| Message ID | 20190401124200.17331-1-andrew.shadura@collabora.co.uk (mailing list archive) |
|---|---|
| State | Mainlined |
| Delegated to: | Jiri Kosina |
| Headers | show |
| Series | [v3] HID: add driver for U2F Zero built-in LED and RNG | expand |
On Mon, 1 Apr 2019, Andrej Shadura wrote: > U2F Zero supports custom commands for blinking the LED and getting data > from the internal hardware RNG. Expose the blinking function as a LED > device, and the internal hardware RNG as an HWRNG so that it can be used > to feed the enthropy pool. So I still am not really happy about this being wired up into generic HID although it's really a USB driver, but we've discussed that already, and I don't see any option that'd work substantially better in this case, especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, thanks!
On Wed, Apr 10, 2019 at 3:14 PM Jiri Kosina <jikos@kernel.org> wrote: > > On Mon, 1 Apr 2019, Andrej Shadura wrote: > > > U2F Zero supports custom commands for blinking the LED and getting data > > from the internal hardware RNG. Expose the blinking function as a LED > > device, and the internal hardware RNG as an HWRNG so that it can be used > > to feed the enthropy pool. > > So I still am not really happy about this being wired up into generic HID > although it's really a USB driver, but we've discussed that already, and I > don't see any option that'd work substantially better in this case, > especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, > thanks! > Hi wish I were CC-ed on these threads. My CI script now fails because hid-u2fzero.ko needs devm_hwrng_register and the Kconfig doesn't force pulling the right dependency. Also, you probably want to add in .probe() a check for the actual transport driver (`hid_is_using_ll_driver(hdev, &usb_hid_driver)`) or you can not use the usbhid functions without crashing the kernel. Cheers, Benjamin
On Thu, 11 Apr 2019, Benjamin Tissoires wrote: > > So I still am not really happy about this being wired up into generic HID > > although it's really a USB driver, but we've discussed that already, and I > > don't see any option that'd work substantially better in this case, > > especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, > > thanks! > > > > Hi wish I were CC-ed on these threads. Gah, for some reason I was convinced you were. > My CI script now fails because hid-u2fzero.ko needs devm_hwrng_register > and the Kconfig doesn't force pulling the right dependency. > > Also, you probably want to add in .probe() a check for the actual > transport driver (`hid_is_using_ll_driver(hdev, &usb_hid_driver)`) or > you can not use the usbhid functions without crashing the kernel. Benjamin, I love your CI :) Andrej, could you please send fixups on top of the applied patch? Thanks,
On 11/04/2019 12:35, Jiri Kosina wrote: > On Thu, 11 Apr 2019, Benjamin Tissoires wrote: > >>> So I still am not really happy about this being wired up into generic HID >>> although it's really a USB driver, but we've discussed that already, and I >>> don't see any option that'd work substantially better in this case, >>> especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, >>> thanks! >>> >> >> Hi wish I were CC-ed on these threads. > > Gah, for some reason I was convinced you were. > >> My CI script now fails because hid-u2fzero.ko needs devm_hwrng_register >> and the Kconfig doesn't force pulling the right dependency. >> Also, you probably want to add in .probe() a check for the actual >> transport driver (`hid_is_using_ll_driver(hdev, &usb_hid_driver)`) or >> you can not use the usbhid functions without crashing the kernel. Thanks, that’s very useful. > Benjamin, I love your CI :) > > Andrej, could you please send fixups on top of the applied patch? Sure, will do.
