I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
If anyone has any ideas where to look, or what might be going on, I'd love to hear from you
Tom Green
Hi Thomas,
On Tue, Jun 21, 2022 at 1:29 PM Thomas Green <[email protected]> wrote:
>
> I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
>
> I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
>
> If anyone has any ideas where to look, or what might be going on, I'd love to hear from you
Do you have any logs? Also what version is this happening, perhaps you
are missing some fixes like:
https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=5ebc2b5a3e48dbe67a2d315979b096a1b4d1eb99
>
> Tom Green
--
Luiz Augusto von Dentz
Luiz,
Thank you for your reply. In my first two attempts I pasted into it the log file created by running:
bluetoothd --experimental -debug -n
The version of bluez is 5.50. I looked at the patch you suggested and it's already included in this version of src/device.c
Are there any further patches I should look at? Would you like a copy of my log again?
Tom
-----Original Message-----
From: Luiz Augusto von Dentz <[email protected]>
Sent: Tuesday, June 21, 2022 2:40 PM
To: Thomas Green <[email protected]>
Cc: [email protected]
Subject: Re: third time
[EXTERNAL]
Hi Thomas,
On Tue, Jun 21, 2022 at 1:29 PM Thomas Green <[email protected]> wrote:
>
> I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
>
> I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
>
> If anyone has any ideas where to look, or what might be going on, I'd
> love to hear from you
Do you have any logs? Also what version is this happening, perhaps you are missing some fixes like:
https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=5ebc2b5a3e48dbe67a2d315979b096a1b4d1eb99
>
> Tom Green
--
Luiz Augusto von Dentz
Hi Thomas,
On Wed, Jun 22, 2022 at 2:22 PM Thomas Green <[email protected]> wrote:
>
> Luiz,
>
> Thank you for your reply. In my first two attempts I pasted into it the log file created by running:
>
> bluetoothd --experimental -debug -n
>
> The version of bluez is 5.50. I looked at the patch you suggested and it's already included in this version of src/device.c
>
> Are there any further patches I should look at? Would you like a copy of my log again?
Ouch, that is quite an old version, have you considered using something newer?
> Tom
>
> -----Original Message-----
> From: Luiz Augusto von Dentz <[email protected]>
> Sent: Tuesday, June 21, 2022 2:40 PM
> To: Thomas Green <[email protected]>
> Cc: [email protected]
> Subject: Re: third time
>
> [EXTERNAL]
>
> Hi Thomas,
>
> On Tue, Jun 21, 2022 at 1:29 PM Thomas Green <[email protected]> wrote:
> >
> > I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
> >
> > I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
> >
> > If anyone has any ideas where to look, or what might be going on, I'd
> > love to hear from you
>
> Do you have any logs? Also what version is this happening, perhaps you are missing some fixes like:
>
> https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=5ebc2b5a3e48dbe67a2d315979b096a1b4d1eb99
>
> >
> > Tom Green
>
>
>
> --
> Luiz Augusto von Dentz
--
Luiz Augusto von Dentz
I will attempt to update our build system to use a newer version
Tom
-----Original Message-----
From: Luiz Augusto von Dentz <[email protected]>
Sent: Wednesday, June 22, 2022 3:32 PM
To: Thomas Green <[email protected]>
Cc: [email protected]
Subject: Re: third time
[EXTERNAL]
Hi Thomas,
On Wed, Jun 22, 2022 at 2:22 PM Thomas Green <[email protected]> wrote:
>
> Luiz,
>
> Thank you for your reply. In my first two attempts I pasted into it the log file created by running:
>
> bluetoothd --experimental -debug -n
>
> The version of bluez is 5.50. I looked at the patch you suggested and
> it's already included in this version of src/device.c
>
> Are there any further patches I should look at? Would you like a copy of my log again?
Ouch, that is quite an old version, have you considered using something newer?
