On October 19, 2026 9:45:58 PM GMT+03:00, bob prohaska <[email protected]>
wrote:
>On Mon, Oct 19, 2026 at 11:31:54AM -0600, Warner Losh wrote:
>>
>> So we're panicking elsewhere. This suggests maybe it's not a storage
>> problem, or a dma/cache problem at all. It suggests maybe a power issue and
>> power drops just enough to cause data corruption. Perhaps you can find a
>> more powerful power supply to test this theory?
>
>A voltmeter across the GPIO +5 and GND pins reports 5.11 volts
>steady-state while a -j3 buildworld is running. The disk is
>connected via its own powered hub, so it shouldn't have much
>effect.
>
>I'll watch voltage during shutdown and reboot to see if anything
>surprising happens.
>
>Thanks for writing!
>
>bob prohaska
>
>
possibly need scope too
this is computer system after all
because what about noise or fast dips
i'm not surprised of hw, that is, sbc, usb-sata adapter, hdd going shit
is there s.m.a.r.t. passthrough in usb bridge? but it might not help either. i
now put my old/new disks through my custom test procedure. maybe this is due
sadly can't run memtest86+ on that platform
the hell, it could be everywhere
i've noticed you run on some tight owner/power budget hw there for long time
for some reason. now it "pays" back
