USB Storage Drive Support

[jasondaming]

I tried testing a couple of different storage drives. The 64gb one formatted as exFAT was not recognized and the 32gb one formatted as Fat32 was. What is the expected behavior and what drives and formats are supported?

Does it make sense to have a eject button on the USB devices section of the System Tab? I realize this wasn't supported by the roboRIO and isn't a required just a nice to have. I also know that I could probably do an unmount manually but that isn't something a typical user would know.

[jwbonner]

I've also observed USB drive mounting being inconsistent. I've tried a 16GB drive formatted with both FAT32 and exFAT. I believe it consistently claims to be mounted when running df, but the mount appears to be functional maybe half of the time (i.e. the contents of the drive actually appear). Unplugging and reconnecting the same drive will often cause different behavior.

[jasondaming]

Additionally I noticed my drives getting rather hot. I don't know if there is anything to be done here but my drive with a metal case especially was very hot to the touch after only being plugged in for a minute.

[bhjelstrom]

Thanks for the notes - this is a pretty new feature and I think we can do better with a custom dedicated service. The current approach just doesn't seem to work consistently.

[bhjelstrom]

@jasondaming can you send the model number of your flash drive? My guess is that there is a broad intermittent issue with the automounting, but it would be good to repro and fix what you're experiencing directly. Did the drive successfully mount? Were you writing to it at all?

Adding UI eject/unmount to the list. I think it's a good feature to have.

[jasondaming]

TLDR:
I tried like 25 drives plus a hub. All 32gb or less worked. The 2 drives I had 64 or greater failed. USB microSD card 128gb worked.
Kingston DTSE9 G2 64gb
SanDisk Ultra SDCZ48-064G

==================================================
@bhjelstrom
Like it isn't even detected. Plug it in, nothing registers plug the Fat32 one after and it goes straight to U and on the UI and works fine.

Changed Fat32 to exFat and it worked fine.

Changed exFat drive to NTFS (Can't do Fat32 too big) and it still didn't see it.

Tried another 4gb drive worked fine.

Tried another 64gb drive different and it also completely unrecognized.

So then I was thinking >=64gb drive might be the issue (the only 2 >64 I could find).

Then I tried a USB microsd card reader (teams please don't do this) and that surprisingly worked great with a 128gb sd card.

Then I tried a USB Hub with Ethernet and 3 already working <32gb drives and all worked fine (it also enumerated the hub and ethernet adapter).

Then I tried another USB drive I found in my backpack that was 128gb that also failed. This one also caused my computer's explorer to crash so I am guessing it has deeper problems.

[jlmcmchl]

We plugged a 32Gb usb drive into systemcore yesterday to collect datalogs and found that robot.service could not write to /U/logs, so logs instead accumulated within /home/systemcore/logs. I noticed that /U was mounted with root ownership, but given that robot code appears to be running as root I'm unclear as to why logs were recorded to internal storage.

[bhjelstrom]

@jasondaming You might see an improvement in the latest release. I can reliably hotplug / boot with my 512GB drive with 3 others attached
Eject was added as well

@jlmcmchl we should have this worked out by release 3

[jasondaming]

I just tried 159. There was no change with the 2 devices that weren't working before.

Is there a good way to verify the version (I am fairly confident it took but just wanted to triple check)?

Here is some logs that might be helpful looked like some interesting things happened?:

