Kernel Userspace Access Prevention utilises a feature of the Radix MMU
which disallows read and write access to userspace addresses. By
utilising this, the kernel is prevented from accessing user data from
outside of trusted paths that perform proper safety checks, such as
copy_{to/from}_user() and friends.

Userspace access is disabled from early boot and is only enabled when
performing an operation like copy_{to/from}_user(). The register that
controls this (AMR) does not prevent userspace from accessing itself,
so there is no need to save and restore when entering and exiting
userspace.

When entering the kernel from the kernel we save AMR and if it is not
blocking user access (because eg. we faulted doing a user access) we
reblock user access for the duration of the exception (ie. the page
fault) and then restore the AMR when returning back to the kernel.

This feature can be tested by using the lkdtm driver (CONFIG_LKDTM=y)
and performing the following:

  # (echo ACCESS_USERSPACE) > [debugfs]/provoke-crash/DIRECT

If enabled, this should send SIGSEGV to the thread.

We also add paranoid checking of AMR in switch and syscall return
under CONFIG_PPC_KUAP_DEBUG.
Co-authored-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

__patch_instruction() is called in early boot, and uses
__put_user_size(), which includes the allow/prevent calls to enforce
KUAP, which could either be called too early, or in the Radix case,
forced to use "early_" versions of functions just to safely handle
this one case.

__put_user_asm() does not do this, and thus is safe to use both in
early boot, and later on since in this case it should only ever be
touching kernel memory.

__patch_instruction() was previously refactored to use
__put_user_size() in order to be able to return -EFAULT, which would
allow the kernel to patch instructions in userspace, which should
never happen. This has the functional change of causing faults on
userspace addresses if KUAP is turned on, which should never happen in
practice.

A future enhancement could be to double check the patch address is
definitely allowed to be tampered with by the kernel.
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Execution protection already exists on radix, this just refactors
the radix init to provide the KUEP setup function instead.

Thus, the only functional change is that it can now be disabled.
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Some platforms (i.e. Radix MMU) need per-CPU initialisation for KUP.

Any platforms that only want to do KUP initialisation once
globally can just check to see if they're running on the boot CPU, or
check if whatever setup they need has already been performed.

Note that this is only for 64-bit.
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

This patch implements a framework for Kernel Userspace Access
Protection.

Then subarches will have the possibility to provide their own
implementation by providing setup_kuap() and
allow/prevent_user_access().

Some platforms will need to know the area accessed and whether it is
accessed from read, write or both. Therefore source, destination and
size and handed over to the two functions.

mpe: Rename to allow/prevent rather than unlock/lock, and add
read/write wrappers. Drop the 32-bit code for now until we have an
implementation for it. Add kuap to pt_regs for 64-bit as well as
32-bit. Don't split strings, use pr_crit_ratelimited().
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

This patch adds a skeleton for Kernel Userspace Execution Prevention.

Then subarches implementing it have to define CONFIG_PPC_HAVE_KUEP
and provide setup_kuep() function.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
[mpe: Don't split strings, use pr_crit_ratelimited()]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

This patch adds a skeleton for Kernel Userspace Protection
functionnalities like Kernel Userspace Access Protection and Kernel
Userspace Execution Prevention

The subsequent implementation of KUAP for radix makes use of a MMU
feature in order to patch out assembly when KUAP is disabled or
unsupported. This won't work unless there's an entry point for KUP
support before the feature magic happens, so for PPC64 setup_kup() is
called early in setup.

On PPC32, feature_fixup() is done too early to allow the same.
Suggested-by: Russell Currey <ruscur@russell.cc>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

In order to implement KUAP (Kernel Userspace Access Protection) on
Power9 we will be using the AMR, and therefore indirectly the
UAMOR/AMOR.

So save/restore these regs in the idle code.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Without restoring the IAMR after idle, execution prevention on POWER9
with Radix MMU is overwritten and the kernel can freely execute
userspace without faulting.

This is necessary when returning from any stop state that modifies
user state, as well as hypervisor state.

To test how this fails without this patch, load the lkdtm driver and
do the following:

  $ echo EXEC_USERSPACE > /sys/kernel/debug/provoke-crash/DIRECT

which won't fault, then boot the kernel with powersave=off, where it
will fault. Applying this patch will fix this.

Fixes: 3b10d009 ("powerpc/mm/radix: Prevent kernel execution of user space")
Cc: stable@vger.kernel.org # v4.10+
Signed-off-by: Russell Currey <ruscur@russell.cc>
Reviewed-by: Akshay Adiga <akshay.adiga@linux.vnet.ibm.com>
Reviewed-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Changing the NUMA associations for CPUs and memory at runtime is
basically unsupported by the core mm, scheduler etc. We see all manner
of crashes, warnings and instability when the pseries code tries to do
this. Disable this behavior by default, and document the switch a bit.
Signed-off-by: Nathan Lynch <nathanl@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

When booted with "topology_updates=no", or when "off" is written to
/proc/powerpc/topology_updates, NUMA reassignments are inhibited for
PRRN and VPHN events. However, migration and suspend unconditionally
re-enable reassignments via start_topology_update(). This is
incoherent.

Check the topology_updates_enabled flag in
start/stop_topology_update() so that callers of those APIs need not be
aware of whether reassignments are enabled. This allows the
administrative decision on reassignments to remain in force across
migrations and suspensions.
Signed-off-by: Nathan Lynch <nathanl@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

sparse complains a lot about opal-call.c:

  arch/powerpc/platforms/powernv/opal-call.c:128:1: warning: symbol 'opal_invalid_call' was not declared. Should it be static?
  arch/powerpc/platforms/powernv/opal-call.c:129:1: warning: symbol 'opal_console_write' was not declared. Should it be static?
  arch/powerpc/platforms/powernv/opal-call.c:130:1: warning: symbol 'opal_console_read' was not declared. Should it be static?

Those symbols are forward declared in opal.h, but we can't include that
because the function signatures in opal.h are different. So instead, just
add an extra forward declaration to the OPAL_CALL macro to shut sparse up.
Signed-off-by: Andrew Donnellan <andrew.donnellan@au1.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

This code was filling a 64K buffer from /dev/urandom in order to
compute a CRC over (on average half of) it by two different methods,
comparing the CRCs, and repeating.

This is not a remotely security-critical application, so use the far
faster and cheaper prandom_u32() generator.

And, while we're at it, only fill as much of the buffer as we plan to use.
Signed-off-by: George Spelvin <lkml@sdf.org>
Acked-by: Daniel Axtens <dja@axtens.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Remove duplicate headers inclusions.
Signed-off-by: Jagadeesh Pagadala <jagdsh.linux@gmail.com>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

The call to of_find_compatible_node() returns a node pointer with
refcount incremented thus it must be explicitly decremented after the
last usage.

irq_domain_add_linear() also calls of_node_get() to increase refcount,
so irq_domain() will not be affected when it is released.

Detected by coccinelle.

Fixes: a8db8cf0 ("irq_domain: Replace irq_alloc_host() with revmap-specific initializers")
Signed-off-by: Wen Yang <wen.yang99@zte.com.cn>
Suggested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Suggested-by: Michael Ellerman <mpe@ellerman.id.au>
Reviewed-by: Peng Hao <peng.hao2@zte.com.cn>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Add support to hwpoison the pages upon hitting machine check
exception.

This patch queues the address where UE is hit to percpu array
and schedules work to plumb it into memory poison infrastructure.
Reviewed-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Ganesh Goudar <ganeshgr@linux.ibm.com>
[mpe: Combine #ifdefs, drop PPC_BIT8(), and empty inline stub]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Add an of_node_put() when a tested device node is not available.

Fixes: c026c987 ("powerpc/83xx: Do not configure or probe disabled FSL DR USB controllers")
Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Paul McKenney attempted to update all email addresses @linux.vnet.ibm.com
to @linux.ibm.com in commit 1dfddcdb
("MAINTAINERS: Update from @linux.vnet.ibm.com to @linux.ibm.com"), but
some still remained.

We update the remaining email addresses in MAINTAINERS, hopefully finally
catching all cases for good.

Fixes: 1dfddcdb ("MAINTAINERS: Update from @linux.vnet.ibm.com to @linux.ibm.com")
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Acked-by: Paul E. McKenney <paulmck@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

The file arch/powerpc/include/uapi/asm/vas.h was considered but
never merged and should be removed from the MAINTAINERS file.

While here, add missing email address.
Reported-by: Joe Perches <joe@perches.com>
Signed-off-by: Sukadev Bhattiprolu <sukadev@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

The commit 4c5d87db ("powerpc/pseries: PAPR persistent memory
support") set a local variable "count" in dlpar_hp_pmem() but never
use it.

  arch/powerpc/platforms/pseries/pmem.c: In function 'dlpar_hp_pmem':
  arch/powerpc/platforms/pseries/pmem.c:109:6: warning: variable 'count' set but not used
Signed-off-by: Qian Cai <cai@lca.pw>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

The commit b7d6bf4f ("powerpc/pseries/pci: Remove obsolete SW
invalidate") left 2 variables unused.

  arch/powerpc/platforms/pseries/iommu.c:108:17: warning: variable 'tces' set but not used
    __be64 *tcep, *tces;
                   ^~~~

  arch/powerpc/platforms/pseries/iommu.c:132:17: warning: variable 'tces' set but not used
    __be64 *tcep, *tces;
                   ^~~~

