linux-brain/drivers/hid/hid-wiimote-core.c

1887 lines
48 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-or-later
/*
* HID driver for Nintendo Wii / Wii U peripherals
* Copyright (c) 2011-2013 David Herrmann <dh.herrmann@gmail.com>
*/
/*
*/
#include <linux/completion.h>
#include <linux/device.h>
#include <linux/hid.h>
#include <linux/input.h>
#include <linux/module.h>
#include <linux/mutex.h>
#include <linux/spinlock.h>
#include "hid-ids.h"
#include "hid-wiimote.h"
/* output queue handling */
static int wiimote_hid_send(struct hid_device *hdev, __u8 *buffer,
size_t count)
{
__u8 *buf;
int ret;
if (!hdev->ll_driver->output_report)
return -ENODEV;
buf = kmemdup(buffer, count, GFP_KERNEL);
if (!buf)
return -ENOMEM;
ret = hid_hw_output_report(hdev, buf, count);
kfree(buf);
return ret;
}
static void wiimote_queue_worker(struct work_struct *work)
{
struct wiimote_queue *queue = container_of(work, struct wiimote_queue,
worker);
struct wiimote_data *wdata = container_of(queue, struct wiimote_data,
queue);
unsigned long flags;
int ret;
spin_lock_irqsave(&wdata->queue.lock, flags);
while (wdata->queue.head != wdata->queue.tail) {
spin_unlock_irqrestore(&wdata->queue.lock, flags);
ret = wiimote_hid_send(wdata->hdev,
wdata->queue.outq[wdata->queue.tail].data,
wdata->queue.outq[wdata->queue.tail].size);
if (ret < 0) {
spin_lock_irqsave(&wdata->state.lock, flags);
wiimote_cmd_abort(wdata);
spin_unlock_irqrestore(&wdata->state.lock, flags);
}
spin_lock_irqsave(&wdata->queue.lock, flags);
wdata->queue.tail = (wdata->queue.tail + 1) % WIIMOTE_BUFSIZE;
}
spin_unlock_irqrestore(&wdata->queue.lock, flags);
}
static void wiimote_queue(struct wiimote_data *wdata, const __u8 *buffer,
size_t count)
{
unsigned long flags;
__u8 newhead;
if (count > HID_MAX_BUFFER_SIZE) {
hid_warn(wdata->hdev, "Sending too large output report\n");
spin_lock_irqsave(&wdata->queue.lock, flags);
goto out_error;
}
/*
* Copy new request into our output queue and check whether the
* queue is full. If it is full, discard this request.
* If it is empty we need to start a new worker that will
* send out the buffer to the hid device.
* If the queue is not empty, then there must be a worker
* that is currently sending out our buffer and this worker
* will reschedule itself until the queue is empty.
*/
spin_lock_irqsave(&wdata->queue.lock, flags);
memcpy(wdata->queue.outq[wdata->queue.head].data, buffer, count);
wdata->queue.outq[wdata->queue.head].size = count;
newhead = (wdata->queue.head + 1) % WIIMOTE_BUFSIZE;
if (wdata->queue.head == wdata->queue.tail) {
wdata->queue.head = newhead;
schedule_work(&wdata->queue.worker);
} else if (newhead != wdata->queue.tail) {
wdata->queue.head = newhead;
} else {
hid_warn(wdata->hdev, "Output queue is full");
goto out_error;
}
goto out_unlock;
out_error:
wiimote_cmd_abort(wdata);
out_unlock:
spin_unlock_irqrestore(&wdata->queue.lock, flags);
}
/*
* This sets the rumble bit on the given output report if rumble is
* currently enabled.
* \cmd1 must point to the second byte in the output report => &cmd[1]
* This must be called on nearly every output report before passing it
* into the output queue!
*/
static inline void wiiproto_keep_rumble(struct wiimote_data *wdata, __u8 *cmd1)
{
if (wdata->state.flags & WIIPROTO_FLAG_RUMBLE)
*cmd1 |= 0x01;
}
void wiiproto_req_rumble(struct wiimote_data *wdata, __u8 rumble)
{
__u8 cmd[2];
rumble = !!rumble;
if (rumble == !!(wdata->state.flags & WIIPROTO_FLAG_RUMBLE))
return;
if (rumble)
wdata->state.flags |= WIIPROTO_FLAG_RUMBLE;
else
wdata->state.flags &= ~WIIPROTO_FLAG_RUMBLE;
cmd[0] = WIIPROTO_REQ_RUMBLE;
cmd[1] = 0;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
void wiiproto_req_leds(struct wiimote_data *wdata, int leds)
{
__u8 cmd[2];
leds &= WIIPROTO_FLAGS_LEDS;
if ((wdata->state.flags & WIIPROTO_FLAGS_LEDS) == leds)
return;
wdata->state.flags = (wdata->state.flags & ~WIIPROTO_FLAGS_LEDS) | leds;
cmd[0] = WIIPROTO_REQ_LED;
cmd[1] = 0;
if (leds & WIIPROTO_FLAG_LED1)
cmd[1] |= 0x10;
if (leds & WIIPROTO_FLAG_LED2)
cmd[1] |= 0x20;
if (leds & WIIPROTO_FLAG_LED3)
cmd[1] |= 0x40;
if (leds & WIIPROTO_FLAG_LED4)
cmd[1] |= 0x80;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
/*
* Check what peripherals of the wiimote are currently
* active and select a proper DRM that supports all of
* the requested data inputs.
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
*
* Not all combinations are actually supported. The following
* combinations work only with limitations:
* - IR cam in extended or full mode disables any data transmission
* of extension controllers. There is no DRM mode that supports
* extension bytes plus extended/full IR.
* - IR cam with accelerometer and extension *_EXT8 is not supported.
* However, all extensions that need *_EXT8 are devices that don't
* support IR cameras. Hence, this shouldn't happen under normal
* operation.
* - *_EXT16 is only supported in combination with buttons and
* accelerometer. No IR or similar can be active simultaneously. As
* above, all modules that require it are mutually exclusive with
* IR/etc. so this doesn't matter.
*/
static __u8 select_drm(struct wiimote_data *wdata)
{
__u8 ir = wdata->state.flags & WIIPROTO_FLAGS_IR;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
bool ext;
ext = (wdata->state.flags & WIIPROTO_FLAG_EXT_USED) ||
(wdata->state.flags & WIIPROTO_FLAG_MP_USED);
/* some 3rd-party balance-boards are hard-coded to KEE, *sigh* */
if (wdata->state.devtype == WIIMOTE_DEV_BALANCE_BOARD) {
if (ext)
return WIIPROTO_REQ_DRM_KEE;
else
return WIIPROTO_REQ_DRM_K;
}
if (ir == WIIPROTO_FLAG_IR_BASIC) {
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (wdata->state.flags & WIIPROTO_FLAG_ACCEL) {
/* GEN10 and ealier devices bind IR formats to DRMs.
* Hence, we cannot use DRM_KAI here as it might be
* bound to IR_EXT. Use DRM_KAIE unconditionally so we
* work with all devices and our parsers can use the
* fixed formats, too. */
return WIIPROTO_REQ_DRM_KAIE;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
} else {
return WIIPROTO_REQ_DRM_KIE;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
}
} else if (ir == WIIPROTO_FLAG_IR_EXT) {
return WIIPROTO_REQ_DRM_KAI;
} else if (ir == WIIPROTO_FLAG_IR_FULL) {
return WIIPROTO_REQ_DRM_SKAI1;
} else {
if (wdata->state.flags & WIIPROTO_FLAG_ACCEL) {
if (ext)
return WIIPROTO_REQ_DRM_KAE;
else
return WIIPROTO_REQ_DRM_KA;
} else {
if (ext)
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
return WIIPROTO_REQ_DRM_KEE;
else
return WIIPROTO_REQ_DRM_K;
}
}
}
void wiiproto_req_drm(struct wiimote_data *wdata, __u8 drm)
{
__u8 cmd[3];
if (wdata->state.flags & WIIPROTO_FLAG_DRM_LOCKED)
drm = wdata->state.drm;
else if (drm == WIIPROTO_REQ_NULL)
drm = select_drm(wdata);
cmd[0] = WIIPROTO_REQ_DRM;
cmd[1] = 0;
cmd[2] = drm;
wdata->state.drm = drm;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
void wiiproto_req_status(struct wiimote_data *wdata)
{
__u8 cmd[2];
cmd[0] = WIIPROTO_REQ_SREQ;
cmd[1] = 0;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
void wiiproto_req_accel(struct wiimote_data *wdata, __u8 accel)
{
accel = !!accel;
if (accel == !!(wdata->state.flags & WIIPROTO_FLAG_ACCEL))
return;
if (accel)
wdata->state.flags |= WIIPROTO_FLAG_ACCEL;
else
wdata->state.flags &= ~WIIPROTO_FLAG_ACCEL;
wiiproto_req_drm(wdata, WIIPROTO_REQ_NULL);
}
void wiiproto_req_ir1(struct wiimote_data *wdata, __u8 flags)
{
__u8 cmd[2];
cmd[0] = WIIPROTO_REQ_IR1;
cmd[1] = flags;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
void wiiproto_req_ir2(struct wiimote_data *wdata, __u8 flags)
{
__u8 cmd[2];
cmd[0] = WIIPROTO_REQ_IR2;
cmd[1] = flags;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
#define wiiproto_req_wreg(wdata, os, buf, sz) \
wiiproto_req_wmem((wdata), false, (os), (buf), (sz))
#define wiiproto_req_weeprom(wdata, os, buf, sz) \
wiiproto_req_wmem((wdata), true, (os), (buf), (sz))
static void wiiproto_req_wmem(struct wiimote_data *wdata, bool eeprom,
__u32 offset, const __u8 *buf, __u8 size)
{
__u8 cmd[22];
if (size > 16 || size == 0) {
hid_warn(wdata->hdev, "Invalid length %d wmem request\n", size);
return;
}
memset(cmd, 0, sizeof(cmd));
cmd[0] = WIIPROTO_REQ_WMEM;
cmd[2] = (offset >> 16) & 0xff;
cmd[3] = (offset >> 8) & 0xff;
cmd[4] = offset & 0xff;
cmd[5] = size;
memcpy(&cmd[6], buf, size);
