linux-stable.git/drivers/firmware/google/coreboot_table.c, branch linux-5.4.y

firmware: coreboot: Register bus in module init

2022-12-08T10:23:00+00:00

[ Upstream commit 65946690ed8d972fdb91a74ee75ac0f0f0d68321 ]

The coreboot_table driver registers a coreboot bus while probing a
"coreboot_table" device representing the coreboot table memory region.
Probing this device (i.e., registering the bus) is a dependency for the
module_init() functions of any driver for this bus (e.g.,
memconsole-coreboot.c / memconsole_driver_init()).

With synchronous probe, this dependency works OK, as the link order in
the Makefile ensures coreboot_table_driver_init() (and thus,
coreboot_table_probe()) completes before a coreboot device driver tries
to add itself to the bus.

With asynchronous probe, however, coreboot_table_probe() may race with
memconsole_driver_init(), and so we're liable to hit one of these two:

1. coreboot_driver_register() eventually hits "[...] the bus was not
   initialized.", and the memconsole driver fails to register; or
2. coreboot_driver_register() gets past #1, but still races with
   bus_register() and hits some other undefined/crashing behavior (e.g.,
   in driver_find() [1])

We can resolve this by registering the bus in our initcall, and only
deferring "device" work (scanning the coreboot memory region and
creating sub-devices) to probe().

[1] Example failure, using 'driver_async_probe=*' kernel command line:

[    0.114217] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000010
...
[    0.114307] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 6.1.0-rc1 #63
[    0.114316] Hardware name: Google Scarlet (DT)
...
[    0.114488] Call trace:
[    0.114494]  _raw_spin_lock+0x34/0x60
[    0.114502]  kset_find_obj+0x28/0x84
[    0.114511]  driver_find+0x30/0x50
[    0.114520]  driver_register+0x64/0x10c
[    0.114528]  coreboot_driver_register+0x30/0x3c
[    0.114540]  memconsole_driver_init+0x24/0x30
[    0.114550]  do_one_initcall+0x154/0x2e0
[    0.114560]  do_initcall_level+0x134/0x160
[    0.114571]  do_initcalls+0x60/0xa0
[    0.114579]  do_basic_setup+0x28/0x34
[    0.114588]  kernel_init_freeable+0xf8/0x150
[    0.114596]  kernel_init+0x2c/0x12c
[    0.114607]  ret_from_fork+0x10/0x20
[    0.114624] Code: 5280002b 1100054a b900092a f9800011 (885ffc01)
[    0.114631] ---[ end trace 0000000000000000 ]---

Fixes: b81e3140e412 ("firmware: coreboot: Make bus registration symmetric")
Cc: 
Signed-off-by: Brian Norris 
Reviewed-by: Guenter Roeck 
Reviewed-by: Stephen Boyd 
Link: https://lore.kernel.org/r/20221019180934.1.If29e167d8a4771b0bf4a39c89c6946ed764817b9@changeid
Signed-off-by: Greg Kroah-Hartman 
Signed-off-by: Sasha Levin

firmware: google: Release devices before unregistering the bus

2022-12-08T10:23:00+00:00

[ Upstream commit cae0970ee9c4527f189aac378c50e2f0ed020418 ]

Fix a bug where the kernel module can't be loaded after it has been
unloaded as the devices are still present and conflicting with the
to be created coreboot devices.

Signed-off-by: Patrick Rudolph 
Link: https://lore.kernel.org/r/20191118101934.22526-2-patrick.rudolph@9elements.com
Signed-off-by: Greg Kroah-Hartman 
Stable-dep-of: 65946690ed8d ("firmware: coreboot: Register bus in module init")
Signed-off-by: Sasha Levin

treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 287

2019-06-05T15:36:37+00:00

Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license v2 0 as
  published by the free software foundation this program is
  distributed in the hope that it will be useful but without any
  warranty without even the implied warranty of merchantability or
  fitness for a particular purpose see the gnu general public license
  for more details

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-only

has been chosen to replace the boilerplate/reference in 23 file(s).

