a report of length zero has been read.
>
> Fix this by appropriately marking the request.
>
> Fixes: b05ff1002a5c ("HID: Add support for Surface Aggregator Module HID transport")
> Signed-off-by: Maximilian Luz <[email protected]>
> ---
> drivers/hid/surface-hid/surface_hid.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/hid/surface-hid/surface_hid.c b/drivers/hid/surface-hid/surface_hid.c
> index 3477b31611ae..a3a70e4f3f6c 100644
> --- a/drivers/hid/surface-hid/surface_hid.c
> +++ b/drivers/hid/surface-hid/surface_hid.c
> @@ -143,7 +143,7 @@ static int ssam_hid_get_raw_report(struct surface_hid_device *shid, u8 rprt_id,
> rqst.target_id = shid->uid.target;
> rqst.instance_id = shid->uid.instance;
> rqst.command_id = SURFACE_HID_CID_GET_FEATURE_REPORT;
> - rqst.flags = 0;
> + rqst.flags = SSAM_REQUEST_HAS_RESPONSE;
> rqst.length = sizeof(rprt_id);
> rqst.payload = &rprt_id;
Queued in for-5.13/upstream-fixes branch, thanks.
--
Jiri Kosina
SUSE Labs