kernel_optimize_test/arch/sh/kernel
Christoph Lameter c473b2c6f6 sh: Replace __get_cpu_var uses
__get_cpu_var() is used for multiple purposes in the kernel source.  One
of them is address calculation via the form &__get_cpu_var(x).  This
calculates the address for the instance of the percpu variable of the
current processor based on an offset.

Other use cases are for storing and retrieving data from the current
processors percpu area.  __get_cpu_var() can be used as an lvalue when
writing data or on the right side of an assignment.

__get_cpu_var() is defined as :

#define __get_cpu_var(var) (*this_cpu_ptr(&(var)))

__get_cpu_var() always only does an address determination.  However, store
and retrieve operations could use a segment prefix (or global register on
other platforms) to avoid the address calculation.

this_cpu_write() and this_cpu_read() can directly take an offset into a
percpu area and use optimized assembly code to read and write per cpu
variables.

This patch converts __get_cpu_var into either an explicit address
calculation using this_cpu_ptr() or into a use of this_cpu operations that
use the offset.  Thereby address calculations are avoided and less
registers are used when code is generated.

At the end of the patch set all uses of __get_cpu_var have been removed so
the macro is removed too.

The patch set includes passes over all arches as well.  Once these
operations are used throughout then specialized macros can be defined in
non -x86 arches as well in order to optimize per cpu access by f.e.  using
a global register that may be set to the per cpu base.

Transformations done to __get_cpu_var()

1. Determine the address of the percpu instance of the current processor.

	DEFINE_PER_CPU(int, y);
	int *x = &__get_cpu_var(y);

    Converts to

	int *x = this_cpu_ptr(&y);

2. Same as  but this time an array structure is involved.

	DEFINE_PER_CPU(int, y[20]);
	int *x = __get_cpu_var(y);

    Converts to

	int *x = this_cpu_ptr(y);

3. Retrieve the content of the current processors instance of a per cpu
variable.

	DEFINE_PER_CPU(int, y);
	int x = __get_cpu_var(y)

   Converts to

	int x = __this_cpu_read(y);

4. Retrieve the content of a percpu struct

	DEFINE_PER_CPU(struct mystruct, y);
	struct mystruct x = __get_cpu_var(y);

   Converts to

	memcpy(&x, this_cpu_ptr(&y), sizeof(x));

5. Assignment to a per cpu variable

	DEFINE_PER_CPU(int, y)
	__get_cpu_var(y) = x;

   Converts to

	__this_cpu_write(y, x);

6. Increment/Decrement etc of a per cpu variable

	DEFINE_PER_CPU(int, y);
	__get_cpu_var(y)++

   Converts to

	__this_cpu_inc(y)

Signed-off-by: Christoph Lameter <cl@linux.com>
Tested-by: Geert Uytterhoeven <geert@linux-m68k.org> [compilation only]
Cc: Paul Mundt <lethal@linux-sh.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-06-04 16:53:52 -07:00
..
cpu sh: Switch to new style MTU2 device 2014-05-11 19:35:28 +09:00
vsyscall
.gitignore
asm-offsets.c
crash_dump.c
debugtraps.S
disassemble.c
dma-nommu.c
dumpstack.c sh: fix format string bug in stack tracer 2014-04-03 16:20:49 -07:00
dwarf.c arch/sh/kernel/dwarf.c: use rbtree postorder iteration helper instead of solution using repeated rb_erase() 2014-01-23 16:37:03 -08:00
entry-common.S sh: push extra copy of r0-r2 for syscall parameters 2014-04-03 16:20:52 -07:00
ftrace.c ftrace: Do not pass data to ftrace_dyn_arch_init 2014-03-07 10:06:14 -05:00
head_32.S
head_64.S
hw_breakpoint.c sh: Replace __get_cpu_var uses 2014-06-04 16:53:52 -07:00
idle.c sched/idle, SH: Remove redundant cpuidle_idle_call() 2014-02-11 09:58:26 +01:00
io_trapped.c Kconfig: rename HAS_IOPORT to HAS_IOPORT_MAP 2014-04-07 16:36:11 -07:00
io.c
iomap.c
ioport.c
irq_32.c
irq_64.c
irq.c sh: Use irq_set_affinity instead of homebrewn code 2014-03-04 17:37:55 +01:00
kdebugfs.c
kgdb.c arch/sh/kernel/kgdb.c: add missing #include <linux/sched.h> 2014-01-21 16:19:42 -08:00
kprobes.c sh: Replace __get_cpu_var uses 2014-06-04 16:53:52 -07:00
localtimer.c sh: Replace __get_cpu_var uses 2014-06-04 16:53:52 -07:00
machine_kexec.c
machvec.c
Makefile Kconfig: rename HAS_IOPORT to HAS_IOPORT_MAP 2014-04-07 16:36:11 -07:00
module.c
nmi_debug.c
perf_callchain.c
perf_event.c sh: Replace __get_cpu_var uses 2014-06-04 16:53:52 -07:00
process_32.c sh: move fpu_counter into ARCH specific thread_struct 2013-11-13 12:09:13 +09:00
process_64.c sh64: kernel: remove useless variable 'regs' 2013-11-13 12:08:59 +09:00
process.c sh: delete __cpuinit usage from all sh files 2013-07-14 19:36:53 -04:00
ptrace_32.c ptrace/sh: revert "hw_breakpoints: Fix racy access to ptrace breakpoints" 2013-07-09 10:33:26 -07:00
ptrace_64.c
ptrace.c
reboot.c
relocate_kernel.S
return_address.c
setup.c memblock: make memblock_set_node() support different memblock_type 2014-01-21 16:19:44 -08:00
sh_bios.c early_printk: consolidate random copies of identical code 2013-04-29 18:28:13 -07:00
sh_ksyms_32.c sh: add EXPORT_SYMBOL(min_low_pfn) and EXPORT_SYMBOL(max_low_pfn) to sh_ksyms_32.c 2014-01-02 14:40:30 -08:00
sh_ksyms_64.c
signal_32.c sh: push extra copy of r0-r2 for syscall parameters 2014-04-03 16:20:52 -07:00
signal_64.c
smp.c sh: Replace __get_cpu_var uses 2014-06-04 16:53:52 -07:00
stacktrace.c
swsusp.c
sys_sh.c
sys_sh32.c sh: push extra copy of r0-r2 for syscall parameters 2014-04-03 16:20:52 -07:00
syscalls_32.S
syscalls_64.S
time.c
topology.c
traps_32.c sh: don't pass saved userspace state to exception handlers 2014-04-03 16:20:52 -07:00
traps_64.c sh: delete __cpuinit usage from all sh files 2013-07-14 19:36:53 -04:00
traps.c
unwinder.c
vmlinux.lds.S