USB Storage Slow?

[DonC52]

I have been wondering why the following occurs on my Windows 10 laptop and the Pi USB. I am using a 128G SanDisk Extreme Plus UHS so I would think it rather fast.

I put in an old "slow" USB stick into my Laptop port and copy a Sentry Mode directory, I get about 20MB/Sec transfer rate.

I put the Pi USB in with power applied from external source and copy the same directory, I get like 1.5MB/s to 2.5MB/s transfer rate.

Looking at top, there is a LOT of wait time going on. Is there anything that can be done to speed things up?

[marcone]

The time you circled in that screenshot is I/O wait time, i.e. time spent waiting for data to be read from the sd card, and time spent transferring data over USB, both of which probably happen via DMA, thus the CPU will be mostly idle during those times.

As to why you get such low transfer rates, no idea. How are you measuring that? I'm getting around 20 MB/sec on both read and write with 'dd' when attached to a Mac or linux machine.

[marcone]

Does dmesg on the Pi show "dwc2 20980000.usb: new device is high-speed" or something else?

[DonC52]

I got the following from dmesg | grep "dwc2 20980000.usb" while connected to the car. It looks like the last timestamps show it as high-speed.
''' root@teslausb: dmesg | grep "dwc2 20980000.usb" [ 3.136821] dwc2 20980000.usb: 20980000.usb supply vusb_d not found, using dummy regulator [ 3.137004] dwc2 20980000.usb: Linked as a consumer to regulator.0 [ 3.137033] dwc2 20980000.usb: 20980000.usb supply vusb_a not found, using dummy regulator [ 3.353750] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm=1 [ 3.353775] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm_clock_gating=1 [ 3.353789] dwc2 20980000.usb: dwc2_check_params: Invalid parameter besl=1 [ 3.353800] dwc2 20980000.usb: dwc2_check_params: Invalid parameter hird_threshold_en=1 [ 3.353866] dwc2 20980000.usb: EPs: 8, dedicated fifos, 4080 entries in SPRAM [ 3.355146] dwc2 20980000.usb: DWC OTG Controller [ 3.355220] dwc2 20980000.usb: new USB bus registered, assigned bus number 1 [ 3.355301] dwc2 20980000.usb: irq 33, io mem 0x20980000 [ 93.619677] dwc2 20980000.usb: bound driver g_mass_storage [ 93.641717] dwc2 20980000.usb: new device is high-speed [ 94.006578] dwc2 20980000.usb: new device is high-speed [ 94.248037] dwc2 20980000.usb: new address 13 [ 756.636039] dwc2 20980000.usb: bound driver g_mass_storage [ 756.762610] dwc2 20980000.usb: new device is high-speed [ 756.836337] dwc2 20980000.usb: new device is high-speed [ 756.897926] dwc2 20980000.usb: new address 14 root@teslausb:~# '''

I test out new things by copying a test event to the Pi USB TeslaCam/SavedClips so it can transfer something without driving the car. I only has like 15 files in it and takes like 6 minutes to copy the file. I use the display that Windows shows when copying a file (details) and it is usually close to whats going on as far as speed. (Thus the 20MB/S to a regular "old" stick

Tomorrow evening I can take out the Pi from the car and plug it into my Laptop. Then take a look at what dmesg has to say.

Sound like a plan?

[DonC52]

Well another day - I first plugged the Pi into power from the wall and let it boot all the way up. I then connected it to he USB port on the computer. Doing a copy of the test file again is like 1.5MB/s rather than the 20 expected. Still a lot of wait going on.

Looking at the windows device manager, I noticed that the Pi is seen as a Linux File-Stor Gadget USB Device. I am wondering if there is an issue with that driver in Windows? When I READ from the Pi TeslaCam/SavedClips I get a flat 20MB/s. When WRITING to the Pi folder, I get between 1.5 - .5MB/s.

