btusb driver need to be unloaded and reloaded after boot & lag issue

btusb driver need to be unloaded and reloaded after boot & lag issue

[help.7ocym]

Hi,
sorry to address both list, but this issue seems related, without knowing where lies the issue
my hardware : https://wiki.gentoo.org/wiki/Lenovo_Yoga_900
I use the pre-built gentoo linux kernel, 
6.2.8-gentoo-dist #1 SMP PREEMPT_DYNAMIC Wed Mar 22 17:15:39 -00 2023 x86_64 Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz GenuineIntel GNU/Linux

since a few update (sadly I didn't noted the latest kernel version that didn't had the issue), after a boot, the bluetooth isn't working, nothing bad in dmesg, I just have to unload btusb module and modprobe it again to have bluetooth working again...

after a suspend to ram, I have to power off-power on the bluetooth to have it work again (bluetoothctl power off; bluetoothctl power on)

bluetooth mouse can also be extremely laggy sometimes,but without error message in dmesg, most of the time `bluetoothctl power` off-on cycling do solve the issue....

I also included the usb mailing list because it might be related to some behavior I noticed :

I have usb3.0 micro sd card reader (SanDisk MobileMate UHS-I microSD Reader/Writer USB 3.0 Reader, Kingston MobileLite Plus (MLPM) microSD Card Reader USB 3.1 microSDHC/SDXC UHS-II, for example) and some extra fast micro sd cards (like sandisk extrem 512G), when transfering data the read rate can be as high as 110Mo/s and write 70Mo/s sustained, nothing impressive but when such rate is achieved for a long time (big file transfer) either reading only access, writing only access or read write, the usb bus become unusable, I can't even use a usb mouse connected to it by wire... even if cpu usage is really low (less than 10%) 
I don't have the issue if I connect a M2 usb3 flash drive, with comparable transfert speed... so not related to some bus over usage...  

so I suspect that there is an issue with the usb driver, and that maybe the bluetooth issue can be related to the usb issue, since the bluetooth controller is on the usb bus on the laptop

the transfer issue of usb is much more older than the bluetooth issue, it's approximative, but :
 - the btusb boot issue is about 3 month old, 
 - the suspend/resume issue of bluetooth is more than a year old
 - the usb transfer issue as more than a year...

I'll gladly provide any useful information, can also do patch tries... 

thanks and regards
JLM


Sent with Proton Mail secure email.

Re: btusb driver need to be unloaded and reloaded after boot & lag issue

[Paul Menzel]

[Cc: +regressions@lists.linux.dev]

Dear JLM,


Am 13.04.23 um 00:06 schrieb help.7ocym@8shield.net:

