nvme drive disappears after about an hour of uptime
#21
(01-17-2021, 02:49 PM)HitsuMaruku Wrote: I meant to answer this before, and had forgotten. Mine is the second revision, I believe, whichever was the first replacement for the original which needed an adjustment to fit (i.e. mine worked out of the box).

Meh, now I cannot recall whether the preproduction revision or the first "official" revision of the NVMe adapter had to be modified to fit the PineBook Pro case. Smile

(01-17-2021, 02:49 PM)HitsuMaruku Wrote: Nice finds! I took a glance at the documentation. What is it that makes you say they shouldn't be lifetime counts? I see nothing claiming for or against, but I do see: "A value of 0h, indicates that this transition has never occurred or this field is not implemented." Particularly, "never occurred" as opposed to something like "has not occurred since uptime" would infer that it's more likely a lifetime count (I'm just basing this on the wording; I'm mostly unfamiliar with this technology).

Those shouldn't be lifetime counts based on this discussion, although the NVMe specification has no mention of the lack of persistence.  That may also be a highly drive-specific violation of the NVMe specification.  Could you, please, try powering your PineBook Pro off and back on, to check if those values will reset to zero?

(01-17-2021, 02:49 PM)HitsuMaruku Wrote: It's also unclear what exactly those thermal thresholds are. I'm not seeing anything in my smart report to indicate what temperatures, so while PS 2 is the only power state available to it currently, I'm guessing it's possible the threshold is such that it's running the function which adjusts those values without actually switching power states.

Interestingly, pages 221-222 of the NVMe specification state that those two thermal throttling thresholds (temp1 and temp2) may be configurable, of course within the respective drive-provided temperature ranges.  The effects of reaching those thresholds, however configured, may be transitions to lower power states, or some other form of vendor-specific controller throttling, as stated on page 322 of the NVMe specification.
  Reply
#22
Can confirm the temp stats DO persist between boots. After a reboot, Smart still reported 4/79 for those two, respectively. So at the very least for the Intel 660p 2TB, these are persistent values.

Finally just finished the full restic restore and verify, and no NVMe disconnects. Regarding power, I was able to just use my barrel power cord for the verify portion without much issue, so seems that writes use significantly more power than reads. Given the new stability after disabling APST, I might just keep the drive instead of returning it.

If I find the time this week, I'll try seeing about individually 1) disabling just the lowest PS (a la your link regarding Samsung vs Toshiba), and 2) restoring the Max Link Speed for PCIe Gen2 speeds. If both works without issue, we can probably call it just an issue with PS 4. If PS 3 still presents an issue, then it's probably APST (either for just this drive, or Linux in general). My suspicion is that the Gen2 PCIe stuff either is a red herring or only exacerbates power state issues, rather than being a core cause of the drive disappearance.
  Reply
#23
Thank you for checking.  Did you actually turn the laptop off and on while testing the behavior of those two SMART parameters?  A reboot alone, without powering off and on, might not be enough for those two parameters to be reset by the NVMe drive.

Speaking about the two available power inputs, the barrel port and the Type-C PD port, PineBook Pro can actually draw a slightly higher current from the barrel port, compared with the Type-C PD port.  That's the way the PineBook Pro's internal charging circuitry is actually configured.  With that in mind, using the barrel port is actually less likely to cause any power-related issues.

It would be great if you could manage to perform the additional testing, as you've described it.  Previous reports on the issues caused by the instability of PCIe Gen2 speed had different error messages than yours, but everything can be checked only through the additional testing.
  Reply
#24
(01-17-2021, 10:01 PM)dsimic Wrote: Thank you for checking.  Did you actually turn the laptop off and on while testing the behavior of those two SMART parameters?  A reboot alone, without powering off and on, might not be enough for those two parameters to be reset by the NVMe drive.


Yep, even fully powering off, this drive retains those values.


(01-17-2021, 10:01 PM)dsimic Wrote: Speaking about the two available power inputs, the barrel port and the Type-C PD port, PineBook Pro can actually draw a slightly higher current from the barrel port, compared with the Type-C PD port.  That's the way the PineBook Pro's internal charging circuitry is actually configured.  With that in mind, using the barrel port is actually less likely to cause any power-related issues.

Good to know! The main reason I'm using USB-C over barrel power is that I don't have a USB-A 15W adapter; the 15W adapter I have is a USB-C to USB-C, and the outlet provides only 2.4A max on the USB-A port, so right now USB-C is going to give me more power. Good to know I should invest in a new adapter, though. Do you know what the specification for the barrel port is? I was under the assumption that both were rated for 5V 3A max.