Anyone else plug the Pi into a Windows 10 machine and try to write to the TeslaCam folder? Here is the dmesg output: ''' pi@teslausb:~ $ dmesg | grep "dwc2 20980000.usb" [ 3.281850] dwc2 20980000.usb: 20980000.usb supply vusb_d not found, using dummy regulator [ 3.282043] dwc2 20980000.usb: Linked as a consumer to regulator.0 [ 3.282074] dwc2 20980000.usb: 20980000.usb supply vusb_a not found, using dummy regulator [ 3.503745] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm=1 [ 3.503770] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm_clock_gating=1 [ 3.503783] dwc2 20980000.usb: dwc2_check_params: Invalid parameter besl=1 [ 3.503795] dwc2 20980000.usb: dwc2_check_params: Invalid parameter hird_threshold_en=1 [ 3.503863] dwc2 20980000.usb: EPs: 8, dedicated fifos, 4080 entries in SPRAM [ 3.505247] dwc2 20980000.usb: DWC OTG Controller [ 3.505316] dwc2 20980000.usb: new USB bus registered, assigned bus number 1 [ 3.505401] dwc2 20980000.usb: irq 33, io mem 0x20980000 [ 92.034137] dwc2 20980000.usb: bound driver g_mass_storage [ 344.999742] dwc2 20980000.usb: new device is high-speed [ 345.052498] dwc2 20980000.usb: new address 2 pi@teslausb:~ $ '''

[DonC52]

Okay - after I sent the last Commment, things changed. Now when WRITING to the Pi I get 120MB/s for a few seconds and then 19.5-20.8MB/s. I have no idea what changed to get the write times better. READING back gives me 20.0-27.7MB/s.

I don't know if I taught the Windows driver something or just what is going on. Much better though.

Still - has any Windows 10 people out there playing around and noticed the same thing?

[DonC52]

Think I have a handle on the speed issue but I need sleep. Will follow-up tomorrow after work.

[curiousgrge]

Okay - after I sent the last Commment, things changed. Now when WRITING to the Pi I get 120MB/s for a few seconds and then 19.5-20.8MB/s. I have no idea what changed to get the write times better. READING back gives me 20.0-27.7MB/s.

I don't know if I taught the Windows driver something or just what is going on. Much better though.

Still - has any Windows 10 people out there playing around and noticed the same thing?

I'm having the exact same problems as you. 1.5-2.5MB/sec. If you have a handle on what you've changed for direct copying, I'm interested in how this is resolved. Took me 2.5+hrs to copy my MUSIC partition back. Even to get 20MB/sec would be dramatic improvement in copy time. For the data connection, I'm connected via USB 3.0. Possibly my issue is my USB cable now that I think about it since it is a thin flat cable that came with my Smart Watch.

[marcone]

You could try adding nofua=1,1 to /etc/modprobe.d/g_mass_storage.conf and see if that helps. (see https://www.kernel.org/doc/Documentation/usb/mass-storage.txt for details)

[curiousgrge]

Just tried the adding nofua=1,1 to the tail end of the conf file and now it is 4-5MB/sec but well shy of 20MB/sec.

[DonC52]

I "Think" I found the solution. Windows 10 tends to want writes to complete before issuing more writes. (NO write caching!) To speed this up, I went into Device Manager, selected Disk Drives and you will see the Linux File-Stor Gadget USB Device. (two of them) Select the policies tab and select Better performance. The write-caching check box will be selected but DO NOT turn on the flushing check-box. (Very bad idea) Do this for BOTH of the devices.

Doing this simple change enabled me to copy 200 music files (about 797MB) in about 30 seconds or so. From the 4-5MB/s it went to about 150MB/s with a short time delay at the end for the system to flush the cache.

Hopefully it will help you as well.

[curiousgrge]

Redoing my TeslaUSB again due to it now not giving me notifications. Tested @DocC52 suggestion and it is at least 3x faster than what it was doing again. After the initial caching, it does slow down quite a bit. Takes about 30 minutes now to copy everything for me rather than 2.5hrs so I guess it improves it 5x.