On Thu, 11 Apr 2019, Andrej Shadura wrote: > >>> So I still am not really happy about this being wired up into generic HID > >>> although it's really a USB driver, but we've discussed that already, and I > >>> don't see any option that'd work substantially better in this case, > >>> especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, > >>> thanks! > >>> > >> > >> Hi wish I were CC-ed on these threads. > > > > Gah, for some reason I was convinced you were. > > > >> My CI script now fails because hid-u2fzero.ko needs devm_hwrng_register > >> and the Kconfig doesn't force pulling the right dependency. > > >> Also, you probably want to add in .probe() a check for the actual > >> transport driver (`hid_is_using_ll_driver(hdev, &usb_hid_driver)`) or > >> you can not use the usbhid functions without crashing the kernel. > > Thanks, that’s very useful. > > > Benjamin, I love your CI :) > > > > Andrej, could you please send fixups on top of the applied patch? > > Sure, will do. I didn't receive any followup; the Kconfig dependency has already been fixed by Mao Wenan, and I've just queued the patch below on top of that as well. From: Jiri Kosina <jkosina@suse.cz> Subject: [PATCH] HID: u2fzero: fail probe if not using USB transport u2fzero driver is USB-only. Therefore we have to give up in ->probe() callback in case we're called with non-USB transport driver bound, otherwise the kernel will crash trying to use USBHID API on a non-USB transport. Fixes: 42337b9d4d958("HID: add driver for U2F Zero built-in LED and RNG") Reported-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz> --- drivers/hid/hid-u2fzero.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/hid/hid-u2fzero.c b/drivers/hid/hid-u2fzero.c index d11a5cb56a0d..95e0807878c7 100644 --- a/drivers/hid/hid-u2fzero.c +++ b/drivers/hid/hid-u2fzero.c @@ -286,6 +286,9 @@ static int u2fzero_probe(struct hid_device *hdev, unsigned int minor; int ret; + if (!hid_is_using_ll_driver(hdev, &usb_hid_driver)) + return -EINVAL; + dev = devm_kzalloc(&hdev->dev, sizeof(*dev), GFP_KERNEL); if (dev == NULL) return -ENOMEM;
On 17/04/2019 16:43, Jiri Kosina wrote: > On Thu, 11 Apr 2019, Andrej Shadura wrote: > >>>>> So I still am not really happy about this being wired up into generic HID >>>>> although it's really a USB driver, but we've discussed that already, and I >>>>> don't see any option that'd work substantially better in this case, >>>>> especially from the UX point of view. Oh well. Applied to for-5.2/u2fzero, >>>>> thanks! >>>>> >>>> >>>> Hi wish I were CC-ed on these threads. >>> >>> Gah, for some reason I was convinced you were. >>> >>>> My CI script now fails because hid-u2fzero.ko needs devm_hwrng_register >>>> and the Kconfig doesn't force pulling the right dependency. >> >>>> Also, you probably want to add in .probe() a check for the actual >>>> transport driver (`hid_is_using_ll_driver(hdev, &usb_hid_driver)`) or >>>> you can not use the usbhid functions without crashing the kernel. >> >> Thanks, that’s very useful. >> >>> Benjamin, I love your CI :) >>> >>> Andrej, could you please send fixups on top of the applied patch? >> >> Sure, will do. > > I didn't receive any followup; the Kconfig dependency has already been > fixed by Mao Wenan, and I've just queued the patch below on top of that as > well. Sorry, I’ve had too much on my plate in the recent days, so I’ve unfortunately been pushing this task off a bit. Thanks to both of you, Mao and Jiří, both things should have really been done by me. I’ll try to get those things right from the beginning the next time. > From: Jiri Kosina <jkosina@suse.cz> > Subject: [PATCH] HID: u2fzero: fail probe if not using USB transport > > u2fzero driver is USB-only. Therefore we have to give up in ->probe() > callback in case we're called with non-USB transport driver bound, > otherwise the kernel will crash trying to use USBHID API on a non-USB > transport. > > Fixes: 42337b9d4d958("HID: add driver for U2F Zero built-in LED and RNG") > Reported-by: Benjamin Tissoires <benjamin.tissoires@redhat.com> > Signed-off-by: Jiri Kosina <jkosina@suse.cz> > --- > drivers/hid/hid-u2fzero.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/hid/hid-u2fzero.c b/drivers/hid/hid-u2fzero.c > index d11a5cb56a0d..95e0807878c7 100644 > --- a/drivers/hid/hid-u2fzero.c > +++ b/drivers/hid/hid-u2fzero.c > @@ -286,6 +286,9 @@ static int u2fzero_probe(struct hid_device *hdev, > unsigned int minor; > int ret; > > + if (!hid_is_using_ll_driver(hdev, &usb_hid_driver)) > + return -EINVAL; > + > dev = devm_kzalloc(&hdev->dev, sizeof(*dev), GFP_KERNEL); > if (dev == NULL) > return -ENOMEM; >
On 01/04/2019 14:42, Andrej Shadura wrote: > U2F Zero supports custom commands for blinking the LED and getting data > from the internal hardware RNG. Expose the blinking function as a LED > device, and the internal hardware RNG as an HWRNG so that it can be used > to feed the enthropy pool. > > Signed-off-by: Andrej Shadura <andrew.shadura@collabora.co.uk> I’ve been testing this with a different modification of U2F Zero, Nitrokey FIDO U2F, and on that device only I’m getting a kernel warning (see below). > +static int u2fzero_recv(struct u2fzero_device *dev, > + struct u2f_hid_report *req, > + struct u2f_hid_msg *resp) > +{ > + int ret; > + struct hid_device *hdev = dev->hdev; > + struct u2fzero_transfer_context ctx; > + > + mutex_lock(&dev->lock); > + > + memcpy(dev->buf_out, req, sizeof(struct u2f_hid_report)); > + > + dev->urb->context = &ctx; > + init_completion(&ctx.done); > + > + ret = usb_submit_urb(dev->urb, GFP_NOIO); Here, usb_submit_urb() returns EBUSY in the first attempt to read random numbers from the device: URB ffff917256d5d540 submitted while active WARNING: CPU: 3 PID: 31 at drivers/usb/core/urb.c:363 usb_submit_urb+0x4c2/0x5b0 <...> Call Trace: u2fzero_rng_read+0x16e/0x340 [hid_u2fzero] ? ttwu_do_activate+0x67/0x90 add_early_randomness+0x53/0xc0 hwrng_register+0x175/0x180 devm_hwrng_register+0x41/0x7e u2fzero_probe+0x2dd/0x350 [hid_u2fzero] hid_device_probe+0x119/0x180 [hid] really_probe+0xfe/0x3b0 driver_probe_device+0xba/0x100 __device_attach_driver+0x97/0x100 I don’t understand why since 1) it’s likely to be the first transmission of this URB, 2) there’s a mutex locked just before it. I received a comment from a colleague mentioning I’m probably not using the mutex correctly, but I don’t understand why. I’m trying to figure this out, so I’d welcome any help with it. Thanks in advance. > + if (unlikely(ret)) { > + hid_err(hdev, "usb_submit_urb failed: %d", ret); > + goto err; > + } > + > + ret = hid_hw_output_report(dev->hdev, dev->buf_out, > + sizeof(struct