In preparation for compat support, we can't assume that user variable
object is represented by a 'struct efi_variable'. Convert the validation
functions to take the variable name as an argument, which is the only
piece of the struct that was ever used anyway.

Cc: Mike Waychison <mikew@google.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Currently, when mounting pstore file system, a read callback of
efi_pstore driver runs mutiple times as below.

- In the first read callback, scan efivar_sysfs_list from head and pass
  a kmsg buffer of a entry to an upper pstore layer.
- In the second read callback, rescan efivar_sysfs_list from the entry
  and pass another kmsg buffer to it.
- Repeat the scan and pass until the end of efivar_sysfs_list.

In this process, an entry is read across the multiple read function
calls. To avoid race between the read and erasion, the whole process
above is protected by a spinlock, holding in open() and releasing in
close().

At the same time, kmemdup() is called to pass the buffer to pstore
filesystem during it. And then, it causes a following lockdep warning.

To make the dynamic memory allocation runnable without taking spinlock,
holding off a deletion of sysfs entry if it happens while scanning it
via efi_pstore, and deleting it after the scan is completed.

To implement it, this patch introduces two flags, scanning and deleting,
to efivar_entry.

On the code basis, it seems that all the scanning and deleting logic is
not needed because __efivars->lock are not dropped when reading from the
EFI variable store.

But, the scanning and deleting logic is still needed because an
efi-pstore and a pstore filesystem works as follows.

In case an entry(A) is found, the pointer is saved to psi->data.  And
efi_pstore_read() passes the entry(A) to a pstore filesystem by
releasing  __efivars->lock.

And then, the pstore filesystem calls efi_pstore_read() again and the
same entry(A), which is saved to psi->data, is used for resuming to scan
a sysfs-list.

So, to protect the entry(A), the logic is needed.

[    1.143710] ------------[ cut here ]------------
[    1.144058] WARNING: CPU: 1 PID: 1 at kernel/lockdep.c:2740 lockdep_trace_alloc+0x104/0x110()
[    1.144058] DEBUG_LOCKS_WARN_ON(irqs_disabled_flags(flags))
[    1.144058] Modules linked in:
[    1.144058] CPU: 1 PID: 1 Comm: systemd Not tainted 3.11.0-rc5 #2
[    1.144058]  0000000000000009 ffff8800797e9ae0 ffffffff816614a5 ffff8800797e9b28
[    1.144058]  ffff8800797e9b18 ffffffff8105510d 0000000000000080 0000000000000046
[    1.144058]  00000000000000d0 00000000000003af ffffffff81ccd0c0 ffff8800797e9b78
[    1.144058] Call Trace:
[    1.144058]  [<ffffffff816614a5>] dump_stack+0x54/0x74
[    1.144058]  [<ffffffff8105510d>] warn_slowpath_common+0x7d/0xa0
[    1.144058]  [<ffffffff8105517c>] warn_slowpath_fmt+0x4c/0x50
[    1.144058]  [<ffffffff8131290f>] ? vsscanf+0x57f/0x7b0
[    1.144058]  [<ffffffff810bbd74>] lockdep_trace_alloc+0x104/0x110
[    1.144058]  [<ffffffff81192da0>] __kmalloc_track_caller+0x50/0x280
[    1.144058]  [<ffffffff815147bb>] ? efi_pstore_read_func.part.1+0x12b/0x170
[    1.144058]  [<ffffffff8115b260>] kmemdup+0x20/0x50
[    1.144058]  [<ffffffff815147bb>] efi_pstore_read_func.part.1+0x12b/0x170
[    1.144058]  [<ffffffff81514800>] ? efi_pstore_read_func.part.1+0x170/0x170
[    1.144058]  [<ffffffff815148b4>] efi_pstore_read_func+0xb4/0xe0
[    1.144058]  [<ffffffff81512b7b>] __efivar_entry_iter+0xfb/0x120
[    1.144058]  [<ffffffff8151428f>] efi_pstore_read+0x3f/0x50
[    1.144058]  [<ffffffff8128d7ba>] pstore_get_records+0x9a/0x150
[    1.158207]  [<ffffffff812af25c>] ? selinux_d_instantiate+0x1c/0x20
[    1.158207]  [<ffffffff8128ce30>] ? parse_options+0x80/0x80
[    1.158207]  [<ffffffff8128ced5>] pstore_fill_super+0xa5/0xc0
[    1.158207]  [<ffffffff811ae7d2>] mount_single+0xa2/0xd0
[    1.158207]  [<ffffffff8128ccf8>] pstore_mount+0x18/0x20
[    1.158207]  [<ffffffff811ae8b9>] mount_fs+0x39/0x1b0
[    1.158207]  [<ffffffff81160550>] ? __alloc_percpu+0x10/0x20
[    1.158207]  [<ffffffff811c9493>] vfs_kern_mount+0x63/0xf0
[    1.158207]  [<ffffffff811cbb0e>] do_mount+0x23e/0xa20
[    1.158207]  [<ffffffff8115b51b>] ? strndup_user+0x4b/0xf0
[    1.158207]  [<ffffffff811cc373>] SyS_mount+0x83/0xc0
[    1.158207]  [<ffffffff81673cc2>] system_call_fastpath+0x16/0x1b
[    1.158207] ---[ end trace 61981bc62de9f6f4 ]---
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Tested-by: NMadper Xie <cxie@redhat.com>
Cc: stable@kernel.org
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

The intent is that if we aren't allowed to block because we're in an
NMI or an emergency then we only take the lock if it is uncontended.