if (!eeprom)
cmd[1] |= 0x04;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
void wiiproto_req_rmem(struct wiimote_data *wdata, bool eeprom, __u32 offset,
__u16 size)
{
__u8 cmd[7];
if (size == 0) {
hid_warn(wdata->hdev, "Invalid length %d rmem request\n", size);
return;
}
cmd[0] = WIIPROTO_REQ_RMEM;
cmd[1] = 0;
cmd[2] = (offset >> 16) & 0xff;
cmd[3] = (offset >> 8) & 0xff;
cmd[4] = offset & 0xff;
cmd[5] = (size >> 8) & 0xff;
cmd[6] = size & 0xff;
if (!eeprom)
cmd[1] |= 0x04;
wiiproto_keep_rumble(wdata, &cmd[1]);
wiimote_queue(wdata, cmd, sizeof(cmd));
}
/* requries the cmd-mutex to be held */
int wiimote_cmd_write(struct wiimote_data *wdata, __u32 offset,
const __u8 *wmem, __u8 size)
{
unsigned long flags;
int ret;
spin_lock_irqsave(&wdata->state.lock, flags);
wiimote_cmd_set(wdata, WIIPROTO_REQ_WMEM, 0);
wiiproto_req_wreg(wdata, offset, wmem, size);
spin_unlock_irqrestore(&wdata->state.lock, flags);
ret = wiimote_cmd_wait(wdata);
if (!ret && wdata->state.cmd_err)
ret = -EIO;
return ret;
}
/* requries the cmd-mutex to be held */
ssize_t wiimote_cmd_read(struct wiimote_data *wdata, __u32 offset, __u8 *rmem,
__u8 size)
{
unsigned long flags;
ssize_t ret;
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.cmd_read_size = size;
wdata->state.cmd_read_buf = rmem;
wiimote_cmd_set(wdata, WIIPROTO_REQ_RMEM, offset & 0xffff);
wiiproto_req_rreg(wdata, offset, size);
spin_unlock_irqrestore(&wdata->state.lock, flags);
ret = wiimote_cmd_wait(wdata);
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.cmd_read_buf = NULL;
spin_unlock_irqrestore(&wdata->state.lock, flags);
if (!ret) {
if (wdata->state.cmd_read_size == 0)
ret = -EIO;
else
ret = wdata->state.cmd_read_size;
}
return ret;
}
/* requires the cmd-mutex to be held */
static int wiimote_cmd_init_ext(struct wiimote_data *wdata)
{
__u8 wmem;
int ret;
/* initialize extension */
wmem = 0x55;
ret = wiimote_cmd_write(wdata, 0xa400f0, &wmem, sizeof(wmem));
if (ret)
return ret;
/* disable default encryption */
wmem = 0x0;
ret = wiimote_cmd_write(wdata, 0xa400fb, &wmem, sizeof(wmem));
if (ret)
return ret;
return 0;
}
/* requires the cmd-mutex to be held */
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
static __u8 wiimote_cmd_read_ext(struct wiimote_data *wdata, __u8 *rmem)
{
int ret;
/* read extension ID */
ret = wiimote_cmd_read(wdata, 0xa400fa, rmem, 6);
if (ret != 6)
return WIIMOTE_EXT_NONE;
hid_dbg(wdata->hdev, "extension ID: %6phC\n", rmem);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (rmem[0] == 0xff && rmem[1] == 0xff && rmem[2] == 0xff &&
rmem[3] == 0xff && rmem[4] == 0xff && rmem[5] == 0xff)
return WIIMOTE_EXT_NONE;
if (rmem[4] == 0x00 && rmem[5] == 0x00)
return WIIMOTE_EXT_NUNCHUK;
if (rmem[4] == 0x01 && rmem[5] == 0x01)
return WIIMOTE_EXT_CLASSIC_CONTROLLER;
if (rmem[4] == 0x04 && rmem[5] == 0x02)
return WIIMOTE_EXT_BALANCE_BOARD;
if (rmem[4] == 0x01 && rmem[5] == 0x20)
return WIIMOTE_EXT_PRO_CONTROLLER;
if (rmem[0] == 0x01 && rmem[1] == 0x00 &&
rmem[4] == 0x01 && rmem[5] == 0x03)
return WIIMOTE_EXT_DRUMS;
if (rmem[0] == 0x00 && rmem[1] == 0x00 &&
rmem[4] == 0x01 && rmem[5] == 0x03)
return WIIMOTE_EXT_GUITAR;
return WIIMOTE_EXT_UNKNOWN;
}
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
/* requires the cmd-mutex to be held */
static int wiimote_cmd_init_mp(struct wiimote_data *wdata)
{
__u8 wmem;
int ret;
/* initialize MP */
wmem = 0x55;
ret = wiimote_cmd_write(wdata, 0xa600f0, &wmem, sizeof(wmem));
if (ret)
return ret;
/* disable default encryption */
wmem = 0x0;
ret = wiimote_cmd_write(wdata, 0xa600fb, &wmem, sizeof(wmem));
if (ret)
return ret;
return 0;
}
/* requires the cmd-mutex to be held */
static bool wiimote_cmd_map_mp(struct wiimote_data *wdata, __u8 exttype)
{
__u8 wmem;
/* map MP with correct pass-through mode */
switch (exttype) {
case WIIMOTE_EXT_CLASSIC_CONTROLLER:
case WIIMOTE_EXT_DRUMS:
case WIIMOTE_EXT_GUITAR:
wmem = 0x07;
break;
case WIIMOTE_EXT_NUNCHUK:
wmem = 0x05;
break;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
default:
wmem = 0x04;
break;
}
return wiimote_cmd_write(wdata, 0xa600fe, &wmem, sizeof(wmem));
}
/* requires the cmd-mutex to be held */
static bool wiimote_cmd_read_mp(struct wiimote_data *wdata, __u8 *rmem)
{
int ret;
/* read motion plus ID */
ret = wiimote_cmd_read(wdata, 0xa600fa, rmem, 6);
if (ret != 6)
return false;
hid_dbg(wdata->hdev, "motion plus ID: %6phC\n", rmem);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (rmem[5] == 0x05)
return true;
hid_info(wdata->hdev, "unknown motion plus ID: %6phC\n", rmem);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
return false;
}
/* requires the cmd-mutex to be held */
static __u8 wiimote_cmd_read_mp_mapped(struct wiimote_data *wdata)
{
int ret;
__u8 rmem[6];
/* read motion plus ID */
ret = wiimote_cmd_read(wdata, 0xa400fa, rmem, 6);
if (ret != 6)
return WIIMOTE_MP_NONE;
hid_dbg(wdata->hdev, "mapped motion plus ID: %6phC\n", rmem);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (rmem[0] == 0xff && rmem[1] == 0xff && rmem[2] == 0xff &&
rmem[3] == 0xff && rmem[4] == 0xff && rmem[5] == 0xff)
return WIIMOTE_MP_NONE;
if (rmem[4] == 0x04 && rmem[5] == 0x05)
return WIIMOTE_MP_SINGLE;
else if (rmem[4] == 0x05 && rmem[5] == 0x05)
return WIIMOTE_MP_PASSTHROUGH_NUNCHUK;
else if (rmem[4] == 0x07 && rmem[5] == 0x05)
return WIIMOTE_MP_PASSTHROUGH_CLASSIC;
return WIIMOTE_MP_UNKNOWN;
}
/* device module handling */
static const __u8 * const wiimote_devtype_mods[WIIMOTE_DEV_NUM] = {
[WIIMOTE_DEV_PENDING] = (const __u8[]){
WIIMOD_NULL,
},
[WIIMOTE_DEV_UNKNOWN] = (const __u8[]){
WIIMOD_NO_MP,
WIIMOD_NULL,
},
[WIIMOTE_DEV_GENERIC] = (const __u8[]){
WIIMOD_KEYS,
WIIMOD_RUMBLE,
WIIMOD_BATTERY,
WIIMOD_LED1,
WIIMOD_LED2,
WIIMOD_LED3,
WIIMOD_LED4,
WIIMOD_ACCEL,
WIIMOD_IR,
WIIMOD_NULL,
},
[WIIMOTE_DEV_GEN10] = (const __u8[]){
WIIMOD_KEYS,
WIIMOD_RUMBLE,
WIIMOD_BATTERY,
WIIMOD_LED1,
WIIMOD_LED2,
WIIMOD_LED3,
WIIMOD_LED4,
WIIMOD_ACCEL,
WIIMOD_IR,
WIIMOD_NULL,
},
[WIIMOTE_DEV_GEN20] = (const __u8[]){
WIIMOD_KEYS,
WIIMOD_RUMBLE,
WIIMOD_BATTERY,
WIIMOD_LED1,
WIIMOD_LED2,
WIIMOD_LED3,
WIIMOD_LED4,
WIIMOD_ACCEL,
WIIMOD_IR,
WIIMOD_BUILTIN_MP,
WIIMOD_NULL,
},
[WIIMOTE_DEV_BALANCE_BOARD] = (const __u8[]) {
WIIMOD_BATTERY,
WIIMOD_LED1,
WIIMOD_NO_MP,
WIIMOD_NULL,
},
[WIIMOTE_DEV_PRO_CONTROLLER] = (const __u8[]) {
WIIMOD_BATTERY,
WIIMOD_LED1,
WIIMOD_LED2,
WIIMOD_LED3,
WIIMOD_LED4,
WIIMOD_NO_MP,
WIIMOD_NULL,
},
};
static void wiimote_modules_load(struct wiimote_data *wdata,
unsigned int devtype)
{
bool need_input = false;
const __u8 *mods, *iter;
const struct wiimod_ops *ops;
int ret;
mods = wiimote_devtype_mods[devtype];
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
if (wiimod_table[*iter]->flags & WIIMOD_FLAG_INPUT) {
need_input = true;
break;
}
}
if (need_input) {
wdata->input = input_allocate_device();
if (!wdata->input)
return;
input_set_drvdata(wdata->input, wdata);
wdata->input->dev.parent = &wdata->hdev->dev;
wdata->input->id.bustype = wdata->hdev->bus;
wdata->input->id.vendor = wdata->hdev->vendor;
wdata->input->id.product = wdata->hdev->product;
wdata->input->id.version = wdata->hdev->version;
wdata->input->name = WIIMOTE_NAME;
}
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
ops = wiimod_table[*iter];
if (!ops->probe)
continue;
ret = ops->probe(ops, wdata);
if (ret)
goto error;
}
if (wdata->input) {
ret = input_register_device(wdata->input);
if (ret)
goto error;
}
spin_lock_irq(&wdata->state.lock);
wdata->state.devtype = devtype;
spin_unlock_irq(&wdata->state.lock);
return;
error:
for ( ; iter-- != mods; ) {
ops = wiimod_table[*iter];
if (ops->remove)
ops->remove(ops, wdata);
}
if (wdata->input) {
input_free_device(wdata->input);
wdata->input = NULL;
}
}
static void wiimote_modules_unload(struct wiimote_data *wdata)
{
const __u8 *mods, *iter;
const struct wiimod_ops *ops;
unsigned long flags;
mods = wiimote_devtype_mods[wdata->state.devtype];
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.devtype = WIIMOTE_DEV_UNKNOWN;
spin_unlock_irqrestore(&wdata->state.lock, flags);
/* find end of list */
for (iter = mods; *iter != WIIMOD_NULL; ++iter)
/* empty */ ;
if (wdata->input) {
input_get_device(wdata->input);
input_unregister_device(wdata->input);
}
for ( ; iter-- != mods; ) {
ops = wiimod_table[*iter];
if (ops->remove)
ops->remove(ops, wdata);
