[v3,28/38] arm64: Avoid cpus_have_const_cap() for ARM64_SPECTRE_V2

Commit Message

Mark Rutland Oct. 10, 2023, 10:31 a.m. UTC

In arm64_apply_bp_hardening() we use cpus_have_const_cap() to check for
ARM64_SPECTRE_V2 , but 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 cpus_have_const_cap() check in arm64_apply_bp_hardening() is
intended to avoid the overhead of looking up and invoking a per-cpu
function pointer when no branch predictor hardening is required. The
arm64_apply_bp_hardening() function itself is called in two distinct
flows:

1) When handling certain exceptions taken from EL0, where the PC could
   be a TTBR1 address and hence might have trained a branch predictor.

   As cpucaps are detected and alternatives are patched long before it
   is possible to execute userspace, it is not necessary to use
   cpus_have_const_cap() for these cases, and cpus_have_final_cap() or
   alternative_has_cap() would be preferable.

2) When switching between tasks in check_and_switch_context().

   This can be called before cpucaps are detected and alternatives are
   patched, but this is long before the kernel mounts filesystems or
   accepts any input. At this stage the kernel hasn't loaded any secrets
   and there is no potential for hostile branch predictor training. Once
   cpucaps have been finalized and alternatives have been patched,
   switching tasks will invalidate any prior predictions. Hence it is
   not necessary to use cpus_have_const_cap() for this case.

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.

Signed-off-by: Mark Rutland <[e-mail]>
Cc: Catalin Marinas <[e-mail]>
Cc: Suzuki K Poulose <[e-mail]>
Cc: Will Deacon <[e-mail]>
---
 arch/arm64/include/asm/spectre.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Patch

diff --git a/arch/arm64/include/asm/spectre.h b/arch/arm64/include/asm/spectre.h
index 9cc501450486d..06c357d83b138 100644
--- a/arch/arm64/include/asm/spectre.h
+++ b/arch/arm64/include/asm/spectre.h
@@ -73,7 +73,7 @@  static __always_inline void arm64_apply_bp_hardening(void)
 {
 	struct bp_hardening_data *d;
 
-	if (!cpus_have_const_cap(ARM64_SPECTRE_V2))
+	if (!alternative_has_cap_unlikely(ARM64_SPECTRE_V2))
 		return;
 
 	d = this_cpu_ptr(&bp_hardening_data);