Commit 9b69d48c7516 ("powerpc/64e: remove implicit soft-masking and
interrupt exit restart logic") limited the implicit soft masking and
restart logic to 64-bit Book3S only. However we are still building
restart_table.c for all 64-bit, ie. Book3E also.

There's no need to build it for 64e, and it also causes missing
prototype warnings for 64e builds, because the prototype is already
behind an #ifdef PPC_BOOK3S_64.

Fixes: 9b69d48c7516 ("powerpc/64e: remove implicit soft-masking and interrupt exit restart logic")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210701125026.292224-1-mpe@ellerman.id.au

The exception table fixup adjusts a failed page fault's interrupt return
location if it was taken at an address specified in the exception table,
to a corresponding fixup handler address.

Introduce a variation of that idea which adds a fixup table for NMIs and
soft-masked asynchronous interrupts. This will be used to protect
certain critical sections that are sensitive to being clobbered by
interrupts coming in (due to using the same SPRs and/or irq soft-mask
state).

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20210617155116.2167984-10-npiggin@gmail.com

Commit 7c95d8893fb5 ("powerpc: Change calling convention for
create_branch() et. al.") complexified the frame of function
do_feature_fixups(), leading to GCC setting up a stack
guard when CONFIG_STACKPROTECTOR is selected.

The problem is that do_feature_fixups() is called very early
while 'current' in r2 is not set up yet and the code is still
not at the final address used at link time.

So, like other instrumentation, stack protection needs to be
deactivated for feature-fixups.c and code-patching.c

Fixes: 7c95d8893fb5 ("powerpc: Change calling convention for create_branch() et. al.")
Cc: stable@vger.kernel.org # v5.8+
Reported-by: Jonathan Neuschaefer <j.neuschaefer@gmx.net>
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Tested-by: Jonathan Neuschaefer <j.neuschaefer@gmx.net>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/b688fe82927b330349d9e44553363fa451ea4d95.1619715114.git.christophe.leroy@csgroup.eu

When probe_kernel_read_inst() was created, there was no good place to
put it, so a file called lib/inst.c was dedicated for it.

Since then, probe_kernel_read_inst() has been renamed
copy_inst_from_kernel_nofault(). And mm/maccess.h didn't exist at that
time. Today, mm/maccess.h is related to copy_from_kernel_nofault().

Move copy_inst_from_kernel_nofault() into mm/maccess.c

Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/9655d8957313906b77b8db5700a0e33ce06f45e5.1618405715.git.christophe.leroy@csgroup.eu

Commit ccbef1674a15 ("Kbuild, lto: add ld-version and ld-ifversion
macros") introduced scripts/ld-version.sh for GCC LTO.

At that time, this script handled 5 version fields because GCC LTO
needed the downstream binutils. (https://lkml.org/lkml/2014/4/8/272)

The code snippet from the submitted patch was as follows:

    # We need HJ Lu's Linux binutils because mainline binutils does not
    # support mixing assembler and LTO code in the same ld -r object.
    # XXX check if the gcc plugin ld is the expected one too
    # XXX some Fedora binutils should also support it. How to check for that?
    ifeq ($(call ld-ifversion,-ge,22710001,y),y)
        ...

However, GCC LTO was not merged into the mainline after all.
(https://lkml.org/lkml/2014/4/8/272)

So, the 4th and 5th fields were never used, and finally removed by
commit 0d61ed17dd30 ("ld-version: Drop the 4th and 5th version
components").

Since then, the last 4-digits returned by this script is always zeros.

Remove the meaningless last 4-digits. This makes the version format
consistent with GCC_VERSION, CLANG_VERSION, LLD_VERSION.

Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Acked-by: Will Deacon <will@kernel.org>
Acked-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de>

In reaction to a proposal to introduce a memcpy_mcsafe_fast()
implementation Linus points out that memcpy_mcsafe() is poorly named
relative to communicating the scope of the interface. Specifically what
addresses are valid to pass as source, destination, and what faults /
exceptions are handled.