u2f_hid_msg)); > + > + if (ret < 0) { > + hid_err(hdev, "hid_hw_output_report failed: %d", ret); > + goto err; > + } > + > + ret = (wait_for_completion_timeout( > + &ctx.done, msecs_to_jiffies(USB_CTRL_SET_TIMEOUT))); > + if (ret < 0) { > + usb_kill_urb(dev->urb); > + hid_err(hdev, "urb submission timed out"); > + } else { > + ret = dev->urb->actual_length; > + memcpy(resp, dev->buf_in, ret); > + } > + > +err: > + mutex_unlock(&dev->lock); > + > + return ret; > +} <...> > +static int u2fzero_rng_read(struct hwrng *rng, void *data, > + size_t max, bool wait) > +{ > + struct u2fzero_device *dev = container_of(rng, > + struct u2fzero_device, hwrng); > + struct u2f_hid_report req = { > + .report_type = 0, > + .msg.cid = CID_BROADCAST, > + .msg.init = { > + .cmd = U2F_CUSTOM_GET_RNG, > + .bcnth = 0, > + .bcntl = 0, > + .data = {0}, > + } > + }; > + struct u2f_hid_msg resp; > + int ret; > + size_t actual_length; > + > + if (!dev->present) { > + hid_dbg(dev->hdev, "device not present"); > + return 0; > + } > + > + ret = u2fzero_recv(dev, &req, &resp); > + if (ret < 0) > + return 0; > + > + /* only take the minimum amount of data it is safe to take */ > + actual_length = min3((size_t)ret - offsetof(struct u2f_hid_msg, > + init.data), U2F_HID_MSG_LEN(resp), max); > + > + memcpy(data, resp.init.data, actual_length); > + > + return actual_length; > +} <...> > +static int u2fzero_init_hwrng(struct u2fzero_device *dev, > + unsigned int minor) > +{ > + dev->rng_name = devm_kasprintf(&dev->hdev->dev, GFP_KERNEL, > + "%s-rng%u", DRIVER_SHORT, minor); > + if (dev->rng_name == NULL) > + return -ENOMEM; > + > + dev->hwrng.name = dev->rng_name; > + dev->hwrng.read = u2fzero_rng_read; > + dev->hwrng.quality = 1; > + > + return devm_hwrng_register(&dev->hdev->dev, &dev->hwrng); > +} > + > +static int u2fzero_fill_in_urb(struct u2fzero_device *dev) > +{ > + struct hid_device *hdev = dev->hdev; > + struct usb_device *udev; > + struct usbhid_device *usbhid = hdev->driver_data; > + unsigned int pipe_in; > + struct usb_host_endpoint *ep; > + > + if (dev->hdev->bus != BUS_USB) > + return -EINVAL; > + > + udev = hid_to_usb_dev(hdev); > + > + if (!usbhid->urbout || !usbhid->urbin) > + return -ENODEV; > + > + ep = usb_pipe_endpoint(udev, usbhid->urbin->pipe); > + if (!ep) > + return -ENODEV; > + > + dev->urb = usb_alloc_urb(0, GFP_KERNEL); > + if (!dev->urb) > + return -ENOMEM; > + > + pipe_in = (usbhid->urbin->pipe & ~(3 << 30)) | (PIPE_INTERRUPT << 30); > + > + usb_fill_int_urb(dev->urb, > + udev, > + pipe_in, > + dev->buf_in, > + HID_REPORT_SIZE, > + u2fzero_read_callback, > + NULL, > + ep->desc.bInterval); > + > + return 0; > +} > + > +static int u2fzero_probe(struct hid_device *hdev, > + const struct hid_device_id *id) > +{ > + struct u2fzero_device *dev; > + unsigned int minor; > + int ret; > + > + dev = devm_kzalloc(&hdev->dev, sizeof(*dev), GFP_KERNEL); > + if (dev == NULL) > + return -ENOMEM; > + > + dev->buf_out = devm_kmalloc(&hdev->dev, > + sizeof(struct u2f_hid_report), GFP_KERNEL); > + if (dev->buf_out == NULL) > + return -ENOMEM; > + > + dev->buf_in = devm_kmalloc(&hdev->dev, > + sizeof(struct u2f_hid_msg), GFP_KERNEL); > + if (dev->buf_in == NULL) > + return -ENOMEM; > + > + ret = hid_parse(hdev); > + if (ret) > + return ret; > + > + dev->hdev = hdev; > + hid_set_drvdata(hdev, dev); > + mutex_init(&dev->lock); > + > + ret = hid_hw_start(hdev, HID_CONNECT_HIDRAW); > + if (ret) > + return ret; > + > + u2fzero_fill_in_urb(dev); > + > + dev->present = true; > + > + minor = ((struct hidraw *) hdev->hidraw)->minor; > + > + ret = u2fzero_init_led(dev, minor); > + if (ret) { > + hid_hw_stop(hdev); > + return ret; > + } > + > + hid_info(hdev, "U2F Zero LED initialised\n"); > + > + ret = u2fzero_init_hwrng(dev, minor); > + if (ret) { > + hid_hw_stop(hdev); > + return ret; > + } > + > + hid_info(hdev, "U2F Zero RNG initialised\n"); > + > + return 0; > +}
On Thu, 3 Oct 2019, Andrej Shadura wrote: > On 01/04/2019 14:42, Andrej Shadura wrote: > > U2F Zero supports custom commands for blinking the LED and getting data > > from the internal hardware RNG. Expose the blinking function as a LED > > device, and the internal hardware RNG as an HWRNG so that it can be used > > to feed the enthropy pool. > > > > Signed-off-by: Andrej Shadura <andrew.shadura@collabora.co.uk> > > I’ve been testing this with a different modification of U2F Zero, > Nitrokey FIDO U2F, and on that device only I’m getting a kernel warning > (see below). > > > +static int u2fzero_recv(struct u2fzero_device *dev, > > + struct u2f_hid_report *req, > > + struct u2f_hid_msg *resp) > > +{ > > + int ret; > > + struct hid_device *hdev = dev->hdev; > > + struct u2fzero_transfer_context ctx; > > + > > + mutex_lock(&dev->lock); > > + > > + memcpy(dev->buf_out, req, sizeof(struct u2f_hid_report)); > > + > > + dev->urb->context = &ctx; > > + init_completion(&ctx.done); > > + > > + ret = usb_submit_urb(dev->urb, GFP_NOIO); > > Here, usb_submit_urb() returns EBUSY in the first attempt to read random > numbers from the device: > > URB ffff917256d5d540 submitted while active > WARNING: CPU: 3 PID: 31 at drivers/usb/core/urb.c:363 > usb_submit_urb+0x4c2/0x5b0 > <...> > Call Trace: > u2fzero_rng_read+0x16e/0x340 [hid_u2fzero] > ? ttwu_do_activate+0x67/0x90 > add_early_randomness+0x53/0xc0 > hwrng_register+0x175/0x180 > devm_hwrng_register+0x41/0x7e > u2fzero_probe+0x2dd/0x350 [hid_u2fzero] > hid_device_probe+0x119/0x180 [hid] > really_probe+0xfe/0x3b0 > driver_probe_device+0xba/0x100 > __device_attach_driver+0x97/0x100 > > I don’t understand why since 1) it’s likely to be the first transmission > of this URB, 2) there’s a mutex locked just before it. I received a > comment from a colleague mentioning I’m probably not using the mutex > correctly, but I don’t understand why. > > I’m trying to figure this out, so I’d welcome any help with it. You can try using usbmon to see exactly what URBs are actually running. Alan Stern
diff --git a/drivers/hid/Kconfig b/drivers/hid/Kconfig index 41e9935fc584..6efb5a43ffc8 100644 --- a/drivers/hid/Kconfig +++ b/drivers/hid/Kconfig @@ -989,6 +989,21 @@ config HID_UDRAW_PS3 Say Y here if you want to use the THQ uDraw gaming tablet for the PS3. +config HID_U2FZERO + tristate "U2F Zero LED and RNG support" + depends on USB_HID + depends on LEDS_CLASS + help + Support for the LED of the U2F Zero device. + + U2F Zero supports custom commands for blinking the LED + and getting data from the internal hardware RNG. + The internal hardware can be used to feed the enthropy pool. + + U2F Zero only supports blinking its LED, so