I didn't test this with the nofua=1,1 so don't know if that makes a difference too.

Also keep in mind if you do this that user will need to eject the drive rather than just being able to pull out. Not sure why the image was unlucky and says 3MB/sec when it was mostly hovering around 15MB/sec or higher.

[ZeFrenchBibster]

Hi,

Not sure if this is related, but I have the same sort of micro SD (128GB Sandisk 'claims' 100 MB/sec) in my pi (Zero-W) behind a 'Zendure 5' powerbanck/HUB.

Once the car wakes up, it works, but after a while, she claims that the USB storage is too slow. I'll fetch it from the car tonight, and try the 'nofua=1,1' thing, as well as any messages for /var/log/messages.

[githubsean]

Sorry, no solutions, just adding a data point.

I'm also having troubles with this, and I wonder if it is some bottleneck somewhere in the raspberry pi. I tested a card on both windows and linux computers - i.e. I plugged the raspberry pi acting as a USB drive into both computers and ran some write speed tests. They both came out to about 2MB/sec. I then took the card out of the pi and tried writing directly to it - about 18MB/sec. I haven't done any more investigation to see if it is:

• on the pi, the raw write speed on the pi to the USB (i.e. slow USB interface)
• on the pi, the write speed to the file system file used as the USB storage (some issue with the file system driver) I have tried the nofua=1,1 thing and I didn't see any significant difference in write speed.

I may try the new "Buster" version to see if things improve.

[mcowger]

I've enabled nofua on my 0W - will see how it goes today.

[mcowger]

I did some quick tests just now. This is a High Endurance SanDisk U3, Class10, V30 card I'd tested well over 25MB/sec writes (its rated for 40). nofua didn't seem to help much.

• Using the Pi0W attached to my Mac as a USB drive, I started a streaming write with 4KB blocks to the drive.
• Initially it was around 20MB/sec, but very quickly (after about 60s) dropped off to less than 1MB/sec, which makes sense that the car would be unhappy with.
• Then I tried it directly on the Pi, mounting the card through to g_mass_storage driver as /mnt/cam. Speeds were inconsistent, but often dipping below 2MB/sec, and then back up to 6.

My views:

1. Given the initial performance of 20MB/sec, I dont think its the RPi USB interface.
2. Given the initial performance and then drop, this feels like overrunning a cache somewhere, and then dropping straight to disk-performance.
3. Given the poor performance when loopback mounted with g_mass_storage, my gut suggests this is a performance problem in the kernel driver or just the RPi0W in general.

Next steps - I'm considering grabbing one of my faster ODroidC2's to see if its a CPU performance issue.

[mcowger]

Update - couldn't quite get my RPi3b+ to work.

[marcone]

Update - couldn't quite get my RPi3b+ to work.

Only the Pi0 and Pi4 support USB gadget mode.

[mcowger]

Well that would explain it :)

I'll pick up a RPi4 and play.

On Mon, Sep 30, 2019 at 4:47 PM marcone [email protected] wrote:

Update - couldn't quite get my RPi3b+ to work.

Only the Pi0 and Pi4 support USB gadget mode.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AAOXCTHJQAOOI7YYKH2AJ3DQMKFYFA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD77OCBA#issuecomment-536797444, or mute the thread https://github.com/notifications/unsubscribe-auth/AAOXCTCBPCBLWPEGGLPFMMLQMKFYFANCNFSM4IUEPZCA .

