Samsung T7 1TB SSD issues

Message

alexbuell · Post by **alexbuell** » Sat Dec 10, 2022 7:02 pm

Anyone else got one of these Samsung T7 1TB SSDs? I'm seeing the following whenever I plug mine in. It still works fine, but performance is bloody horrible. It works just fine on my other machine. This machine is an Asrock X399 motherboard rocking a threadripper 2970wx w/128GB ECC RAM.

Code: Select all

Nov 10 22:54:17 titanium kernel: usb 2-8.2: new SuperSpeed USB device number 15 using xhci_hcd
Nov 10 22:54:17 titanium kernel: usb 2-8.2: New USB device found, idVendor=04e8, idProduct=4001, bcdDevice= 1.00
Nov 10 22:54:17 titanium kernel: usb 2-8.2: New USB device strings: Mfr=2, Product=3, SerialNumber=1
Nov 10 22:54:17 titanium kernel: usb 2-8.2: Product: PSSD T7
Nov 10 22:54:17 titanium kernel: usb 2-8.2: Manufacturer: Samsung
Nov 10 22:54:17 titanium kernel: usb 2-8.2: SerialNumber: S6U8NS0RC31147R
Nov 10 22:54:17 titanium mtp-probe[47454]: checking bus 2, device 15: "/sys/devices/pci0000:00/0000:00:01.1/0000:01:00.0/usb2/2-8/2-8.2"
Nov 10 22:54:17 titanium mtp-probe[47454]: bus: 2, device: 15 was not an MTP device
Nov 10 22:54:17 titanium kernel: scsi host10: uas
Nov 10 22:54:17 titanium kernel: scsi 10:0:0:0: Direct-Access     Samsung  PSSD T7          0    PQ: 0 ANSI: 6
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] 1953525168 512-byte logical blocks: (1.00 TB/932 GiB)
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] Write Protect is off
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] Mode Sense: 43 00 00 00
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] Preferred minimum I/O size 512 bytes
Nov 10 22:54:17 titanium kernel: sd 10:0:0:0: [sde] Optimal transfer size 33553920 bytes
Nov 10 22:54:17 titanium mtp-probe[47467]: checking bus 2, device 15: "/sys/devices/pci0000:00/0000:00:01.1/0000:01:00.0/usb2/2-8/2-8.2"
Nov 10 22:54:17 titanium mtp-probe[47467]: bus: 2, device: 15 was not an MTP device
Nov 10 22:54:18 titanium kernel:  sde: sde1
Nov 10 22:54:18 titanium kernel: sd 10:0:0:0: [sde] Attached SCSI disk
Nov 10 22:54:18 titanium dbus-daemon[5825]: [session uid=1000 pid=5823] Activating service name='org.gnome.Shell.HotplugSniffer' requested by ':1.778' (uid=1000 pid=5897 comm="/usr/bin/gnome-shell")
Nov 10 22:54:18 titanium dbus-daemon[5825]: [session uid=1000 pid=5823] Successfully activated service 'org.gnome.Shell.HotplugSniffer'
Nov 10 22:54:48 titanium kernel: sd 10:0:0:0: [sde] tag#13 uas_eh_abort_handler 0 uas-tag 1 inflight: IN 
Nov 10 22:54:48 titanium kernel: sd 10:0:0:0: [sde] tag#13 CDB: opcode=0x28 28 00 54 01 00 00 00 00 08 00
Nov 10 22:54:48 titanium kernel: scsi host10: uas_eh_device_reset_handler start
Nov 10 22:54:49 titanium kernel: usb 2-8.2: reset SuperSpeed USB device number 15 using xhci_hcd
Nov 10 22:54:49 titanium kernel: scsi host10: uas_eh_device_reset_handler success
Nov 10 23:06:53 titanium kernel: usb 2-8.2: new SuperSpeed USB device number 16 using xhci_hcd
Nov 10 23:06:53 titanium kernel: usb 2-8.2: New USB device found, idVendor=04e8, idProduct=4001, bcdDevice= 1.00
Nov 10 23:06:53 titanium kernel: usb 2-8.2: New USB device strings: Mfr=2, Product=3, SerialNumber=1
Nov 10 23:06:53 titanium kernel: usb 2-8.2: Product: PSSD T7
Nov 10 23:06:53 titanium kernel: usb 2-8.2: Manufacturer: Samsung
Nov 10 23:06:53 titanium kernel: usb 2-8.2: SerialNumber: S6U8NS0RC31147R
Nov 10 23:06:53 titanium dhcpcd[4873]: eth0: part of a Router Advertisement expired
Nov 10 23:06:53 titanium dhcpcd[4870]: eth1: part of a Router Advertisement expired
Nov 10 23:06:53 titanium mtp-probe[48188]: checking bus 2, device 16: "/sys/devices/pci0000:00/0000:00:01.1/0000:01:00.0/usb2/2-8/2-8.2"
Nov 10 23:06:53 titanium mtp-probe[48188]: bus: 2, device: 16 was not an MTP device
Nov 10 23:06:53 titanium kernel: scsi host10: uas
Nov 10 23:06:53 titanium kernel: scsi 10:0:0:0: Direct-Access     Samsung  PSSD T7          0    PQ: 0 ANSI: 6
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] 1953525168 512-byte logical blocks: (1.00 TB/932 GiB)
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] Write Protect is off
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] Mode Sense: 43 00 00 00
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] Preferred minimum I/O size 512 bytes
Nov 10 23:06:53 titanium kernel: sd 10:0:0:0: [sde] Optimal transfer size 33553920 bytes
Nov 10 23:06:53 titanium mtp-probe[48232]: checking bus 2, device 16: "/sys/devices/pci0000:00/0000:00:01.1/0000:01:00.0/usb2/2-8/2-8.2"
Nov 10 23:06:53 titanium mtp-probe[48232]: bus: 2, device: 16 was not an MTP device
Nov 10 23:06:54 titanium kernel:  sde: sde1
Nov 10 23:06:54 titanium kernel: sd 10:0:0:0: [sde] Attached SCSI disk
Nov 10 23:06:54 titanium dbus-daemon[5825]: [session uid=1000 pid=5823] Activating service name='org.gnome.Shell.HotplugSniffer' requested by ':1.778' (uid=1000 pid=5897 comm="/usr/bin/gnome-shell")
Nov 10 23:06:54 titanium dbus-daemon[5825]: [session uid=1000 pid=5823] Successfully activated service 'org.gnome.Shell.HotplugSniffer'
Nov 10 23:06:54 titanium kernel: xhci_hcd 0000:01:00.0: ERROR Transfer event TRB DMA ptr not part of current TD ep_index 2 comp_code 1
Nov 10 23:06:54 titanium kernel: xhci_hcd 0000:01:00.0: Looking for event-dma 00000000afe4f310 trb-start 00000000afe4f320 trb-end 00000000afe4f390 seg-start 00000000afe4f000 seg-end 00000000afe4fff0
Nov 10 23:07:25 titanium kernel: sd 10:0:0:0: [sde] tag#22 uas_eh_abort_handler 0 uas-tag 1 inflight: IN
Nov 10 23:07:25 titanium kernel: sd 10:0:0:0: [sde] tag#22 CDB: opcode=0x28 28 00 54 2d 40 20 00 04 00 00
Nov 10 23:07:25 titanium kernel: scsi host10: uas_eh_device_reset_handler start
Nov 10 23:07:25 titanium kernel: usb 2-8.2: reset SuperSpeed USB device number 16 using xhci_hcd
Nov 10 23:07:25 titanium kernel: scsi host10: uas_eh_device_reset_handler success