this driver doesn't + allow setting the brightness to anything but 1, which will + trigger a single blink and immediately reset to back 0. + config HID_WACOM tristate "Wacom Intuos/Graphire tablet support (USB)" depends on USB_HID diff --git a/drivers/hid/Makefile b/drivers/hid/Makefile index 896a51ce7ce0..d9724ab8df5f 100644 --- a/drivers/hid/Makefile +++ b/drivers/hid/Makefile @@ -108,6 +108,7 @@ obj-$(CONFIG_HID_THRUSTMASTER) += hid-tmff.o obj-$(CONFIG_HID_TIVO) += hid-tivo.o obj-$(CONFIG_HID_TOPSEED) += hid-topseed.o obj-$(CONFIG_HID_TWINHAN) += hid-twinhan.o +obj-$(CONFIG_HID_U2FZERO) += hid-u2fzero.o obj-$(CONFIG_HID_UCLOGIC) += hid-uclogic.o obj-$(CONFIG_HID_UDRAW_PS3) += hid-udraw-ps3.o obj-$(CONFIG_HID_LED) += hid-led.o diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 24f846d67478..36f898eab36f 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -321,6 +321,7 @@ #define USB_DEVICE_ID_CYGNAL_RADIO_SI470X 0x818a #define USB_DEVICE_ID_FOCALTECH_FTXXXX_MULTITOUCH 0x81b9 #define USB_DEVICE_ID_CYGNAL_CP2112 0xea90 +#define USB_DEVICE_ID_U2F_ZERO 0x8acf #define USB_DEVICE_ID_CYGNAL_RADIO_SI4713 0x8244 diff --git a/drivers/hid/hid-u2fzero.c b/drivers/hid/hid-u2fzero.c new file mode 100644 index 000000000000..d11a5cb56a0d --- /dev/null +++ b/drivers/hid/hid-u2fzero.c @@ -0,0 +1,371 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * U2F Zero LED and RNG driver + * + * Copyright 2018 Andrej Shadura <andrew@shadura.me> + * Loosely based on drivers/hid/hid-led.c + * and drivers/usb/misc/chaoskey.c + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation, version 2. + */ + +#include <linux/hid.h> +#include <linux/hidraw.h> +#include <linux/hw_random.h> +#include <linux/leds.h> +#include <linux/module.h> +#include <linux/mutex.h> +#include <linux/usb.h> + +#include "usbhid/usbhid.h" +#include "hid-ids.h" + +#define DRIVER_SHORT "u2fzero" + +#define HID_REPORT_SIZE 64 + +/* We only use broadcast (CID-less) messages */ +#define CID_BROADCAST 0xffffffff + +struct u2f_hid_msg { + u32 cid; + union { + struct { + u8 cmd; + u8 bcnth; + u8 bcntl; + u8 data[HID_REPORT_SIZE - 7]; + } init; + struct { + u8 seq; + u8 data[HID_REPORT_SIZE - 5]; + } cont; + }; +} __packed; + +struct u2f_hid_report { + u8 report_type; + struct u2f_hid_msg msg; +} __packed; + +#define U2F_HID_MSG_LEN(f) (size_t)(((f).init.bcnth << 8) + (f).init.bcntl) + +/* Custom extensions to the U2FHID protocol */ +#define U2F_CUSTOM_GET_RNG 0x21 +#define U2F_CUSTOM_WINK 0x24 + +struct u2fzero_device { + struct hid_device *hdev; + struct urb *urb; /* URB for the RNG data */ + struct led_classdev ldev; /* Embedded struct for led */ + struct hwrng hwrng; /* Embedded struct for hwrng */ + char *led_name; + char *rng_name; + u8 *buf_out; + u8 *buf_in; + struct mutex lock; + bool present; +}; + +static int u2fzero_send(struct u2fzero_device *dev, struct u2f_hid_report *req) +{ + int ret; + + mutex_lock(&dev->lock); + + memcpy(dev->buf_out, req, sizeof(struct u2f_hid_report)); + + ret = hid_hw_output_report(dev->hdev, dev->buf_out, + sizeof(struct u2f_hid_msg)); + + mutex_unlock(&dev->lock); + + if (ret < 0) + return ret; + + return ret == sizeof(struct u2f_hid_msg) ? 