On Thursday, August 11, 2022 3:17 AM, Jia He wrote:
> Commit dc4e8c07e9e2 ("ACPI: APEI: explicit init of HEST and GHES in
> apci_init()") introduced a bug that ghes_edac_register() would be invoked
> before edac_init(). Because at that time, the bus "edac" hasn't been even
> registered, this created sysfs /devices/mc0 instead of
> /sys/devices/system/edac/mc/mc0 on an Ampere eMag server.
>
> The solution is modularizing the ghes_edac driver.
> Use a list to save the probing devices in ghes_probe(), and defer the
> ghes_edac_register() to the new ghes_edac module_init() by iterating over
> the devices list.
Since this approach defers the ghes_edac registration, the series needs
to include a way to prevent chipset-specific edac drivers from registering
before ghes_edac as discussed in the original email thread.
Toshi
