Message-ID: <CAO-hwJJmVA7fcPBUpVZqKQmaJfk1AMqdh2-ZWXBVWtG2KgHyDw@mail.gmail.com>
Date:   Fri, 10 Feb 2023 10:21:29 +0100
From:   Benjamin Tissoires <benjamin.tissoires@...hat.com>
To:     Aditya Garg <gargaditya08@...e.com>
Cc:     Thomas Weißschuh <thomas@...ch.de>,
        Jiri Kosina <jikos@...nel.org>,
        "jkosina@...e.cz" <jkosina@...e.cz>,
        Andy Shevchenko <andy@...radead.org>,
        "andy.shevchenko@...il.com" <andy.shevchenko@...il.com>,
        LKML <linux-kernel@...r.kernel.org>,
        "linux-input@...r.kernel.org" <linux-input@...r.kernel.org>,
        "ronald@...ovation.ch" <ronald@...ovation.ch>,
        "kekrby@...il.com" <kekrby@...il.com>,
        Orlando Chamberlain <orlandoch.dev@...il.com>
Subject: Re: [PATCH 1/3] HID: apple-ibridge: Add Apple iBridge HID driver for
 T1 chip.

On Fri, Feb 10, 2023 at 9:54 AM Aditya Garg <gargaditya08@...e.com> wrote:
>
>
>
> > On 10-Feb-2023, at 2:09 PM, Benjamin Tissoires <benjamin.tissoires@...hat.com> wrote:
> >
> > On Fri, Feb 10, 2023 at 9:30 AM Aditya Garg <gargaditya08@...e.com> wrote:
> >>
> >>
> >>
> >>> On 10-Feb-2023, at 10:26 AM, Thomas Weißschuh <thomas@...ch.de> wrote:
> >>>
> >>> Hi,
> >>>
> >>> some comments inline.
> >>>
> >>> On Fri, Feb 10, 2023 at 03:43:24AM +0000, Aditya Garg wrote:
> >>>> From: Ronald Tschalär <ronald@...ovation.ch>
> >>>>
> >>>> The iBridge device provides access to several devices, including:
> >>>> - the Touch Bar
> >>>> - the iSight webcam
> >>>> - the light sensor
> >>>> - the fingerprint sensor
> >>>>
> >>>> This driver provides the core support for managing the iBridge device
> >>>> and the access to the underlying devices. In particular, the
> >>>> functionality for the touch bar and light sensor is exposed via USB HID
> >>>> interfaces, and on devices with the T1 chip one of the HID devices is
> >>>> used for both functions. So this driver creates virtual HID devices, one
> >>>> per top-level report collection on each HID device (for a total of 3
> >>>> virtual HID devices). The sub-drivers then bind to these virtual HID
> >>>> devices.
> >>>>
> >>>> This way the Touch Bar and ALS drivers can be kept in their own modules,
> >>>> while at the same time making them look very much like as if they were
> >>>> connected to the real HID devices. And those drivers then work (mostly)
> >>>> without further changes on MacBooks with the T2 chip that don't need
> >>>> this driver.
> >>>>
> >>>> Signed-off-by: Ronald Tschalär <ronald@...ovation.ch>
> >>>> [Kerem Karabay: convert to a platform driver]
> >>>> [Kerem Karabay: fix appleib_forward_int_op]
> >>>> [Kerem Karabay: rely on HID core's parsing in appleib_add_device]
> >>>> Signed-off-by: Kerem Karabay <kekrby@...il.com>
> >>>> Signed-off-by: Aditya Garg <gargaditya08@...e.com>
> >>>> ---
> >>>> drivers/hid/Kconfig         |  15 +
> >>>> drivers/hid/Makefile        |   1 +
> >>>> drivers/hid/apple-ibridge.c | 610 ++++++++++++++++++++++++++++++++++++
> >>>> drivers/hid/apple-ibridge.h |  15 +
> >>>> drivers/hid/hid-ids.h       |   1 +
> >>>> drivers/hid/hid-quirks.c    |   3 +
> >>>> 6 files changed, 645 insertions(+)
> >>>> create mode 100644 drivers/hid/apple-ibridge.c
> >>>> create mode 100644 drivers/hid/apple-ibridge.h
> >>>>
> >>>
> >>>>
> >>>> +}
> >>>> +
> >>>> +static int appleib_ll_raw_request(struct hid_device *hdev,
> >>>> +   unsigned char reportnum, __u8 *buf,
> >>>> +   size_t len, unsigned char rtype, int reqtype)
> >>>> +{
> >>>> + struct appleib_hid_dev_info *hdev_info = hdev->driver_data;
> >>>> +
> >>>> + return hid_hw_raw_request(hdev_info->hdev, reportnum, buf, len, rtype,
