Uprobes uses emulate_step in sstep.c, but we haven't explicitly specified
the dependency. On pseries HAVE_HW_BREAKPOINT protects us, but 44x has no
such luxury.

Consolidate other users that depend on sstep and create a new config option.

Signed-off-by: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Signed-off-by: Suzuki K. Poulose <suzuki@in.ibm.com>
Cc: linuxppc-dev@ozlabs.org
Cc: stable@vger.kernel.org
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Finally remove the two level TOC and build with -mcmodel=medium.

Unfortunately we can't build modules with -mcmodel=medium due to
the tricks the kernel module loader plays with percpu data:

# -mcmodel=medium breaks modules because it uses 32bit offsets from
# the TOC pointer to create pointers where possible. Pointers into the
# percpu data area are created by this method.
#
# The kernel module loader relocates the percpu data section from the
# original location (starting with 0xd...) to somewhere in the base
# kernel percpu data space (starting with 0xc...). We need a full
# 64bit relocation for this to work, hence -mcmodel=large.

On older kernels we fall back to the two level TOC (-mminimal-toc)

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Implement a POWER7 optimised memcpy using VMX and enhanced prefetch
instructions.

This is a copy of the POWER7 optimised copy_to_user/copy_from_user
loop. Detailed implementation and performance details can be found in
commit a66086b8197d (powerpc: POWER7 optimised
copy_to_user/copy_from_user using VMX).

I noticed memcpy issues when profiling a RAID6 workload:

	.memcpy
	.async_memcpy
	.async_copy_data
	.__raid_run_ops
	.handle_stripe
	.raid5d
	.md_thread

I created a simplified testcase by building a RAID6 array with 4 1GB
ramdisks (booting with brd.rd_size=1048576):

# mdadm -CR -e 1.2 /dev/md0 --level=6 -n4 /dev/ram[0-3]

I then timed how long it took to write to the entire array:

# dd if=/dev/zero of=/dev/md0 bs=1M

Before: 892 MB/s
After:  999 MB/s

A 12% improvement.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Implement a POWER7 optimised copy_page using VMX and enhanced
prefetch instructions. We use enhanced prefetch hints to prefetch
both the load and store side. We copy a cacheline at a time and
fall back to regular loads and stores if we are unable to use VMX
(eg we are in an interrupt).

The following microbenchmark was used to assess the impact of
the patch:

http://ozlabs.org/~anton/junkcode/page_fault_file.c

We test MAP_PRIVATE page faults across a 1GB file, 100 times:

# time ./page_fault_file -p -l 1G -i 100

Before: 22.25s
After:  18.89s

17% faster

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Subsequent patches will add more VMX library functions and it makes
sense to keep all the c-code helper functions in the one file.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

I noticed __clear_user high up in a profile of one of my RAID stress
tests. The testcase was doing a dd from /dev/zero which ends up
calling __clear_user.

__clear_user is basically a loop with a single 4 byte store which
is horribly slow. We can do much better by aligning the desination
and doing 32 bytes of 8 byte stores in a loop.

The following testcase was used to verify the patch:

http://ozlabs.org/~anton/junkcode/stress_clear_user.c

To show the improvement in performance I ran a dd from /dev/zero
to /dev/null on a POWER7 box:

Before:

# dd if=/dev/zero of=/dev/null bs=1M count=10000
10485760000 bytes (10 GB) copied, 3.72379 s, 2.8 GB/s

After:

# time dd if=/dev/zero of=/dev/null bs=1M count=10000
10485760000 bytes (10 GB) copied, 0.728318 s, 14.4 GB/s

Over 5x faster.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Implement a POWER7 optimised copy_to_user/copy_from_user using VMX.
For large aligned copies this new loop is over 10% faster, and for
large unaligned copies it is over 200% faster.

If we take a fault we fall back to the old version, this keeps
things relatively simple and easy to verify.

On POWER7 unaligned stores rarely slow down - they only flush when
a store crosses a 4KB page boundary. Furthermore this flush is
handled completely in hardware and should be 20-30 cycles.

Unaligned loads on the other hand flush much more often - whenever
crossing a 128 byte cache line, or a 32 byte sector if either sector
is an L1 miss.

Considering this information we really want to get the loads aligned
and not worry about the alignment of the stores. Microbenchmarks
confirm that this approach is much faster than the current unaligned
copy loop that uses shifts and rotates to ensure both loads and
stores are aligned.

We also want to try and do the stores in cacheline aligned, cacheline
sized chunks. If the store queue is unable to merge an entire
cacheline of stores then the L2 cache will have to do a
read/modify/write. Even worse, we will serialise this with the stores
in the next iteration of the copy loop since both iterations hit
the same cacheline.