I did have some issues today while plugged into USB-C, similar to what is described in this thread, where my screen appeared unresponsive, and subsequent attempts to boot resulted in the same power LED symptoms as the video. After leaving it to cool for an hour, same symptoms. I then charged it for an hour or two, and all was well. I'm guessing I either overheated it and the battery drained too quickly; was fine an hour beforehand, but had it in a less cooling-friendly spot.


(01-17-2021, 10:01 PM)dsimic Wrote: It would be great if you could manage to perform the additional testing, as you've described it.  Previous reports on the issues caused by the instability of PCIe Gen2 speed had different error messages than yours, but everything can be checked only through the additional testing.


Until the above happened, I again ran a restic restore overnight, this time with 5500 max exlat set. The result was still no disappearing NVMe, so I think for this drive, it's just APST with the PS 4 that's the issue; PS 3 seems to work.

Going to try leaving my PBP on overnight with Gen 1 and no I/O this time, to see if the drive disappears if it goes down to idle PS 3 without heavy load. If that works fine, I'll give Gen 2 speeds a shot.
  Reply
#25
(01-18-2021, 08:11 PM)HitsuMaruku Wrote: Until the above happened, I again ran a restic restore overnight, this time with 5500 max exlat set. The result was still no disappearing NVMe, so I think for this drive, it's just APST with the PS 4 that's the issue; PS 3 seems to work.

Going to try leaving my PBP on overnight with Gen 1 and no I/O this time, to see if the drive disappears if it goes down to idle PS 3 without heavy load. If that works fine, I'll give Gen 2 speeds a shot.

No issues to report.

Left PBP overnight idling with 5500 max exlat (presumably leaving PS 3 and APST enabled and PS 4 disabled); didn't see any output in feature "0x0c -H" about new state transitions (previously shown 3 entires with 500ms each, unsure when those occurred, or whether they're static or being populated on startup each time, but happens with APST disabled, so make of it what you will).

After, I replaced the device tree back to Gen 2 speeds and did another restore. Everything copied fine, seemed to actually take less power (maybe because it wasn't running at max bandwidth for that gen? idk, I'm speculating).

Thus, for my findings, I'm considering PS 4 the issue. That may be due to this model in particular incompatible with that low power draw for the PBP to sustain, or it may be an issue in compatibility with PCIe power states which work in tandem with NVMe power states. Again, make of it what you will, but for me for this model, disabling PS 4 fixes the disappearing issues.
  Reply
#26
My apologies for the delayed response.

It's good to hear that your NVMe drive now works as expected.  Thank your for confirming that the volatile nature of drive stats is not the general rule.  By the way, based on user reports here on the forum, the need to restrict the PCIe interface to Gen1 speed seems not to be affecting all RK3399 chips, so it's even better that yours seems not to be affected.

Regarding your question about the barrel port charger, I'd suggest that you buy the official PineBook Pro charger from the Pine64 store.  It's inexepensive and has solid build quality, at least based on the way its exterior looks and feels.  Trying to source an "aftermarket" charger with the right barrel plug and right plug polarity would take a lot of time and, at best, would result in a charger that requires some sort of an ugly, bulky barrel plug adapter.  Trust me, I've tried. Smile

The lockup issues you've described seem pretty bad.  The only thing that comes to my mind is some sort of power or overheating issue, although I cannot see what could be the actual culprit.  It might also be some manufacturing defect.  You might want to create an official ticket, if the issue persists.
  Reply


Possibly Related Threads…
Thread Author Replies Views Last Post
  NVMe-related crashes and instability, plus a solution simonsouth 9 4,268 01-24-2021, 05:37 PM
Last Post: HitsuMaruku
  NVME SPI Update not booting SD Card WZ9V 5 2,563 10-18-2020, 08:36 PM
Last Post: wdt
  Possible to upgrade PBP with NVMe M.2 1TB SSD from Sabrent? kkdao 15 6,584 08-22-2020, 05:20 AM
Last Post: kkdao
  LTE adapter via NVMe grego 2 1,732 07-08-2020, 01:37 PM
Last Post: manawyrm
  Pro NVMe adapter Does not fit ! ? bcnaz 110 45,995 07-03-2020, 10:51 PM
Last Post: xmixahlx
  NVME problem chaoskampf 2 1,684 07-01-2020, 09:10 AM
Last Post: chaoskampf
  PBP with NVMe installed - Issues booting to SD eluno1 2 1,576 06-05-2020, 09:33 AM
Last Post: eluno1
Lightbulb Hacking the Ill-Fitting NVMe Adapter diodelass 0 994 06-03-2020, 03:15 PM
Last Post: diodelass
  NVMe SSD Interface Adapter linuxfreund 1 1,303 06-02-2020, 06:31 AM
Last Post: tophneal
Question New Design NVMe Adaptor? Feakster 6 3,146 05-16-2020, 03:21 AM
Last Post: Feakster

Forum Jump:


Users browsing this thread: 1 Guest(s)