Re: [PATCH] HID: rmi: Scan the report descriptor to determine if the device is suitable for the hid-rmi driver
From: Andrew Duggan
Date: Tue Oct 07 2014 - 20:36:51 EST
Hi Benjamin,
Thanks for reviewing.
On 10/03/2014 02:25 PM, Benjamin Tissoires wrote:
Hi Andrew,
On Oct 02 2014 or thereabouts, Andrew Duggan wrote:
On composite HID devices there may be multiple HID devices on separate interfaces, but hid-rmi
should only bind to the touchpad. Commit e19ff99f256aeeff6c07b373e01883b72e049552 simply checked
that the interface protocol was set to mouse. Unfortuately, it is not always the case that the
touchpad has the mouse interface protocol set. This patch takes a different approach and scans
the report descriptor looking for the Vendor Specific Top Level Collection and the associated
usages and report IDs needed by the hid-rmi driver to interface with the device.
I am going to leave for XDC 2014 next week and then I'll have one week
off, so I prefer sending my impressions now (or it will have to wait 2
more weeks).
I am not sure this solution is sustainable in the long term. If every
driver starts checking each vendor collection in core, there is no point
in having core and drivers separated.
Unfortunately, I can not think of an easier way of doing it. So I think
(but Jiri can choose to ignore me) that we should at least hold this
until 3.19 development starts.
I agree checking for specific vendor's reports in core isn't ideal, but
I couldn't figure out a better way to uniquely identify our touchpad on
certain composite USB devices.
Specifically, I am looking at the Razer Blade 14" laptop which has a
composite USB device with 4 HID devices on separate interfaces. One is
the touchpad, two others report to be keyboards, and the forth device
shows up as a mouse. Since the VID and PID are identical, scanning the
report descriptor is the only way to determine which device. There are
slight differences in the report descriptors which we could use to make
a guess. For instance our touchpad uses Application
GenericDesktop.Pointer while the other device uses GenericDesktop.Mouse.
The fact that only one has a vendor specific collection is also an
indication. But, these indicators don't seem particularly conclusive.
Also, the mouse device has it's interface protocol set to mouse instead
of the touchpad, so the current code which checks that value would
attach to the wrong device (if you add the Razer VID and PID to the
Vendor specific handling check in hid_scan_report).
One or two more comments inlined:
Signed-off-by: Andrew Duggan <aduggan@xxxxxxxxxxxxx>
---
drivers/hid/hid-core.c | 42 +++++++++++++++++++++++++++++++++++++-----
include/linux/hid.h | 8 +++++++-
2 files changed, 44 insertions(+), 6 deletions(-)
diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c
index 12b6e67..498f674 100644
--- a/drivers/hid/hid-core.c
+++ b/drivers/hid/hid-core.c
@@ -686,6 +686,23 @@ static void hid_scan_input_usage(struct hid_parser *parser, u32 usage)
if (usage == HID_DG_CONTACTID)
hid->group = HID_GROUP_MULTITOUCH;
+
+ if (usage == 0xff000004 && parser->global.report_id == 0xb)
+ parser->scan_flags |= HID_SCAN_FLAG_RMI_INPUT_READ;
+
+ if (usage == 0xff000005 && parser->global.report_id == 0xc)
+ parser->scan_flags |= HID_SCAN_FLAG_RMI_ATTN;
+}
+
+static void hid_scan_output_usage(struct hid_parser *parser, u32 usage)
+{
+ struct hid_device *hid = parser->device;
+
+ if (usage == 0xff000002 && parser->global.report_id == 0x9)
+ parser->scan_flags |= HID_SCAN_FLAG_RMI_WRITE;
+
+ if (usage == 0xff000003 && parser->global.report_id == 0xa)