Part of the problem is the test is reversed so we return -EBUSY if we
acquire the lock.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Seiji reported getting empty dmesg-* files, because the data was never
actually read in efi_pstore_read_func(), and so the memcpy() was copying
garbage data.

This patch necessitated adding __efivar_entry_get() which is callable
between efivar_entry_iter_{begin,end}(). We can also delete
__efivar_entry_size() because efi_pstore_read_func() was the only
caller.
Reported-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Tested-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Matthew Garrett <matthew.garrett@nebula.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

This registers /sys/firmware/efi/{,systab,efivars/} whenever EFI is enabled
and the system is booted with EFI.

This allows
 *) userspace to check for the existence of /sys/firmware/efi as a way
    to determine whether or it is running on an EFI system.
 *) 'mount -t efivarfs none /sys/firmware/efi/efivars' without manually
    loading any modules.

[ Also, move the efivar API into vars.c and unconditionally compile it.
  This allows us to move efivars.c, which now only contains the sysfs
  variable code, into the firmware/efi directory. Note that the efivars.c
  filename is kept to maintain backwards compatability with the old
  efivars.ko module. With this patch it is now possible for efivarfs
  to be built without CONFIG_EFI_VARS - Matt ]

Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Mike Waychison <mikew@google.com>
Cc: Kay Sievers <kay@vrfy.org>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Chun-Yi Lee <jlee@suse.com>
Cc: Andy Whitcroft <apw@canonical.com>
Cc: Tobias Powalowski <tpowa@archlinux.org>
Signed-off-by: NTom Gundersen <teg@jklm.no>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Now that efivarfs uses the efivar API, move it out of efivars.c and
into fs/efivarfs where it belongs. This move will eventually allow us
to enable the efivarfs code without having to also enable
CONFIG_EFI_VARS built, and vice versa.

Furthermore, things like,

    mount -t efivarfs none /sys/firmware/efi/efivars

will now work if efivarfs is built as a module without requiring the
use of MODULE_ALIAS(), which would have been necessary when the
efivarfs code was part of efivars.c.

Cc: Matthew Garrett <matthew.garrett@nebula.com>
Cc: Jeremy Kerr <jk@ozlabs.org>
Reviewed-by: NTom Gundersen <teg@jklm.no>
Tested-by: NTom Gundersen <teg@jklm.no>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

efivars.c has grown far too large and needs to be divided up. Create a
new directory and move the persistence storage code to efi-pstore.c now
that it uses the new efivar API. This helps us to greatly reduce the
size of efivars.c and paves the way for moving other code out of
efivars.c.

Note that because CONFIG_EFI_VARS can be built as a module efi-pstore
must also include support for building as a module.
Reviewed-by: NTom Gundersen <teg@jklm.no>
Tested-by: NTom Gundersen <teg@jklm.no>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: Anton Vorontsov <cbouatmailru@gmail.com>
Cc: Colin Cross <ccross@android.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Tony Luck <tony.luck@intel.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

There isn't really a formal interface for dealing with EFI variables
or struct efivar_entry. Historically, this has led to various bits of
code directly accessing the generic EFI variable ops, which inherently
ties it to specific EFI variable operations instead of indirectly
using whatever ops were registered with register_efivars(). This lead
to the efivarfs code only working with the generic EFI variable ops
and not CONFIG_GOOGLE_SMI.

Encapsulate everything that needs to access '__efivars' inside an
efivar_entry_* API and use the new API in the pstore, sysfs and
efivarfs code.

Much of the efivars code had to be rewritten to use this new API. For
instance, it is now up to the users of the API to build the initial
list of EFI variables in their efivar_init() callback function. The
variable list needs to be passed to efivar_init() which allows us to
keep work arounds for things like implementation bugs in
GetNextVariable() in a central location.

Allowing users of the API to use a callback function to build the list
greatly benefits the efivarfs code which needs to allocate inodes and
dentries for every variable.  It previously did this in a racy way
because the code ran without holding the variable spinlock. Both the
sysfs and efivarfs code maintain their own lists which means the two
interfaces can be running simultaneously without interference, though
it should be noted that because no synchronisation is performed it is
very easy to create inconsistencies. efibootmgr doesn't currently use
efivarfs and users are likely to also require the old sysfs interface,
so it makes sense to allow both to be built.
Reviewed-by: NTom Gundersen <teg@jklm.no>
Tested-by: NTom Gundersen <teg@jklm.no>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Mike Waychison <mikew@google.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Some machines have an EFI variable interface that does not conform to
the UEFI specification, e.g. CONFIG_GOOGLE_SMI. Add the necessary code
so that it's only possible to use one implementation of EFI variable
operations at runtime. This allows us to keep a single (file-scope)
global pointer 'struct efivars', which simplifies access. This will
hopefully dissuade developers from accessing the generic operations
struct directly in the future, as was done in the efivarfs and pstore
code, thereby allowing future code to work with both the generic efivar
ops and the google SMI ops.