[ 6.101072] macb 1f00100000.ethernet eth0: PHY [1f00100000.ethernet-ffffffff:00] driver [Broadcom BCM54213PE] (irq=POLL)
[ 6.101080] macb 1f00100000.ethernet eth0: configuring for phy/rgmii-id link mode
[ 6.104220] pps pps0: new PPS source ptp1
[ 6.104477] macb 1f00100000.ethernet: gem-ptp-timer ptp clock registered.
[ 6.110759] mcp251xfd spi2.0 can_s0: renamed from can4
[ 6.147766] mcp251xfd spi2.1 can_s1: renamed from can3
[ 6.165300] mcp251xfd spi1.0 can_s2: renamed from can2
[ 6.185211] mcp251xfd spi1.1 can_s3: renamed from can1
[ 6.256784] mcp251xfd spi1.2 can_s4: bit-timing not yet defined
[ 6.278408] mcp251xfd spi2.0 can_s0: bit-timing not yet defined
[ 6.297267] mcp251xfd spi2.1 can_s1: bit-timing not yet defined
[ 6.313952] mcp251xfd spi1.0 can_s2: bit-timing not yet defined
[ 6.331483] mcp251xfd spi1.1 can_s3: bit-timing not yet defined
[ 6.375637] brcmfmac: brcmf_cfg80211_set_power_mgmt: power save enabled
[ 6.377749] mcp251xfd spi1.2 can_s4: bit-timing not yet defined
[ 6.378854] mcp251xfd spi2.0 can_s0: bit-timing not yet defined
[ 6.379520] mcp251xfd spi1.1 can_s3: bit-timing not yet defined
[ 6.381196] mcp251xfd spi2.1 can_s1: bit-timing not yet defined
[ 6.381830] mcp251xfd spi1.0 can_s2: bit-timing not yet defined
[ 6.398177] bridge: filtering via arp/ip/ip6tables is no longer available by default. Update your scripts to load br_netfilter if you need this.
[ 6.399468] Bridge firewalling registered
[ 6.421898] audit: type=1325 audit(1706136355.967:7): table=filter family=2 entries=0 op=xt_register pid=648 comm="iptables"
[ 6.421944] audit: type=1300 audit(1706136355.967:7): arch=c00000b7 syscall=209 success=yes exit=0 a0=4 a1=0 a2=40 a3=7fc324e520 items=0 ppid=513 pid=648 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=(none) ses=4294967295 comm="iptables" exe="/usr/sbin/xtables-legacy-multi" key=(null)
[ 6.421964] audit: type=1327 audit(1706136355.967:7): proctitle=2F7573722F7362696E2F69707461626C6573002D2D77616974002D4C002D6E
[ 6.437025] audit: type=1325 audit(1706136355.979:8): table=nat family=2 entries=0 op=xt_register pid=651 comm="iptables"
[ 6.437135] audit: type=1300 audit(1706136355.979:8): arch=c00000b7 syscall=209 success=yes exit=0 a0=4 a1=0 a2=40 a3=7fedddb8e0 items=0 ppid=513 pid=651 auid=4294967295 uid=0 gid=0 euid=0 suid=0 fsuid=0 egid=0 sgid=0 fsgid=0 tty=(none) ses=4294967295 comm="iptables" exe="/usr/sbin/xtables-legacy-multi" key=(null)
[ 6.437140] audit: type=1327 audit(1706136355.979:8): proctitle=2F7573722F7362696E2F69707461626C6573002D2D77616974002D74006E6174002D4400505245524F5554494E47002D6D006164647274797065002D2D6473742D74797065004C4F43414C002D6A00444F434B4552
[ 6.454681] audit: type=1325 audit(1706136356.007:9): table=nat family=2 entries=5 op=xt_replace pid=669 comm="iptables"
[ 6.515224] Initializing XFRM netlink socket
[ 6.880931] rp1-firmware: probe of rp1_firmware failed with error -110
[ 7.167760] configfs-gadget.libComposite gadget.0: HOST MAC 00:1a:2c:a2:bb:46
[ 7.167768] configfs-gadget.libComposite gadget.0: MAC 00:1a:2b:a2:bb:46
[ 7.168163] configfs-gadget.libComposite gadget.0: HOST MAC 00:1a:2e:a2:bb:46
[ 7.168165] configfs-gadget.libComposite gadget.0: MAC 00:1a:2d:a2:bb:46
[ 7.171043] dwc2 1000480000.usb: bound driver configfs-gadget.libComposite
[ 8.405906] usb 1-1: USB disconnect, device number 2
[ 8.760938] usb 1-1: new full-speed USB device number 3 using xhci-hcd
[ 9.084867] CAN Heartbeat: Initializing module
[ 9.084880] CAN Heartbeat: Searching for CAN devices with pattern can_s*
[ 9.084882] CAN Heartbeat: Found matching CAN device can_s4
[ 9.084885] CAN Heartbeat: Found matching CAN device can_s3
[ 9.084886] CAN Heartbeat: Found matching CAN device can_s2
[ 9.084888] CAN Heartbeat: Found matching CAN device can_s1
[ 9.084890] CAN Heartbeat: Found matching CAN device can_s0
[ 9.084893] CAN Heartbeat: Found 5 matching CAN devices
[ 9.084919] CAN Heartbeat: Module initialized successfully
[ 9.088564] i2c_dev: i2c /dev entries driver
[ 9.684387] usb 1-1: New USB device found, idVendor=cafe, idProduct=4011, bcdDevice= 1.00
[ 9.684396] usb 1-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[ 9.684399] usb 1-1: Product: Limelight RT Subsystem
[ 9.684402] usb 1-1: Manufacturer: Limelight Vision
[ 9.684404] usb 1-1: SerialNumber: 76D98A400020F5FF
[ 9.721516] cdc_acm 1-1:1.0: ttyACM0: USB ACM device
[ 9.721575] usbcore: registered new interface driver cdc_acm
[ 9.721577] cdc_acm: USB Abstract Control Model driver for USB modems and ISDN adapters
[ 10.166362] usb 1-1: usbfs: interface 0 claimed by cdc_acm while 'iodaemon' sets config #0
[ 10.206401] macb 1f00100000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off
[ 19.933329] platform 1f00178000.pio: deferred probe pending
[ 21.069718] brcmfmac: brcmf_set_channel: set chanspec 0x100c fail, reason -52
[ 21.069943] brcmfmac: brcmf_set_channel: set chanspec 0x100d fail, reason -52

[bhjelstrom]

@jasondaming Thanks, are you booting with devices plugged in? Or hotplugging?

We will have your drives here soon for a more direct repro of what you are seeing.

[jasondaming]

I was just hotplugging. I didn't think to test on boot up but am happy to report that both devices now show up on boot up (I am pretty sure I tested that in my testing on the previous version).

Unfortunately once unplugged I can't get the device back without rebooting.

[bhjelstrom]

Ok, this makes sense and could actually help us fix hotplugging, can you boot without the drives, then hotplug, then grab the complete dmesg log?

[jasondaming]

Hopefully this helps

dmesg_output.txt

[bhjelstrom]

@jasondaming

[   40.091716] ntfs: (device sdb1): load_system_files(): $LogFile is not clean.  Mounting read-only.  Mount in Windows.
[   40.917848] ntfs: volume version 3.1.
[   40.920110] ntfs: (device sda1): load_system_files(): $LogFile is not clean.  Mounting read-only.  Mount in Windows.

https://forensafe.com/blogs/windowslogfile.html

Looks like removing without ejection on windows can cause this to happen. I think Windows can clean this up automatically, and there is a way to automatically clean this up in our OS as well.

---

Is this correct:

1. You booted, hotplugged two NTFS drives, and neither of them mounted properly. They are mounted read-only?
2. These two drives successfully mount if they are plugged in before booting?

Or are we missing events for another drive?

[jasondaming]

@bhjelstrom
I am really unsure of the state my systemcore is in. I updated to 159. I now see a higher voltage. However I don't see any of the UI improvements like the eject USB. Should we troubleshoot this more or should I just go to 160 and see if that resolves things?