kernel/linux.git/include/linux/intel-ish-client-if.h, branch v6.18.21

HID: intel-ish-hid: Use dedicated unbound workqueues to prevent resume blocking

2026-01-23T10:21:36+00:00

commit 0d30dae38fe01cd1de358c6039a0b1184689fe51 upstream. During suspend/resume tests with S2IDLE, some ISH functional failures were observed because of delay in executing ISH resume handler. Here schedule_work() is used from resume handler to do actual work. schedule_work() uses system_wq, which is a per CPU work queue. Although the queuing is not bound to a CPU, but it prefers local CPU of the caller, unless prohibited. Users of this work queue are not supposed to queue long running work. But in practice, there are scenarios where long running work items are queued on other unbound workqueues, occupying the CPU. As a result, the ISH resume handler may not get a chance to execute in a timely manner. In one scenario, one of the ish_resume_handler() executions was delayed nearly 1 second because another work item on an unbound workqueue occupied the same CPU. This delay causes ISH functionality failures. A similar issue was previously observed where the ISH HID driver timed out while getting the HID descriptor during S4 resume in the recovery kernel, likely caused by the same workqueue contention problem. Create dedicated unbound workqueues for all ISH operations to allow work items to execute on any available CPU, eliminating CPU-specific bottlenecks and improving resume reliability under varying system loads. Also ISH has three different components, a bus driver which implements ISH protocols, a PCI interface layer and HID interface. Use one dedicated work queue for all of them. Signed-off-by: Zhang Lixu Signed-off-by: Jiri Kosina Signed-off-by: Greg Kroah-Hartman

HID: intel-ish-hid: Remove unused ishtp_cl_tx_empty

2025-01-09T09:09:11+00:00

ishtp_cl_tx_empty() was added in 2018 by commit a1c40ce62fd2 ("HID: intel-ish-hid: ishtp: add helper functions for client buffer operation") but has remained unused. Remove it. Signed-off-by: Dr. David Alan Gilbert Acked-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: Intel-ish-hid: Ishtp: Add helper functions for client connection

2023-12-06T10:33:45+00:00

For every ishtp client driver during initialization state, the flow is: 1 - Allocate an ISHTP client instance 2 - Reserve a host id and link the client instance 3 - Search a firmware client using UUID and get related client information 4 - Bind firmware client id to the ISHTP client instance 5 - Set the state the ISHTP client instance to CONNECTING 6 - Send connect request to firmware 7 - Register event callback for messages from the firmware During deinitizalization state, the flow is: 9 - Set the state the ISHTP client instance to ISHTP_CL_DISCONNECTING 10 - Issue disconnect request to firmware 11 - Unlike the client instance 12 - Flush message queue 13 - Free ISHTP client instance Step 2-7 are identical to the steps of client driver initialization and driver reset flow, but reallocation of the RX/TX ring buffers can be avoided in reset flow. Also for step 9-12, they are identical to the steps of client driver failure handling after connect request, driver reset flow and driver removing. So, add two helper functions to simplify client driver code. ishtp_cl_establish_connection() ishtp_cl_destroy_connection() No functional changes are expected. Signed-off-by: Even Xu Acked-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: intel-ish-hid: fix module device-id handling

2021-11-11T21:45:40+00:00

A late addititon to the intel-ish-hid framework caused a build failure with clang, and introduced an ABI to the module loader that stops working if any driver ever needs to bind to more than one UUID: drivers/hid/intel-ish-hid/ishtp-fw-loader.c:1067:4: error: initializer element is not a compile-time constant Change the ishtp_device_id to have correct documentation and a driver_data field like all the other ones, and change the drivers to use the ID table as the primary identification in a way that works with all compilers and avoids duplciating the identifiers. Fixes: f155dfeaa4ee ("platform/x86: isthp_eclite: only load for matching devices") Fixes: facfe0a4fdce ("platform/chrome: chros_ec_ishtp: only load for matching devices") Fixes: 0d0cccc0fd83 ("HID: intel-ish-hid: hid-client: only load for matching devices") Fixes: 44e2a58cb880 ("HID: intel-ish-hid: fw-loader: only load for matching devices") Fixes: cb1a2c6847f7 ("HID: intel-ish-hid: use constants for modaliases") Fixes: fa443bc3c1e4 ("HID: intel-ish-hid: add support for MODULE_DEVICE_TABLE()") Signed-off-by: Arnd Bergmann Reviewed-by: Hans de Goede [jkosina@suse.cz: fix ecl_ishtp_cl_driver.id initialization] [jkosina@suse.cz: fix conflict with already fixed kerneldoc] Signed-off-by: Jiri Kosina

