On Tuesday, 17 May 2022 13:43:55 CEST Сергей Фёдоров wrote:
> On the advice of Diederik de Haas ([email protected] ) on May 16 at
> 22:41 I am sending you my message in
> https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1001286
That bug report and https://bugs.debian.org/1001286 do contain the full
history, but it looks like the most interesting/actionable part of that
conversation wasn't shared with upstream.
I have no specific knowledge about this issue, but will try to summarize it.
/var/log/kern.log
[ 1.341085] i801_smbus 0000:00:1f.3: SMBus using PCI interrupt
[ 1.341467] i2c i2c-0: 8/8 memory slots populated (from DMI)
[ 1.341470] i2c i2c-0: Systems with more than 4 memory slots not supported yet, not instantiating SPD
On Monday, 16 May 2022 20:55:20 CEST Сергей Фёдоров wrote:
> 1. Systems with more than 4 memory slots not supported yet, not
> instantiating SPD. Now I have made it so that the system sees 8 slots, as
> described below, but I do not see SPD.
>
> 2. Kingston memory slots do not appear on the I2C bus, although I see them
> in the BIOS and can change their parameters. Nevertheless, Linux works
> fine.
> I rebuilt the kernel by changing to
> "./linux-source-5.17.6-1/drivers/i2c/i2c-smbus.c"
>
> line 358:
> if (slot_count > 4) {
> dev_warn(&adap->dev,
> "Systems with more than 4 memory slots not
> supported yet, not instantiating SPD\n"); return;
> }
>
> and replace with
>
> if (slot_count > 8) {
> dev_warn(&adap->dev,
> "Systems with more than 8 memory slots not
> supported yet, not instantiating SPD\n"); return;
> }
To which I responded:
The 4 slot limit was specified in 5ace60859e84113b7a185c117fbf2c429d381b59
(upstream commit ID) and the secondary commit message had this:
"Start with just DDR2, DDR3 and DDR4 on x86 for now, and only for systems with
no more than 4 memory slots. These limitations may be lifted later."
That commit by Jean Delvare was from ~2 years ago and it appears that there
may now be a reason to lift/change that limit of 4 (to 8)
As such a change should be discussed (and possibly implemented) upstream,
I asked to forward the request there/here.
Hopefully you now do have the needed information.
Cheers,
Diederik