> >>>> +   reqtype);
> >>>> +}
> >>>> +
> >>>> +static int appleib_ll_output_report(struct hid_device *hdev, __u8 *buf,
> >>>> +     size_t len)
> >>>> +{
> >>>> + struct appleib_hid_dev_info *hdev_info = hdev->driver_data;
> >>>> +
> >>>> + return hid_hw_output_report(hdev_info->hdev, buf, len);
> >>>> +}
> >>>> +
> >>>> +static struct hid_ll_driver appleib_ll_driver = {
> >>>> + .start = appleib_ll_start,
> >>>> + .stop = appleib_ll_stop,
> >>>> + .open = appleib_ll_open,
> >>>> + .close = appleib_ll_close,
> >>>> + .power = appleib_ll_power,
> >>>> + .parse = appleib_ll_parse,
> >>>> + .request = appleib_ll_request,
> >>>> + .wait = appleib_ll_wait,
> >>>> + .raw_request = appleib_ll_raw_request,
> >>>> + .output_report = appleib_ll_output_report,
> >>>> +};
> >>>
> >>> const
> >>
> >> Are you sure about const here?
> >>
> >>>
> >>>> +
> >>>> + if (udev->actconfig->desc.bConfigurationValue != APPLEIB_BASIC_CONFIG) {
> >>>> + rc = usb_driver_set_configuration(udev, APPLEIB_BASIC_CONFIG);
> >>>> + return rc ? rc : -ENODEV;
> >>>> + }
> >>>> +
> >>>> + rc = hid_parse(hdev);
> >>>> + if (rc) {
> >>>> + hid_err(hdev, "ib: hid parse failed (%d)\n", rc);
> >>>> + goto error;
> >>>> + }
> >>>> +
> >>>> + rc = hid_hw_start(hdev, HID_CONNECT_DRIVER);
> >>>> + if (rc) {
> >>>> + hid_err(hdev, "ib: hw start failed (%d)\n", rc);
> >>>> + goto error;
> >>>> + }
> >>>> +
> >>>> + hdev_info = appleib_add_device(hdev);
> >>>> + if (IS_ERR(hdev_info)) {
> >>>> + rc = PTR_ERR(hdev_info);
> >>>> + goto stop_hw;
> >>>> + }
> >>>> +
> >>>> + hid_set_drvdata(hdev, hdev_info);
> >>>> +
> >>>> + rc = hid_hw_open(hdev);
> >>>> + if (rc) {
> >>>> + hid_err(hdev, "ib: failed to open hid: %d\n", rc);
> >>>> + goto remove_dev;
> >>>> + }
> >>>> +
> >>>> + return 0;
> >>>> +
> >>>> +remove_dev:
> >>>> + appleib_remove_device(hdev);
> >>>> +stop_hw:
> >>>> + hid_hw_stop(hdev);
> >>>> +error:
> >>>> + return rc;
> >>>> +}
> >>>> +
> >>>> +static void appleib_hid_remove(struct hid_device *hdev)
> >>>> +{
> >>>> + hid_hw_close(hdev);
> >>>> + appleib_remove_device(hdev);
> >>>> + hid_hw_stop(hdev);
> >>>> +}
> >>>> +
> >>>> +static const struct hid_device_id appleib_hid_ids[] = {
> >>>> + { HID_USB_DEVICE(USB_VENDOR_ID_APPLE, USB_DEVICE_ID_APPLE_IBRIDGE) },
> >>>> + { },
> >>>> +};
> >>>> +
> >>>> +static struct hid_driver appleib_hid_driver = {
> >>>> + .name = "apple-ibridge-hid",
> >>>> + .id_table = appleib_hid_ids,
> >>>> + .probe = appleib_hid_probe,
> >>>> + .remove = appleib_hid_remove,
> >>>> + .raw_event = appleib_hid_raw_event,
> >>>> + .report_fixup = appleib_report_fixup,
> >>>> +#ifdef CONFIG_PM
> >>>> + .suspend = appleib_hid_suspend,
> >>>> + .resume = appleib_hid_resume,
> >>>> + .reset_resume = appleib_hid_reset_resume,
> >>>> +#endif
> >>>> +};
> >>>
> >>> const
> >>
> >> Are you sure you want to do const here, cause other hid-drivers are NOT using const.
> >
> > It is scheduled for 6.3:
> > https://git.kernel.org/pub/scm/linux/kernel/git/hid/hid.git/log/?h=for-6.3/hid-core
> >
> > So yes, please make them const.
>
> Makes sense for `hid_ll_driver` as per the tree you send me, but I don’t see such a thing for
> `static struct hid_driver`. Is that scheduled as well?

Good point. I don't recall having seen such series now that you pinpoint this.

Thomas, is this something you have planned to submit?

Cheers,
Benjamin

>
> >
> > Cheers,
> > Benjamin
> >
> >>
> >>>> +
> >>>> +static struct appleib_device *appleib_alloc_device(struct platform_device *pdev)
> >>>> +{
> >>>> + struct appleib_device *ib_dev;
> >>>> + acpi_status sts;
>
>

Powered by blists - more mailing lists