We’ll see. I set up the live USB last night and will retry today with all ports, as requested. I learned my lesson about blindly trusting various program options, though…
I’ll reply back with details once I have them
Which program did you have problems with “blindly trusting various program options” ?
As I said, I used badblocks and it was a destructive test since I didn’t look into the -w option. Either way, I set up the live USB again and will attempt to run some tests this evening
Well, here are the results of dd if=/dev/sda of=/dev/null bs=4k status=progress
First, let’s start with the known-good USB (4 GB, USB 2.0, Ubuntu Sway Remix)
Port 6 (port with suspected issue):
4017344512 bytes (4.0 GB, 3.7 GiB) copied, 120 s, 33.5 MB/s
983040+0 records in
983040+0 records out
4026531840 bytes (4.0 GB, 3.8 GiB) copied, 120.525 s, 33.4 MB/s
Port 3 (no issues expected on either drive):
4019834880 bytes (4.0 GB, 3.7 GiB) copied, 120 s, 33.5 MB/s
983040+0 records in
983040+0 records out
4026531840 bytes (4.0 GB, 3.8 GiB) copied, 120.451 s, 33.4 MB/s
Now for the other USB drive (16 GB, USB 3.0, Ubuntu 22.04 LTS)
Port 6:
5217570816 bytes (5.2 GB, 4.9 GiB) copied, 73 s, 71.5 MB/s
dd: error reading ‘/dev/sda’: Input/output error
1279516+0 records in
1279516+0 records out
5240897536 bytes (5.2 GB, 4.9 GiB) copied, 85.2236 s, 61.5 MB/s
After running this test, no USB drives were detected in either USB-A port until I rebooted. The other ports (in numerical order, not chronological) are below
Port 1:
15447478272 bytes (15 GB, 14 GiB) copied, 145 s, 107 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 146.523 s, 106 MB/s
Port 2:
15496892416 bytes (15 GB, 14 GiB) copied, 397 s, 39.0 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 398.323 s, 38.9 MB/s
Port 3:
15485358080 bytes (15 GB, 14 GiB) copied, 397 s, 39.0 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 398.613 s, 38.9 MB/s
Port 4:
15429390336 bytes (15 GB, 14 GiB) copied, 145 s, 106 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 146.711 s, 106 MB/s
Port 5:
15481819136 bytes (15 GB, 14 GiB) copied, 397 s, 39.0 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 398.743 s, 38.9 MB/s
I also tried swapping my USB-A expansion cards, to interesting effect:
15499644928 bytes (15 GB, 14 GiB) copied, 233 s, 66.5 MB/s
3785376+0 records in
3785376+0 records out
15504900096 bytes (16 GB, 14 GiB) copied, 234.181 s, 66.2 MB/s
After plugging in my USB drive, my headphones (USB-C expansion card → Google USB-C to AUX dongle) stopped working until I unplugged them and plugged them back in.
There’s a lot going in with the read speeds. Ports 1 & 4 seem to run at max speed (~100 MB/s), with 2 & 5 running much slower, 39 MB/s. 5 & 6 do some strange stuff. In my initial configuration, port 6 tried to run at 71.5 MB/s and failed. In the last test, with the swapped USB-A expansion cards, port 6 ran slower, at 66.5 MP/s (still faster than 2 & 5…), but also succeeded.
I don’t really know what to make of this, but I also tried booting from my USB. With the results above, I didn’t test every port exhaustively, (I expect 1, 2, 4, and 5 to all work similarly), and in the original configuration, I was able to replicate my issue:
- The 16 GB, USB 3.0, Ubuntu 22.04 drive would not boot on port 6
- Sector 5B86C0 failed both times I attempted it – Notably, this is not the same sector as had previously shown issues, see original post
- The same drive would boot on port 3
- The same drive would boot on port 5
- The 4 GB, USB 2.0 Ubuntu Sway remix drive would boot on 3 and 6
Also unexpectedly, when I disconnected all of my other expansion cards (from 1, 2, 3, 4, and 5), I was able to boot from port 6 when I also swapped by USB-A expansion card, but when I returned to the original configuration, it failed to boot.
I think there’s something more to which devices are generally connected since I also occasionally get some acpi kernel messages about “partner alt mode failed to register” for USBC000, iirc.
I will try to put together a better test table, but I am thinking that the issue is a combination of factors, including which other USB devices/expansion cards are connected and which specific expansion card is installed, so I will need to organize my next set of tests better.
At the very least, the fact that I can boot the drive from every other port in all tested circumstances seems to indicate to me that there’s nothing fundamentally wrong with the drive itself. Suggestions for further tests are welcome, but I currently plan to do a parametric test on ports 3 and 6 with the expansion cards in original and swapped positions, and with/without other expansion cards, and with/without other devices connected to the expansion cards. I don’t think testing the ability to boot on ports 1, 2, 4, and 5 has much further value since they seem to boot fine
1 Like
At this point the best step is to open a support ticket so we can take a closer look.
1 Like
Following up here with the reply from support: It seems there is likely a mainboard issue, but I opted not to replace it since I can work around the issue (now that I know it exists) and I didn’t want to deal with the additional hassle.