This may seem like a step backwards in terms of modularity, but we don't
need to track more than one 'struct efivars' at one time. There is no
synchronisation done between multiple EFI variable operations, and
according to Mike no one is using both the generic EFI var ops and
CONFIG_GOOGLE_SMI simultaneously, though a single kernel build _does_
need to able to support both. It also helps to clearly highlight which
functions form the core of the efivars interface - those that require
access to __efivars.
Reviewed-by: NTom Gundersen <teg@jklm.no>
Tested-by: NTom Gundersen <teg@jklm.no>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

There are currently two implementations of the utf16 string functions.
Somewhat confusingly, they've got different names.

Centralise the functions in efi.h.
Reviewed-by: NTom Gundersen <teg@jklm.no>
Tested-by: NTom Gundersen <teg@jklm.no>
Reviewed-by: NMike Waychison <mikew@google.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

We want to be able to use the utf16 functions that are currently present
in the EFI variables code in platform-specific code as well. Move them to
the kernel core, and in the process rename them to accurately describe what
they do - they don't handle UTF16, only UCS2.
Signed-off-by: NMatthew Garrett <matthew.garrett@nebula.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Let's not burden ia64 with checks in the common efivars code that we're not
writing too much data to the variable store. That kind of thing is an x86
firmware bug, plain and simple.

efi_query_variable_store() provides platforms with a wrapper in which they can
perform checks and workarounds for EFI variable storage bugs.

Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

The 'CONFIG_' prefix is not implicit in IS_ENABLED().
Signed-off-by: NBen Hutchings <ben@decadent.org.uk>
Cc: Seth Forshee <seth.forshee@canonical.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Some firmware exhibits a bug where the same VariableName and
VendorGuid values are returned on multiple invocations of
GetNextVariableName(). See,

    https://bugzilla.kernel.org/show_bug.cgi?id=47631

As a consequence of such a bug, Andre reports hitting the following
WARN_ON() in the sysfs code after updating the BIOS on his, "Gigabyte
Technology Co., Ltd. To be filled by O.E.M./Z77X-UD3H, BIOS F19e
11/21/2012)" machine,

[    0.581554] EFI Variables Facility v0.08 2004-May-17
[    0.584914] ------------[ cut here ]------------
[    0.585639] WARNING: at /home/andre/linux/fs/sysfs/dir.c:536 sysfs_add_one+0xd4/0x100()
[    0.586381] Hardware name: To be filled by O.E.M.
[    0.587123] sysfs: cannot create duplicate filename '/firmware/efi/vars/SbAslBufferPtrVar-01f33c25-764d-43ea-aeea-6b5a41f3f3e8'
[    0.588694] Modules linked in:
[    0.589484] Pid: 1, comm: swapper/0 Not tainted 3.8.0+ #7
[    0.590280] Call Trace:
[    0.591066]  [<ffffffff81208954>] ? sysfs_add_one+0xd4/0x100
[    0.591861]  [<ffffffff810587bf>] warn_slowpath_common+0x7f/0xc0
[    0.592650]  [<ffffffff810588bc>] warn_slowpath_fmt+0x4c/0x50
[    0.593429]  [<ffffffff8134dd85>] ? strlcat+0x65/0x80
[    0.594203]  [<ffffffff81208954>] sysfs_add_one+0xd4/0x100
[    0.594979]  [<ffffffff81208b78>] create_dir+0x78/0xd0
[    0.595753]  [<ffffffff81208ec6>] sysfs_create_dir+0x86/0xe0
[    0.596532]  [<ffffffff81347e4c>] kobject_add_internal+0x9c/0x220
[    0.597310]  [<ffffffff81348307>] kobject_init_and_add+0x67/0x90
[    0.598083]  [<ffffffff81584a71>] ? efivar_create_sysfs_entry+0x61/0x1c0
[    0.598859]  [<ffffffff81584b2b>] efivar_create_sysfs_entry+0x11b/0x1c0
[    0.599631]  [<ffffffff8158517e>] register_efivars+0xde/0x420
[    0.600395]  [<ffffffff81d430a7>] ? edd_init+0x2f5/0x2f5
[    0.601150]  [<ffffffff81d4315f>] efivars_init+0xb8/0x104
[    0.601903]  [<ffffffff8100215a>] do_one_initcall+0x12a/0x180
[    0.602659]  [<ffffffff81d05d80>] kernel_init_freeable+0x13e/0x1c6
[    0.603418]  [<ffffffff81d05586>] ? loglevel+0x31/0x31
[    0.604183]  [<ffffffff816a6530>] ? rest_init+0x80/0x80
[    0.604936]  [<ffffffff816a653e>] kernel_init+0xe/0xf0
[    0.605681]  [<ffffffff816ce7ec>] ret_from_fork+0x7c/0xb0
[    0.606414]  [<ffffffff816a6530>] ? rest_init+0x80/0x80
[    0.607143] ---[ end trace 1609741ab737eb29 ]---

There's not much we can do to work around and keep traversing the
variable list once we hit this firmware bug. Our only solution is to
terminate the loop because, as Lingzhu reports, some machines get
stuck when they encounter duplicate names,

  > I had an IBM System x3100 M4 and x3850 X5 on which kernel would
  > get stuck in infinite loop creating duplicate sysfs files because,
  > for some reason, there are several duplicate boot entries in nvram
  > getting GetNextVariableName into a circle of iteration (with
  > period > 2).