Based on this, the new loop does the following things:

1 - 127 bytes
Get the source 8 byte aligned and use 8 byte loads and stores. Pretty
boring and similar to how the current loop works.

128 - 4095 bytes
Get the source 8 byte aligned and use 8 byte loads and stores,
1 cacheline at a time. We aren't doing the stores in cacheline
aligned chunks so we will potentially serialise once per cacheline.
Even so it is much better than the loop we have today.

4096 - bytes
If both source and destination have the same alignment get them both
16 byte aligned, then get the destination cacheline aligned. Do
cacheline sized loads and stores using VMX.

If source and destination do not have the same alignment, we get the
destination cacheline aligned, and use permute to do aligned loads.

In both cases the VMX loop should be optimal - we always do aligned
loads and stores and are always doing stores in cacheline aligned,
cacheline sized chunks.

To be able to use VMX we must be careful about interrupts and
sleeping. We don't use the VMX loop when in an interrupt (which should
be rare anyway) and we wrap the VMX loop in disable/enable_pagefault
and fall back to the existing copy_tofrom_user loop if we do need to
sleep.

The VMX breakpoint of 4096 bytes was chosen using this microbenchmark:

http://ozlabs.org/~anton/junkcode/copy_to_user.c

Since we are using VMX and there is a cost to saving and restoring
the user VMX state there are two broad cases we need to benchmark:

- Best case - userspace never uses VMX

- Worst case - userspace always uses VMX

In reality a userspace process will sit somewhere between these two
extremes. Since we need to test both aligned and unaligned copies we
end up with 4 combinations. The point at which the VMX loop begins to
win is:

0% VMX
aligned		2048 bytes
unaligned	2048 bytes

100% VMX
aligned		16384 bytes
unaligned	8192 bytes

Considering this is a microbenchmark, the data is hot in cache and
the VMX loop has better store queue merging properties we set the
breakpoint to 4096 bytes, a little below the unaligned breakpoints.

Some future optimisations we can look at:

- Looking at the perf data, a significant part of the cost when a
  task is always using VMX is the extra exception we take to restore
  the VMX state. As such we should do something similar to the x86
  optimisation that restores FPU state for heavy users. ie:

        /*
         * If the task has used fpu the last 5 timeslices, just do a full
         * restore of the math state immediately to avoid the trap; the
         * chances of needing FPU soon are obviously high now
         */
        preload_fpu = tsk_used_math(next_p) && next_p->fpu_counter > 5;

  and

        /*
         * fpu_counter contains the number of consecutive context switches
         * that the FPU is used. If this is over a threshold, the lazy fpu
         * saving becomes unlazy to save the trap. This is an unsigned char
         * so that after 256 times the counter wraps and the behavior turns
         * lazy again; this to deal with bursty apps that only use FPU for
         * a short time
         */

- We could create a paca bit to mirror the VMX enabled MSR bit and check
  that first, avoiding multiple calls to calling enable_kernel_altivec.
  That should help with iovec based system calls like readv.

- We could have two VMX breakpoints, one for when we know the user VMX
  state is loaded into the registers and one when it isn't. This could
  be a second bit in the paca so we can calculate the break points quickly.

- One suggestion from Ben was to save and restore the VSX registers
  we use inline instead of using enable_kernel_altivec.

[BenH: Fixed a problem with preempt and fixed build without CONFIG_ALTIVEC]

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

POWER5 added popcntb, and POWER7 added popcntw and popcntd. As a first step
this patch does all the work out of line, but it would be nice to implement
them as inlines with an out of line fallback.

The performance issue with hweight was noticed when disabling SMT on a large
(192 thread) POWER7 box. The patch improves that testcase by about 8%.

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

Replace EXTRA_CFLAGS with ccflags-y and EXTRA_AFLAGS with asflags-y.

Signed-off-by: matt mooney <mfm@muteddisk.com>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>

We use the same core loop as the new csum_partial, adding in the
stores and exception handling code. To keep things simple we do all the
exception fixup in csum_and_copy_from_user. This wrapper function is
modelled on the generic checksum code and is careful to always calculate
a complete checksum even if we only copied part of the data to userspace.

To test this I forced checksumming on over loopback and ran socklib (a
simple TCP benchmark). On a POWER6 575 throughput improved by 19% with
this patch. If I forced both the sender and receiver onto the same cpu
(with the hope of shifting the benchmark from being cache bandwidth limited
to cpu limited), adding this patch improved performance by 55%

Signed-off-by: Anton Blanchard <anton@samba.org>
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>