> sorry to address both list, but this issue seems related, without
> knowing where lies the issue > my hardware : https://wiki.gentoo.org/wiki/Lenovo_Yoga_900
> I use the pre-built gentoo linux kernel,
> 6.2.8-gentoo-dist #1 SMP PREEMPT_DYNAMIC Wed Mar 22 17:15:39 -00 2023 x86_64 Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz GenuineIntel GNU/Linux
> 
> since a few update (sadly I didn't noted the latest kernel version
> that didn't had the issue), after a boot, the bluetooth isn't
> working, nothing bad in dmesg, I just have to unload btusb module and
> modprobe it again to have bluetooth working again...
> 
> after a suspend to ram, I have to power off-power on the bluetooth to
> have it work again (bluetoothctl power off; bluetoothctl power on)
> 
> bluetooth mouse can also be extremely laggy sometimes,but without
> error message in dmesg, most of the time `bluetoothctl power` off-on
> cycling do solve the issue....
> 
> I also included the usb mailing list because it might be related to
> some behavior I noticed :
> 
> I have usb3.0 micro sd card reader (SanDisk MobileMate UHS-I microSD
> Reader/Writer USB 3.0 Reader, Kingston MobileLite Plus (MLPM) microSD
> Card Reader USB 3.1 microSDHC/SDXC UHS-II, for example) and some
> extra fast micro sd cards (like sandisk extrem 512G), when
> transferring data the read rate can be as high as 110Mo/s and write
> 70Mo/s sustained, nothing impressive but when such rate is achieved
> for a long time (big file transfer) either reading only access,
> writing only access or read write, the usb bus become unusable, I
> can't even use a usb mouse connected to it by wire... even if cpu
> usage is really low (less than 10%) I don't have the issue if I
> connect a M2 usb3 flash drive, with comparable transfert speed... so
> not related to some bus over usage...
> 
> so I suspect that there is an issue with the usb driver, and that
> maybe the bluetooth issue can be related to the usb issue, since the
> bluetooth controller is on the usb bus on the laptop >
> the transfer issue of usb is much more older than the bluetooth
> issue, it's approximative, but : >   - the btusb boot issue is about 3 month old,
>   - the suspend/resume issue of bluetooth is more than a year old
>   - the usb transfer issue as more than a year...
> 
> I'll gladly provide any useful information, can also do patch tries...
As you use Gentoo and are able to build your own Linux kernel, the 
fastest way to get these issues addressed is to bisect them. To shorten 
the test cycles, I recommend to try, if you can reproduce the issues in 
QEMU and passing through the problematic devices to the VM [1][2].

I also recommend to start a separate thread for each issue and, as these 
seem to be regressions, also keep the regression folks in the loop [3].


Kind regards,

Paul


[1]: 
https://lore.kernel.org/all/5891f0d5-8d51-9da5-7663-718f301490b1@molgen.mpg.de/
      (The commands were working for after all, and the device didn’t 
show up due to a (second) Linux kernel regression.)
[2]: https://station.eciton.net/qemu-usb-host-device-pass-through.html
[3]: 
https://www.kernel.org/doc/html/latest/admin-guide/reporting-regressions.html

Re: btusb driver need to be unloaded and reloaded after boot & lag issue

[help.7ocym]

[-- Attachment #1.1: Type: text/plain, Size: 4188 bytes --]

Hi,

------- Original Message -------
On Thursday, April 13th, 2023 at 7:39 AM, Paul Menzel - pmenzel at molgen.mpg.de <pmenzel_at_molgen_mpg_de_rislbjit@simplelogin.co> wrote:


> [Cc: +regressions@lists.linux.dev]
>
> Dear JLM,
>
>
> Am 13.04.23 um 00:06 schrieb help.7ocym@8shield.net:
>
> > sorry to address both list, but this issue seems related, without
> > knowing where lies the issue > my hardware : https://wiki.gentoo.org/wiki/Lenovo_Yoga_900
> > I use the pre-built gentoo linux kernel,
> > 6.2.8-gentoo-dist #1 SMP PREEMPT_DYNAMIC Wed Mar 22 17:15:39 -00 2023 x86_64 Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz GenuineIntel GNU/Linux
> >
> > since a few update (sadly I didn't noted the latest kernel version
> > that didn't had the issue), after a boot, the bluetooth isn't
> > working, nothing bad in dmesg, I just have to unload btusb module and
> > modprobe it again to have bluetooth working again...
> >
> > after a suspend to ram, I have to power off-power on the bluetooth to
> > have it work again (bluetoothctl power off; bluetoothctl power on)
> >
> > bluetooth mouse can also be extremely laggy sometimes,but without
> > error message in dmesg, most of the time `bluetoothctl power` off-on
> > cycling do solve the issue....
> >
> > I also included the usb mailing list because it might be related to
> > some behavior I noticed :
> >
> > I have usb3.0 micro sd card reader (SanDisk MobileMate UHS-I microSD
> > Reader/Writer USB 3.0 Reader, Kingston MobileLite Plus (MLPM) microSD
> > Card Reader USB 3.1 microSDHC/SDXC UHS-II, for example) and some
> > extra fast micro sd cards (like sandisk extrem 512G), when
> > transferring data the read rate can be as high as 110Mo/s and write
> > 70Mo/s sustained, nothing impressive but when such rate is achieved
> > for a long time (big file transfer) either reading only access,
> > writing only access or read write, the usb bus become unusable, I
> > can't even use a usb mouse connected to it by wire... even if cpu
> > usage is really low (less than 10%) I don't have the issue if I
> > connect a M2 usb3 flash drive, with comparable transfert speed... so
> > not related to some bus over usage...
> >
> > so I suspect that there is an issue with the usb driver, and that
> > maybe the bluetooth issue can be related to the usb issue, since the
> > bluetooth controller is on the usb bus on the laptop >
> > the transfer issue of usb is much more older than the bluetooth
> > issue, it's approximative, but : > - the btusb boot issue is about 3 month old,
> > - the suspend/resume issue of bluetooth is more than a year old
> > - the usb transfer issue as more than a year...
> >
> > I'll gladly provide any useful information, can also do patch tries...
>
> As you use Gentoo and are able to build your own Linux kernel, the
> fastest way to get these issues addressed is to bisect them. To shorten
> the test cycles, I recommend to try, if you can reproduce the issues in
> QEMU and passing through the problematic devices to the VM [1][2].
>
> I also recommend to start a separate thread for each issue and, as these
> seem to be regressions, also keep the regression folks in the loop [3].
>
>
> Kind regards,
>
> Paul
>
>
> [1]:
> https://lore.kernel.org/all/5891f0d5-8d51-9da5-7663-718f301490b1@molgen.mpg.de/
> (The commands were working for after all, and the device didn’t
> show up due to a (second) Linux kernel regression.)
> [2]: https://station.eciton.net/qemu-usb-host-device-pass-through.html
> [3]:
> https://www.kernel.org/doc/html/latest/admin-guide/reporting-regressions.html

thanks for the reply,
after upgrade to 6.2.11 version, 

- seems that the issue at boot of btusb module is gone... I don't have anymore to unload btusb module after a boot... so that's great news! 

- the bluetooth resumes also after a suspend S3 (suspend to ram)! so great news again! (I've to check for suspend to disk, but I'm confident about success)

I still have to do some check with the micro usb readers to see if the bug "bus monopolization" is still present, so far so good!



Sent with Proton Mail secure email.

[-- Attachment #1.2: publickey - jl.malet@protonmail.com - 0xA25F20CF.asc --]
[-- Type: application/pgp-keys, Size: 657 bytes --]

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 249 bytes --]

Re: btusb driver need to be unloaded and reloaded after boot & lag issue

[help.7ocym]

Hi,

Sent with Proton Mail secure email.

------- Original Message -------
On Wednesday, April 19th, 2023 at 10:46 AM, jl.malet <jl.malet@protonmail.com> wrote:


