Samsung 860/870 SSDs Continue Causing Problems For Linux Users

[phoronix]

Phoronix: Samsung 860/870 SSDs Continue Causing Problems For Linux Users

While Samsung has explicitly stated before that queued TRIM works for Samsung 860 SSDs on Linux and thus leading to only older Samsung 840/850 drives being blocked from queued TRIM usage, that turns out to be inaccurate and now more quirks are added for the Samsung 860 and 870 series SSDs on Linux...

https://www.phoronix.com/scan.php?page=news_item&px=Samsung-860-870-More-Quirks

[obri]

Hmm,

I use an 850 Evo and an 860 Evo and I am not aware of any problems.
How can I find out, if my drives/controller will bring issues to me?

I use a B350 Motherboard, but MSI has no Info, what exact controller is in use.

[mlau]

What does Windows do with these drives, does it also disable these features on these host controllers?

[linner]

obri Your 850 is probably already blacklisted from queued TRIM and your 860 will be soon if not already. You'll probably want to disable active TRIM and just do batched TRIM at set intervals like when your computer is not in use.

Kind of glad I didn't go with Samsung on my last SSD's. Although my HP drives are probably slower and don't support queued TRIM anyway so it's a wash.

Wow, can't say I'm surprised that AMD chipsets are causing yet more problems. I have an AMD laptop that "supports" USB3 but it's total garbage. As I sit here typing on my AMD Radeon driven displays that are flickering giant garbled graphic blocks in several places on the screen. AMD... pffft, yuck

[kpedersen]

It reminds me a little of the Intel "SandForce" SSDs. They would randomly disappear from the host in every OS (apart from Windows). It turns out it was a power saving problem with the firmware and other operating systems were doing it correctly (and thus exposing it). Windows would not sleep the disks (presumably so it can constantly scan and leak data to Microsoft's *actual* customers )

The solution was to bite the bullet and disable the power saving ability entirely in the firmware with a slightly ratty tool (built ontop of Yocto if I recall)

https://community.intel.com/t5/Solid-State-Drives/Intel-SSD-535-180-GB-SATA3-shows-as-quot-SandForce-2000-quot/td-p/565310

https://www.intel.com/content/www/us/en/download/18405/25348/intel-solid-state-drive-firmware-update-tool-for-devsleep-issue.html

Amazing how companies specializing in disk drives can't get this right. Did they not test them for more than 5 mins?

[user1]

According to Wkipedia, Linux is the only OS with queued trim support as of 1 July 2015.

[skeevy420]

What does Windows do ... ?

I feel like that's the computer version of WWJD.

[Melcar]

Glad I stopped bothering with Samsung drives a while ago. And yes, AMD chipsets have always had "issues". SATA and more recently USB.

[bash2bash]

Interesting... while I don't know about TRIM support, I do know that my samsung drives have been the most reliable ones, take a peek:

Samsung SSD 850 PRO, current power-on uptime: 7 years, 2 months and 7 days
used reserved blocks: 0
program fail count: 0
erase fail count: 0
runtime bad blocks: 0
reported uncorrectable errors: 0
hardware ECC recovered: 0


I am slowly upgrading all 850 PRO with new Samsung SSD 980 PRO. Hopefully, the 980 will be as reliable as the 850 is.

[Adarion]

Strange, I got some Samsung Pro SSDs (for the MLC) and I haven't been aware of problems. All of them on AMD chipsets (E-350, Kabini and Ryzen/B-350).
Instead of having to disable all sorts of things via Kernel it would be better if the cause of issues could be fixed by Samsung via FW updates. Also, on their webpage (products) Samsung stated that the devices should be running well with the Penguin. I wonder what's amess here.