For 32-bit ARM devices align method headers and code to 64-byte
boundaries.
For other architectures, move back to instruction set alignment.
Rename kJitCodeCacheAlignment to kJitCodeAccountingBytes since it's
now only used for accounting rather than alignment.
Bug: 132205399
Test: art/test.py --host --jit
Test: manual (described in bug)
Change-Id: I88f5f39381bf0331ce8540a929c6a68b3b5e0c75
diff --git a/runtime/jit/jit_code_cache.h b/runtime/jit/jit_code_cache.h
index 9683b48..8a9c010 100644
--- a/runtime/jit/jit_code_cache.h
+++ b/runtime/jit/jit_code_cache.h
@@ -74,7 +74,9 @@
class MarkCodeClosure;
-using CodeCacheBitmap = gc::accounting::MemoryRangeBitmap<kJitCodeAlignment>;
+// Type of bitmap used for tracking live functions in the JIT code cache for the purposes
+// of garbage collecting code.
+using CodeCacheBitmap = gc::accounting::MemoryRangeBitmap<kJitCodeAccountingBytes>;
class JitCodeCache {
public: