arm64: Avoid cpus_have_const_cap() for ARM64_HAS_CNP

In system_supports_cnp() we use cpus_have_const_cap() to check for
ARM64_HAS_CNP, but this is only necessary so that the cpu_enable_cnp()
callback can run prior to alternatives being patched, and otherwise this
is not necessary and alternative_has_cap_*() would be preferable.

For historical reasons, cpus_have_const_cap() is more complicated than
it needs to be. Before cpucaps are finalized, it will perform a bitmap
test of the system_cpucaps bitmap, and once cpucaps are finalized it
will use an alternative branch. This used to be necessary to handle some
race conditions in the window between cpucap detection and the
subsequent patching of alternatives and static branches, where different
branches could be out-of-sync with one another (or w.r.t. alternative
sequences). Now that we use alternative branches instead of static
branches, these are all patched atomically w.r.t. one another, and there
are only a handful of cases that need special care in the window between
cpucap detection and alternative patching.

Due to the above, it would be nice to remove cpus_have_const_cap(), and
migrate callers over to alternative_has_cap_*(), cpus_have_final_cap(),
or cpus_have_cap() depending on when their requirements. This will
remove redundant instructions and improve code generation, and will make
it easier to determine how each callsite will behave before, during, and
after alternative patching.

The cpu_enable_cnp() callback is run immediately after the ARM64_HAS_CNP
cpucap is detected system-wide under setup_system_capabilities(), prior
to alternatives being patched. During this window cpu_enable_cnp() uses
cpu_replace_ttbr1() to set the CNP bit for the swapper_pg_dir in TTBR1.
No other users of the ARM64_HAS_CNP cpucap need the up-to-date value
during this window:

* As KVM isn't initialized yet, kvm_get_vttbr() isn't reachable.

* As cpuidle isn't initialized yet, __cpu_suspend_exit() isn't
  reachable.

* At this point all CPUs are using the swapper_pg_dir with a reserved
  ASID in TTBR1, and the idmap_pg_dir in TTBR0, so neither
  check_and_switch_context() nor cpu_do_switch_mm() need to do anything
  special.

This patch replaces the use of cpus_have_const_cap() with
alternative_has_cap_unlikely(), which will avoid generating code to test
the system_cpucaps bitmap and should be better for all subsequent calls
at runtime. To allow cpu_enable_cnp() to function prior to alternatives
being patched, cpu_replace_ttbr1() is split into cpu_replace_ttbr1() and
cpu_enable_swapper_cnp(), with the former only used for early TTBR1
replacement, and the latter used by both cpu_enable_cnp() and
__cpu_suspend_exit().

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Vladimir Murzin <vladimir.murzin@arm.com>
Cc: Will Deacon <will@kernel.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>

arch/arm64/include/asm/cpufeature.h

@@ -801,7 +801,7 @@ static __always_inline bool system_supports_tpidr2(void)
 
 static __always_inline bool system_supports_cnp(void)
 {
-	return cpus_have_const_cap(ARM64_HAS_CNP);
+	return alternative_has_cap_unlikely(ARM64_HAS_CNP);
 }
 
 static inline bool system_supports_address_auth(void)

arch/arm64/include/asm/mmu_context.h

@@ -152,7 +152,7 @@ static inline void cpu_install_ttbr0(phys_addr_t ttbr0, unsigned long t0sz)
  * Atomically replaces the active TTBR1_EL1 PGD with a new VA-compatible PGD,
  * avoiding the possibility of conflicting TLB entries being allocated.
  */
-static inline void cpu_replace_ttbr1(pgd_t *pgdp, pgd_t *idmap)
+static inline void __cpu_replace_ttbr1(pgd_t *pgdp, pgd_t *idmap, bool cnp)
 {
 	typedef void (ttbr_replace_func)(phys_addr_t);
 	extern ttbr_replace_func idmap_cpu_replace_ttbr1;
@@ -162,17 +162,8 @@ static inline void cpu_replace_ttbr1(pgd_t *pgdp, pgd_t *idmap)
 	/* phys_to_ttbr() zeros lower 2 bits of ttbr with 52-bit PA */
 	phys_addr_t ttbr1 = phys_to_ttbr(virt_to_phys(pgdp));
 
-	if (system_supports_cnp() && !WARN_ON(pgdp != lm_alias(swapper_pg_dir))) {
-		/*
-		 * cpu_replace_ttbr1() is used when there's a boot CPU
-		 * up (i.e. cpufeature framework is not up yet) and
-		 * latter only when we enable CNP via cpufeature's
-		 * enable() callback.
-		 * Also we rely on the system_cpucaps bit being set before
-		 * calling the enable() function.
-		 */
+	if (cnp)
 		ttbr1 |= TTBR_CNP_BIT;
-	}
 
 	replace_phys = (void *)__pa_symbol(idmap_cpu_replace_ttbr1);
 
@@ -189,6 +180,21 @@ static inline void cpu_replace_ttbr1(pgd_t *pgdp, pgd_t *idmap)
 	cpu_uninstall_idmap();
 }
 
+static inline void cpu_enable_swapper_cnp(void)
+{
+	__cpu_replace_ttbr1(lm_alias(swapper_pg_dir), idmap_pg_dir, true);
+}
+
+static inline void cpu_replace_ttbr1(pgd_t *pgdp, pgd_t *idmap)
+{
+	/*
+	 * Only for early TTBR1 replacement before cpucaps are finalized and
+	 * before we've decided whether to use CNP.
+	 */
+	WARN_ON(system_capabilities_finalized());
+	__cpu_replace_ttbr1(pgdp, idmap, false);
+}
+
 /*
  * It would be nice to return ASIDs back to the allocator, but unfortunately
  * that introduces a race with a generation rollover where we could erroneously

arch/arm64/kernel/cpufeature.c

@@ -3458,7 +3458,7 @@ subsys_initcall_sync(init_32bit_el0_mask);
 
 static void __maybe_unused cpu_enable_cnp(struct arm64_cpu_capabilities const *cap)
 {
-	cpu_replace_ttbr1(lm_alias(swapper_pg_dir), idmap_pg_dir);
+	cpu_enable_swapper_cnp();
 }
 
 /*

arch/arm64/kernel/suspend.c

@@ -55,7 +55,7 @@ void notrace __cpu_suspend_exit(void)
 
 	/* Restore CnP bit in TTBR1_EL1 */
 	if (system_supports_cnp())
-		cpu_replace_ttbr1(lm_alias(swapper_pg_dir), idmap_pg_dir);
+		cpu_enable_swapper_cnp();
 
 	/*
 	 * PSTATE was not saved over suspend/resume, re-enable any detected