M.2 ssd is not recognized

When I connect my Integral 120 GB ssd both directly or via M.2 extension I get the following in dmesg:
rockchip-pcie: probe of f8000000.pcie failed with error -110.

Does it mean faulty board or faulty drive? Currently I don’t have a way to connect SSD drive to any other device to see if it is faulty.

I got M.2 reader. So SSD is fine and not broken. I saw (but lost a link) a post where someone recommended to check voltage with V-meter. Can someone guide me where to check it?

I measured voltage as advised in this post by @jack.
Both pads give 3.33v. So voltage is not a problem I guess. What else can I do to understand root cause of a problem?

Which model is the SSD? Recently we got report from other user that some KingSpec SSD can not be detected even in kernel. But from your log:

rockchip-pcie: probe of f8000000.pcie failed with error -110.

This might be just a connection issue instead of a link/training issue. I suggest you direct plug the SSD to the ROCK Pi 4 M2 slot and test.

1 Like

Sorry, what do you mean by mode? Also I see in FAQ that

- Known working under Linux kernel but not directly in bootloader(uSD/eMMC + NVMe booting) -
Integral 120GB M.2 2280 PCIE NVME INSSD120GM280N

Does this quote say it is not possible to boot from this SSD?

This one

  • Integral 120GB M.2 2280 PCIE NVME INSSD120GM280N

is added by users, not tested by us. Usually we only recommend certain brand because some brands sometimes change the controller without changing the model nor notice.

As i said, from your log, it’s more like a connection issue. Did you try direct connect the SSD to ROCK Pi 4? Also, we don’t support hot plug.

I connected SSD directly and I get the same -110 error. Would any additional logs help? It would be helpful for me to understand if connection is faulty or SSD is not supported so I can buy an alternative one.

Yes, please paste the log here.

dmesg | grep pcie

That is what I get when connected directly

[    1.518776] of_get_named_gpiod_flags: parsed 'gpio' property of node '/vcc3v3-pcie-regulator[0]' - status (0)
[    1.518843] reg-fixed-voltage vcc3v3-pcie-regulator: Looking up vin-supply from device tree
[    1.518879] vcc3v3_pcie: supplied by vcc3v3_sys
[    1.519315] vcc3v3_pcie: at 3300 mV 
[    1.519480] reg-fixed-voltage vcc3v3-pcie-regulator: vcc3v3_pcie supplying 0uV
[    1.652183] phy phy-pcie-phy.9: Looking up phy-supply from device tree
[    1.652193] phy phy-pcie-phy.9: Looking up phy-supply property in node /pcie-phy failed
[    1.653634] rockchip-pcie f8000000.pcie: GPIO lookup for consumer ep
[    1.653643] rockchip-pcie f8000000.pcie: using device tree for GPIO lookup
[    1.653671] of_get_named_gpiod_flags: parsed 'ep-gpios' property of node '/pcie@f8000000[0]' - status (0)
[    1.653906] rockchip-pcie f8000000.pcie: Looking up vpcie3v3-supply from device tree
[    1.653917] rockchip-pcie f8000000.pcie: Looking up vpcie3v3-supply property in node /pcie@f8000000 failed
[    1.653929] rockchip-pcie f8000000.pcie: no vpcie3v3 regulator found
[    1.654529] rockchip-pcie f8000000.pcie: Looking up vpcie1v8-supply from device tree
[    1.654540] rockchip-pcie f8000000.pcie: Looking up vpcie1v8-supply property in node /pcie@f8000000 failed
[    1.654551] rockchip-pcie f8000000.pcie: no vpcie1v8 regulator found
[    1.655112] rockchip-pcie f8000000.pcie: Looking up vpcie0v9-supply from device tree
[    1.655121] rockchip-pcie f8000000.pcie: Looking up vpcie0v9-supply property in node /pcie@f8000000 failed
[    1.655132] rockchip-pcie f8000000.pcie: no vpcie0v9 regulator found
[    2.178219] rockchip-pcie f8000000.pcie: PCIe link training gen1 timeout!
[    2.178944] rockchip-pcie: probe of f8000000.pcie failed with error -110

Were my logs helpful? Can they say if the problem is most likely in my SDD or in Rock PI hardware?

1 Like

It might be a hardware connection issue, try to attach the SSD directly on ROCK Pi 4. Check if the m2 adapter board FPC is connected well.

My logs are actually from when I connected SSD directly to a slot (no adapter was used)

Can I change it to a different board from a supplier I purchased it? Is it a guarantee case?

Yes, you can. But I think you should try with another SSD first before you ask for an RMA.

Hey,

I do have exactly the same problem but actually using a ‘tested by team’ HP EX900 M.2 with 120GB. demsg output is exactly the same than in Druid’s case. Tested with Ubuntu, Armbian, Debian both directly connected to the RockPi and with the extension board. Manjaro seems to give a slightly different output only looking for a 12V supply including a driver error exiting also with -110.
Shall I also try with another SSD? What do you recommend for a smaller SPI bootable SSD?

I also found the German FAQ where it is stated that the HP EX900 is exactly the one that doesn’t work. Compare here (German) vs. here (en)

Thanks folks!

I am waiting for Acer AT500M which is marked as good in English FAQ and see how it goes. If it also doesn’t work I will ask for a replacement. I think it is fair. @krang where did you buy your board? I ordered mine from Germany.

Board + extra stuff is from innet. They replied quickly when I had a question about the Germany only weird U.LF adapter /w and w/o… And good prize, though.
SSD and case from Conrad, the only ones with a smaller kind of tested SSD. I really don’t need 250GB for network controller, maybe pfsense, pihole and openvpn. :smiley:

It looks to be similar problem reported here:

@krang and @Druid did you try the Radxa Debian image? The u-boot in that image is reported to work.

Hey aaditya,

I don’t think, it is the same problem here. In my case it is absolutely not recognized. There is no /dev/nvme* entry, lspcie is empty, lsblk lists the sd as mmc* with the different partitions only.
And yes, I used the official Radxa images, both Ubuntu and Debian. Kernel Versions are 4.4, I think, for Armbian, I used 4.4 and 5.4 and for Manjaro 5.4 and 5.6. The latter says the also updated u-boot…

1 Like

Ok. For some users /dev/nvme* entries dont appear, as witnessed here: