Inspired by x86 commit 864b4355("x86/jump_label: Mark arguments as
const to satisfy asm constraints"), constify alternative_has_feature_*
argument to satisfy asm constraints. And Steven in [1] also pointed
out that "The "i" constraint needs to be a constant."

Tested with building a simple external kernel module with "O0".

Before the patch, got similar gcc warnings and errors as below:

In file included from <command-line>:
In function ‘alternative_has_feature_likely’,
    inlined from ‘system_capabilities_finalized’ at
arch/arm64/include/asm/cpufeature.h:440:9,
    inlined from ‘arm64_preempt_schedule_irq’ at
arch/arm64/kernel/entry-common.c:264:6:
include/linux/compiler_types.h:285:33: warning:
‘asm’ operand 0 probably does not match constraints
  285 | #define asm_volatile_goto(x...) asm goto(x)
      |                                 ^~~
arch/arm64/include/asm/alternative-macros.h:232:9:
note: in expansion of macro ‘asm_volatile_goto’
  232 |         asm_volatile_goto(
      |         ^~~~~~~~~~~~~~~~~
include/linux/compiler_types.h:285:33: error:
impossible constraint in ‘asm’
  285 | #define asm_volatile_goto(x...) asm goto(x)
      |                                 ^~~
arch/arm64/include/asm/alternative-macros.h:232:9:
note: in expansion of macro ‘asm_volatile_goto’
  232 |         asm_volatile_goto(
      |         ^~~~~~~~~~~~~~~~~

After the patch, the simple external test kernel module is built fine
with "-O0".

[1]https://lore.kernel.org/all/20210212094059.5f8d05e8@gandalf.local.home/Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Link: https://lore.kernel.org/r/20221006075542.2658-3-jszhang@kernel.orgSigned-off-by: Will Deacon <will@kernel.org>