}
if (wdata->input) {
input_put_device(wdata->input);
wdata->input = NULL;
}
}
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
/* device extension handling */
static void wiimote_ext_load(struct wiimote_data *wdata, unsigned int ext)
{
unsigned long flags;
const struct wiimod_ops *ops;
int ret;
ops = wiimod_ext_table[ext];
if (ops->probe) {
ret = ops->probe(ops, wdata);
if (ret)
ext = WIIMOTE_EXT_UNKNOWN;
}
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.exttype = ext;
spin_unlock_irqrestore(&wdata->state.lock, flags);
}
static void wiimote_ext_unload(struct wiimote_data *wdata)
{
unsigned long flags;
const struct wiimod_ops *ops;
ops = wiimod_ext_table[wdata->state.exttype];
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.exttype = WIIMOTE_EXT_UNKNOWN;
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_USED;
spin_unlock_irqrestore(&wdata->state.lock, flags);
if (ops->remove)
ops->remove(ops, wdata);
}
static void wiimote_mp_load(struct wiimote_data *wdata)
{
unsigned long flags;
const struct wiimod_ops *ops;
int ret;
__u8 mode = 2;
ops = &wiimod_mp;
if (ops->probe) {
ret = ops->probe(ops, wdata);
if (ret)
mode = 1;
}
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.mp = mode;
spin_unlock_irqrestore(&wdata->state.lock, flags);
}
static void wiimote_mp_unload(struct wiimote_data *wdata)
{
unsigned long flags;
const struct wiimod_ops *ops;
if (wdata->state.mp < 2)
return;
ops = &wiimod_mp;
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.mp = 0;
wdata->state.flags &= ~WIIPROTO_FLAG_MP_USED;
spin_unlock_irqrestore(&wdata->state.lock, flags);
if (ops->remove)
ops->remove(ops, wdata);
}
/* device (re-)initialization and detection */
static const char *wiimote_devtype_names[WIIMOTE_DEV_NUM] = {
[WIIMOTE_DEV_PENDING] = "Pending",
[WIIMOTE_DEV_UNKNOWN] = "Unknown",
[WIIMOTE_DEV_GENERIC] = "Generic",
[WIIMOTE_DEV_GEN10] = "Nintendo Wii Remote (Gen 1)",
[WIIMOTE_DEV_GEN20] = "Nintendo Wii Remote Plus (Gen 2)",
[WIIMOTE_DEV_BALANCE_BOARD] = "Nintendo Wii Balance Board",
[WIIMOTE_DEV_PRO_CONTROLLER] = "Nintendo Wii U Pro Controller",
};
/* Try to guess the device type based on all collected information. We
* first try to detect by static extension types, then VID/PID and the
* device name. If we cannot detect the device, we use
* WIIMOTE_DEV_GENERIC so all modules will get probed on the device. */
static void wiimote_init_set_type(struct wiimote_data *wdata,
__u8 exttype)
{
__u8 devtype = WIIMOTE_DEV_GENERIC;
__u16 vendor, product;
const char *name;
vendor = wdata->hdev->vendor;
product = wdata->hdev->product;
name = wdata->hdev->name;
if (exttype == WIIMOTE_EXT_BALANCE_BOARD) {
devtype = WIIMOTE_DEV_BALANCE_BOARD;
goto done;
} else if (exttype == WIIMOTE_EXT_PRO_CONTROLLER) {
devtype = WIIMOTE_DEV_PRO_CONTROLLER;
goto done;
}
if (!strcmp(name, "Nintendo RVL-CNT-01")) {
devtype = WIIMOTE_DEV_GEN10;
goto done;
} else if (!strcmp(name, "Nintendo RVL-CNT-01-TR")) {
devtype = WIIMOTE_DEV_GEN20;
goto done;
} else if (!strcmp(name, "Nintendo RVL-WBC-01")) {
devtype = WIIMOTE_DEV_BALANCE_BOARD;
goto done;
} else if (!strcmp(name, "Nintendo RVL-CNT-01-UC")) {
devtype = WIIMOTE_DEV_PRO_CONTROLLER;
goto done;
}
if (vendor == USB_VENDOR_ID_NINTENDO) {
if (product == USB_DEVICE_ID_NINTENDO_WIIMOTE) {
devtype = WIIMOTE_DEV_GEN10;
goto done;
} else if (product == USB_DEVICE_ID_NINTENDO_WIIMOTE2) {
devtype = WIIMOTE_DEV_GEN20;
goto done;
}
}
done:
if (devtype == WIIMOTE_DEV_GENERIC)
hid_info(wdata->hdev, "cannot detect device; NAME: %s VID: %04x PID: %04x EXT: %04x\n",
name, vendor, product, exttype);
else
hid_info(wdata->hdev, "detected device: %s\n",
wiimote_devtype_names[devtype]);
wiimote_modules_load(wdata, devtype);
}
static void wiimote_init_detect(struct wiimote_data *wdata)
{
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
__u8 exttype = WIIMOTE_EXT_NONE, extdata[6];
bool ext;
int ret;
wiimote_cmd_acquire_noint(wdata);
spin_lock_irq(&wdata->state.lock);
wdata->state.devtype = WIIMOTE_DEV_UNKNOWN;
wiimote_cmd_set(wdata, WIIPROTO_REQ_SREQ, 0);
wiiproto_req_status(wdata);
spin_unlock_irq(&wdata->state.lock);
ret = wiimote_cmd_wait_noint(wdata);
if (ret)
goto out_release;
spin_lock_irq(&wdata->state.lock);
ext = wdata->state.flags & WIIPROTO_FLAG_EXT_PLUGGED;
spin_unlock_irq(&wdata->state.lock);
if (!ext)
goto out_release;
wiimote_cmd_init_ext(wdata);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
exttype = wiimote_cmd_read_ext(wdata, extdata);
out_release:
wiimote_cmd_release(wdata);
wiimote_init_set_type(wdata, exttype);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
/* schedule MP timer */
spin_lock_irq(&wdata->state.lock);
if (!(wdata->state.flags & WIIPROTO_FLAG_BUILTIN_MP) &&
!(wdata->state.flags & WIIPROTO_FLAG_NO_MP))
mod_timer(&wdata->timer, jiffies + HZ * 4);
spin_unlock_irq(&wdata->state.lock);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
}
/*
* MP hotplug events are not generated by the wiimote. Therefore, we need
* polling to detect it. We use a 4s interval for polling MP registers. This
* seems reasonable considering applications can trigger it manually via
* sysfs requests.
*/
static void wiimote_init_poll_mp(struct wiimote_data *wdata)
{
bool mp;
__u8 mpdata[6];
wiimote_cmd_acquire_noint(wdata);
wiimote_cmd_init_mp(wdata);
mp = wiimote_cmd_read_mp(wdata, mpdata);
wiimote_cmd_release(wdata);
/* load/unload MP module if it changed */
if (mp) {
if (!wdata->state.mp) {
hid_info(wdata->hdev, "detected extension: Nintendo Wii Motion Plus\n");
wiimote_mp_load(wdata);
}
} else if (wdata->state.mp) {
wiimote_mp_unload(wdata);
}
mod_timer(&wdata->timer, jiffies + HZ * 4);
}
/*
* Check whether the wiimote is in the expected state. The extension registers
* may change during hotplug and initialization so we might get hotplug events
* that we caused by remapping some memory.
* We use some heuristics here to check known states. If the wiimote is in the
* expected state, we can ignore the hotplug event.
*
* Returns "true" if the device is in expected state, "false" if we should
* redo hotplug handling and extension initialization.
*/
static bool wiimote_init_check(struct wiimote_data *wdata)
{
__u32 flags;
__u8 type, data[6];
bool ret, poll_mp;
spin_lock_irq(&wdata->state.lock);
flags = wdata->state.flags;
spin_unlock_irq(&wdata->state.lock);
wiimote_cmd_acquire_noint(wdata);
/* If MP is used and active, but the extension is not, we expect:
* read_mp_mapped() == WIIMOTE_MP_SINGLE
* state.flags == !EXT_ACTIVE && !MP_PLUGGED && MP_ACTIVE
* We do not check EXT_PLUGGED because it might change during
* initialization of MP without extensions.
* - If MP is unplugged/replugged, read_mp_mapped() fails
* - If EXT is plugged, MP_PLUGGED will get set */
if (wdata->state.exttype == WIIMOTE_EXT_NONE &&
wdata->state.mp > 0 && (flags & WIIPROTO_FLAG_MP_USED)) {
type = wiimote_cmd_read_mp_mapped(wdata);
ret = type == WIIMOTE_MP_SINGLE;
spin_lock_irq(&wdata->state.lock);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_EXT_ACTIVE);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_MP_PLUGGED);
ret = ret && (wdata->state.flags & WIIPROTO_FLAG_MP_ACTIVE);
spin_unlock_irq(&wdata->state.lock);
if (!ret)
hid_dbg(wdata->hdev, "state left: !EXT && MP\n");
/* while MP is mapped, we get EXT_PLUGGED events */
poll_mp = false;
goto out_release;
}
/* If MP is unused, but the extension port is used, we expect:
* read_ext == state.exttype
* state.flags == !MP_ACTIVE && EXT_ACTIVE
* - If MP is plugged/unplugged, our timer detects it
* - If EXT is unplugged/replugged, EXT_ACTIVE will become unset */
if (!(flags & WIIPROTO_FLAG_MP_USED) &&
wdata->state.exttype != WIIMOTE_EXT_NONE) {
type = wiimote_cmd_read_ext(wdata, data);
ret = type == wdata->state.exttype;
spin_lock_irq(&wdata->state.lock);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_MP_ACTIVE);
ret = ret && (wdata->state.flags & WIIPROTO_FLAG_EXT_ACTIVE);
spin_unlock_irq(&wdata->state.lock);
if (!ret)
hid_dbg(wdata->hdev, "state left: EXT && !MP\n");
/* poll MP for hotplug events */
poll_mp = true;
goto out_release;
}
/* If neither MP nor an extension are used, we expect:
* read_ext() == WIIMOTE_EXT_NONE
* state.flags == !MP_ACTIVE && !EXT_ACTIVE && !EXT_PLUGGED
* No need to perform any action in this case as everything is
* disabled already.