Also disable the workqueue, as efivar_update_sysfs_entries() uses
GetNextVariableName() to figure out which variables have been created
since the last iteration. That algorithm isn't going to work if
GetNextVariableName() returns duplicates. Note that we don't disable
EFI variable creation completely on the affected machines, it's just
that any pstore dump-* files won't appear in sysfs until the next
boot.
Reported-by: NAndre Heider <a.heider@gmail.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Tested-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

It's not wise to assume VariableNameSize represents the length of
VariableName, as not all firmware updates VariableNameSize in the same
way (some don't update it at all if EFI_SUCCESS is returned). There
are even implementations out there that update VariableNameSize with
values that are both larger than the string returned in VariableName
and smaller than the buffer passed to GetNextVariableName(), which
resulted in the following bug report from Michael Schroeder,

  > On HP z220 system (firmware version 1.54), some EFI variables are
  > incorrectly named :
  >
  > ls -d /sys/firmware/efi/vars/*8be4d* | grep -v -- -8be returns
  > /sys/firmware/efi/vars/dbxDefault-pport8be4df61-93ca-11d2-aa0d-00e098032b8c
  > /sys/firmware/efi/vars/KEKDefault-pport8be4df61-93ca-11d2-aa0d-00e098032b8c
  > /sys/firmware/efi/vars/SecureBoot-pport8be4df61-93ca-11d2-aa0d-00e098032b8c
  > /sys/firmware/efi/vars/SetupMode-Information8be4df61-93ca-11d2-aa0d-00e098032b8c

The issue here is that because we blindly use VariableNameSize without
verifying its value, we can potentially read garbage values from the
buffer containing VariableName if VariableNameSize is larger than the
length of VariableName.

Since VariableName is a string, we can calculate its size by searching
for the terminating NULL character.
Reported-by: NFrederic Crozat <fcrozat@suse.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: Michael Schroeder <mls@suse.com>
Cc: Lee, Chun-Yi <jlee@suse.com>
Cc: Lingzhu Xiang <lxiang@redhat.com>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

We know that with some firmware implementations writing too much data to
UEFI variables can lead to bricking machines. Recent changes attempt to
address this issue, but for some it may still be prudent to avoid
writing large amounts of data until the solution has been proven on a
wide variety of hardware.

Crash dumps or other data from pstore can potentially be a large data
source. Add a pstore_module parameter to efivars to allow disabling its
use as a backend for pstore. Also add a config option,
CONFIG_EFI_VARS_PSTORE_DEFAULT_DISABLE, to allow setting the default
value of this paramter to true (i.e. disabled by default).
Signed-off-by: NSeth Forshee <seth.forshee@canonical.com>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Add a new option, CONFIG_EFI_VARS_PSTORE, which can be set to N to
avoid using efivars as a backend to pstore, as some users may want to
compile out the code completely.

Set the default to Y to maintain backwards compatability, since this
feature has always been enabled until now.
Signed-off-by: NSeth Forshee <seth.forshee@canonical.com>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Seiji Aguchi <seiji.aguchi@hds.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Joseph was hitting a failure case when mounting efivarfs which
resulted in an incorrect error message,

  $ sudo mount -v /sys/firmware/efi/efivars mount: Cannot allocate memory

triggered when efivarfs_valid_name() returned -EINVAL.

Make sure we pass accurate return values up the stack if
efivarfs_fill_super() fails to build inodes for EFI variables.
Reported-by: NJoseph Yasi <joe.yasi@gmail.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: <stable@vger.kernel.org> # v3.8
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Stricter validation was introduced with commit da27a243
("efivarfs: guid part of filenames are case-insensitive") and commit
47f531e8 ("efivarfs: Validate filenames much more aggressively"),
which is necessary for the guid portion of efivarfs filenames, but we
don't need to be so strict with the first part, the variable name. The
UEFI specification doesn't impose any constraints on variable names
other than they be a NULL-terminated string.

The above commits caused a regression that resulted in users seeing
the following message,

  $ sudo mount -v /sys/firmware/efi/efivars mount: Cannot allocate memory

whenever pstore EFI variables were present in the variable store,
since their variable names failed to pass the following check,

    /* GUID should be right after the first '-' */
    if (s - 1 != strchr(str, '-'))

as a typical pstore filename is of the form, dump-type0-10-1-<guid>.
The fix is trivial since the guid portion of the filename is GUID_LEN
bytes, we can use (len - GUID_LEN) to ensure the '-' character is
where we expect it to be.

(The bogus ENOMEM error value will be fixed in a separate patch.)
Reported-by: NJoseph Yasi <joe.yasi@gmail.com>
Tested-by: NJoseph Yasi <joe.yasi@gmail.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: <stable@vger.kernel.org> # v3.8
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

UEFI variables are typically stored in flash. For various reasons, avaiable
space is typically not reclaimed immediately upon the deletion of a
variable - instead, the system will garbage collect during initialisation
after a reboot.

