linux-input.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Benjamin Tissoires <bentiss@kernel.org>
To: Jiri Kosina <jikos@kernel.org>,
	Alexei Starovoitov <ast@kernel.org>,
	 Shuah Khan <shuah@kernel.org>, Jonathan Corbet <corbet@lwn.net>
Cc: linux-input@vger.kernel.org, linux-kernel@vger.kernel.org,
	 bpf@vger.kernel.org, linux-kselftest@vger.kernel.org,
	 linux-doc@vger.kernel.org,
	Benjamin Tissoires <bentiss@kernel.org>
Subject: [PATCH HID 03/12] HID: bpf: add HID-BPF hooks for hid_hw_raw_requests
Date: Fri, 21 Jun 2024 10:55:48 +0200	[thread overview]
Message-ID: <20240621-hid_hw_req_bpf-v1-3-d7ab8b885a0b@kernel.org> (raw)
In-Reply-To: <20240621-hid_hw_req_bpf-v1-0-d7ab8b885a0b@kernel.org>

This allows to intercept and prevent or change the behavior of
hid_hw_raw_request() from a bpf program.

The intent is to solve a couple of use case:
- firewalling a HID device: a firewall can monitor who opens the hidraw
  nodes and then prevent or allow access to write operations on that
  hidraw node.
- change the behavior of a device and emulate a new HID feature request

The hook is allowed to be run as sleepable so it can itself call
hid_bpf_hw_request(), which allows to "convert" one feature request into
another or even call the feature request on a different HID device on the
same physical device.

Signed-off-by: Benjamin Tissoires <bentiss@kernel.org>

---

Here checkpatch complains about:
WARNING: use of RCU tasks trace is incorrect outside BPF or core RCU code

However, we are jumping in BPF code, so I think this is correct, but I'd
like to have the opinion on the BPF folks.
---
 drivers/hid/bpf/hid_bpf_dispatch.c   | 36 ++++++++++++++++++++++++++++++++++++
 drivers/hid/bpf/hid_bpf_struct_ops.c |  1 +
 drivers/hid/hid-core.c               |  6 ++++++
 include/linux/hid_bpf.h              | 35 +++++++++++++++++++++++++++++++++++
 4 files changed, 78 insertions(+)

diff --git a/drivers/hid/bpf/hid_bpf_dispatch.c b/drivers/hid/bpf/hid_bpf_dispatch.c
index 2df31decaac3..813b53119b85 100644
--- a/drivers/hid/bpf/hid_bpf_dispatch.c
+++ b/drivers/hid/bpf/hid_bpf_dispatch.c
@@ -74,6 +74,42 @@ dispatch_hid_bpf_device_event(struct hid_device *hdev, enum hid_report_type type
 }
 EXPORT_SYMBOL_GPL(dispatch_hid_bpf_device_event);
 
+int dispatch_hid_bpf_raw_requests(struct hid_device *hdev,
+				  unsigned char reportnum, u8 *buf,
+				  u32 size, enum hid_report_type rtype,
+				  enum hid_class_request reqtype,
+				  u64 source)
+{
+	struct hid_bpf_ctx_kern ctx_kern = {
+		.ctx = {
+			.hid = hdev,
+			.allocated_size = size,
+			.size = size,
+		},
+		.data = buf,
+	};
+	struct hid_bpf_ops *e;
+	int ret;
+
+	if (rtype >= HID_REPORT_TYPES)
+		return -EINVAL;
+
+	rcu_read_lock_trace();
+	list_for_each_entry_rcu(e, &hdev->bpf.prog_list, list) {
+		if (e->hid_hw_request) {
+			ret = e->hid_hw_request(&ctx_kern.ctx, reportnum, rtype, reqtype, source);
+			if (ret)
+				goto out;
+		}
+	}
+	ret = 0;
+
+out:
+	rcu_read_unlock_trace();
+	return ret;
+}
+EXPORT_SYMBOL_GPL(dispatch_hid_bpf_raw_requests);
+
 u8 *call_hid_bpf_rdesc_fixup(struct hid_device *hdev, u8 *rdesc, unsigned int *size)
 {
 	int ret;
diff --git a/drivers/hid/bpf/hid_bpf_struct_ops.c b/drivers/hid/bpf/hid_bpf_struct_ops.c
index 8063db1c8d62..93c824ba6a65 100644
--- a/drivers/hid/bpf/hid_bpf_struct_ops.c
+++ b/drivers/hid/bpf/hid_bpf_struct_ops.c
@@ -44,6 +44,7 @@ static int hid_bpf_ops_check_member(const struct btf_type *t,
 
 	switch (moff) {
 	case offsetof(struct hid_bpf_ops, hid_rdesc_fixup):
+	case offsetof(struct hid_bpf_ops, hid_hw_request):
 		break;
 	default:
 		if (prog->sleepable)
diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c
index 0775a32f5272..d839cfdf1624 100644
--- a/drivers/hid/hid-core.c
+++ b/drivers/hid/hid-core.c
@@ -2406,6 +2406,7 @@ int __hid_hw_raw_request(struct hid_device *hdev,
 			 __u64 source)
 {
 	unsigned int max_buffer_size = HID_MAX_BUFFER_SIZE;
+	int ret;
 
 	if (hdev->ll_driver->max_buffer_size)
 		max_buffer_size = hdev->ll_driver->max_buffer_size;
@@ -2413,6 +2414,11 @@ int __hid_hw_raw_request(struct hid_device *hdev,
 	if (len < 1 || len > max_buffer_size || !buf)
 		return -EINVAL;
 