> Tom
>
> -----Original Message-----
> From: Luiz Augusto von Dentz <[email protected]>
> Sent: Tuesday, June 21, 2022 2:40 PM
> To: Thomas Green <[email protected]>
> Cc: [email protected]
> Subject: Re: third time
>
> [EXTERNAL]
>
> Hi Thomas,
>
> On Tue, Jun 21, 2022 at 1:29 PM Thomas Green <[email protected]> wrote:
> >
> > I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
> >
> > I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
> >
> > If anyone has any ideas where to look, or what might be going on,
> > I'd love to hear from you
>
> Do you have any logs? Also what version is this happening, perhaps you are missing some fixes like:
>
> https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=5ebc2b5a
> 3e48dbe67a2d315979b096a1b4d1eb99
>
> >
> > Tom Green
>
>
>
> --
> Luiz Augusto von Dentz
--
Luiz Augusto von Dentz
Luiz,
I've updated (at least my testing system) to version 5.64 and are having other problems now too. For example the first time I attach one of these devices it works well, subsequent attempts to pair other of the same type of device fails with and AuthenticationFailed error. On 5.50, I could pair as many of these devices as I wanted. Should we take this offline and perhaps we could discuss this more?
Tom
-----Original Message-----
From: Thomas Green <[email protected]>
Sent: Wednesday, June 22, 2022 3:43 PM
To: [email protected]
Subject: RE: third time
[EXTERNAL]
I will attempt to update our build system to use a newer version
Tom
-----Original Message-----
From: Luiz Augusto von Dentz <[email protected]>
Sent: Wednesday, June 22, 2022 3:32 PM
To: Thomas Green <[email protected]>
Cc: [email protected]
Subject: Re: third time
[EXTERNAL]
Hi Thomas,
On Wed, Jun 22, 2022 at 2:22 PM Thomas Green <[email protected]> wrote:
>
> Luiz,
>
> Thank you for your reply. In my first two attempts I pasted into it the log file created by running:
>
> bluetoothd --experimental -debug -n
>
> The version of bluez is 5.50. I looked at the patch you suggested and
> it's already included in this version of src/device.c
>
> Are there any further patches I should look at? Would you like a copy of my log again?
Ouch, that is quite an old version, have you considered using something newer?
> Tom
>
> -----Original Message-----
> From: Luiz Augusto von Dentz <[email protected]>
> Sent: Tuesday, June 21, 2022 2:40 PM
> To: Thomas Green <[email protected]>
> Cc: [email protected]
> Subject: Re: third time
>
> [EXTERNAL]
>
> Hi Thomas,
>
> On Tue, Jun 21, 2022 at 1:29 PM Thomas Green <[email protected]> wrote:
> >
> > I've twice before written about this issue I'm having, but never heard anything back. It regards a GATT device that connects and operates correctly. The issue is when we call the Adapter1.RemoveDevice method for this device. Some times the device is removed completely, and everything works fine. Some times it fails and bluetoothd becomes unusable. Even when I shutdown and restart bluetoothd it doesn't operate correctly when restarting. The difference is that when it works, at the end of the removal process, it reconnects the device, tries to recreate the device, then removes it. All internally. When things fail, those steps don't happen. I've posted before the salient part of the Bluetooth log. The only thing that will make bluetoothd function correctly is to shut it down, remove all of the similar GATT devices, then restart bluetoothd.
> >
> > I have yet to hear if anyone has -any- idea regarding this, not even to tell me it's not something that needs to be looked at. It is something that renders bluetoothd unusable, so I would think that it is something that might be interesting to look into.
> >
> > If anyone has any ideas where to look, or what might be going on,
> > I'd love to hear from you
>
> Do you have any logs? Also what version is this happening, perhaps you are missing some fixes like:
>
> https://git.kernel.org/pub/scm/bluetooth/bluez.git/commit/?id=5ebc2b5a
> 3e48dbe67a2d315979b096a1b4d1eb99
>
> >
> > Tom Green
>
>
>
> --
> Luiz Augusto von Dentz
--
Luiz Augusto von Dentz
Dear Thomas,
Am 23.06.22 um 22:00 schrieb Thomas Green:
> I've updated (at least my testing system) to version 5.64 and are
> having other problems now too. For example the first time I attach
> one of these devices it works well, subsequent attempts to pair other
> of the same type of device fails with and AuthenticationFailed error.
> On 5.50, I could pair as many of these devices as I wanted. Should
> we take this offline and perhaps we could discuss this more?
Can you please (git) bisect the issue? It’s often the fastest way to
find the cause, and then figuring out a fix.
Kind regards,
Paul