On 6/21/2023 1:09 AM, Mario Limonciello wrote:
> A crash was reported in amd-sfh related to hid core initialization
> before SFH initialization has run.
>
> ```
> amdtp_hid_request+0x36/0x50 [amd_sfh
> 2e3095779aada9fdb1764f08ca578ccb14e41fe4]
> sensor_hub_get_feature+0xad/0x170 [hid_sensor_hub
> d6157999c9d260a1bfa6f27d4a0dc2c3e2c5654e]
> hid_sensor_parse_common_attributes+0x217/0x310 [hid_sensor_iio_common
> 07a7935272aa9c7a28193b574580b3e953a64ec4]
> hid_gyro_3d_probe+0x7f/0x2e0 [hid_sensor_gyro_3d
> 9f2eb51294a1f0c0315b365f335617cbaef01eab]
> platform_probe+0x44/0xa0
> really_probe+0x19e/0x3e0
> ```
>
> Ensure that sensors have been set up before calling into
> amd_sfh_get_report() or amd_sfh_set_report().
>
> Cc: stable@xxxxxxxxxxxxxxx
> Cc: Linux regression tracking (Thorsten Leemhuis) <regressions@xxxxxxxxxxxxx>
> Fixes: 7bcfdab3f0c6 ("HID: amd_sfh: if no sensors are enabled, clean up")
> Reported-by: Haochen Tong <linux@xxxxxxxxxxxx>
> Link: https://lore.kernel.org/all/3250319.ancTxkQ2z5@zen/T/
> Signed-off-by: Mario Limonciello <mario.limonciello@xxxxxxx>
> ---
> drivers/hid/amd-sfh-hid/amd_sfh_client.c | 6 ++++++
> 1 file changed, 6 insertions(+)
>
> diff --git a/drivers/hid/amd-sfh-hid/amd_sfh_client.c b/drivers/hid/amd-sfh-hid/amd_sfh_client.c
> index d9b7b01900b5..88f3d913eaa1 100644
> --- a/drivers/hid/amd-sfh-hid/amd_sfh_client.c
> +++ b/drivers/hid/amd-sfh-hid/amd_sfh_client.c
> @@ -25,6 +25,9 @@ void amd_sfh_set_report(struct hid_device *hid, int report_id,
> struct amdtp_cl_data *cli_data = hid_data->cli_data;
> int i;
>
> + if (!cli_data->is_any_sensor_enabled)
> + return;
> +
can we check below patch series which solves this issue by initializing HID only
if is_any_sensor_enabled
https://lore.kernel.org/all/nycvar.YFH.7.76.2305231559000.29760@xxxxxxxxxxxxx/
> for (i = 0; i < cli_data->num_hid_devices; i++) {
> if (cli_data->hid_sensor_hubs[i] == hid) {
> cli_data->cur_hid_dev = i;
> @@ -41,6 +44,9 @@ int amd_sfh_get_report(struct hid_device *hid, int report_id, int report_type)
> struct request_list *req_list = &cli_data->req_list;
> int i;
>
> + if (!cli_data->is_any_sensor_enabled)
> + return -ENODEV;
> +
can we check below patch series which solves this issue by initializing HID only
if is_any_sensor_enabled.
https://lore.kernel.org/all/nycvar.YFH.7.76.2305231559000.29760@xxxxxxxxxxxxx/
Thanks,
--
Basavaraj
> for (i = 0; i < cli_data->num_hid_devices; i++) {
> if (cli_data->hid_sensor_hubs[i] == hid) {
> struct request_list *new = kzalloc(sizeof(*new), GFP_KERNEL);