+ parser->scan_flags |= HID_SCAN_FLAG_RMI_OUTPUT_READ;
}
static void hid_scan_feature_usage(struct hid_parser *parser, u32 usage)
@@ -693,6 +710,9 @@ static void hid_scan_feature_usage(struct hid_parser *parser, u32 usage)
if (usage == 0xff0000c5 && parser->global.report_count == 256 &&
parser->global.report_size == 8)
parser->scan_flags |= HID_SCAN_FLAG_MT_WIN_8;
+
+ if (usage == 0xff000006 && parser->global.report_id == 0xf)
+ parser->scan_flags |= HID_SCAN_FLAG_RMI_MODE;
}
static void hid_scan_collection(struct hid_parser *parser, unsigned type)
@@ -702,6 +722,9 @@ static void hid_scan_collection(struct hid_parser *parser, unsigned type)
if (((parser->global.usage_page << 16) == HID_UP_SENSOR) &&
type == HID_COLLECTION_PHYSICAL)
hid->group = HID_GROUP_SENSOR_HUB;
+
+ if ((parser->global.usage_page << 16) == HID_UP_MSVENDOR)
+ parser->scan_flags |= HID_SCAN_FLAG_VENDOR_SPECIFIC;
}
static int hid_scan_main(struct hid_parser *parser, struct hid_item *item)
@@ -725,6 +748,10 @@ static int hid_scan_main(struct hid_parser *parser, struct hid_item *item)
hid_scan_input_usage(parser, parser->local.usage[i]);
break;
case HID_MAIN_ITEM_TAG_OUTPUT:
+ if (data & HID_MAIN_ITEM_CONSTANT)
+ break;
+ for (i = 0; i < parser->local.usage_index; i++)
+ hid_scan_output_usage(parser, parser->local.usage[i]);
break;
case HID_MAIN_ITEM_TAG_FEATURE:
for (i = 0; i < parser->local.usage_index; i++)
@@ -783,11 +810,16 @@ static int hid_scan_report(struct hid_device *hid)
* Vendor specific handlings
*/
if ((hid->vendor == USB_VENDOR_ID_SYNAPTICS) &&
- (hid->group == HID_GROUP_GENERIC) &&
- /* only bind to the mouse interface of composite USB devices */
- (hid->bus != BUS_USB || hid->type == HID_TYPE_USBMOUSE))
- /* hid-rmi should take care of them, not hid-generic */
- hid->group = HID_GROUP_RMI;
+ (hid->group == HID_GROUP_GENERIC)) {
+ if ((parser->scan_flags & HID_SCAN_FLAG_VENDOR_SPECIFIC)
+ && (parser->scan_flags & HID_SCAN_FLAG_RMI_WRITE)
+ && (parser->scan_flags & HID_SCAN_FLAG_RMI_INPUT_READ)
+ && (parser->scan_flags & HID_SCAN_FLAG_RMI_OUTPUT_READ)
+ && (parser->scan_flags & HID_SCAN_FLAG_RMI_ATTN)
+ && (parser->scan_flags & HID_SCAN_FLAG_RMI_MODE))
Maybe you should consider defining:
#define HID_SCAN_FLAGS_RMI HID_SCAN_FLAG_VENDOR_SPECIFIC | \
HID_SCAN_FLAG_RMI_WRITE | \
HID_SCAN_FLAG_RMI_INPUT_READ | \
HID_SCAN_FLAG_RMI_OUTPUT_READ | \
HID_SCAN_FLAG_RMI_ATTN | \
HID_SCAN_FLAG_RMI_MODE
and then just test:
if ((parser->scan_flags & HID_SCAN_FLAGS_RMI) == HID_SCAN_FLAGS_RMI)
+ /* hid-rmi should take care of them, not hid-generic */
+ hid->group = HID_GROUP_RMI;
+ }
/*
* Vendor specific handlings
diff --git a/include/linux/hid.h b/include/linux/hid.h
index f53c4a9..82d1f82 100644
--- a/include/linux/hid.h
+++ b/include/linux/hid.h
@@ -547,7 +547,13 @@ static inline void hid_set_drvdata(struct hid_device *hdev, void *data)
#define HID_GLOBAL_STACK_SIZE 4
#define HID_COLLECTION_STACK_SIZE 4
-#define HID_SCAN_FLAG_MT_WIN_8 0x00000001
+#define HID_SCAN_FLAG_MT_WIN_8 BIT(0)
+#define HID_SCAN_FLAG_VENDOR_SPECIFIC BIT(1)
+#define HID_SCAN_FLAG_RMI_WRITE BIT(2)
+#define HID_SCAN_FLAG_RMI_INPUT_READ BIT(3)
+#define HID_SCAN_FLAG_RMI_OUTPUT_READ BIT(3)
I guess this one should be BIT(4) and the two after should be shifted by
one too.
Yep, this is a copy and paste mistake. I'll fix it if I end up
submitting a v2.
+#define HID_SCAN_FLAG_RMI_ATTN BIT(4)
+#define HID_SCAN_FLAG_RMI_MODE BIT(5)
struct hid_parser {
struct hid_global global;
--
1.9.1
Cheers,
Benjamin
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/