* - If MP is plugged/unplugged, our timer detects it
* - If EXT is plugged, EXT_PLUGGED will be set */
if (!(flags & WIIPROTO_FLAG_MP_USED) &&
wdata->state.exttype == WIIMOTE_EXT_NONE) {
type = wiimote_cmd_read_ext(wdata, data);
ret = type == wdata->state.exttype;
spin_lock_irq(&wdata->state.lock);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_EXT_ACTIVE);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_MP_ACTIVE);
ret = ret && !(wdata->state.flags & WIIPROTO_FLAG_EXT_PLUGGED);
spin_unlock_irq(&wdata->state.lock);
if (!ret)
hid_dbg(wdata->hdev, "state left: !EXT && !MP\n");
/* poll MP for hotplug events */
poll_mp = true;
goto out_release;
}
/* The trickiest part is if both EXT and MP are active. We cannot read
* the EXT ID, anymore, because MP is mapped over it. However, we use
* a handy trick here:
* - EXT_ACTIVE is unset whenever !MP_PLUGGED is sent
* MP_PLUGGED might be re-sent again before we are scheduled, but
* EXT_ACTIVE will stay unset.
* So it is enough to check for mp_mapped() and MP_ACTIVE and
* EXT_ACTIVE. EXT_PLUGGED is a sanity check. */
if (wdata->state.exttype != WIIMOTE_EXT_NONE &&
wdata->state.mp > 0 && (flags & WIIPROTO_FLAG_MP_USED)) {
type = wiimote_cmd_read_mp_mapped(wdata);
ret = type != WIIMOTE_MP_NONE;
ret = ret && type != WIIMOTE_MP_UNKNOWN;
ret = ret && type != WIIMOTE_MP_SINGLE;
spin_lock_irq(&wdata->state.lock);
ret = ret && (wdata->state.flags & WIIPROTO_FLAG_EXT_PLUGGED);
ret = ret && (wdata->state.flags & WIIPROTO_FLAG_EXT_ACTIVE);
ret = ret && (wdata->state.flags & WIIPROTO_FLAG_MP_ACTIVE);
spin_unlock_irq(&wdata->state.lock);
if (!ret)
hid_dbg(wdata->hdev, "state left: EXT && MP\n");
/* while MP is mapped, we get EXT_PLUGGED events */
poll_mp = false;
goto out_release;
}
/* unknown state */
ret = false;
out_release:
wiimote_cmd_release(wdata);
/* only poll for MP if requested and if state didn't change */
if (ret && poll_mp && !(flags & WIIPROTO_FLAG_BUILTIN_MP) &&
!(flags & WIIPROTO_FLAG_NO_MP))
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
wiimote_init_poll_mp(wdata);
return ret;
}
static const char *wiimote_exttype_names[WIIMOTE_EXT_NUM] = {
[WIIMOTE_EXT_NONE] = "None",
[WIIMOTE_EXT_UNKNOWN] = "Unknown",
[WIIMOTE_EXT_NUNCHUK] = "Nintendo Wii Nunchuk",
[WIIMOTE_EXT_CLASSIC_CONTROLLER] = "Nintendo Wii Classic Controller",
[WIIMOTE_EXT_BALANCE_BOARD] = "Nintendo Wii Balance Board",
[WIIMOTE_EXT_PRO_CONTROLLER] = "Nintendo Wii U Pro Controller",
[WIIMOTE_EXT_DRUMS] = "Nintendo Wii Drums",
[WIIMOTE_EXT_GUITAR] = "Nintendo Wii Guitar",
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
};
/*
* Handle hotplug events
* If we receive an hotplug event and the device-check failed, we deinitialize
* the extension ports, re-read all extension IDs and set the device into
* the desired state. This involves mapping MP into the main extension
* registers, setting up extension passthrough modes and initializing the
* requested extensions.
*/
static void wiimote_init_hotplug(struct wiimote_data *wdata)
{
__u8 exttype, extdata[6], mpdata[6];
__u32 flags;
bool mp;
hid_dbg(wdata->hdev, "detect extensions..\n");
wiimote_cmd_acquire_noint(wdata);
spin_lock_irq(&wdata->state.lock);
/* get state snapshot that we will then work on */
flags = wdata->state.flags;
/* disable event forwarding temporarily */
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_ACTIVE;
wdata->state.flags &= ~WIIPROTO_FLAG_MP_ACTIVE;
spin_unlock_irq(&wdata->state.lock);
/* init extension and MP (deactivates current extension or MP) */
wiimote_cmd_init_ext(wdata);
if (flags & WIIPROTO_FLAG_NO_MP) {
mp = false;
} else {
wiimote_cmd_init_mp(wdata);
mp = wiimote_cmd_read_mp(wdata, mpdata);
}
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
exttype = wiimote_cmd_read_ext(wdata, extdata);
wiimote_cmd_release(wdata);
/* load/unload extension module if it changed */
if (exttype != wdata->state.exttype) {
/* unload previous extension */
wiimote_ext_unload(wdata);
if (exttype == WIIMOTE_EXT_UNKNOWN) {
hid_info(wdata->hdev, "cannot detect extension; %6phC\n",
extdata);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
} else if (exttype == WIIMOTE_EXT_NONE) {
spin_lock_irq(&wdata->state.lock);
wdata->state.exttype = WIIMOTE_EXT_NONE;
spin_unlock_irq(&wdata->state.lock);
} else {
hid_info(wdata->hdev, "detected extension: %s\n",
wiimote_exttype_names[exttype]);
/* try loading new extension */
wiimote_ext_load(wdata, exttype);
}
}
/* load/unload MP module if it changed */
if (mp) {
if (!wdata->state.mp) {
hid_info(wdata->hdev, "detected extension: Nintendo Wii Motion Plus\n");
wiimote_mp_load(wdata);
}
} else if (wdata->state.mp) {
wiimote_mp_unload(wdata);
}
/* if MP is not used, do not map or activate it */
if (!(flags & WIIPROTO_FLAG_MP_USED))
mp = false;
/* map MP into main extension registers if used */
if (mp) {
wiimote_cmd_acquire_noint(wdata);
wiimote_cmd_map_mp(wdata, exttype);
wiimote_cmd_release(wdata);
/* delete MP hotplug timer */
del_timer_sync(&wdata->timer);
} else {
/* reschedule MP hotplug timer */
if (!(flags & WIIPROTO_FLAG_BUILTIN_MP) &&
!(flags & WIIPROTO_FLAG_NO_MP))
mod_timer(&wdata->timer, jiffies + HZ * 4);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
}
spin_lock_irq(&wdata->state.lock);
/* enable data forwarding again and set expected hotplug state */
if (mp) {
wdata->state.flags |= WIIPROTO_FLAG_MP_ACTIVE;
if (wdata->state.exttype == WIIMOTE_EXT_NONE) {
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_PLUGGED;
wdata->state.flags &= ~WIIPROTO_FLAG_MP_PLUGGED;
} else {
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_PLUGGED;
wdata->state.flags |= WIIPROTO_FLAG_MP_PLUGGED;
wdata->state.flags |= WIIPROTO_FLAG_EXT_ACTIVE;
}
} else if (wdata->state.exttype != WIIMOTE_EXT_NONE) {
wdata->state.flags |= WIIPROTO_FLAG_EXT_ACTIVE;
}
/* request status report for hotplug state updates */
wiiproto_req_status(wdata);
spin_unlock_irq(&wdata->state.lock);
hid_dbg(wdata->hdev, "detected extensions: MP: %d EXT: %d\n",
wdata->state.mp, wdata->state.exttype);
}
static void wiimote_init_worker(struct work_struct *work)
{
struct wiimote_data *wdata = container_of(work, struct wiimote_data,
init_worker);
bool changed = false;
if (wdata->state.devtype == WIIMOTE_DEV_PENDING) {
wiimote_init_detect(wdata);
changed = true;
}
if (changed || !wiimote_init_check(wdata))
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
wiimote_init_hotplug(wdata);
if (changed)
kobject_uevent(&wdata->hdev->dev.kobj, KOBJ_CHANGE);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
}
void __wiimote_schedule(struct wiimote_data *wdata)
{
if (!(wdata->state.flags & WIIPROTO_FLAG_EXITING))
schedule_work(&wdata->init_worker);
}
static void wiimote_schedule(struct wiimote_data *wdata)
{
unsigned long flags;
spin_lock_irqsave(&wdata->state.lock, flags);
__wiimote_schedule(wdata);
spin_unlock_irqrestore(&wdata->state.lock, flags);
}
treewide: setup_timer() -> timer_setup() This converts all remaining cases of the old setup_timer() API into using timer_setup(), where the callback argument is the structure already holding the struct timer_list. These should have no behavioral changes, since they just change which pointer is passed into the callback with the same available pointers after conversion. It handles the following examples, in addition to some other variations. Casting from unsigned long: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... setup_timer(&ptr->my_timer, my_callback, ptr); and forced object casts: void my_callback(struct something *ptr) { ... } ... setup_timer(&ptr->my_timer, my_callback, (unsigned long)ptr); become: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... timer_setup(&ptr->my_timer, my_callback, 0); Direct function assignments: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... ptr->my_timer.function = my_callback; have a temporary cast added, along with converting the args: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... ptr->my_timer.function = (TIMER_FUNC_TYPE)my_callback; And finally, callbacks without a data assignment: void my_callback(unsigned long data) { ... } ... setup_timer(&ptr->my_timer, my_callback, 0); have their argument renamed to verify they're unused during conversion: void my_callback(struct timer_list *unused) { ... } ... timer_setup(&ptr->my_timer, my_callback, 0); The conversion is done with the following Coccinelle script: spatch --very-quiet --all-includes --include-headers \ -I ./arch/x86/include -I ./arch/x86/include/generated \ -I ./include -I ./arch/x86/include/uapi \ -I ./arch/x86/include/generated/uapi -I ./include/uapi \ -I ./include/generated/uapi --include ./include/linux/kconfig.h \ --dir . \ --cocci-file ~/src/data/timer_setup.cocci @fix_address_of@ expression e; @@ setup_timer( -&(e) +&e , ...) // Update any raw setup_timer() usages that have a NULL callback, but // would otherwise match change_timer_function_usage, since the latter // will update all function assignments done in the face of a NULL // function initialization in setup_timer(). @change_timer_function_usage_NULL@ expression _E; identifier _timer; type _cast_data; @@ ( -setup_timer(&_E->_timer, NULL, _E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E->_timer, NULL, (_cast_data)_E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E._timer, NULL, &_E); +timer_setup(&_E._timer, NULL, 0); | -setup_timer(&_E._timer, NULL, (_cast_data)&_E); +timer_setup(&_E._timer, NULL, 0); ) @change_timer_function_usage@ expression _E; identifier _timer; struct timer_list _stl; identifier _callback; type _cast_func, _cast_data; @@ ( -setup_timer(&_E->_timer, _callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | _E->_timer@_stl.function = _callback; | _E->_timer@_stl.function = &_callback; | _E->_timer@_stl.function = (_cast_func)_callback; | _E->_timer@_stl.function = (_cast_func)&_callback; | _E._timer@_stl.function = _callback; | _E._timer@_stl.function = &_callback; | _E._timer@_stl.function = (_cast_func)_callback; | _E._timer@_stl.function = (_cast_func)&_callback; ) // callback(unsigned long arg) @change_callback_handle_cast depends on change_timer_function_usage@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; identifier _handle; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { ( ... when != _origarg _handletype *_handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg ) } // callback(unsigned long arg) without existing variable @change_callback_handle_cast_no_arg depends on change_timer_function_usage && !change_callback_handle_cast@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { + _handletype *_origarg = from_timer(_origarg, t, _timer); + ... when != _origarg - (_handletype *)_origarg + _origarg ... when != _origarg } // Avoid already converted callbacks. @match_callback_converted depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier t; @@ void _callback(struct timer_list *t) { ... } // callback(struct something *handle) @change_callback_handle_arg depends on change_timer_function_usage && !match_callback_converted && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; @@ void _callback( -_handletype *_handle +struct timer_list *t ) { + _handletype *_handle = from_timer(_handle, t, _timer); ... } // If change_callback_handle_arg ran on an empty function, remove // the added handler. @unchange_callback_handle_arg depends on change_timer_function_usage && change_callback_handle_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; identifier t; @@ void _callback(struct timer_list *t) { - _handletype *_handle = from_timer(_handle, t, _timer); } // We only want to refactor the setup_timer() data argument if we've found // the matching callback. This undoes changes in change_timer_function_usage. @unchange_timer_function_usage depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg && !change_callback_handle_arg@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type change_timer_function_usage._cast_data; @@ ( -timer_setup(&_E->_timer, _callback, 0); +setup_timer(&_E->_timer, _callback, (_cast_data)_E); | -timer_setup(&_E._timer, _callback, 0); +setup_timer(&_E._timer, _callback, (_cast_data)&_E); ) // If we fixed a callback from a .function assignment, fix the // assignment cast now. @change_timer_function_assignment depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_func; typedef TIMER_FUNC_TYPE; @@ ( _E->_timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -&_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)_callback; +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -&_callback; +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; ) // Sometimes timer functions are called directly. Replace matched args. @change_timer_function_calls depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression _E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_data; @@ _callback( ( -(_cast_data)_E +&_E->_timer | -(_cast_data)&_E +&_E._timer | -_E +&_E->_timer ) ) // If a timer has been configured without a data argument, it can be // converted without regard to the callback argument, since it is unused. @match_timer_function_unused_data@ expression _E; identifier _timer; identifier _callback; @@ ( -setup_timer(&_E->_timer, _callback, 0); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0L); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0UL); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0L); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0UL); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_timer, _callback, 0); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0L); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0UL); +timer_setup(&_timer, _callback, 0); | -setup_timer(_timer, _callback, 0); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0L); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0UL); +timer_setup(_timer, _callback, 0); ) @change_callback_unused_data depends on match_timer_function_unused_data@ identifier match_timer_function_unused_data._callback; type _origtype; identifier _origarg; @@ void _callback( -_origtype _origarg +struct timer_list *unused ) { ... when != _origarg } Signed-off-by: Kees Cook <keescook@chromium.org>
2017-10-17 06:43:17 +09:00
static void wiimote_init_timeout(struct timer_list *t)
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
{
treewide: setup_timer() -> timer_setup() This converts all remaining cases of the old setup_timer() API into using timer_setup(), where the callback argument is the structure already holding the struct timer_list. These should have no behavioral changes, since they just change which pointer is passed into the callback with the same available pointers after conversion. It handles the following examples, in addition to some other variations. Casting from unsigned long: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... setup_timer(&ptr->my_timer, my_callback, ptr); and forced object casts: void my_callback(struct something *ptr) { ... } ... setup_timer(&ptr->my_timer, my_callback, (unsigned long)ptr); become: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... timer_setup(&ptr->my_timer, my_callback, 0); Direct function assignments: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... ptr->my_timer.function = my_callback; have a temporary cast added, along with converting the args: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... ptr->my_timer.function = (TIMER_FUNC_TYPE)my_callback; And finally, callbacks without a data assignment: void my_callback(unsigned long data) { ... } ... setup_timer(&ptr->my_timer, my_callback, 0); have their argument renamed to verify they're unused during conversion: void my_callback(struct timer_list *unused) { ... } ... timer_setup(&ptr->my_timer, my_callback, 0); The conversion is done with the following Coccinelle script: spatch --very-quiet --all-includes --include-headers \ -I ./arch/x86/include -I ./arch/x86/include/generated \ -I ./include -I ./arch/x86/include/uapi \ -I ./arch/x86/include/generated/uapi -I ./include/uapi \ -I ./include/generated/uapi --include ./include/linux/kconfig.h \ --dir . \ --cocci-file ~/src/data/timer_setup.cocci @fix_address_of@ expression e; @@ setup_timer( -&(e) +&e , ...) // Update any raw setup_timer() usages that have a NULL callback, but // would otherwise match change_timer_function_usage, since the latter // will update all function assignments done in the face of a NULL // function initialization in setup_timer(). @change_timer_function_usage_NULL@ expression _E; identifier _timer; type _cast_data; @@ ( -setup_timer(&_E->_timer, NULL, _E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E->_timer, NULL, (_cast_data)_E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E._timer, NULL, &_E); +timer_setup(&_E._timer, NULL, 0); | -setup_timer(&_E._timer, NULL, (_cast_data)&_E); +timer_setup(&_E._timer, NULL, 0); ) @change_timer_function_usage@ expression _E; identifier _timer; struct timer_list _stl; identifier _callback; type _cast_func, _cast_data; @@ ( -setup_timer(&_E->_timer, _callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | _E->_timer@_stl.function = _callback; | _E->_timer@_stl.function = &_callback; | _E->_timer@_stl.function = (_cast_func)_callback; | _E->_timer@_stl.function = (_cast_func)&_callback; | _E._timer@_stl.function = _callback; | _E._timer@_stl.function = &_callback; | _E._timer@_stl.function = (_cast_func)_callback; | _E._timer@_stl.function = (_cast_func)&_callback; ) // callback(unsigned long arg) @change_callback_handle_cast depends on change_timer_function_usage@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; identifier _handle; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { ( ... when != _origarg _handletype *_handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg ) } // callback(unsigned long arg) without existing variable @change_callback_handle_cast_no_arg depends on change_timer_function_usage && !change_callback_handle_cast@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { + _handletype *_origarg = from_timer(_origarg, t, _timer); + ... when != _origarg - (_handletype *)_origarg + _origarg ... when != _origarg } // Avoid already converted callbacks. @match_callback_converted depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier t; @@ void _callback(struct timer_list *t) { ... } // callback(struct something *handle) @change_callback_handle_arg depends on change_timer_function_usage && !match_callback_converted && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; @@ void _callback( -_handletype *_handle +struct timer_list *t ) { + _handletype *_handle = from_timer(_handle, t, _timer); ... } // If change_callback_handle_arg ran on an empty function, remove // the added handler. @unchange_callback_handle_arg depends on change_timer_function_usage && change_callback_handle_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; identifier t; @@ void _callback(struct timer_list *t) { - _handletype *_handle = from_timer(_handle, t, _timer); } // We only want to refactor the setup_timer() data argument if we've found // the matching callback. This undoes changes in change_timer_function_usage. @unchange_timer_function_usage depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg && !change_callback_handle_arg@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type change_timer_function_usage._cast_data; @@ ( -timer_setup(&_E->_timer, _callback, 0); +setup_timer(&_E->_timer, _callback, (_cast_data)_E); | -timer_setup(&_E._timer, _callback, 0); +setup_timer(&_E._timer, _callback, (_cast_data)&_E); ) // If we fixed a callback from a .function assignment, fix the // assignment cast now. @change_timer_function_assignment depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_func; typedef TIMER_FUNC_TYPE; @@ ( _E->_timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -&_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)_callback; +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -&_callback; +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; ) // Sometimes timer functions are called directly. Replace matched args. @change_timer_function_calls depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression _E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_data; @@ _callback( ( -(_cast_data)_E +&_E->_timer | -(_cast_data)&_E +&_E._timer | -_E +&_E->_timer ) ) // If a timer has been configured without a data argument, it can be // converted without regard to the callback argument, since it is unused. @match_timer_function_unused_data@ expression _E; identifier _timer; identifier _callback; @@ ( -setup_timer(&_E->_timer, _callback, 0); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0L); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0UL); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0L); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0UL); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_timer, _callback, 0); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0L); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0UL); +timer_setup(&_timer, _callback, 0); | -setup_timer(_timer, _callback, 0); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0L); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0UL); +timer_setup(_timer, _callback, 0); ) @change_callback_unused_data depends on match_timer_function_unused_data@ identifier match_timer_function_unused_data._callback; type _origtype; identifier _origarg; @@ void _callback( -_origtype _origarg +struct timer_list *unused ) { ... when != _origarg } Signed-off-by: Kees Cook <keescook@chromium.org>