Post by **NeddySeagoon** » Mon Dec 12, 2022 10:22 am

alexbuell,

What does

Code: Select all

smartctl -x /dev/sde

tell about the drive.
As its a USB attached drive, it may not be useful but we need to look.

alexbuell · Post by **alexbuell** » Tue Dec 13, 2022 12:17 pm

Here's the output from smartctl:

Code: Select all

$ smartctl -x /dev/sde
smartctl 7.3 2022-02-28 r5338 [x86_64-linux-6.0.1-gentoo-x86_64] (local build)
Copyright (C) 2002-22, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       Samsung Portable SSD T7
Serial Number:                      S6U8NS0RC31147R
Firmware Version:                   FXG42P2Q
PCI Vendor/Subsystem ID:            0x144d
IEEE OUI Identifier:                0x002538
Total NVM Capacity:                 1,000,204,886,016 [1.00 TB]
Unallocated NVM Capacity:           0
Controller ID:                      5
NVMe Version:                       1.3
Number of Namespaces:               1
Namespace 1 Size/Capacity:          1,000,204,886,016 [1.00 TB]
Namespace 1 Utilization:            858,306,785,280 [858 GB]
Namespace 1 Formatted LBA Size:     512
Local Time is:                      Tue Dec 13 12:16:37 2022 GMT
Firmware Updates (0x16):            3 Slots, no Reset required
Optional Admin Commands (0x0007):   Security Format Frmw_DL
Optional NVM Commands (0x001f):     Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat
Log Page Attributes (0x03):         S/H_per_NS Cmd_Eff_Lg
Maximum Data Transfer Size:         512 Pages
Warning  Comp. Temp. Threshold:     54 Celsius
Critical Comp. Temp. Threshold:     56 Celsius

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +     4.83W       -        -    0  0  0  0        0       0
 1 +     3.54W       -        -    1  1  1  1        0       0
 2 +     3.04W       -        -    2  2  2  2        0       0
 3 -   0.0500W       -        -    3  3  3  3    26000    1000

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         0

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        20 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    0%
Data Units Read:                    3,213,856 [1.64 TB]
Data Units Written:                 3,373,677 [1.72 TB]
Host Read Commands:                 21,947,939
Host Write Commands:                14,704,437
Controller Busy Time:               161
Power Cycles:                       202
Power On Hours:                     39
Unsafe Shutdowns:                   196
Media and Data Integrity Errors:    0
Error Information Log Entries:      0
Warning  Comp. Temperature Time:    0
Critical Comp. Temperature Time:    0
Temperature Sensor 1:               20 Celsius
Temperature Sensor 2:               23 Celsius

Warning: NVMe Get Log truncated to 0x200 bytes, 0x200 bytes zero filled
Error Information (NVMe Log 0x01, 16 of 64 entries)
No Errors Logged

I just realised I've been on the forums for more than 20 years!

Post by **NeddySeagoon** » Tue Dec 13, 2022 12:44 pm

alexbuell,

Its a 1TB NVMe drive on the end of USB3.x.
With only 1.72 TB written its no where near worn out. That's only the second time round the drive.
There are no issues in the smartctl output.

From the images on the Samsung website it appears to have a detachable cable with a micro-USB3 connector on the drive end.
I've always had issues with them becoming detached from the PCB inside the case, so micro-USB3 connectors are a no-no for me.
They are very difficult to hand solder too.

Some things to try.

Try another USB3 port.
Try another system.
... another cable.

This last one is a bit cheeky ....
Look closely at the micro-USB3 connector on the drive. It looks like two connectors side by side. That's because it is.
One part is a micro USB2 connector. The other (bigger) part is USB3.
If the drive will fall back to USB2. It should. Connect the drive with a micro USB2 cable and leave the USB3 part empty.
It should work at USB2 speeds.
If that works, the inference is that there is a problem inside the drive with the USB3 workings.

-- edit --

Congratulations on your 20 years with Gentoo :)

alexbuell · Post by **alexbuell** » Tue Dec 13, 2022 2:23 pm

Hmm, this SSD definitely has a USB-C socket and I have two different cables for it, USB-C to USB-C and USB-C to USB-A. I certainly don't see a micro-USB socket inside the SSD's USB-C socket. From what I can see of the dmesg output its's most likely to be an issue with the UAS driver. I had similar issues with a raspberry pi and some usb enclosures. If it wasn't for the fact it works flawlessnessly under Windows, macOS and Linux on my laptops, I'd think it was an hardware issue.

Post by **NeddySeagoon** » Tue Dec 13, 2022 2:30 pm

alexbuell,

USB-C is robust. I was looking at the Samsung provided image on the web.
Its unlikely to be connector issues then.

Try unloading the UAS module. I've seen UAS issues on Raspberry Pi too.
Without UAS you will lose a lot of the go faster goodies.

alexbuell · Post by **alexbuell** » Tue Dec 13, 2022 3:41 pm

I finally ended up creating a file uas.conf in /etc/modprobe.d/

Code: Select all

blacklist uas
options usb-storage quirks=04e8:4001:u

That has done the trick, no more errors, but now it's not as fast any more.

I think I should be asking the kernel folks if they have a fix for this.

Also, I came across some docs that says that this external drive is internally a NVME device.

Post by **NeddySeagoon** » Tue Dec 13, 2022 4:10 pm

alexbuell,

It is advertised as NVMe internally.

Its worth trying the 6.1 kernel before you raise a bug. The kernel people will ask you to anyway.
Maybe start with a gentoo kernel bug.

alexbuell · Post by **alexbuell** » Tue Dec 13, 2022 4:48 pm

I agree. I tested the drive with the quirks fix and without the quirks fix.
it was pitifully slow with quirks fix, about 180MB/s, took out the quirks fix, and it flew at 400MBs.
So, yes, I think it's time to get the kernel folks looking into this one.

I'll try upgrading to 6.1 and see if there is any improvement.