Of particular concern is that even though x86 might be able to handle
the semantics of copy_mc_to_user() with its common copy_user_generic()
implementation other archs likely need / want an explicit path for this
case:

  On Fri, May 1, 2020 at 11:28 AM Linus Torvalds <torvalds@linux-foundation.org> wrote:
  >
  > On Thu, Apr 30, 2020 at 6:21 PM Dan Williams <dan.j.williams@intel.com> wrote:
  > >
  > > However now I see that copy_user_generic() works for the wrong reason.
  > > It works because the exception on the source address due to poison
  > > looks no different than a write fault on the user address to the
  > > caller, it's still just a short copy. So it makes copy_to_user() work
  > > for the wrong reason relative to the name.
  >
  > Right.
  >
  > And it won't work that way on other architectures. On x86, we have a
  > generic function that can take faults on either side, and we use it
  > for both cases (and for the "in_user" case too), but that's an
  > artifact of the architecture oddity.
  >
  > In fact, it's probably wrong even on x86 - because it can hide bugs -
  > but writing those things is painful enough that everybody prefers
  > having just one function.

Replace a single top-level memcpy_mcsafe() with either
copy_mc_to_user(), or copy_mc_to_kernel().

Introduce an x86 copy_mc_fragile() name as the rename for the
low-level x86 implementation formerly named memcpy_mcsafe(). It is used
as the slow / careful backend that is supplanted by a fast
copy_mc_generic() in a follow-on patch.

One side-effect of this reorganization is that separating copy_mc_64.S
to its own file means that perf no longer needs to track dependencies
for its memcpy_64.S benchmarks.

 [ bp: Massage a bit. ]

Signed-off-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Reviewed-by: Tony Luck <tony.luck@intel.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Cc: <stable@vger.kernel.org>
Link: http://lore.kernel.org/r/CAHk-=wjSqtXAqfUJxFtWNwmguFASTgB0dz1dT3V-78Quiezqbg@mail.gmail.com
Link: https://lkml.kernel.org/r/160195561680.2163339.11574962055305783722.stgit@dwillia2-desk3.amr.corp.intel.com

These have shown significantly improved performance and fairness when
spinlock contention is moderate to high on very large systems.

With this series including subsequent patches, on a 16 socket 1536
thread POWER9, a stress test such as same-file open/close from all
CPUs gets big speedups, 11620op/s aggregate with simple spinlocks vs
384158op/s (33x faster), where the difference in throughput between
the fastest and slowest thread goes from 7x to 1.4x.

Thanks to the fast path being identical in terms of atomics and
barriers (after a subsequent optimisation patch), single threaded
performance is not changed (no measurable difference).

On smaller systems, performance and fairness seems to be generally
improved. Using dbench on tmpfs as a test (that starts to run into
kernel spinlock contention), a 2-socket OpenPOWER POWER9 system was
tested with bare metal and KVM guest configurations. Results can be
found here:

https://github.com/linuxppc/issues/issues/305#issuecomment-663487453

Observations are:

- Queued spinlocks are equal when contention is insignificant, as
  expected and as measured with microbenchmarks.

- When there is contention, on bare metal queued spinlocks have better
  throughput and max latency at all points.

- When virtualised, queued spinlocks are slightly worse approaching
  peak throughput, but significantly better throughput and max latency
  at all points beyond peak, until queued spinlock maximum latency
  rises when clients are 2x vCPUs.

The regressions haven't been analysed very well yet, there are a lot
of things that can be tuned, particularly the paravirtualised locking,
but the numbers already look like a good net win even on relatively
small systems.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Acked-by: Waiman Long <longman@redhat.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200724131423.1362108-4-npiggin@gmail.com

Expand the code-patching self-tests to includes tests for patching
prefixed instructions.

Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
[mpe: Use CONFIG_PPC64 not __powerpc64__]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200506034050.24806-25-jniethe5@gmail.com

Introduce a probe_user_read_inst() function to use in cases where
probe_user_read() is used for getting an instruction. This will be
more useful for prefixed instructions.

Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Reviewed-by: Alistair Popple <alistair@popple.id.au>
[mpe: Don't write to *inst on error, fold in __user annotations]
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200506034050.24806-14-jniethe5@gmail.com

The pmem infrastructure uses memcpy_mcsafe in the pmem layer so as to
convert machine check exceptions into a return value on failure in case
a machine check exception is encountered during the memcpy. The return
value is the number of bytes remaining to be copied.

This patch largely borrows from the copyuser_power7 logic and does not add
the VMX optimizations, largely to keep the patch simple. If needed those
optimizations can be folded in.

Signed-off-by: Balbir Singh <bsingharora@gmail.com>
[arbab@linux.ibm.com: Added symbol export]
Co-developed-by: Santosh Sivaraj <santosh@fossix.org>
Signed-off-by: Santosh Sivaraj <santosh@fossix.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20190820081352.8641-7-santosh@fossix.org