2017-10-17 06:43:17 +09:00
struct wiimote_data *wdata = from_timer(wdata, t, timer);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
wiimote_schedule(wdata);
}
/* protocol handlers */
static void handler_keys(struct wiimote_data *wdata, const __u8 *payload)
{
const __u8 *iter, *mods;
const struct wiimod_ops *ops;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
ops = wiimod_ext_table[wdata->state.exttype];
if (ops->in_keys) {
ops->in_keys(wdata, payload);
return;
}
mods = wiimote_devtype_mods[wdata->state.devtype];
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
ops = wiimod_table[*iter];
if (ops->in_keys) {
ops->in_keys(wdata, payload);
break;
}
}
}
static void handler_accel(struct wiimote_data *wdata, const __u8 *payload)
{
const __u8 *iter, *mods;
const struct wiimod_ops *ops;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
ops = wiimod_ext_table[wdata->state.exttype];
if (ops->in_accel) {
ops->in_accel(wdata, payload);
return;
}
mods = wiimote_devtype_mods[wdata->state.devtype];
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
ops = wiimod_table[*iter];
if (ops->in_accel) {
ops->in_accel(wdata, payload);
break;
}
}
}
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
static bool valid_ext_handler(const struct wiimod_ops *ops, size_t len)
{
if (!ops->in_ext)
return false;
if ((ops->flags & WIIMOD_FLAG_EXT8) && len < 8)
return false;
if ((ops->flags & WIIMOD_FLAG_EXT16) && len < 16)
return false;
return true;
}
static void handler_ext(struct wiimote_data *wdata, const __u8 *payload,
size_t len)
{
static const __u8 invalid[21] = { 0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff,
0xff };
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
const __u8 *iter, *mods;
const struct wiimod_ops *ops;
bool is_mp;
if (len > 21)
len = 21;
if (len < 6 || !memcmp(payload, invalid, len))
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
return;
/* if MP is active, track MP slot hotplugging */
if (wdata->state.flags & WIIPROTO_FLAG_MP_ACTIVE) {
/* this bit is set for invalid events (eg. during hotplug) */
if (payload[5] & 0x01)
return;
if (payload[4] & 0x01) {
if (!(wdata->state.flags & WIIPROTO_FLAG_MP_PLUGGED)) {
hid_dbg(wdata->hdev, "MP hotplug: 1\n");
wdata->state.flags |= WIIPROTO_FLAG_MP_PLUGGED;
__wiimote_schedule(wdata);
}
} else {
if (wdata->state.flags & WIIPROTO_FLAG_MP_PLUGGED) {
hid_dbg(wdata->hdev, "MP hotplug: 0\n");
wdata->state.flags &= ~WIIPROTO_FLAG_MP_PLUGGED;
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_ACTIVE;
__wiimote_schedule(wdata);
}
}
/* detect MP data that is sent interleaved with EXT data */
is_mp = payload[5] & 0x02;
} else {
is_mp = false;
}
/* ignore EXT events if no extension is active */
if (!(wdata->state.flags & WIIPROTO_FLAG_EXT_ACTIVE) && !is_mp)
return;
/* try forwarding to extension handler, first */
ops = wiimod_ext_table[wdata->state.exttype];
if (is_mp && ops->in_mp) {
ops->in_mp(wdata, payload);
return;
} else if (!is_mp && valid_ext_handler(ops, len)) {
ops->in_ext(wdata, payload);
return;
}
/* try forwarding to MP handler */
ops = &wiimod_mp;
if (is_mp && ops->in_mp) {
ops->in_mp(wdata, payload);
return;
} else if (!is_mp && valid_ext_handler(ops, len)) {
ops->in_ext(wdata, payload);
return;
}
/* try forwarding to loaded modules */
mods = wiimote_devtype_mods[wdata->state.devtype];
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
ops = wiimod_table[*iter];
if (is_mp && ops->in_mp) {
ops->in_mp(wdata, payload);
return;
} else if (!is_mp && valid_ext_handler(ops, len)) {
ops->in_ext(wdata, payload);
return;
}
}
}
#define ir_to_input0(wdata, ir, packed) handler_ir((wdata), (ir), (packed), 0)
#define ir_to_input1(wdata, ir, packed) handler_ir((wdata), (ir), (packed), 1)
#define ir_to_input2(wdata, ir, packed) handler_ir((wdata), (ir), (packed), 2)
#define ir_to_input3(wdata, ir, packed) handler_ir((wdata), (ir), (packed), 3)
static void handler_ir(struct wiimote_data *wdata, const __u8 *payload,
bool packed, unsigned int id)
{
const __u8 *iter, *mods;
const struct wiimod_ops *ops;
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
ops = wiimod_ext_table[wdata->state.exttype];
if (ops->in_ir) {
ops->in_ir(wdata, payload, packed, id);
return;
}
mods = wiimote_devtype_mods[wdata->state.devtype];
for (iter = mods; *iter != WIIMOD_NULL; ++iter) {
ops = wiimod_table[*iter];
if (ops->in_ir) {
ops->in_ir(wdata, payload, packed, id);
break;
}
}
}
/* reduced status report with "BB BB" key data only */
static void handler_status_K(struct wiimote_data *wdata,
const __u8 *payload)
{
handler_keys(wdata, payload);
/* on status reports the drm is reset so we need to resend the drm */
wiiproto_req_drm(wdata, WIIPROTO_REQ_NULL);
}
/* extended status report with "BB BB LF 00 00 VV" data */
static void handler_status(struct wiimote_data *wdata, const __u8 *payload)
{
handler_status_K(wdata, payload);
/* update extension status */
if (payload[2] & 0x02) {
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (!(wdata->state.flags & WIIPROTO_FLAG_EXT_PLUGGED)) {
hid_dbg(wdata->hdev, "EXT hotplug: 1\n");
wdata->state.flags |= WIIPROTO_FLAG_EXT_PLUGGED;
__wiimote_schedule(wdata);
}
} else {
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
if (wdata->state.flags & WIIPROTO_FLAG_EXT_PLUGGED) {
hid_dbg(wdata->hdev, "EXT hotplug: 0\n");
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_PLUGGED;
wdata->state.flags &= ~WIIPROTO_FLAG_MP_PLUGGED;
wdata->state.flags &= ~WIIPROTO_FLAG_EXT_ACTIVE;
wdata->state.flags &= ~WIIPROTO_FLAG_MP_ACTIVE;
__wiimote_schedule(wdata);
}
}
wdata->state.cmd_battery = payload[5];
if (wiimote_cmd_pending(wdata, WIIPROTO_REQ_SREQ, 0))
wiimote_cmd_complete(wdata);
}
/* reduced generic report with "BB BB" key data only */
static void handler_generic_K(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
}
static void handler_data(struct wiimote_data *wdata, const __u8 *payload)
{
__u16 offset = payload[3] << 8 | payload[4];
__u8 size = (payload[2] >> 4) + 1;
__u8 err = payload[2] & 0x0f;
handler_keys(wdata, payload);
if (wiimote_cmd_pending(wdata, WIIPROTO_REQ_RMEM, offset)) {
if (err)
size = 0;
else if (size > wdata->state.cmd_read_size)
size = wdata->state.cmd_read_size;
wdata->state.cmd_read_size = size;
if (wdata->state.cmd_read_buf)
memcpy(wdata->state.cmd_read_buf, &payload[5], size);
wiimote_cmd_complete(wdata);
}
}
static void handler_return(struct wiimote_data *wdata, const __u8 *payload)
{
__u8 err = payload[3];
__u8 cmd = payload[2];
handler_keys(wdata, payload);
if (wiimote_cmd_pending(wdata, cmd, 0)) {
wdata->state.cmd_err = err;
wiimote_cmd_complete(wdata);
} else if (err) {
hid_warn(wdata->hdev, "Remote error %hhu on req %hhu\n", err,
cmd);
}
}
static void handler_drm_KA(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
handler_accel(wdata, payload);
}
static void handler_drm_KE(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, &payload[2], 8);
}
static void handler_drm_KAI(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
handler_accel(wdata, payload);
ir_to_input0(wdata, &payload[5], false);
ir_to_input1(wdata, &payload[8], false);
ir_to_input2(wdata, &payload[11], false);
ir_to_input3(wdata, &payload[14], false);
}
static void handler_drm_KEE(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, &payload[2], 19);
}
static void handler_drm_KIE(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
ir_to_input0(wdata, &payload[2], false);
ir_to_input1(wdata, &payload[4], true);
ir_to_input2(wdata, &payload[7], false);
ir_to_input3(wdata, &payload[9], true);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, &payload[12], 9);
}
static void handler_drm_KAE(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
handler_accel(wdata, payload);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, &payload[5], 16);
}
static void handler_drm_KAIE(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
handler_accel(wdata, payload);
ir_to_input0(wdata, &payload[5], false);
ir_to_input1(wdata, &payload[7], true);
ir_to_input2(wdata, &payload[10], false);
ir_to_input3(wdata, &payload[12], true);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, &payload[15], 6);
}
static void handler_drm_E(struct wiimote_data *wdata, const __u8 *payload)
{
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
handler_ext(wdata, payload, 21);
}
static void handler_drm_SKAI1(struct wiimote_data *wdata, const __u8 *payload)
{
handler_keys(wdata, payload);
wdata->state.accel_split[0] = payload[2];
wdata->state.accel_split[1] = (payload[0] >> 1) & (0x10 | 0x20);
wdata->state.accel_split[1] |= (payload[1] << 1) & (0x40 | 0x80);
ir_to_input0(wdata, &payload[3], false);
ir_to_input1(wdata, &payload[12], false);
}
static void handler_drm_SKAI2(struct wiimote_data *wdata, const __u8 *payload)
{
__u8 buf[5];
handler_keys(wdata, payload);
wdata->state.accel_split[1] |= (payload[0] >> 5) & (0x01 | 0x02);
wdata->state.accel_split[1] |= (payload[1] >> 3) & (0x04 | 0x08);
buf[0] = 0;
buf[1] = 0;
buf[2] = wdata->state.accel_split[0];
buf[3] = payload[2];
