[PATCH] bitmap: Fix optimization of bitmap_set/clear for big-endian machines

[Michael Ellerman]

Paul Mackerras <paulus at ozlabs.org> writes:

> On Wed, Oct 25, 2017 at 07:39:48AM +0000, Matthew Wilcox wrote:
>> Hang on, don't tell me you found this by inspection. Are you not
>> running the bitmap testcase, enabled by CONFIG_TEST_BITMAP? Either
>> that should be producing an error, or there's a missing test case, or
>> your inspection is wrong ...

Oops.

  [    1.907533] test_bitmap: test 13: input is '0-2047:128/256' OK, Time: 1359
  [    1.907596] set not equal 0 8
  [    1.907640] set not __equal 0 8
  [    1.907684] clear not equal 0 8
  [    1.907728] clear not __equal 0 8
  [    1.907772] set not equal 0 16
  [    1.907816] set not __equal 0 16
  [    1.907861] clear not equal 0 16
  [    1.907905] clear not __equal 0 16
  [    1.907949] set not equal 0 24
  [    1.907993] set not __equal 0 24
  [    1.908038] clear not equal 0 24
  [    1.908082] clear not __equal 0 24
  ...
  [ snip ~30,000 lines ]
  ...
  [    3.345729] set not equal 1008 8
  [    3.345773] set not __equal 1008 8
  [    3.345818] clear not equal 1008 8
  [    3.345863] clear not __equal 1008 8

  [    3.345909] test_bitmap: all 460506 tests passed
                              ^^^^^^^^^^^^^^^^^^^^^^^
                              O_o

With the patch:

  [    1.904393] test_bitmap: test 13: input is '0-2047:128/256' OK, Time: 1359
  [    1.916270] test_bitmap: all 460506 tests passed

cheers