+	ret = dispatch_hid_bpf_raw_requests(hdev, reportnum, buf, len, rtype,
+					    reqtype, source);
+	if (ret)
+		return ret;
+
 	return hdev->ll_driver->raw_request(hdev, reportnum, buf, len,
 					    rtype, reqtype);
 }
diff --git a/include/linux/hid_bpf.h b/include/linux/hid_bpf.h
index b88fa2df9f2c..5c3f179dc2ab 100644
--- a/include/linux/hid_bpf.h
+++ b/include/linux/hid_bpf.h
@@ -128,6 +128,31 @@ struct hid_bpf_ops {
 	 */
 	int (*hid_rdesc_fixup)(struct hid_bpf_ctx *ctx);
 
+	/**
+	 * @hid_hw_request: called whenever a hid_hw_raw_request() call is emitted
+	 * on the HID device
+	 *
+	 * It has the following arguments:
+	 *
+	 * ``ctx``: The HID-BPF context as &struct hid_bpf_ctx
+	 * ``reportnum``: the report number, as in hid_hw_raw_request()
+	 * ``rtype``: the report type (``HID_INPUT_REPORT``, ``HID_FEATURE_REPORT``,
+	 *            ``HID_OUTPUT_REPORT``)
+	 * ``reqtype``: the request
+	 * ``source``: a u64 referring to a uniq but identifiable source. If %0, the
+	 *             kernel itself emitted that call. For hidraw, ``source`` is set
+	 *             to the associated ``struct file *``.
+	 *
+	 * Return: %0 to keep processing the request by hid-core; any other value
+	 * stops hid-core from processing that event. A positive value should be
+	 * returned with the number of bytes returned in the incoming buffer; a
+	 * negative error code interrupts the processing of this call.
+	 */
+	int (*hid_hw_request)(struct hid_bpf_ctx *ctx, unsigned char reportnum,
+			       enum hid_report_type rtype, enum hid_class_request reqtype,
+			       __u64 source);
+
+
 	/* private: do not show up in the docs */
 	struct hid_device *hdev;
 };
@@ -149,6 +174,11 @@ struct hid_bpf {
 #ifdef CONFIG_HID_BPF
 u8 *dispatch_hid_bpf_device_event(struct hid_device *hid, enum hid_report_type type, u8 *data,
 				  u32 *size, int interrupt, u64 source);
+int dispatch_hid_bpf_raw_requests(struct hid_device *hdev,
+				  unsigned char reportnum, __u8 *buf,
+				  u32 size, enum hid_report_type rtype,
+				  enum hid_class_request reqtype,
+				  __u64 source);
 int hid_bpf_connect_device(struct hid_device *hdev);
 void hid_bpf_disconnect_device(struct hid_device *hdev);
 void hid_bpf_destroy_device(struct hid_device *hid);
@@ -158,6 +188,11 @@ u8 *call_hid_bpf_rdesc_fixup(struct hid_device *hdev, u8 *rdesc, unsigned int *s
 static inline u8 *dispatch_hid_bpf_device_event(struct hid_device *hid, enum hid_report_type type,
 						u8 *data, u32 *size, int interrupt,
 						u64 source) { return data; }
+static inline int dispatch_hid_bpf_raw_requests(struct hid_device *hdev,
+						unsigned char reportnum, u8 *buf,
+						u32 size, enum hid_report_type rtype,
+						enum hid_class_request reqtype,
+						u64 source) { return 0; }
 static inline int hid_bpf_connect_device(struct hid_device *hdev) { return 0; }
 static inline void hid_bpf_disconnect_device(struct hid_device *hdev) {}
 static inline void hid_bpf_destroy_device(struct hid_device *hid) {}

-- 
2.44.0


  parent reply	other threads:[~2024-06-21  8:56 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-06-21  8:55 [PATCH HID 00/12] HID: bpf_struct_ops, part 2 Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 01/12] HID: bpf: fix dispatch_hid_bpf_device_event uninitialized ret value Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 02/12] HID: add source argument to HID low level functions Benjamin Tissoires
2024-06-21  8:55 ` Benjamin Tissoires [this message]
2024-06-21 15:35   ` [PATCH HID 03/12] HID: bpf: add HID-BPF hooks for hid_hw_raw_requests Alexei Starovoitov
2024-06-21  8:55 ` [PATCH HID 04/12] HID: bpf: prevent infinite recursions with hid_hw_raw_requests hooks Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 05/12] selftests/hid: add tests for hid_hw_raw_request HID-BPF hooks Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 06/12] HID: bpf: add HID-BPF hooks for hid_hw_output_report Benjamin Tissoires
2024-06-21 15:37   ` Alexei Starovoitov
2024-06-21 16:08     ` Benjamin Tissoires
2024-06-21 16:26       ` Alexei Starovoitov
2024-06-24  8:52         ` Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 07/12] selftests/hid: add tests for hid_hw_output_report HID-BPF hooks Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 08/12] HID: bpf: make hid_bpf_input_report() sleep until the device is ready Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 09/12] selftests/hid: add wq test for hid_bpf_input_report() Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 10/12] HID: bpf: allow hid_device_event hooks to inject input reports on self Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 11/12] selftests/hid: add another test for injecting an event from an event hook Benjamin Tissoires
2024-06-21  8:55 ` [PATCH HID 12/12] selftests/hid: add an infinite loop test for hid_bpf_try_input_report Benjamin Tissoires

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20240621-hid_hw_req_bpf-v1-3-d7ab8b885a0b@kernel.org \
    --to=bentiss@kernel.org \
    --cc=ast@kernel.org \
    --cc=bpf@vger.kernel.org \
    --cc=corbet@lwn.net \
    --cc=jikos@kernel.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-input@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-kselftest@vger.kernel.org \
    --cc=shuah@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).