Some systems appear to handle this garbage collection extremely poorly,
failing if more than 50% of the system flash is in use. This can result in
the machine refusing to boot. The safest thing to do for the moment is to
forbid writes if they'd end up using more than half of the storage space.
We can make this more finegrained later if we come up with a method for
identifying the broken machines.
Signed-off-by: NMatthew Garrett <matthew.garrett@nebula.com>
Cc: Josh Boyer <jwboyer@redhat.com>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Modify the request_module to prefix the file system type with "fs-"
and add aliases to all of the filesystems that can be built as modules
to match.

A common practice is to build all of the kernel code and leave code
that is not commonly needed as modules, with the result that many
users are exposed to any bug anywhere in the kernel.

Looking for filesystems with a fs- prefix limits the pool of possible
modules that can be loaded by mount to just filesystems trivially
making things safer with no real cost.

Using aliases means user space can control the policy of which
filesystem modules are auto-loaded by editing /etc/modprobe.d/*.conf
with blacklist and alias directives.  Allowing simple, safe,
well understood work-arounds to known problematic software.

This also addresses a rare but unfortunate problem where the filesystem
name is not the same as it's module name and module auto-loading
would not work.  While writing this patch I saw a handful of such
cases.  The most significant being autofs that lives in the module
autofs4.

This is relevant to user namespaces because we can reach the request
module in get_fs_type() without having any special permissions, and
people get uncomfortable when a user specified string (in this case
the filesystem type) goes all of the way to request_module.

After having looked at this issue I don't think there is any
particular reason to perform any filtering or permission checks beyond
making it clear in the module request that we want a filesystem
module.  The common pattern in the kernel is to call request_module()
without regards to the users permissions.  In general all a filesystem
module does once loaded is call register_filesystem() and go to sleep.
Which means there is not much attack surface exposed by loading a
filesytem module unless the filesystem is mounted.  In a user
namespace filesystems are not mounted unless .fs_flags = FS_USERNS_MOUNT,
which most filesystems do not set today.
Acked-by: NSerge Hallyn <serge.hallyn@canonical.com>
Acked-by: NKees Cook <keescook@chromium.org>
Reported-by: NKees Cook <keescook@google.com>
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>

This removes an open coded simple_open() function and
replaces file operations references to the function
with simple_open() instead.
Signed-off-by: NWei Yongjun <yongjun_wei@trendmicro.com.cn>
Cc: Jeremy Kerr <jk@ozlabs.org>
Cc: Matthew Garret <mjg59@srcf.ucam.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

[Problem]
efi_pstore creates sysfs entries, which enable users to access to NVRAM,
in a write callback. If a kernel panic happens in an interrupt context,
it may fail because it could sleep due to dynamic memory allocations during
creating sysfs entries.

[Patch Description]
This patch removes sysfs operations from a write callback by introducing
a workqueue updating sysfs entries which is scheduled after the write
callback is called.

Also, the workqueue is kicked in a just oops case.
A system will go down in other cases such as panic, clean shutdown and emergency
restart. And we don't need to create sysfs entries because there is no chance for
users to access to them.

efi_pstore will be robust against a kernel panic in an interrupt context with this patch.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMatt Fleming <matt.fleming@intel.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Problem]
There is a scenario which efi_pstore fails to log messages in a panic case.

 - CPUA holds an efi_var->lock in either efivarfs parts
   or efi_pstore with interrupt enabled.
 - CPUB panics and sends IPI to CPUA in smp_send_stop().
 - CPUA stops with holding the lock.
 - CPUB kicks efi_pstore_write() via kmsg_dump(KSMG_DUMP_PANIC)
   but it returns without logging messages.

[Patch Description]
This patch disables an external interruption while holding efivars->lock
as follows.

In efi_pstore_write() and get_var_data(), spin_lock/spin_unlock is
replaced by spin_lock_irqsave/spin_unlock_irqrestore because they may
be called in an interrupt context.

In other functions, they are replaced by spin_lock_irq/spin_unlock_irq.
because they are all called from a process context.

By applying this patch, we can avoid the problem above with
a following senario.

 - CPUA holds an efi_var->lock with interrupt disabled.
 - CPUB panics and sends IPI to CPUA in smp_send_stop().
 - CPUA receives the IPI after releasing the lock because it is
   disabling interrupt while holding the lock.
 - CPUB waits for one sec until CPUA releases the lock.
 - CPUB kicks efi_pstore_write() via kmsg_dump(KSMG_DUMP_PANIC)
   And it can hold the lock successfully.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Acked-by: NMatt Fleming <matt.fleming@intel.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

It makes no sense to treat the following filenames as unique,

	VarName-abcdefab-abcd-abcd-abcd-abcdefabcdef
	VarName-ABCDEFAB-ABCD-ABCD-ABCD-ABCDEFABCDEF
	VarName-ABcDEfAB-ABcD-ABcD-ABcD-ABcDEfABcDEf
	VarName-aBcDEfAB-aBcD-aBcD-aBcD-aBcDEfaBcDEf
	... etc ...

since the guid will be converted into a binary representation, which
has no case.