buf[4] = wdata->state.accel_split[1];
handler_accel(wdata, buf);
ir_to_input2(wdata, &payload[3], false);
ir_to_input3(wdata, &payload[12], false);
}
struct wiiproto_handler {
__u8 id;
size_t size;
void (*func)(struct wiimote_data *wdata, const __u8 *payload);
};
static struct wiiproto_handler handlers[] = {
{ .id = WIIPROTO_REQ_STATUS, .size = 6, .func = handler_status },
{ .id = WIIPROTO_REQ_STATUS, .size = 2, .func = handler_status_K },
{ .id = WIIPROTO_REQ_DATA, .size = 21, .func = handler_data },
{ .id = WIIPROTO_REQ_DATA, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_RETURN, .size = 4, .func = handler_return },
{ .id = WIIPROTO_REQ_RETURN, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_K, .size = 2, .func = handler_keys },
{ .id = WIIPROTO_REQ_DRM_KA, .size = 5, .func = handler_drm_KA },
{ .id = WIIPROTO_REQ_DRM_KA, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KE, .size = 10, .func = handler_drm_KE },
{ .id = WIIPROTO_REQ_DRM_KE, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KAI, .size = 17, .func = handler_drm_KAI },
{ .id = WIIPROTO_REQ_DRM_KAI, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KEE, .size = 21, .func = handler_drm_KEE },
{ .id = WIIPROTO_REQ_DRM_KEE, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KAE, .size = 21, .func = handler_drm_KAE },
{ .id = WIIPROTO_REQ_DRM_KAE, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KIE, .size = 21, .func = handler_drm_KIE },
{ .id = WIIPROTO_REQ_DRM_KIE, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_KAIE, .size = 21, .func = handler_drm_KAIE },
{ .id = WIIPROTO_REQ_DRM_KAIE, .size = 2, .func = handler_generic_K },
{ .id = WIIPROTO_REQ_DRM_E, .size = 21, .func = handler_drm_E },
{ .id = WIIPROTO_REQ_DRM_SKAI1, .size = 21, .func = handler_drm_SKAI1 },
{ .id = WIIPROTO_REQ_DRM_SKAI2, .size = 21, .func = handler_drm_SKAI2 },
{ .id = 0 }
};
static int wiimote_hid_event(struct hid_device *hdev, struct hid_report *report,
u8 *raw_data, int size)
{
struct wiimote_data *wdata = hid_get_drvdata(hdev);
struct wiiproto_handler *h;
int i;
unsigned long flags;
if (size < 1)
return -EINVAL;
spin_lock_irqsave(&wdata->state.lock, flags);
for (i = 0; handlers[i].id; ++i) {
h = &handlers[i];
if (h->id == raw_data[0] && h->size < size) {
h->func(wdata, &raw_data[1]);
break;
}
}
if (!handlers[i].id)
hid_warn(hdev, "Unhandled report %hhu size %d\n", raw_data[0],
size);
spin_unlock_irqrestore(&wdata->state.lock, flags);
return 0;
}
static ssize_t wiimote_ext_show(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct wiimote_data *wdata = dev_to_wii(dev);
__u8 type;
unsigned long flags;
spin_lock_irqsave(&wdata->state.lock, flags);
type = wdata->state.exttype;
spin_unlock_irqrestore(&wdata->state.lock, flags);
switch (type) {
case WIIMOTE_EXT_NONE:
return sprintf(buf, "none\n");
case WIIMOTE_EXT_NUNCHUK:
return sprintf(buf, "nunchuk\n");
case WIIMOTE_EXT_CLASSIC_CONTROLLER:
return sprintf(buf, "classic\n");
case WIIMOTE_EXT_BALANCE_BOARD:
return sprintf(buf, "balanceboard\n");
case WIIMOTE_EXT_PRO_CONTROLLER:
return sprintf(buf, "procontroller\n");
case WIIMOTE_EXT_DRUMS:
return sprintf(buf, "drums\n");
case WIIMOTE_EXT_GUITAR:
return sprintf(buf, "guitar\n");
case WIIMOTE_EXT_UNKNOWN:
/* fallthrough */
default:
return sprintf(buf, "unknown\n");
}
}
static ssize_t wiimote_ext_store(struct device *dev,
struct device_attribute *attr,
const char *buf, size_t count)
{
struct wiimote_data *wdata = dev_to_wii(dev);
if (!strcmp(buf, "scan")) {
wiimote_schedule(wdata);
} else {
return -EINVAL;
}
return strnlen(buf, PAGE_SIZE);
}
static DEVICE_ATTR(extension, S_IRUGO | S_IWUSR | S_IWGRP, wiimote_ext_show,
wiimote_ext_store);
static ssize_t wiimote_dev_show(struct device *dev,
struct device_attribute *attr,
char *buf)
{
struct wiimote_data *wdata = dev_to_wii(dev);
__u8 type;
unsigned long flags;
spin_lock_irqsave(&wdata->state.lock, flags);
type = wdata->state.devtype;
spin_unlock_irqrestore(&wdata->state.lock, flags);
switch (type) {
case WIIMOTE_DEV_GENERIC:
return sprintf(buf, "generic\n");
case WIIMOTE_DEV_GEN10:
return sprintf(buf, "gen10\n");
case WIIMOTE_DEV_GEN20:
return sprintf(buf, "gen20\n");
case WIIMOTE_DEV_BALANCE_BOARD:
return sprintf(buf, "balanceboard\n");
case WIIMOTE_DEV_PRO_CONTROLLER:
return sprintf(buf, "procontroller\n");
case WIIMOTE_DEV_PENDING:
return sprintf(buf, "pending\n");
case WIIMOTE_DEV_UNKNOWN:
/* fallthrough */
default:
return sprintf(buf, "unknown\n");
}
}
static DEVICE_ATTR(devtype, S_IRUGO, wiimote_dev_show, NULL);
static struct wiimote_data *wiimote_create(struct hid_device *hdev)
{
struct wiimote_data *wdata;
wdata = kzalloc(sizeof(*wdata), GFP_KERNEL);
if (!wdata)
return NULL;
wdata->hdev = hdev;
hid_set_drvdata(hdev, wdata);
spin_lock_init(&wdata->queue.lock);
INIT_WORK(&wdata->queue.worker, wiimote_queue_worker);
spin_lock_init(&wdata->state.lock);
init_completion(&wdata->state.ready);
mutex_init(&wdata->state.sync);
wdata->state.drm = WIIPROTO_REQ_DRM_K;
wdata->state.cmd_battery = 0xff;
INIT_WORK(&wdata->init_worker, wiimote_init_worker);
treewide: setup_timer() -> timer_setup() This converts all remaining cases of the old setup_timer() API into using timer_setup(), where the callback argument is the structure already holding the struct timer_list. These should have no behavioral changes, since they just change which pointer is passed into the callback with the same available pointers after conversion. It handles the following examples, in addition to some other variations. Casting from unsigned long: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... setup_timer(&ptr->my_timer, my_callback, ptr); and forced object casts: void my_callback(struct something *ptr) { ... } ... setup_timer(&ptr->my_timer, my_callback, (unsigned long)ptr); become: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... timer_setup(&ptr->my_timer, my_callback, 0); Direct function assignments: void my_callback(unsigned long data) { struct something *ptr = (struct something *)data; ... } ... ptr->my_timer.function = my_callback; have a temporary cast added, along with converting the args: void my_callback(struct timer_list *t) { struct something *ptr = from_timer(ptr, t, my_timer); ... } ... ptr->my_timer.function = (TIMER_FUNC_TYPE)my_callback; And finally, callbacks without a data assignment: void my_callback(unsigned long data) { ... } ... setup_timer(&ptr->my_timer, my_callback, 0); have their argument renamed to verify they're unused during conversion: void my_callback(struct timer_list *unused) { ... } ... timer_setup(&ptr->my_timer, my_callback, 0); The conversion is done with the following Coccinelle script: spatch --very-quiet --all-includes --include-headers \ -I ./arch/x86/include -I ./arch/x86/include/generated \ -I ./include -I ./arch/x86/include/uapi \ -I ./arch/x86/include/generated/uapi -I ./include/uapi \ -I ./include/generated/uapi --include ./include/linux/kconfig.h \ --dir . \ --cocci-file ~/src/data/timer_setup.cocci @fix_address_of@ expression e; @@ setup_timer( -&(e) +&e , ...) // Update any raw setup_timer() usages that have a NULL callback, but // would otherwise match change_timer_function_usage, since the latter // will update all function assignments done in the face of a NULL // function initialization in setup_timer(). @change_timer_function_usage_NULL@ expression _E; identifier _timer; type _cast_data; @@ ( -setup_timer(&_E->_timer, NULL, _E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E->_timer, NULL, (_cast_data)_E); +timer_setup(&_E->_timer, NULL, 0); | -setup_timer(&_E._timer, NULL, &_E); +timer_setup(&_E._timer, NULL, 0); | -setup_timer(&_E._timer, NULL, (_cast_data)&_E); +timer_setup(&_E._timer, NULL, 0); ) @change_timer_function_usage@ expression _E; identifier _timer; struct timer_list _stl; identifier _callback; type _cast_func, _cast_data; @@ ( -setup_timer(&_E->_timer, _callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, &_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, _E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, &_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)_E); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, (_cast_func)&_callback, (_cast_data)&_E); +timer_setup(&_E._timer, _callback, 0); | _E->_timer@_stl.function = _callback; | _E->_timer@_stl.function = &_callback; | _E->_timer@_stl.function = (_cast_func)_callback; | _E->_timer@_stl.function = (_cast_func)&_callback; | _E._timer@_stl.function = _callback; | _E._timer@_stl.function = &_callback; | _E._timer@_stl.function = (_cast_func)_callback; | _E._timer@_stl.function = (_cast_func)&_callback; ) // callback(unsigned long arg) @change_callback_handle_cast depends on change_timer_function_usage@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; identifier _handle; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { ( ... when != _origarg _handletype *_handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(_handletype *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg | ... when != _origarg _handletype *_handle; ... when != _handle _handle = -(void *)_origarg; +from_timer(_handle, t, _timer); ... when != _origarg ) } // callback(unsigned long arg) without existing variable @change_callback_handle_cast_no_arg depends on change_timer_function_usage && !change_callback_handle_cast@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _origtype; identifier _origarg; type _handletype; @@ void _callback( -_origtype _origarg +struct timer_list *t ) { + _handletype *_origarg = from_timer(_origarg, t, _timer); + ... when != _origarg - (_handletype *)_origarg + _origarg ... when != _origarg } // Avoid already converted callbacks. @match_callback_converted depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier t; @@ void _callback(struct timer_list *t) { ... } // callback(struct something *handle) @change_callback_handle_arg depends on change_timer_function_usage && !match_callback_converted && !change_callback_handle_cast && !change_callback_handle_cast_no_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; @@ void _callback( -_handletype *_handle +struct timer_list *t ) { + _handletype *_handle = from_timer(_handle, t, _timer); ... } // If change_callback_handle_arg