EDAC/amd64: Fix PCI component registration
authorBorislav Petkov <bp@suse.de>
Sun, 22 Nov 2020 14:57:21 +0000 (15:57 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 30 Dec 2020 10:54:11 +0000 (11:54 +0100)
commiteae95da7fcc6fcc01e4325f461576aefe8a670ba
tree3170c1a27e63da1eef7b0163299994d8340da50d
parent3a881be1b4e9f7485de4bc1a2f5968763df8a789
EDAC/amd64: Fix PCI component registration

commit 706657b1febf446a9ba37dc51b89f46604f57ee9 upstream.

In order to setup its PCI component, the driver needs any node private
instance in order to get a reference to the PCI device and hand that
into edac_pci_create_generic_ctl(). For convenience, it uses the 0th
memory controller descriptor under the assumption that if any, the 0th
will be always present.

However, this assumption goes wrong when the 0th node doesn't have
memory and the driver doesn't initialize an instance for it:

  EDAC amd64: F17h detected (node 0).
  ...
  EDAC amd64: Node 0: No DIMMs detected.

But looking up node instances is not really needed - all one needs is
the pointer to the proper device which gets discovered during instance
init.

So stash that pointer into a variable and use it when setting up the
EDAC PCI component.

Clear that variable when the driver needs to unwind due to some
instances failing init to avoid any registration imbalance.

Cc: <stable@vger.kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/20201122150815.13808-1-bp@alien8.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/edac/amd64_edac.c