0 : -EMSGSIZE; +} + +struct u2fzero_transfer_context { + struct completion done; + int status; +}; + +static void u2fzero_read_callback(struct urb *urb) +{ + struct u2fzero_transfer_context *ctx = urb->context; + + ctx->status = urb->status; + complete(&ctx->done); +} + +static int u2fzero_recv(struct u2fzero_device *dev, + struct u2f_hid_report *req, + struct u2f_hid_msg *resp) +{ + int ret; + struct hid_device *hdev = dev->hdev; + struct u2fzero_transfer_context ctx; + + mutex_lock(&dev->lock); + + memcpy(dev->buf_out, req, sizeof(struct u2f_hid_report)); + + dev->urb->context = &ctx; + init_completion(&ctx.done); + + ret = usb_submit_urb(dev->urb, GFP_NOIO); + if (unlikely(ret)) { + hid_err(hdev, "usb_submit_urb failed: %d", ret); + goto err; + } + + ret = hid_hw_output_report(dev->hdev, dev->buf_out, + sizeof(struct u2f_hid_msg)); + + if (ret < 0) { + hid_err(hdev, "hid_hw_output_report failed: %d", ret); + goto err; + } + + ret = (wait_for_completion_timeout( + &ctx.done, msecs_to_jiffies(USB_CTRL_SET_TIMEOUT))); + if (ret < 0) { + usb_kill_urb(dev->urb); + hid_err(hdev, "urb submission timed out"); + } else { + ret = dev->urb->actual_length; + memcpy(resp, dev->buf_in, ret); + } + +err: + mutex_unlock(&dev->lock); + + return ret; +} + +static int u2fzero_blink(struct led_classdev *ldev) +{ + struct u2fzero_device *dev = container_of(ldev, + struct u2fzero_device, ldev); + struct u2f_hid_report req = { + .report_type = 0, + .msg.cid = CID_BROADCAST, + .msg.init = { + .cmd = U2F_CUSTOM_WINK, + .bcnth = 0, + .bcntl = 0, + .data = {0}, + } + }; + return u2fzero_send(dev, &req); +} + +static int u2fzero_brightness_set(struct led_classdev *ldev, + enum led_brightness brightness) +{ + ldev->brightness = LED_OFF; + if (brightness) + return u2fzero_blink(ldev); + else + return 0; +} + +static int u2fzero_rng_read(struct hwrng *rng, void *data, + size_t max, bool wait) +{ + struct u2fzero_device *dev = container_of(rng, + struct u2fzero_device, hwrng); + struct u2f_hid_report req = { + .report_type = 0, + .msg.cid = CID_BROADCAST, + .msg.init = { + .cmd = U2F_CUSTOM_GET_RNG, + .bcnth = 0, + .bcntl = 0, + .data = {0}, + } + }; + struct u2f_hid_msg resp; + int ret; + size_t actual_length; + + if (!dev->present) { + hid_dbg(dev->hdev, "device not present"); + return 0; + } + + ret = u2fzero_recv(dev, &req, &resp); + if (ret < 0) + return 0; + + /* only take the minimum amount of data it is safe to take */ + actual_length = min3((size_t)ret - offsetof(struct u2f_hid_msg, + init.data), U2F_HID_MSG_LEN(resp), max); + + memcpy(data, resp.init.data, actual_length); + + return actual_length; +} + +static int u2fzero_init_led(struct u2fzero_device *dev, + unsigned int minor) +{ + dev->led_name = devm_kasprintf(&dev->hdev->dev, GFP_KERNEL, + "%s%u", DRIVER_SHORT, minor); + if (dev->led_name == NULL) + return -ENOMEM; + + dev->ldev.name = dev->led_name; + dev->ldev.max_brightness = LED_ON; + dev->ldev.flags = LED_HW_PLUGGABLE; + dev->ldev.brightness_set_blocking = u2fzero_brightness_set; + + return devm_led_classdev_register(&dev->hdev->dev, &dev->ldev); +} + +static int u2fzero_init_hwrng(struct u2fzero_device *dev, + unsigned int minor) +{ + dev->rng_name = devm_kasprintf(&dev->hdev->dev, GFP_KERNEL, + "%s-rng%u", DRIVER_SHORT, minor); + if (dev->rng_name == NULL) + return -ENOMEM; + + dev->hwrng.name = dev->rng_name; + dev->hwrng.read = u2fzero_rng_read; + dev->hwrng.quality = 1; + + return