Roll our own dentry operations so that we can treat the variable name
part of filenames ("VarName" in the above example) as case-sensitive,
but the guid portion as case-insensitive. That way, efivarfs will
refuse to create the above files if any one already exists.
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Jeremy Kerr <jeremy.kerr@canonical.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

The only thing that efivarfs does to enforce a valid filename is
ensure that the name isn't too short. We need to strongly sanitise any
filenames, not least because variable creation is delayed until
efivarfs_file_write(), which means we can't rely on the firmware to
inform us of an invalid name, because if the file is never written to
we'll never know it's invalid.

Perform a couple of steps before agreeing to create a new file,

  * hex_to_bin() returns a value indicating whether or not it was able
    to convert its arguments to a binary representation - we should
    check it.

  * Ensure that the GUID portion of the filename is the correct length
    and format.

  * The variable name portion of the filename needs to be at least one
    character in size.
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Jeremy Kerr <jeremy.kerr@canonical.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Instead of adding a magic 4 to the variable size, use sizeof() to make
it explicitly clear what the quantity represents (the variable's
attributes).

CC: Jeremy Kerr <jeremy.kerr@canonical.com>
Cc: Chun-Yi Lee <joeyli.kernel@gmail.com>
Cc: Andy Whitcroft <apw@canonical.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Originally 'efi_enabled' indicated whether a kernel was booted from
EFI firmware. Over time its semantics have changed, and it now
indicates whether or not we are booted on an EFI machine with
bit-native firmware, e.g. 64-bit kernel with 64-bit firmware.

The immediate motivation for this patch is the bug report at,

    https://bugs.launchpad.net/ubuntu-cdimage/+bug/1040557

which details how running a platform driver on an EFI machine that is
designed to run under BIOS can cause the machine to become
bricked. Also, the following report,

    https://bugzilla.kernel.org/show_bug.cgi?id=47121

details how running said driver can also cause Machine Check
Exceptions. Drivers need a new means of detecting whether they're
running on an EFI machine, as sadly the expression,

    if (!efi_enabled)

hasn't been a sufficient condition for quite some time.

Users actually want to query 'efi_enabled' for different reasons -
what they really want access to is the list of available EFI
facilities.

For instance, the x86 reboot code needs to know whether it can invoke
the ResetSystem() function provided by the EFI runtime services, while
the ACPI OSL code wants to know whether the EFI config tables were
mapped successfully. There are also checks in some of the platform
driver code to simply see if they're running on an EFI machine (which
would make it a bad idea to do BIOS-y things).

This patch is a prereq for the samsung-laptop fix patch.

Cc: David Airlie <airlied@linux.ie>
Cc: Corentin Chary <corentincj@iksaif.net>
Cc: Matthew Garrett <mjg59@srcf.ucam.org>
Cc: Dave Jiang <dave.jiang@intel.com>
Cc: Olof Johansson <olof@lixom.net>
Cc: Peter Jones <pjones@redhat.com>
Cc: Colin Ian King <colin.king@canonical.com>
Cc: Steve Langasek <steve.langasek@canonical.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Konrad Rzeszutek Wilk <konrad@kernel.org>
Cc: Rafael J. Wysocki <rjw@sisk.pl>
Cc: <stable@vger.kernel.org>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>
Signed-off-by: NH. Peter Anvin <hpa@linux.intel.com>

Unlike the unlink path that is called from the VFS layer, we need to
call d_delete() ourselves when a variable is deleted in
efivarfs_file_write().

Failure to do so means we can access a stale struct efivar_entry when
reading/writing the file, which can result in the following oops,

