Retry opening of wiimote channels on initial failure #4940

sjnewbury · 2017-02-20T11:56:57Z

It's not uncommon on some Bluetooth controllers for a
peripheral device to be "present" before it's "ready".
There is a window where connection can fail because the
device communication didn't succeed initially, but
will always work on retry after a short delay.

I'm not the original author of this patch and I can't
remember where I found it. I can confirm that it
works perfectly for me though...

JosJuice · 2017-02-20T14:41:20Z

Please run the change through clang-format (or fix the lint errors manually, as shown here: https://buildbot.dolphin-emu.org/builders/lint/builds/5017/steps/shell/logs/stdio)

sjnewbury · 2017-02-20T16:06:18Z

@JosJuice Oops! Okay.

BhaaLseN · 2017-02-20T18:13:35Z

Can you update that comment to include a brief summary of what you added to your commit message? We can see that you wait and try again from what is done; but it isn't obvious why this is done.

sjnewbury · 2017-02-20T19:13:05Z

@BhaaLseN I'm not sure what I can add. There seems to be a race condition between a peripheral device connecting to the bluetooth controller and it being ready to use. It's very short and it depends upon the controller, some appear to connect synchronously and block until the device is ready, others report the device upon discovery but do not allow communication straight away. I don't know which is the correct behaviour, or whether it depends on the peripheral, controller or both. Anyway, Dolphin waits for a remote to appear and immediately attempts to open the communication channels, this can fail because the device isn't ready yet, delay, try again, and it works.

I'm willing to bet this actually makes wiimotes work with most bluetooth controllers which otherwise fail.

BhaaLseN · 2017-02-20T19:14:56Z

When in doubt, just add that part of your commit message. Otherwise nobody knows why you do the sleep there, and might remove it thinking it has no purpose.

sjnewbury · 2017-02-20T22:31:35Z

@BhaaLseN I've updated the commit message based on my reply above.

Parlane · 2017-02-23T22:46:36Z

@dolphin-emu-bot rebuild

PatrickFerry · 2017-02-23T23:21:21Z

There are other (unlikely) chances the device is busy at random moments after this initial race condition.
Could you loop around try to reconnect 2-3 times, it would be more robust for the chaotic and fragmented world of bluetooth adaptors/devices/driver stacks?

sjnewbury · 2017-02-24T21:38:02Z

On Thu, 2017-02-23 at 15:21 -0800, Patrick A. Ferry wrote: There are other (unlikely) chances the device is busy at random moments after this initial race condition. Could you loop around try to reconnect 2-3 times, it would be more robust for the chaotic and fragmented world of bluetooth adaptors/devices/driver stacks?

Sure, it makes sense. I'll update the patch.

sjnewbury · 2017-02-28T14:54:43Z

Updated patch with loops instead of in-line retries.

sjnewbury · 2017-03-01T19:21:51Z

@ZephyrSurfer I've put in multiple retries as you suggested.

Source/Core/Core/HW/WiimoteReal/IOLinux.cpp


  // Output channel
  addr.l2_psm = htobs(WM_OUTPUT_CHANNEL);
-  if ((m_cmd_sock = socket(AF_BLUETOOTH, SOCK_SEQPACKET, BTPROTO_L2CAP)) == -1 ||
-      connect(m_cmd_sock, (sockaddr*)&addr, sizeof(addr)) < 0)
+  retry = 0;


PatrickFerry · 2017-03-01T21:10:31Z

@sjnewbury It seems the original patch is from here

Could you bump the attempts up one more time?

sjnewbury · 2017-03-02T15:08:15Z

@ZephyrSurfer I'll put that reference into the commit message.

sjnewbury · 2017-03-02T15:14:25Z

@ZephyrSurfer As requested, one more attempt.

leoetlino · 2017-04-01T23:13:27Z

Looks like this will need a rebase + some (quick and easy) lint fixes.

sjnewbury · 2017-04-12T16:14:29Z

Sorry for the delay, I've just been distracted with other things. I'll take a look now.

Source/Core/Core/HW/WiimoteReal/IOLinux.cpp

+      m_cmd_sock = -1;
+      return false;
+    }
+    retry++;


Source/Core/Core/HW/WiimoteReal/IOLinux.cpp

+      return false;
+    }
+    retry++;
+    sleep(1);


sjnewbury · 2017-04-16T13:29:39Z

There's no point in retrying opening of the socket, it either works or it doesn't, so I've changed the code to just try opening the sockets, if it succeeds, then retry the connections.

Will push after testing...