ran on an empty function, remove // the added handler. @unchange_callback_handle_arg depends on change_timer_function_usage && change_callback_handle_arg@ identifier change_timer_function_usage._callback; identifier change_timer_function_usage._timer; type _handletype; identifier _handle; identifier t; @@ void _callback(struct timer_list *t) { - _handletype *_handle = from_timer(_handle, t, _timer); } // We only want to refactor the setup_timer() data argument if we've found // the matching callback. This undoes changes in change_timer_function_usage. @unchange_timer_function_usage depends on change_timer_function_usage && !change_callback_handle_cast && !change_callback_handle_cast_no_arg && !change_callback_handle_arg@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type change_timer_function_usage._cast_data; @@ ( -timer_setup(&_E->_timer, _callback, 0); +setup_timer(&_E->_timer, _callback, (_cast_data)_E); | -timer_setup(&_E._timer, _callback, 0); +setup_timer(&_E._timer, _callback, (_cast_data)&_E); ) // If we fixed a callback from a .function assignment, fix the // assignment cast now. @change_timer_function_assignment depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression change_timer_function_usage._E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_func; typedef TIMER_FUNC_TYPE; @@ ( _E->_timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -&_callback +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)_callback; +(TIMER_FUNC_TYPE)_callback ; | _E->_timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -&_callback; +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)_callback +(TIMER_FUNC_TYPE)_callback ; | _E._timer.function = -(_cast_func)&_callback +(TIMER_FUNC_TYPE)_callback ; ) // Sometimes timer functions are called directly. Replace matched args. @change_timer_function_calls depends on change_timer_function_usage && (change_callback_handle_cast || change_callback_handle_cast_no_arg || change_callback_handle_arg)@ expression _E; identifier change_timer_function_usage._timer; identifier change_timer_function_usage._callback; type _cast_data; @@ _callback( ( -(_cast_data)_E +&_E->_timer | -(_cast_data)&_E +&_E._timer | -_E +&_E->_timer ) ) // If a timer has been configured without a data argument, it can be // converted without regard to the callback argument, since it is unused. @match_timer_function_unused_data@ expression _E; identifier _timer; identifier _callback; @@ ( -setup_timer(&_E->_timer, _callback, 0); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0L); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E->_timer, _callback, 0UL); +timer_setup(&_E->_timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0L); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_E._timer, _callback, 0UL); +timer_setup(&_E._timer, _callback, 0); | -setup_timer(&_timer, _callback, 0); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0L); +timer_setup(&_timer, _callback, 0); | -setup_timer(&_timer, _callback, 0UL); +timer_setup(&_timer, _callback, 0); | -setup_timer(_timer, _callback, 0); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0L); +timer_setup(_timer, _callback, 0); | -setup_timer(_timer, _callback, 0UL); +timer_setup(_timer, _callback, 0); ) @change_callback_unused_data depends on match_timer_function_unused_data@ identifier match_timer_function_unused_data._callback; type _origtype; identifier _origarg; @@ void _callback( -_origtype _origarg +struct timer_list *unused ) { ... when != _origarg } Signed-off-by: Kees Cook <keescook@chromium.org>
2017-10-17 06:43:17 +09:00
timer_setup(&wdata->timer, wiimote_init_timeout, 0);
return wdata;
}
static void wiimote_destroy(struct wiimote_data *wdata)
{
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
unsigned long flags;
wiidebug_deinit(wdata);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
/* prevent init_worker from being scheduled again */
spin_lock_irqsave(&wdata->state.lock, flags);
wdata->state.flags |= WIIPROTO_FLAG_EXITING;
spin_unlock_irqrestore(&wdata->state.lock, flags);
cancel_work_sync(&wdata->init_worker);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
del_timer_sync(&wdata->timer);
device_remove_file(&wdata->hdev->dev, &dev_attr_devtype);
device_remove_file(&wdata->hdev->dev, &dev_attr_extension);
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
wiimote_mp_unload(wdata);
wiimote_ext_unload(wdata);
wiimote_modules_unload(wdata);
cancel_work_sync(&wdata->queue.worker);
hid_hw_close(wdata->hdev);
hid_hw_stop(wdata->hdev);
kfree(wdata);
}
static int wiimote_hid_probe(struct hid_device *hdev,
const struct hid_device_id *id)
{
struct wiimote_data *wdata;
int ret;
hdev->quirks |= HID_QUIRK_NO_INIT_REPORTS;
wdata = wiimote_create(hdev);
if (!wdata) {
hid_err(hdev, "Can't alloc device\n");
return -ENOMEM;
}
ret = hid_parse(hdev);
if (ret) {
hid_err(hdev, "HID parse failed\n");
goto err;
}
ret = hid_hw_start(hdev, HID_CONNECT_HIDRAW);
if (ret) {
hid_err(hdev, "HW start failed\n");
goto err;
}
ret = hid_hw_open(hdev);
if (ret) {
hid_err(hdev, "cannot start hardware I/O\n");
goto err_stop;
}
ret = device_create_file(&hdev->dev, &dev_attr_extension);
if (ret) {
hid_err(hdev, "cannot create sysfs attribute\n");
goto err_close;
}
ret = device_create_file(&hdev->dev, &dev_attr_devtype);
if (ret) {
hid_err(hdev, "cannot create sysfs attribute\n");
goto err_ext;
}
ret = wiidebug_init(wdata);
if (ret)
goto err_free;
hid_info(hdev, "New device registered\n");
/* schedule device detection */
HID: wiimote: add extension hotplug support The Wii Remote has several extension ports. The first port (EXT) provides hotplug events whenever an extension is plugged. The second port (MP) does not provide hotplug events by default. Instead, we have to map MP into EXT to get events for it. This patch introduces hotplug support for extensions. It is fairly complicated to get this right because the Wii Remote sends a lot of noise-hotplug events while activating extension ports. We need to filter the events and only handle the events that are real hotplug events. Mapping MP into EXT is easy. But if we want both, MP _and_ EXT at the same time, we need to map MP into EXT and enable a passthrough-mode. This will then send real EXT events through the mapped MP interleaved with real MP events. But once MP is mapped, we no longer have access to the real EXT registers so we need to perform setup _before_ mapping MP. Furthermore, we no longer can read EXT IDs so we cannot verify if EXT is still the same extension that we expect it to be. We deal with this by unmapping MP whenever we got into a situation where EXT might have changed. We then re-read EXT and MP and remap everything. The real Wii Console takes a fairly easy approach: It simply reconnects to the device on hotplug events that it didn't expect. So if a program wants MP events, but MP is disconnected, it fails and reconnects so it can wait for MP hotplug events again. This simplifies hotplugging a lot because we just react on PLUG events and ignore UNPLUG events. The more sophisticated Wii applications avoid reconnection (well, they still reconnect during many weird events, but at least not during UNPLUG) but they start polling the device. This allows them to disable the device, poll for the extension ports to settle and then initialize them again. Unfortunately, this approach fails whenever an extension is replugged while it is initialized. We would loose UNPLUG events and polling the device later will give unreliable results because the extension port might be in some weird state, even though it's actually unplugged. Our approach is a real HOTPLUG approch. We keep track of the EXT and mapped MP hotplug events whenever they occur. We then re-evaluate the device state and initialize any possible new extension or deinitialize any gone extension. Only during initialization, we set an extension port ACTIVE. However, during an unplug event we mark them as INACTIVE. This guarantess that a fast UNPLUG -> PLUG event sequence doesn't keep them marked as PLUGGED+ACTIVE but only PLUGGED. To deal with annoying noise-hotplug events during extension mapping, we simply rescan the device before performing any mapping. This allows us to ignore all the noise events as long as the device is in the correct state. Long story short: EXT and MP registers are sparsely known and we need to jump through hoops to get reliable HOTPLUG working. But while Nintendo needs *FOUR* Bluetooth reconnections for the shortest imaginable boot->menu->game->menu->shutdown sequence, we now need *ZERO*. As always, 3rd party devices tend to break whenever we behave differently than the original Wii. So there are also devices which _expect_ a disconnect after UNPLUG. Obviously, these devices won't benefit from this patch. But all official devices were tested extensively and work great during any hotplug sequence. Yay! Signed-off-by: David Herrmann <dh.herrmann@gmail.com> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2013-05-06 06:12:57 +09:00
wiimote_schedule(wdata);
return 0;
err_free:
wiimote_destroy(wdata);
return ret;
err_ext:
device_remove_file(&wdata->hdev->dev, &dev_attr_extension);
err_close:
hid_hw_close(hdev);
err_stop:
hid_hw_stop(hdev);
err:
input_free_device(wdata->ir);
input_free_device(wdata->accel);
kfree(wdata);
return ret;
}
static void wiimote_hid_remove(struct hid_device *hdev)
{
struct wiimote_data *wdata = hid_get_drvdata(hdev);
hid_info(hdev, "Device removed\n");
wiimote_destroy(wdata);
}
static const struct hid_device_id wiimote_hid_devices[] = {
{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_NINTENDO,
USB_DEVICE_ID_NINTENDO_WIIMOTE) },
{ HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_NINTENDO,
USB_DEVICE_ID_NINTENDO_WIIMOTE2) },
{ }
};
MODULE_DEVICE_TABLE(hid, wiimote_hid_devices);
static struct hid_driver wiimote_hid_driver = {
.name = "wiimote",
.id_table = wiimote_hid_devices,
.probe = wiimote_hid_probe,
.remove = wiimote_hid_remove,
.raw_event = wiimote_hid_event,
};
module_hid_driver(wiimote_hid_driver);
MODULE_LICENSE("GPL");
MODULE_AUTHOR("David Herrmann <dh.herrmann@gmail.com>");
MODULE_DESCRIPTION("Driver for Nintendo Wii / Wii U peripherals");