> Hi,
> 
> ------- Original Message -------
> On Thursday, April 13th, 2023 at 7:39 AM, Paul Menzel - pmenzel at molgen.mpg.de pmenzel_at_molgen_mpg_de_rislbjit@simplelogin.co wrote:
> 
> 
> 
> > [Cc: +regressions@lists.linux.dev]
> > 
> > Dear JLM,
> > 
> > Am 13.04.23 um 00:06 schrieb help.7ocym@8shield.net:
> > 
> > > sorry to address both list, but this issue seems related, without
> > > knowing where lies the issue > my hardware : https://wiki.gentoo.org/wiki/Lenovo_Yoga_900
> > > I use the pre-built gentoo linux kernel,
> > > 6.2.8-gentoo-dist #1 SMP PREEMPT_DYNAMIC Wed Mar 22 17:15:39 -00 2023 x86_64 Intel(R) Core(TM) i7-6500U CPU @ 2.50GHz GenuineIntel GNU/Linux
> > > 
> > > since a few update (sadly I didn't noted the latest kernel version
> > > that didn't had the issue), after a boot, the bluetooth isn't
> > > working, nothing bad in dmesg, I just have to unload btusb module and
> > > modprobe it again to have bluetooth working again...
> > > 
> > > after a suspend to ram, I have to power off-power on the bluetooth to
> > > have it work again (bluetoothctl power off; bluetoothctl power on)
> > > 
> > > bluetooth mouse can also be extremely laggy sometimes,but without
> > > error message in dmesg, most of the time `bluetoothctl power` off-on
> > > cycling do solve the issue....
> > > 
> > > I also included the usb mailing list because it might be related to
> > > some behavior I noticed :
> > > 
> > > I have usb3.0 micro sd card reader (SanDisk MobileMate UHS-I microSD
> > > Reader/Writer USB 3.0 Reader, Kingston MobileLite Plus (MLPM) microSD
> > > Card Reader USB 3.1 microSDHC/SDXC UHS-II, for example) and some
> > > extra fast micro sd cards (like sandisk extrem 512G), when
> > > transferring data the read rate can be as high as 110Mo/s and write
> > > 70Mo/s sustained, nothing impressive but when such rate is achieved
> > > for a long time (big file transfer) either reading only access,
> > > writing only access or read write, the usb bus become unusable, I
> > > can't even use a usb mouse connected to it by wire... even if cpu
> > > usage is really low (less than 10%) I don't have the issue if I
> > > connect a M2 usb3 flash drive, with comparable transfert speed... so
> > > not related to some bus over usage...
> > > 
> > > so I suspect that there is an issue with the usb driver, and that
> > > maybe the bluetooth issue can be related to the usb issue, since the
> > > bluetooth controller is on the usb bus on the laptop >
> > > the transfer issue of usb is much more older than the bluetooth
> > > issue, it's approximative, but : > - the btusb boot issue is about 3 month old,
> > > - the suspend/resume issue of bluetooth is more than a year old
> > > - the usb transfer issue as more than a year...
> > > 
> > > I'll gladly provide any useful information, can also do patch tries...
> > 
> > As you use Gentoo and are able to build your own Linux kernel, the
> > fastest way to get these issues addressed is to bisect them. To shorten
> > the test cycles, I recommend to try, if you can reproduce the issues in
> > QEMU and passing through the problematic devices to the VM [1][2].
> > 
> > I also recommend to start a separate thread for each issue and, as these
> > seem to be regressions, also keep the regression folks in the loop [3].
> > 
> > Kind regards,
> > 
> > Paul
> > 
> > (The commands were working for after all, and the device didn’t
> > show up due to a (second) Linux kernel regression.)
> > [2]: https://station.eciton.net/qemu-usb-host-device-pass-through.html
> > [3]:
> > https://www.kernel.org/doc/html/latest/admin-guide/reporting-regressions.html
> 
> 
> thanks for the reply,
> after upgrade to 6.2.11 version,
> - seems that the issue at boot of btusb module is gone... I don't have anymore to unload btusb module after a boot... so that's great news!
> - the bluetooth resumes also after a suspend S3 (suspend to ram)! so great news again! (I've to check for suspend to disk, but I'm confident about success)
> 
> I still have to do some check with the micro usb readers to see if the bug "bus monopolization" is still present, so far so good!
> 
> 
> 
> Sent with Proton Mail secure email.

so far, so good... the usb congestion seems also to be gone... tried {usb micro flash reader}->{disk}, {disk}->{usb micro flash reader}, {usb micro flash reader}->{usb micro flash reader}, all CPU 100% due to gentoo upgrade... and the {bluetooth mouse} connected to the {bluetooth usb host} had no lag... before (2~3 years) I used the trackpad because it was soooo laggy or not even moving... so, there is something between  6.2.8 and 6.2.11 that did fixed the issue.

sorry for the inconvenience, seems that just reporting the issue (that was annoying me for years) did fixed the issue...
this can be closed, thanks and regards
JLM