Source/Core/Core/HW/WiimoteReal/IOLinux.cpp

-      connect(m_cmd_sock, (sockaddr*)&addr, sizeof(addr)) < 0)
+  if (m_cmd_sock = socket(AF_BLUETOOTH, SOCK_SEQPACKET, BTPROTO_L2CAP))
+  {
+    retry = 0;


Source/Core/Core/HW/WiimoteReal/IOLinux.cpp

-      connect(m_int_sock, (sockaddr*)&addr, sizeof(addr)) < 0)
+  if (m_int_sock = socket(AF_BLUETOOTH, SOCK_SEQPACKET, BTPROTO_L2CAP))
+  {
+    retry = 0;


sjnewbury · 2017-06-19T07:45:36Z

Looks like I need to rebase again ...

There seems to be a race condition between a peripheral device connecting to the bluetooth controller and it being ready to use. It's very short and it depends upon the controller, some appear to connect synchronously and block until the device is ready, others report the device upon discovery but do not allow communication straight away. I don't know which is the correct behaviour, or whether it depends on the peripheral, controller or both. Anyway, Dolphin waits for a remote to appear and immediately attempts to open the communication channels, this can fail because the device isn't ready yet, delay, try again, and it works. There are other (unlikely) chances the device is busy at random moments after this initial race condition so it loops around try to reconnect. This was inspired by an earlier patch, see here: https://bugs.dolphin-emu.org/issues/5997#note-20 I can confirm that it works perfectly for me on a bluetooth controller where otherwise it's impossible to connect (Dell 380 Bluetooth 4.0).

sjnewbury · 2017-06-19T08:23:01Z

Rebased

sjnewbury force-pushed the wiimote-connect branch from 9fa82c6 to c7c53d9 Compare February 20, 2017 16:11

sjnewbury force-pushed the wiimote-connect branch from c7c53d9 to 988e635 Compare February 20, 2017 21:09

sjnewbury force-pushed the wiimote-connect branch from 988e635 to 5c204d6 Compare February 28, 2017 14:50

leoetlino reviewed Mar 1, 2017

View reviewed changes

sjnewbury force-pushed the wiimote-connect branch from 5c204d6 to f152981 Compare March 2, 2017 15:13

riking suggested changes Apr 13, 2017

View reviewed changes

sjnewbury force-pushed the wiimote-connect branch from f152981 to c33d6c8 Compare April 16, 2017 13:34

leoetlino reviewed Jun 18, 2017

View reviewed changes

sjnewbury force-pushed the wiimote-connect branch from c33d6c8 to e692fa6 Compare June 19, 2017 07:44

sjnewbury force-pushed the wiimote-connect branch from e692fa6 to e9a696b Compare June 19, 2017 08:22

leoetlino approved these changes Jun 19, 2017

View reviewed changes

leoetlino merged commit 1be10e7 into dolphin-emu:master Jun 19, 2017

This pull request was closed.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Retry opening of wiimote channels on initial failure #4940

Retry opening of wiimote channels on initial failure #4940

sjnewbury commented Feb 20, 2017

JosJuice commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

BhaaLseN commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

BhaaLseN commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

Parlane commented Feb 23, 2017

PatrickFerry commented Feb 23, 2017

sjnewbury commented Feb 24, 2017 via email

sjnewbury commented Feb 28, 2017

sjnewbury commented Mar 1, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

PatrickFerry commented Mar 1, 2017

sjnewbury commented Mar 2, 2017

sjnewbury commented Mar 2, 2017

leoetlino commented Apr 1, 2017

sjnewbury commented Apr 12, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

sjnewbury commented Apr 16, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

sjnewbury commented Jun 19, 2017

sjnewbury commented Jun 19, 2017

Retry opening of wiimote channels on initial failure #4940

Retry opening of wiimote channels on initial failure #4940

Conversation

sjnewbury commented Feb 20, 2017

JosJuice commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

BhaaLseN commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

BhaaLseN commented Feb 20, 2017

sjnewbury commented Feb 20, 2017

Parlane commented Feb 23, 2017

PatrickFerry commented Feb 23, 2017

sjnewbury commented Feb 24, 2017 via email

sjnewbury commented Feb 28, 2017

sjnewbury commented Mar 1, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

PatrickFerry commented Mar 1, 2017

sjnewbury commented Mar 2, 2017

sjnewbury commented Mar 2, 2017

leoetlino commented Apr 1, 2017

sjnewbury commented Apr 12, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

sjnewbury commented Apr 16, 2017

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

This comment was marked as off-topic.

sjnewbury commented Jun 19, 2017

sjnewbury commented Jun 19, 2017