platform/x86: wmi: Call acpi_wmi_init() later
authorRafael J. Wysocki <rafael.j.wysocki@intel.com>
Wed, 3 Jan 2018 11:49:29 +0000 (12:49 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 17 Jan 2018 08:45:18 +0000 (09:45 +0100)
commit203c1e538eb6abf6298c2a4214c92022a204fdbc
treeef6676e6020a5e7222f21c4234ebc7ceebcc1ba6
parent38244295d9e7663919d606cb9a005f7dee3b7799
platform/x86: wmi: Call acpi_wmi_init() later

commit 98b8e4e5c17bf87c1b18ed929472051dab39878c upstream.

Calling acpi_wmi_init() at the subsys_initcall() level causes ordering
issues to appear on some systems and they are difficult to reproduce,
because there is no guaranteed ordering between subsys_initcall()
calls, so they may occur in different orders on different systems.

In particular, commit 86d9f48534e8 (mm/slab: fix kmemcg cache
creation delayed issue) exposed one of these issues where genl_init()
and acpi_wmi_init() are both called at the same initcall level, but
the former must run before the latter so as to avoid a NULL pointer
dereference.

For this reason, move the acpi_wmi_init() invocation to the
initcall_sync level which should still be early enough for things
to work correctly in the WMI land.

Link: https://marc.info/?t=151274596700002&r=1&w=2
Reported-by: Jonathan McDowell <noodles@earth.li>
Reported-by: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Tested-by: Jonathan McDowell <noodles@earth.li>
Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com>
Signed-off-by: Darren Hart (VMware) <dvhart@infradead.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/platform/x86/wmi.c