Oh interesting.
Also adding to what @Mario_Limonciello said, here’s a guide for WD drives on the Arch Wiki.
1 Like
I update the firmware of my lexar ssd and ended up not being able to use it on most linux distros as they wpuld ship older kernels that did not yet support that new firmware…
To update WD SSD firmware from linux, check this thread: Western Digital Drive Update Guide Without Windows/WD Dashboard - #48 by nos1609
I followed the manual directions (roughly) to update my SN770 1TB:
$ sudo smartctl -a /dev/nvme0n1
...
=== START OF INFORMATION SECTION ===
Model Number: WD_BLACK SN770 1TB
Firmware Version: 731120WD
...
Yes it is the 4tb nm790 it showed up in fwupd, i was using debian back then and went on with the update. Afterwards i could not boot into my system anymore. Fedora seems to have the necessary patches included already, i am using it since then.
A bit of a late reply to this, I was temporarily “separated” from my laptop.
Taking inspiration from your approach I ran a bunch of tests on the fan behaviour and I believe it’s unlikely to be due to any changes in 3.05. I saw a few other threads, even with 3.03, mentioning the fan behaviour, so I thought I’d do a write up in a separate thread to keep things more consolidated.
My experiments appear to match your results. If I get a chance I will revert to 3.03 and have another go, purely out of curiosity, but thought it might be best to keep any discussion in another thread that is independent of the BIOS release unless an actual regression can be identified.
Update:
USB ports are working again! 
I was recommended on the discord to perform a mainboard reset and that did the trick.
I will update again if the issue reappears.
1 Like
I had done some performance testing with the my Ryzen 7040 board in a Cooler Master case with rather disappoint results. Max all core clocks reaching max of 2100Mhz. Can anyone report their results with the new BIOS in performance mode?
Just wanted to throw into this thread that I updated to 3.05 BIOS two days after the blog post went up via fwupdmgr and haven’t seen any regressions in behavior. After updating the BIOS, I did have to power the system off completely and then boot for it to pick up my drive correctly after I had re-enabled secure boot, but otherwise the update process itself worked about as flawlessly as expected for my setup and I’ve had no issues. I’ve had several kernel/driver updates since, and no stability issues, currently on Linux 6.8.8
Weird. I just doublechecked, and I definitely didn’t.
Thanks for the input!
I got this as well. I’m on EndeavorOS running KDE.
It looks like, from my limited messing around, that / is being mounted as readonly at some point, which causes the OS to no longer be able to open file handles, and since linux uses them for everything the whole OS collapses on itself in really weird ways. No longer able to open file handles, or create child processes, etc.
You get errors like those talked abut here: [Solved] input/output error when running commands / Kernel & Hardware / Arch Linux Forums, which is why I suspect it’s / being remounted as RO, though I couldn’t work out a command that would be helpful in debugging that.
If someone can walk me through what it would be helpful to run, I don’t mind going back to 3.05 and chipping through debugging it.
I’m experiencing a long delay when resuming from suspend, which did not happen on BIOS 3.03. The display will remain black for ~30 seconds before coming on, even though the power and keyboard lights transition into run state immediately.
I see this in the kernel logs after resume:
nvme 0000:02:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x000e address=0x58e5e000 flags=0x0000]
Also sometimes I see a CPU stall error in my logs too.
rcu: INFO: rcu_preempt self-detected stall on CPU
rcu: 0-...!: (2 ticks this GP) idle=4e8c/1/0x4000000000000000 softirq=43128/43129 fqs=0
rcu: (t=9172 jiffies g=92725 q=28 ncpus=12)
rcu: rcu_preempt kthread timer wakeup didn't happen for 9171 jiffies! g92725 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x200
rcu: Possible timer handling issue on cpu=7 timer-softirq=3767
rcu: rcu_preempt kthread starved for 9172 jiffies! g92725 f0x0 RCU_GP_WAIT_FQS(5) ->state=0x200 ->cpu=6
rcu: Unless rcu_preempt kthread gets sufficient CPU time, OOM is now expected behavior.
rcu: RCU grace-period kthread stack dump:
task:rcu_preempt state:R stack:0 pid:17 tgid:17 ppid:2 flags:0x00004000
Reproducible on Fedora Workstation 40 live ISO image with Linux 6.8.5, and also my Debian installation with Kernel 6.8.9. My SSD is the WD Blue SN580 1TB with the latest firmware (281040WD).
Thank you @Steve-Tech! Took a look and will likely need to wait until the 6.10 kernel rolls out.
At least fan data is a want rather than a need.
1 Like
Enable Z state support.
This update completely fixed the biggest issue I had with this laptop: standby power consumption. It went from discharging by 40% in one night to a reasonable 7%, comparable to S3 sleep. I think the power-saver PPD mode improved too.
2 Likes
Been following this thread for a bit since it came out. Got the AMD 13 and it came shipped with 3.03 i believe, upgraded to 3.03b then to the latest 3.05 once it went stable.
*original minor version could be wrong but 3.03b is correct.
fresh install of 22.04 with the suggestions outlined by framework. Seemed to work fine until I moved to 3.05 using fwupd. Noticing the following regressions, so far but as firmware updates are close together and me buying/setting up the laptop its been difficult to rule out any consquences of changing the system layout.
-System fails to suspend/resumes from suspend if plugged into charger with lid closed [fixed]
-If suspended/hibernated and resumed the charging power usage appeared to bounce up and down from 50-90w from a usb-c charger (readout from external batt pack) with it appearing to cut out and renegotiate possibly after trying to draw too much current. Reboot appears to fix this.
-Wifi Adaptor will occasionally fail to work correctly when resuming from suspend
-suspend generally seems intermittent and sometimes failing to suspend correctly on closing lid (sound notifications are the giveaway)
-Enabled hibernation recently to deal with powerdraw of suspend, device will inconsistently hibernate.
Some of these issues appeared to be known but not as a direct result of this firmware. What are the options for rolling back firmware/bios as the frame.work support pages appear to not list previous versions. Guessing grab it from the community forums and flash using usb?
Have any of the above been confirmed to affect this firmware version/ will be fixed in a later release?
Linux framework 6.5.0-1018-oem 22.04.4
Thanks both, suspend issue seems fixed, i’ll update the others if it has solved those too. Very pleased to get this suspend working again.
I have the same problem with freezes/crashes as described by @Ablomme, @Matias and @J3d, so I downgraded to 3.03 and the crashes were gone.
I am on Fedora 40 but experienced the crashes with Fedora 39 as well. Does this bug only affect Linux users? Are there any solutions for it yet?
Cheers
1 Like
I stopped having problems even with 3.05 after updating my SSD’s firmware (following this comment.)
2 Likes