Re: i915 Unstable <- no it a hardware fault
I have seen the dimm pins bend and cause really weird stuff. You
might look at the dimm slot pins closely, it may only be one slot, the
ones I saw had the dimm pins end up bent across the bottom of the slot
so they were visually obvious if you looked at the slot.
The machines i saw it on was bunch of machines/motherboards made by
one company in a short window so that cause must have been that the
dimm slot metal was bad or miss-installed.
On Sat, Nov 12, 2022 at 11:06 AM Ken Smith <kens(a)kensnet.org> wrote:
>
>
>
> Roger Heflin wrote:
>
> The dmesg in the dump where it starts having issues is often enough to have an idea.
>
> On Fri, Oct 28, 2022, 9:09 AM Ken Smith via users
<users(a)lists.fedoraproject.org> wrote:
>>
>>
>>
>> Felix Miata wrote:
>> > Ken Smith via users composed on 2022-10-26 13:05 (UTC+0100):
>> >
>> >> This machine has a KabbyLake processor (Intel Core i5-7200U 2.50GHz),
8G
>> >> memory, an Intel i915 (HD Graphics 620) GPU and SSD storage. I've
read
>> >> on various forums that this combination of CPU/GPU can be problematic
>> >> and some mentions that an upcoming Kernel release may contain a remedy
>> >> for this.
>
>
> {Snip}
>
> To wrap this thread up, its a hardware problem. memtest fails in SMP mode between 4
& 6G. If I swap in a 4G memory, memtest ran fine for a couple of days. Beyond 4G it
fails, memtest itself crashes. Suspect an address line is flaky, perhaps 4-6G is showing
up lower in the address space and memtest overwrites itself. (don't know enough about
the architecture of DDR4 memory)
>
> Thanks
>
> Ken
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.