devm_hwrng_register(&dev->hdev->dev, &dev->hwrng); +} + +static int u2fzero_fill_in_urb(struct u2fzero_device *dev) +{ + struct hid_device *hdev = dev->hdev; + struct usb_device *udev; + struct usbhid_device *usbhid = hdev->driver_data; + unsigned int pipe_in; + struct usb_host_endpoint *ep; + + if (dev->hdev->bus != BUS_USB) + return -EINVAL; + + udev = hid_to_usb_dev(hdev); + + if (!usbhid->urbout || !usbhid->urbin) + return -ENODEV; + + ep = usb_pipe_endpoint(udev, usbhid->urbin->pipe); + if (!ep) + return -ENODEV; + + dev->urb = usb_alloc_urb(0, GFP_KERNEL); + if (!dev->urb) + return -ENOMEM; + + pipe_in = (usbhid->urbin->pipe & ~(3 << 30)) | (PIPE_INTERRUPT << 30); + + usb_fill_int_urb(dev->urb, + udev, + pipe_in, + dev->buf_in, + HID_REPORT_SIZE, + u2fzero_read_callback, + NULL, + ep->desc.bInterval); + + return 0; +} + +static int u2fzero_probe(struct hid_device *hdev, + const struct hid_device_id *id) +{ + struct u2fzero_device *dev; + unsigned int minor; + int ret; + + dev = devm_kzalloc(&hdev->dev, sizeof(*dev), GFP_KERNEL); + if (dev == NULL) + return -ENOMEM; + + dev->buf_out = devm_kmalloc(&hdev->dev, + sizeof(struct u2f_hid_report), GFP_KERNEL); + if (dev->buf_out == NULL) + return -ENOMEM; + + dev->buf_in = devm_kmalloc(&hdev->dev, + sizeof(struct u2f_hid_msg), GFP_KERNEL); + if (dev->buf_in == NULL) + return -ENOMEM; + + ret = hid_parse(hdev); + if (ret) + return ret; + + dev->hdev = hdev; + hid_set_drvdata(hdev, dev); + mutex_init(&dev->lock); + + ret = hid_hw_start(hdev, HID_CONNECT_HIDRAW); + if (ret) + return ret; + + u2fzero_fill_in_urb(dev); + + dev->present = true; + + minor = ((struct hidraw *) hdev->hidraw)->minor; + + ret = u2fzero_init_led(dev, minor); + if (ret) { + hid_hw_stop(hdev); + return ret; + } + + hid_info(hdev, "U2F Zero LED initialised\n"); + + ret = u2fzero_init_hwrng(dev, minor); + if (ret) { + hid_hw_stop(hdev); + return ret; + } + + hid_info(hdev, "U2F Zero RNG initialised\n"); + + return 0; +} + +static void u2fzero_remove(struct hid_device *hdev) +{ + struct u2fzero_device *dev = hid_get_drvdata(hdev); + + mutex_lock(&dev->lock); + dev->present = false; + mutex_unlock(&dev->lock); + + hid_hw_stop(hdev); + usb_poison_urb(dev->urb); + usb_free_urb(dev->urb); +} + +static const struct hid_device_id u2fzero_table[] = { + { HID_USB_DEVICE(USB_VENDOR_ID_CYGNAL, + USB_DEVICE_ID_U2F_ZERO) }, + { } +}; +MODULE_DEVICE_TABLE(hid, u2fzero_table); + +static struct hid_driver u2fzero_driver = { + .name = "hid-" DRIVER_SHORT, + .probe = u2fzero_probe, + .remove = u2fzero_remove, + .id_table = u2fzero_table, +}; + +module_hid_driver(u2fzero_driver); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Andrej Shadura <andrew@shadura.me>"); +MODULE_DESCRIPTION("U2F Zero LED and RNG driver");
U2F Zero supports custom commands for blinking the LED and getting data from the internal hardware RNG. Expose the blinking function as a LED device, and the internal hardware RNG as an HWRNG so that it can be used to feed the enthropy pool. Signed-off-by: Andrej Shadura <andrew.shadura@collabora.co.uk> --- drivers/hid/Kconfig | 15 ++ drivers/hid/Makefile | 1 + drivers/hid/hid-ids.h | 1 + drivers/hid/hid-u2fzero.c | 371 ++++++++++++++++++++++++++++++++++++++ 4 files changed, 388 insertions(+) create mode 100644 drivers/hid/hid-u2fzero.c