HID: intel-ish-hid: use async resume function

2021-07-15T18:49:09+00:00

ISH IPC driver uses asynchronous workqueue to do resume now, but there is a potential timing issue: when child devices resume before bus driver, it will cause child devices resume failed and cannot be recovered until reboot. The current implementation in this case do wait for IPC to resume but fail to accommodate for a case when there is no ISH reboot and soft resume is taking time. This issue is apparent on Tiger Lake platform with 5.11.13 kernel when doing suspend to idle then resume(s0ix) test. To resolve this issue, we change ISHTP HID client to use asynchronous resume callback too. In the asynchronous resume callback, it waits for the ISHTP resume done event, and then notify ISHTP HID client link ready. Signed-off-by: Ye Xiang Acked-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: ishtp-hid-client: Fix 'suggest-attribute=format' compiler warning

2021-04-07T15:56:34+00:00

Fixes the following W=1 kernel build warning(s): drivers/hid/intel-ish-hid/ishtp/bus.c: In function ‘ishtp_trace_callback’: drivers/hid/intel-ish-hid/ishtp/bus.c:876:29: warning: return type might be a candidate for a format attribute [-Wsuggest-attribute=format] 876 | return cl_device->ishtp_dev->print_log; | ~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~ Cc: Srinivas Pandruvada Cc: Jiri Kosina Cc: Benjamin Tissoires Cc: Daniel Drubin Cc: linux-input@vger.kernel.org Suggested-by: Arnd Bergmann Signed-off-by: Lee Jones Acked-by: Srinivas Pandruvada Signed-off-by: Benjamin Tissoires

HID: intel-ish-hid: Make remove callback return void

2021-03-08T16:16:03+00:00

The driver core ignores the return value of struct bus_type::remove() because there is only little that can be done. To simplify the quest to make this function return void, let struct ishtp_cl_driver::remove() return void, too. All users already unconditionally return 0, this commit makes it obvious that returning an error value is a bad idea. Signed-off-by: Uwe Kleine-König Acked-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: intel-ish-hid: fix wrong driver_data usage

2019-06-26T12:08:11+00:00

Currently, in suspend() and resume(), ishtp client drivers are using driver_data to get "struct ishtp_cl_device" object which is set by bus driver. It's wrong since the driver_data should not be owned bus. driver_data should be owned by the corresponding ishtp client driver. Due to this, some ishtp client driver like cros_ec_ishtp which uses its driver_data to transfer its data to its child doesn't work correctly. So this patch removes setting driver_data in bus drier and instead of using driver_data to get "struct ishtp_cl_device", since "struct device" is embedded in "struct ishtp_cl_device", we introduce a helper function that returns "struct ishtp_cl_device" from "struct device". Signed-off-by: Hyungwoo Yang Acked-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: intel-ish-hid: Add interface function for PCI device pointer

2019-03-19T10:57:24+00:00

Instead of directly accessing PCI device poitner via struct ishtp_cl, create interface function for same. This is required for DMA transfer. Signed-off-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina

HID: intel-ish-hid: Move functions related to bus and device

2019-03-19T10:57:23+00:00

Move function idefinitions related to bus and device to common header file. Also create new function to get fw client id and move ish_hw_reset() from inline to exported function. Signed-off-by: Srinivas Pandruvada Signed-off-by: Jiri Kosina