  [   59.978216] general protection fault: 0000 [#1] SMP
  [   60.038660] CPU 9
  [   60.040501] Pid: 1001, comm: cat Not tainted 3.7.0-2.fc19.x86_64 #1 IBM System x3550 M3 -[7944I21]-/69Y4438
  [   60.050840] RIP: 0010:[<ffffffff810d5d1e>]  [<ffffffff810d5d1e>] __lock_acquire+0x5e/0x1bb0
  [   60.059198] RSP: 0018:ffff880270595ce8  EFLAGS: 00010046
  [   60.064500] RAX: 0000000000000046 RBX: 0000000000000002 RCX: 0000000000000000
  [   60.071617] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 6b6b6b6b6b6b6b83
  [   60.078735] RBP: ffff880270595dd8 R08: 0000000000000002 R09: 0000000000000000
  [   60.085852] R10: 6b6b6b6b6b6b6b83 R11: 0000000000000000 R12: 0000000000000000
  [   60.092971] R13: ffff88027170cd20 R14: 0000000000000000 R15: 0000000000000000
  [   60.100091] FS:  00007fc0c8ff3740(0000) GS:ffff880277000000(0000) knlGS:0000000000000000
  [   60.108164] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  [   60.113899] CR2: 0000000001520000 CR3: 000000026d594000 CR4: 00000000000007e0
  [   60.121016] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  [   60.128135] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
  [   60.135254] Process cat (pid: 1001, threadinfo ffff880270594000, task ffff88027170cd20)
  [   60.143239] Stack:
  [   60.145251]  ffff880270595cf8 ffffffff81021da3 ffff880270595d08 ffffffff81021e19
  [   60.152714]  ffff880270595d38 ffffffff810acdb5 ffff880200000168 0000000000000086
  [   60.160175]  ffff88027170d5e8 ffffffff810d25ed ffff880270595d58 ffffffff810ace7f
  [   60.167638] Call Trace:
  [   60.170088]  [<ffffffff81021da3>] ? native_sched_clock+0x13/0x80
  [   60.176085]  [<ffffffff81021e19>] ? sched_clock+0x9/0x10
  [   60.181389]  [<ffffffff810acdb5>] ? sched_clock_cpu+0xc5/0x120
  [   60.187211]  [<ffffffff810d25ed>] ? trace_hardirqs_off+0xd/0x10
  [   60.193121]  [<ffffffff810ace7f>] ? local_clock+0x6f/0x80
  [   60.198513]  [<ffffffff810d2f6f>] ? lock_release_holdtime.part.26+0xf/0x180
  [   60.205465]  [<ffffffff810d7b57>] ? lock_release_non_nested+0x2e7/0x320
  [   60.212073]  [<ffffffff815638bb>] ? efivarfs_file_write+0x5b/0x280
  [   60.218242]  [<ffffffff810d7f41>] lock_acquire+0xa1/0x1f0
  [   60.223633]  [<ffffffff81563971>] ? efivarfs_file_write+0x111/0x280
  [   60.229892]  [<ffffffff8118b47c>] ? might_fault+0x5c/0xb0
  [   60.235287]  [<ffffffff816f1bf6>] _raw_spin_lock+0x46/0x80
  [   60.240762]  [<ffffffff81563971>] ? efivarfs_file_write+0x111/0x280
  [   60.247018]  [<ffffffff81563971>] efivarfs_file_write+0x111/0x280
  [   60.253103]  [<ffffffff811d307f>] vfs_write+0xaf/0x190
  [   60.258233]  [<ffffffff811d33d5>] sys_write+0x55/0xa0
  [   60.263278]  [<ffffffff816fbd19>] system_call_fastpath+0x16/0x1b
  [   60.269271] Code: 41 0f 45 d8 4c 89 75 f0 4c 89 7d f8 85 c0 0f 84 09 01 00 00 8b 05 a3 f9 ff 00 49 89 fa 41 89 f6 41 89 d3 85 c0 0f 84 12 01 00 00 <49> 8b 02 ba 01 00 00 00 48 3d a0 07 14 82 0f 44 da 41 83 fe 01
  [   60.289431] RIP  [<ffffffff810d5d1e>] __lock_acquire+0x5e/0x1bb0
  [   60.295444]  RSP <ffff880270595ce8>
  [   60.298928] ---[ end trace 1bbfd41a2cf6a0d8 ]---

Cc: Josh Boyer <jwboyer@redhat.com>
Acked-by: NJeremy Kerr <jeremy.kerr@canonical.com>
Cc: Lee, Chun-Yi <jlee@suse.com>
Cc: Andy Whitcroft <apw@canonical.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

Files are created in efivarfs_create() before a corresponding variable
is created in the firmware. This leads to users being able to
read/write to the file without the variable existing in the
firmware. Reading a non-existent variable currently returns -ENOENT,
which is confusing because the file obviously *does* exist.

Convert EFI_NOT_FOUND into -EIO which is the closest thing to "error
while interacting with firmware", and should hopefully indicate to the
caller that the variable is in some uninitialised state.

Cc: Josh Boyer <jwboyer@redhat.com>
Acked-by: NJeremy Kerr <jeremy.kerr@canonical.com>
Cc: Lee, Chun-Yi <jlee@suse.com>
Cc: Andy Whitcroft <apw@canonical.com>
Reported-by: NLingzhu Xiang <lxiang@redhat.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

efivarfs_unlink() should drop the file's link count, not the directory's.
Signed-off-by: NLingzhu Xiang <lxiang@redhat.com>
Cc: Jeremy Kerr <jeremy.kerr@canonical.com>
Tested-by: NLee, Chun-Yi <jlee@suse.com>
Signed-off-by: NMatt Fleming <matt.fleming@intel.com>

[Issue]

There is a scenario which efi_pstore may hang up:

 - cpuA grabs efivars->lock
 - cpuB panics and calls smp_send_stop
 - smp_send_stop sends IRQ to cpuA
 - after 1 second, cpuB gives up on cpuA and sends an NMI instead
 - cpuA is now in an NMI handler while still holding efivars->lock
 - cpuB is deadlocked

This case may happen if a firmware has a bug and
cpuA is stuck talking with it.

[Solution]

This patch changes a spin_lock to a spin_trylock in non-blocking paths.
and if the spin_lock has already taken by another cpu,
it returns without accessing to a firmware to avoid the deadlock.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NDon Zickus <dzickus@redhat.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

When compiling efivars.c the build fails with:

   CC      drivers/firmware/efivars.o
  drivers/firmware/efivars.c: In function ‘efivarfs_get_inode’:
  drivers/firmware/efivars.c:886:31: error: incompatible types when assigning to type ‘kgid_t’ from type ‘int’
  make[2]: *** [drivers/firmware/efivars.o] Error 1
  make[1]: *** [drivers/firmware/efivars.o] Error 2

Fix the build error by removing the duplicate initialization of i_uid and
i_gid inode_init_always has already initialized them to 0.
Signed-off-by: N"Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

[Issue]

a format of variable name has been updated to type, id, count and ctime
to support holding multiple logs.

