Re: [PATCH 1/3] x86/MCE, EDAC/mce_amd: Add support for new MCA_SYND{1,2} registers

[Yazen Ghannam]

On Mon, Jul 18, 2022 at 03:50:46PM +0200, Borislav Petkov wrote:
> On Mon, Jul 18, 2022 at 10:57:19AM +0200, Borislav Petkov wrote:
> > Lemme talk to rostedt.
> 
> Right, he says __dynamic_array(). Grep the tree for examples.
> 
> I'm thinking we can add a field called vendor_info or so, at the end of
> the TP and then dump whatever fields we want there.
> 
> We can even slap a flag in front of the whole thing saying what the
> vendor info type is. But we don't have to get ahead of ourselves and
> keep it simple first.
> 
> How does that sound?
>

Sounds okay to me. But how should this look in the internal kernel data
structures?

struct mce {
    ...
    void *vendor_info; /* Points to a vendor-defined struct. */
};

..or..

struct mce_ext {
    struct mce *m;
    void *vendor_info;
};

I ask because struct mce is UAPI. But I think this is just for /dev/mcelog,
and this has been deprecated for a while. So on a related note, should
/dev/mcelog be removed and struct mce moved out of UAPI? Then changes to
struct mce won't affect user space, and we can just consider the mce trace
event when reporting to user space.

Thanks,
Yazen