UPSTREAM: kcsan: Avoid READ_ONCE() in read_instrumented_memory()

Haibo Li reported:

 | Unable to handle kernel paging request at virtual address
 |   ffffff802a0d8d7171
 | Mem abort info:o:
 |   ESR = 0x9600002121
 |   EC = 0x25: DABT (current EL), IL = 32 bitsts
 |   SET = 0, FnV = 0 0
 |   EA = 0, S1PTW = 0 0
 |   FSC = 0x21: alignment fault
 | Data abort info:o:
 |   ISV = 0, ISS = 0x0000002121
 |   CM = 0, WnR = 0 0
 | swapper pgtable: 4k pages, 39-bit VAs, pgdp=000000002835200000
 | [ffffff802a0d8d71] pgd=180000005fbf9003, p4d=180000005fbf9003,
 | pud=180000005fbf9003, pmd=180000005fbe8003, pte=006800002a0d8707
 | Internal error: Oops: 96000021 [#1] PREEMPT SMP
 | Modules linked in:
 | CPU: 2 PID: 45 Comm: kworker/u8:2 Not tainted
 |   5.15.78-android13-8-g63561175bbda-dirty #1
 | ...
 | pc : kcsan_setup_watchpoint+0x26c/0x6bc
 | lr : kcsan_setup_watchpoint+0x88/0x6bc
 | sp : ffffffc00ab4b7f0
 | x29: ffffffc00ab4b800 x28: ffffff80294fe588 x27: 0000000000000001
 | x26: 0000000000000019 x25: 0000000000000001 x24: ffffff80294fdb80
 | x23: 0000000000000000 x22: ffffffc00a70fb68 x21: ffffff802a0d8d71
 | x20: 0000000000000002 x19: 0000000000000000 x18: ffffffc00a9bd060
 | x17: 0000000000000001 x16: 0000000000000000 x15: ffffffc00a59f000
 | x14: 0000000000000001 x13: 0000000000000000 x12: ffffffc00a70faa0
 | x11: 00000000aaaaaaab x10: 0000000000000054 x9 : ffffffc00839adf8
 | x8 : ffffffc009b4cf00 x7 : 0000000000000000 x6 : 0000000000000007
 | x5 : 0000000000000000 x4 : 0000000000000000 x3 : ffffffc00a70fb70
 | x2 : 0005ff802a0d8d71 x1 : 0000000000000000 x0 : 0000000000000000
 | Call trace:
 |  kcsan_setup_watchpoint+0x26c/0x6bc
 |  __tsan_read2+0x1f0/0x234
 |  inflate_fast+0x498/0x750
 |  zlib_inflate+0x1304/0x2384
 |  __gunzip+0x3a0/0x45c
 |  gunzip+0x20/0x30
 |  unpack_to_rootfs+0x2a8/0x3fc
 |  do_populate_rootfs+0xe8/0x11c
 |  async_run_entry_fn+0x58/0x1bc
 |  process_one_work+0x3ec/0x738
 |  worker_thread+0x4c4/0x838
 |  kthread+0x20c/0x258
 |  ret_from_fork+0x10/0x20
 | Code: b8bfc2a8 2a0803f7 14000007 d503249f (78bfc2a8) )
 | ---[ end trace 613a943cb0a572b6 ]-----

The reason for this is that on certain arm64 configuration since
e35123d ("arm64: lto: Strengthen READ_ONCE() to acquire when
CONFIG_LTO=y"), READ_ONCE() may be promoted to a full atomic acquire
instruction which cannot be used on unaligned addresses.

Fix it by avoiding READ_ONCE() in read_instrumented_memory(), and simply
forcing the compiler to do the required access by casting to the
appropriate volatile type. In terms of generated code this currently
only affects architectures that do not use the default READ_ONCE()
implementation.

The only downside is that we are not guaranteed atomicity of the access
itself, although on most architectures a plain load up to machine word
size should still be atomic (a fact the default READ_ONCE() still relies
on itself).

BUG: 285794521
(cherry picked from commit 8dec88070d964bfeb4198f34cb5956d89dd1f557)
Reported-by: Haibo Li <[email protected]>
Tested-by: Haibo Li <[email protected]>
Cc: <[email protected]> # 5.17+
Change-Id: I16c9f83c3b4e28021a936249cafb1501760aa59d
Signed-off-by: Marco Elver <[email protected]>
Signed-off-by: Paul E. McKenney <[email protected]>
Signed-off-by: 杨辉 <[email protected]>

kernel/kcsan/core.c

@@ -337,11 +337,20 @@ static void delay_access(int type)
  */
 static __always_inline u64 read_instrumented_memory(const volatile void *ptr, size_t size)
 {
+	/*
+	 * In the below we don't necessarily need the read of the location to
+	 * be atomic, and we don't use READ_ONCE(), since all we need for race
+	 * detection is to observe 2 different values.
+	 *
+	 * Furthermore, on certain architectures (such as arm64), READ_ONCE()
+	 * may turn into more complex instructions than a plain load that cannot
+	 * do unaligned accesses.
+	 */
 	switch (size) {
-	case 1:  return READ_ONCE(*(const u8 *)ptr);
-	case 2:  return READ_ONCE(*(const u16 *)ptr);
-	case 4:  return READ_ONCE(*(const u32 *)ptr);
-	case 8:  return READ_ONCE(*(const u64 *)ptr);
+	case 1:  return *(const volatile u8 *)ptr;
+	case 2:  return *(const volatile u16 *)ptr;
+	case 4:  return *(const volatile u32 *)ptr;
+	case 8:  return *(const volatile u64 *)ptr;
 	default: return 0; /* Ignore; we do not diff the values. */
 	}
 }