Format of current variable name
  dump-type0-1-2-12345678

  type:0
  id:1
  count:2
  ctime:12345678

On the other hand, if an old variable name before being updated
remains, users can't erase it via /dev/pstore.

Format of old variable name
  dump-type0-1-12345678

  type:0
  id:1
  ctime:12345678

[Solution]

This patch add a format check for the old variable name in a erase callback to make it erasable.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

a format of variable name has been updated to type, id, count and ctime
to support holding multiple logs.

Format of current variable name
  dump-type0-1-2-12345678

  type:0
  id:1
  count:2
  ctime:12345678

On the other hand, if an old variable name before being updated
remains, users can't read it via /dev/pstore.

Format of old variable name
  dump-type0-1-12345678

  type:0
  id:1
  ctime:12345678

[Solution]

This patch add a format check for the old variable name in a read callback
to make it readable.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

Currently, a variable name, which identifies each entry, consists of type, id and ctime.
But if multiple events happens in a short time, a second/third event may fail to log because
efi_pstore can't distinguish each event with current variable name.

[Solution]

A reasonable way to identify all events precisely is introducing a sequence counter to
the variable name.

The sequence counter has already supported in a pstore layer with "oopscount".
So, this patch adds it to a variable name.
Also, it is passed to read/erase callbacks of platform drivers in accordance with
the modification of the variable name.

  <before applying this patch>
 a variable name of first event: dump-type0-1-12345678
 a variable name of second event: dump-type0-1-12345678

  type:0
  id:1
  ctime:12345678

 If multiple events happen in a short time, efi_pstore can't distinguish them because
 variable names are same among them.

  <after applying this patch>

 it can be distinguishable by adding a sequence counter as follows.

 a variable name of first event: dump-type0-1-1-12345678
 a variable name of Second event: dump-type0-1-2-12345678

  type:0
  id:1
  sequence counter: 1(first event), 2(second event)
  ctime:12345678

In case of a write callback executed in pstore_console_write(), "0" is added to
an argument of the write callback because it just logs all kernel messages and
doesn't need to care about multiple events.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NRafael J. Wysocki <rafael.j.wysocki@intel.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

Currently, a variable name, which is used to identify each log entry, consists of type,
id and ctime. But an erase callback does not use ctime.

If efi_pstore supported just one log, type and id were enough.
However, in case of supporting multiple logs, it doesn't work because
it can't distinguish each entry without ctime at erasing time.

 <Example>

 As you can see below, efi_pstore can't differentiate first event from second one without ctime.

 a variable name of first event: dump-type0-1-12345678
 a variable name of second event: dump-type0-1-23456789

  type:0
  id:1
  ctime:12345678, 23456789

[Solution]

This patch adds ctime to an argument of an erase callback.

It works across reboots because ctime of pstore means the date that the record was originally stored.
To do this, efi_pstore saves the ctime to variable name at writing time and passes it to pstore
at reading time.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

Currently, efi_pstore driver simply overwrites existing panic messages in NVRAM.
So, in the following scenario, we will lose 1st panic messages.

1. kernel panics.
2. efi_pstore is kicked and writes panic messages to NVRAM.
3. system reboots.
4. kernel panics again before a user checks the 1st panic messages in NVRAM.

[Solution]

A reasonable solution to fix the issue is just holding multiple logs without erasing
existing entries.
This patch removes a logic erasing existing entries in a write callback
because the logic is not needed in the write callback to support holding multiple logs.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

Currently, efi_pstore driver simply overwrites existing panic messages in NVRAM.
So, in the following scenario, we will lose 1st panic messages.

 1. kernel panics.
 2. efi_pstore is kicked and writes panic messages to NVRAM.
 3. system reboots.
 4. kernel panics again before a user checks the 1st panic messages in NVRAM.

[Solution]

A reasonable solution to fix the issue is just holding multiple logs without erasing
existing entries.

This patch freshly adds a logic erasing existing entries, which shared with a write callback,
to an erase callback.
To support holding multiple logs, the write callback doesn't need to erase any entries and
it will be removed in a subsequent patch.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>

[Issue]

As discussed in a thread below, Running out of space in EFI isn't a well-tested scenario.
And we wouldn't expect all firmware to handle it gracefully.
http://marc.info/?l=linux-kernel&m=134305325801789&w=2

On the other hand, current efi_pstore doesn't check a remaining space of storage at writing time.
Therefore, efi_pstore may not work if it tries to write a large amount of data.

[Patch Description]

To avoid handling the situation above, this patch checks if there is a space enough to log with
QueryVariableInfo() before writing data.
Signed-off-by: NSeiji Aguchi <seiji.aguchi@hds.com>
Acked-by: NMike Waychison <mikew@google.com>
Signed-off-by: NTony Luck <tony.luck@intel.com>