Also, the commit 68c0449e ("powerpc/pseries/iommu: Use memory@
nodes in max RAM address calculation") set "ranges" in
ddw_memory_hotplug_max() but never use it.

  arch/powerpc/platforms/pseries/iommu.c: In function 'ddw_memory_hotplug_max':
  arch/powerpc/platforms/pseries/iommu.c:948:7: warning: variable 'ranges' set but not used
     int ranges, n_mem_addr_cells, n_mem_size_cells, len;
         ^~~~~~
Signed-off-by: Qian Cai <cai@lca.pw>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

pte_unmap() compiles away on some powerpc platforms, so silence the
warnings below by making it a static inline function.

  mm/memory.c: In function 'copy_pte_range':
  mm/memory.c:820:24: warning: variable 'orig_dst_pte' set but not used
  mm/memory.c:820:9: warning: variable 'orig_src_pte' set but not used
  mm/madvise.c: In function 'madvise_free_pte_range':
  mm/madvise.c:318:9: warning: variable 'orig_pte' set but not used
  mm/swap_state.c: In function 'swap_ra_info':
  mm/swap_state.c:634:15: warning: variable 'orig_pte' set but not used
Suggested-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Qian Cai <cai@lca.pw>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

resize_hpt_for_hotplug() reports a warning when it cannot
resize the hash page table ("Unable to resize hash page
table to target order") but in some cases it's not a problem
and can make user thinks something has not worked properly.

This patch moves the warning to arch_remove_memory() to
only report the problem when it is needed.
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Signed-off-by: Laurent Vivier <lvivier@redhat.com>
Reviewed-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Add radix_enabled() check to avoid SLB preload with radix translation.
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Recent versions of QEMU provide a XHCI device by default these
days instead of an old-fashioned OHCI device:

 https://git.qemu.org/?p=qemu.git;a=commitdiff;h=57040d451315320b7d27

So to get the keyboard working in the graphical console there again,
we should now include XHCI support in the kernel by default, too.
Signed-off-by: Thomas Huth <thuth@redhat.com>
Reviewed-by: David Gibson <david@gibson.dropbear.id.au>
Acked-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

This is a follow up to the patch that fixed misleading print for TLB
mutlihit due to wrongly populated mc_err_types[] array. Convert all the
static array initialization to '[x] = val' style for better
readability of array indexing and avoid any further confusion.
Suggested-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Mahesh Salgaonkar <mahesh@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

With STRICT_KERNEL_RWX enabled anything marked __init is placed at a 16M
boundary.  This is necessary so that it can be repurposed later with
different permissions.  However, in kernels with text larger than 16M,
this pushes early_setup past 32M, incapable of being reached by the
branch instruction.

Fix this by setting the CTR and branching there instead.

Fixes: 1e0fc9d1 ("powerpc/Kconfig: Enable STRICT_KERNEL_RWX for some configs")
Signed-off-by: Russell Currey <ruscur@russell.cc>
[mpe: Fix it to work on BE by using DOTSYM()]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Silence the following warnings triggered using W=1:

  arch/powerpc/platforms/embedded6xx/holly.c:236:6: error: no previous prototype for 'holly_power_off'
  arch/powerpc/platforms/embedded6xx/holly.c:243:6: error: no previous prototype for 'holly_halt'
Signed-off-by: Mathieu Malaterre <malat@debian.org>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

In commit cb9e4d10 ("[POWERPC] Add support for 750CL Holly board")
new functions were added. Since most of these functions can be made
static, make it so.

Both holly_power_off and holly_halt functions were not changed since
they are unused, making them static would have triggered the following
warning (treated as error):

  arch/powerpc/platforms/embedded6xx/holly.c:244:13: error: 'holly_halt' defined but not used

Silence the following warnings triggered using W=1:

  arch/powerpc/platforms/embedded6xx/holly.c:47:5: error: no previous prototype for 'holly_exclude_device'
  arch/powerpc/platforms/embedded6xx/holly.c:190:6: error: no previous prototype for 'holly_show_cpuinfo'
  arch/powerpc/platforms/embedded6xx/holly.c:196:17: error: no previous prototype for 'holly_restart'
Signed-off-by: Mathieu Malaterre <malat@debian.org>
Reviewed-by: Mukesh Ojha <mojha@codeaurora.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

The 32-bit vDSO is not needed and not normally built for 64-bit
little-endian configurations.  However, the vdso_install target still
builds and installs it.  Add the same config condition as is normally
used for the build.

Fixes: e0d00591 ("powerpc/vdso: Disable building the 32-bit VDSO ...")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Add comments describing the size in bytes of the various levels of the
page table tree, and the size of the virtual address space mapped by
each level, to make it clear what the sizes are without having to also
look up other definitions.