Signed-off-by: Thomas Gleixner 
Reviewed-by: Alexios Zavras 
Reviewed-by: Allison Randal 
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190529141901.115786599@linutronix.de
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Only populate devices in coreboot_table_init()

2018-09-14T13:37:23+00:00

This function checks the header for sanity, registers a bus, and
populates devices for each coreboot table entry. Let's just populate
devices here and pull the other bits up into the caller so that this
function can be repurposed for pure device creation and registration.

Cc: Wei-Ning Huang 
Cc: Julius Werner 
Cc: Brian Norris 
Cc: Samuel Holland 
Suggested-by: Julius Werner 
Signed-off-by: Stephen Boyd 
Reviewed-by: Julius Werner 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Remap RAM with memremap() instead of ioremap()

2018-09-14T13:37:23+00:00

This is all system memory, so we shouldn't be mapping this all with
ioremap() as these aren't I/O regions. Instead, they're memory regions
so we should use memremap(). Pick MEMREMAP_WB so we can map memory from
RAM directly if that's possible, otherwise it falls back to
ioremap_cache() like is being done here already. This also nicely
silences the sparse warnings in this code and reduces the need to copy
anything around anymore.

Cc: Wei-Ning Huang 
Cc: Julius Werner 
Cc: Brian Norris 
Cc: Samuel Holland 
Signed-off-by: Stephen Boyd 
Reviewed-by: Julius Werner 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Collapse platform drivers into bus core

2018-09-14T13:37:23+00:00

The DT based and ACPI based platform drivers here do the same thing; map
some memory and hand it over to the coreboot bus to populate devices.
The only major difference is that the DT based driver doesn't map the
coreboot table header to figure out how large of a region to map for the
whole coreboot table and it uses of_iomap() instead of ioremap_cache().
A cached or non-cached mapping shouldn't matter here and mapping some
smaller region first before mapping the whole table is just more work
but should be OK. In the end, we can remove two files and combine the
code all in one place making it easier to reason about things.

We leave the old Kconfigs in place for a little while longer but make
them hidden and select the previously hidden config option. This way
users can upgrade without having to know to reselect this config in the
future. Later on we can remove the old hidden configs.

Cc: Wei-Ning Huang 
Cc: Julius Werner 
Cc: Brian Norris 
Cc: Samuel Holland 
Signed-off-by: Stephen Boyd 
Reviewed-by: Julius Werner 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Make bus registration symmetric

2018-09-14T13:37:23+00:00

The bus is registered in module_init() but is unregistered when the
platform driver remove() function calls coreboot_table_exit(). That
isn't symmetric and it causes the bus to appear on systems that compile
this code in, even when there isn't any coreboot firmware on the device.
Let's move the registration to the coreboot_table_init() function so
that it matches the exit path.

Cc: Wei-Ning Huang 
Cc: Julius Werner 
Cc: Brian Norris 
Cc: Samuel Holland 
Signed-off-by: Stephen Boyd 
Reviewed-by: Julius Werner 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Unmap ioregion after device population

2018-09-14T13:37:23+00:00

Both callers of coreboot_table_init() ioremap the pointer that comes in
but they don't unmap the memory on failure. Both of them also fail probe
immediately with the return value of coreboot_table_init(), leaking a
mapping when it fails. The mapping isn't necessary at all after devices
are populated either, so we can just drop the mapping here when we exit
the function. Let's do that to simplify the code a bit and plug the leak.

Cc: Wei-Ning Huang 
Cc: Julius Werner 
Cc: Brian Norris 
Cc: Samuel Holland 
Fixes: 570d30c2823f ("firmware: coreboot: Expose the coreboot table as a bus")
Signed-off-by: Stephen Boyd 
Reviewed-by: Julius Werner 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Remove unused coreboot_table_find

2018-04-23T11:37:19+00:00

Now that all users of the coreboot_table_find function have been updated
to hang off the coreboot table bus instead, remove it.

Signed-off-by: Samuel Holland 
Signed-off-by: Greg Kroah-Hartman

firmware: coreboot: Expose the coreboot table as a bus

2018-04-23T11:37:19+00:00

This simplifies creating device drivers for hardware or information
described in the coreboot table. It also avoids needing to search
through the table every time a driver is loaded.

Signed-off-by: Samuel Holland 
Signed-off-by: Greg Kroah-Hartman