-- -- Matt

[gacevedo]

I'll try the nofua=1,1 setting next. Here's the output from dmesg:

root@teslausb:~# dmesg | grep 'dwc2 20980000.usb' [ 2.974130] dwc2 20980000.usb: 20980000.usb supply vusb_d not found, using dummy regulator [ 2.974288] dwc2 20980000.usb: Linked as a consumer to regulator.0 [ 2.974317] dwc2 20980000.usb: 20980000.usb supply vusb_a not found, using dummy regulator [ 3.194131] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm=1 [ 3.194158] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm_clock_gating=1 [ 3.194171] dwc2 20980000.usb: dwc2_check_params: Invalid parameter besl=1 [ 3.194183] dwc2 20980000.usb: dwc2_check_params: Invalid parameter hird_threshold_en=1 [ 3.194250] dwc2 20980000.usb: EPs: 8, dedicated fifos, 4080 entries in SPRAM [ 3.195656] dwc2 20980000.usb: DWC OTG Controller [ 3.195726] dwc2 20980000.usb: new USB bus registered, assigned bus number 1 [ 3.195811] dwc2 20980000.usb: irq 33, io mem 0x20980000 [ 1845.177686] dwc2 20980000.usb: bound driver g_mass_storage [ 1845.434266] dwc2 20980000.usb: new device is high-speed [ 1845.490515] dwc2 20980000.usb: new device is high-speed [ 1845.545863] dwc2 20980000.usb: new device is high-speed [ 1845.601519] dwc2 20980000.usb: new device is high-speed [ 1845.656760] dwc2 20980000.usb: new device is high-speed [ 1845.712490] dwc2 20980000.usb: new device is high-speed [ 1845.768239] dwc2 20980000.usb: new device is high-speed [ 1845.823636] dwc2 20980000.usb: new device is high-speed [ 1845.878855] dwc2 20980000.usb: new device is high-speed [ 1845.970411] dwc2 20980000.usb: new address 10

[itsrainingben]

Apologies no solutions on my side either, just throwing my hat into this issue. Was not something I ran into prior to V10 & updating the pi following instructions here.

I mounted the rPi (stem enclosure if that matters) to a macOS host and tested read/write with Aja System Test Lite and saw between 17-19MB/s write. The software would error out after writing about 1GB. New to dmesg, so I tailed it and this is what I was able to pull -

disk3s1: I/O error. disk3s1: I/O error. **** [IOBluetoothDevice][decrementNumberOfOutstandingPacketsBy] -- Handle 0x000c -- remove ACL Timer 0xffffff8069112d20 for ACL packet 0xffffff806d2bb40 0 -- numOutstandingHWPackets = 1, mNumNotDequeuedAckedPackets = 0, mDestroyDeviceCalled = FALSE -- hostController = 0x1080 ****

**** [IOBluetoothDevice][decrementNumberOfOutstandingPacketsBy] -- Handle 0x000c -- remove ACL Timer 0xffffff8069113e60 for ACL packet 0xffffff806d1c6c0 0 -- numOutstandingHWPackets = 1, mNumNotDequeuedAckedPackets = 0, mDestroyDeviceCalled = FALSE -- hostController = 0x1080 **** disk3s1: I/O error. disk3s1: I/O error. disk3s1: I/O error. **** [IOBluetoothDevice][decrementNumberOfOutstandingPacketsBy] -- Handle 0x000c -- remove ACL Timer 0xffffff8057255900 for ACL packet 0xffffff806d15520 0 -- numOutstandingHWPackets = 1, mNumNotDequeuedAckedPackets = 0, mDestroyDeviceCalled = FALSE -- hostController = 0x1080 ****

If there are any recommended commands for dmesg feel free to throw them here and I'm happy to run. Otherwise I'll be following closely and thank you @marcone for all your work on this project.

[mcowger]

@marcone @gacevedo - Got my RPi4 today and installed the latest buster based release on it.

Native card write performance: 58MB/sec Thru RPI as USB (like the car would): 20.1MB/sec

This is vastly faster. Also, I watched the power consumption using an inline meter - even with default settings, the RPi4 maxed at a momentary 4.2W draw during install, and after settling down during normal running conditions draws about 2.9W....well within the power delivery available on at least the Model 3.

I've got it running in the car right now and will report back.

[DMBlakeley]

Curious what RAM size of RPi4 people are using?

[mcowger]

Mine is 1GB (which is 2x that of the RPi0W), the smallest. memory usage on the device shows 65MB.