The code that calculates the sizes actually uses sizeof(pgd_t) etc.,
so in theory these comments could skew vs the code, but the size of
pgd_t etc. is unlikely to change very often.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

In arch/powerpc/mm/highmem.c, BUG_ON() is called only when
CONFIG_DEBUG_HIGHMEM is selected, this means the BUG_ON() is not vital
and can be replaced by a a WARN_ON().

At the same time, use IS_ENABLED() instead of #ifdef to clean a bit.
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Our logic for choosing defconfig doesn't work well in some situations.

For example if you're on a ppc64le machine but you specify a non-empty
CROSS_COMPILE, in order to use a non-default toolchain, then defconfig
will give you ppc64_defconfig (big endian):

  $ make CROSS_COMPILE=~/toolchains/gcc-8/bin/powerpc-linux- defconfig
  *** Default configuration is based on 'ppc64_defconfig'

This is because we assume that CROSS_COMPILE being set means we
can't be on a ppc machine and rather than checking we just default to
ppc64_defconfig.

We should just ignore CROSS_COMPILE, instead check the machine with
uname and if it's one of ppc, ppc64 or ppc64le then use that
defconfig. If it's none of those then we fall back to ppc64_defconfig.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Add a generic 32-bit defconfig called ppc_defconfig. This means we'll
have a defconfig matching "uname -m" for all cases.

This config is mostly intended for build testing but if someone wants
to tweak it to get it booting on something that would be fine too.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Tested-by: Mathieu Malaterre <malat@debian.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Merge our fixes branch. In particular the radix segment exception
handling fix is necessary to avoid odd crashes.

The recent commit 8bc08689 ("powerpc/mm: Only define
MAX_PHYSMEM_BITS in SPARSEMEM configurations") removed our definition
of MAX_PHYSMEM_BITS when SPARSEMEM is disabled.

This inadvertently broke some 64-bit FLATMEM using configs with eg:

  arch/powerpc/include/asm/book3s/64/mmu-hash.h:584:6: error: "MAX_PHYSMEM_BITS" is not defined, evaluates to 0
   #if (MAX_PHYSMEM_BITS > MAX_EA_BITS_PER_CONTEXT)
        ^~~~~~~~~~~~~~~~

Fix it by making sure we define MAX_PHYSMEM_BITS for all 64-bit
configs regardless of SPARSEMEM.

Fixes: 8bc08689 ("powerpc/mm: Only define MAX_PHYSMEM_BITS in SPARSEMEM configurations")
Reported-by: Andreas Schwab <schwab@linux-m68k.org>
Reported-by: Hugh Dickins <hughd@google.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Commit 48e7b769 ("powerpc/64s/hash: Convert SLB miss handlers to C")
broke the radix-mode segment exception handler. In radix mode, this is
exception is not an SLB miss, rather it signals that the EA is outside
the range translated by any page table.

The commit lost the radix feature alternate code patch, which can
cause faults to some EAs to kernel BUG at arch/powerpc/mm/slb.c:639!

The original radix code would send faults to slb_miss_large_addr,
which would end up faulting due to slb_addr_limit being 0. This patch
sends radix directly to do_bad_slb_fault, which is a bit clearer.

Fixes: 48e7b769 ("powerpc/64s/hash: Convert SLB miss handlers to C")
Cc: stable@vger.kernel.org # v4.20+
Reported-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Commit b5b4453e ("powerpc/vdso64: Fix CLOCK_MONOTONIC
inconsistencies across Y2038") changed the type of wtom_clock_sec
to s64 on PPC64. Therefore, VDSO32 needs to read it with a 4 bytes
shift in order to retrieve the lower part of it.

Fixes: b5b4453e ("powerpc/vdso64: Fix CLOCK_MONOTONIC inconsistencies across Y2038")
Reported-by: Christian Zigotzky <chzigotzky@xenosoft.de>
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>

Commit 0df977ea ("powerpc/6xx: Don't use SPRN_SPRG2 for storing
stack pointer while in RTAS") changes the code to use a field in
thread struct to store the stack pointer while in RTAS instead of
using SPRN_SPRG2. It therefore converts all places which were
manipulating SPRN_SPRG2 to use that field. During early startup, the
zeroing of SPRN_SPRG2 has been replaced by a zeroing of that field in
thread struct. But at least in start_here, that's done wrongly because
it used the physical address of the fields while MMU is on at that
time.

So the virtual address of the field should be used instead, but in
the meantime, thread struct has already been zeroed and initialised
so we can just drop this initialisation.
Reported-by: Larry Finger <Larry.Finger@lwfinger.net>
Fixes: 0df977ea ("powerpc/6xx: Don't use SPRN_SPRG2 for storing stack pointer while in RTAS")
Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
Tested-by: Larry Finger <Larry.Finger@lwfinger.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>