[DMBlakeley]

Thanks! Didn’t want to go overboard.

[itsrainingben]

Hey Mat - 20MB/s is what I was get on the zeroW over USB (like the car does).

On Oct 1, 2019, at 5:24 PM, Matt Cowger [email protected] wrote:

 @marcone @gacevedo - Got my RPi4 today and installed the latest buster based release on it.

Native card write performance: 58MB/sec Thru RPI as USB (like the car would): 20.1MB/sec

This is vastly faster. Also, I watched the power consumption using an inline meter - even with default settings, the RPi4 maxed at a momentary 4.2W draw during install, and after settling down during normal running conditions draws about 2.9W....well within the power delivery available on at least the Model 3.

I've got it running in the car right now and will report back.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

[mcowger]

I got 20mb/s at first l, then it throttled down to like 2. This is sustained.

On Tue, Oct 1, 2019 at 5:59 PM Ben Tucker [email protected] wrote:

Hey Mat - 20MB/s is what I was get on the zeroW over USB (like the car does).

On Oct 1, 2019, at 5:24 PM, Matt Cowger [email protected] wrote:

 @marcone @gacevedo - Got my RPi4 today and installed the latest buster based release on it.

Native card write performance: 58MB/sec Thru RPI as USB (like the car would): 20.1MB/sec

This is vastly faster. Also, I watched the power consumption using an inline meter - even with default settings, the RPi4 maxed at a momentary 4.2W draw during install, and after settling down during normal running conditions draws about 2.9W....well within the power delivery available on at least the Model 3.

I've got it running in the car right now and will report back.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AAOXCTDLWYE4KPJKSQ3BGBDQMPW7JA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADGSMA#issuecomment-537291056, or mute the thread https://github.com/notifications/unsubscribe-auth/AAOXCTBU77GYF2YOVVA46MTQMPW7JANCNFSM4IUEPZCA .

-- -- Matt

[itsrainingben]

How long does that write speed sustain for? Mine errors out after a minute (roughly 1GB written to disk).

On Tue, Oct 1, 2019 at 6:03 PM Matt Cowger [email protected] wrote:

I got 20mb/s at first l, then it throttled down to like 2. This is sustained.

On Tue, Oct 1, 2019 at 5:59 PM Ben Tucker [email protected] wrote:

Hey Mat - 20MB/s is what I was get on the zeroW over USB (like the car does).

On Oct 1, 2019, at 5:24 PM, Matt Cowger [email protected] wrote:

 @marcone @gacevedo - Got my RPi4 today and installed the latest buster based release on it.

Native card write performance: 58MB/sec Thru RPI as USB (like the car would): 20.1MB/sec

This is vastly faster. Also, I watched the power consumption using an inline meter - even with default settings, the RPi4 maxed at a momentary 4.2W draw during install, and after settling down during normal running conditions draws about 2.9W....well within the power delivery available on at least the Model 3.

I've got it running in the car right now and will report back.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AAOXCTDLWYE4KPJKSQ3BGBDQMPW7JA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADGSMA#issuecomment-537291056 , or mute the thread < https://github.com/notifications/unsubscribe-auth/AAOXCTBU77GYF2YOVVA46MTQMPW7JANCNFSM4IUEPZCA

.

-- -- Matt

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AGE3NGRHSRMFX3T2FUPJYQ3QMPXMTA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADGYAI#issuecomment-537291777, or mute the thread https://github.com/notifications/unsubscribe-auth/AGE3NGVQPYNFJTEBCMYCRTDQMPXMTANCNFSM4IUEPZCA .

[mcowger]

On rpi0w it’s about the same.

On my 4 I sustained it for as long as I tested (5ish mins).

On Tue, Oct 1, 2019 at 6:10 PM Ben Tucker [email protected] wrote:

How long does that write speed sustain for? Mine errors out after a minute (roughly 1GB written to disk).

On Tue, Oct 1, 2019 at 6:03 PM Matt Cowger [email protected] wrote:

I got 20mb/s at first l, then it throttled down to like 2. This is sustained.

On Tue, Oct 1, 2019 at 5:59 PM Ben Tucker [email protected] wrote:

Hey Mat - 20MB/s is what I was get on the zeroW over USB (like the car does).

On Oct 1, 2019, at 5:24 PM, Matt Cowger [email protected] wrote:

 @marcone @gacevedo - Got my RPi4 today and installed the latest buster based release on it.

Native card write performance: 58MB/sec Thru RPI as USB (like the car would): 20.1MB/sec

This is vastly faster. Also, I watched the power consumption using an inline meter - even with default settings, the RPi4 maxed at a momentary 4.2W draw during install, and after settling down during normal running conditions draws about 2.9W....well within the power delivery available on at least the Model 3.

I've got it running in the car right now and will report back.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or mute the thread.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub <

https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AAOXCTDLWYE4KPJKSQ3BGBDQMPW7JA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADGSMA#issuecomment-537291056

, or mute the thread <

https://github.com/notifications/unsubscribe-auth/AAOXCTBU77GYF2YOVVA46MTQMPW7JANCNFSM4IUEPZCA

.

-- -- Matt

— You are receiving this because you commented. Reply to this email directly, view it on GitHub < https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AGE3NGRHSRMFX3T2FUPJYQ3QMPXMTA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADGYAI#issuecomment-537291777 , or mute the thread < https://github.com/notifications/unsubscribe-auth/AGE3NGVQPYNFJTEBCMYCRTDQMPXMTANCNFSM4IUEPZCA

.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/marcone/teslausb/issues/239?email_source=notifications&email_token=AAOXCTACWGY5435PBVBX44DQMPYIVA5CNFSM4IUEPZCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEADHC6Q#issuecomment-537293178, or mute the thread https://github.com/notifications/unsubscribe-auth/AAOXCTCZPMP36KRHWDJZADDQMPYIVANCNFSM4IUEPZCA .

-- -- Matt

[itsrainingben]

Just got home, able to add my dmesg for @marcone 's request -

root@teslapi:/home/pi# dmesg | grep "dwc2 20980000.usb" [ 3.426141] dwc2 20980000.usb: 20980000.usb supply vusb_d not found, using dummy regulator [ 3.426308] dwc2 20980000.usb: Linked as a consumer to regulator.0 [ 3.426337] dwc2 20980000.usb: 20980000.usb supply vusb_a not found, using dummy regulator [ 3.643940] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm=1 [ 3.643965] dwc2 20980000.usb: dwc2_check_params: Invalid parameter lpm_clock_gating=1 [ 3.643980] dwc2 20980000.usb: dwc2_check_params: Invalid parameter besl=1 [ 3.643993] dwc2 20980000.usb: dwc2_check_params: Invalid parameter hird_threshold_en=1 [ 3.644061] dwc2 20980000.usb: EPs: 8, dedicated fifos, 4080 entries in SPRAM [ 3.645655] dwc2 20980000.usb: DWC OTG Controller [ 3.645765] dwc2 20980000.usb: new USB bus registered, assigned bus number 1 [ 3.645850] dwc2 20980000.usb: irq 33, io mem 0x20980000 [ 74.977688] dwc2 20980000.usb: bound driver g_mass_storage [ 75.001183] dwc2 20980000.usb: new device is high-speed [ 75.365197] dwc2 20980000.usb: new device is high-speed [ 75.617993] dwc2 20980000.usb: new device is high-speed [ 75.682544] dwc2 20980000.usb: new address 10 [ 1057.171063] dwc2 20980000.usb: bound driver g_mass_storage [ 1057.314384] dwc2 20980000.usb: new device is high-speed [ 1057.376563] dwc2 20980000.usb: new device is high-speed [ 1057.439266] dwc2 20980000.usb: new address 8

[marcone]

That looks OK to me