[sanitizer_common] Fix UnwindFast on SPARC (llvm#101634)

```
  UBSan-Standalone-sparc :: TestCases/Misc/Linux/diag-stacktrace.cpp
```
`FAIL`s on 32 and 64-bit Linux/sparc64 (and on Solaris/sparcv9, too: the
test isn't Linux-specific at all). With
`UBSAN_OPTIONS=fast_unwind_on_fatal=1`, the stack trace shows a
duplicate innermost frame:
```
compiler-rt/test/ubsan/TestCases/Misc/Linux/diag-stacktrace.cpp:14:31: runtime error: execution reached the end of a value-returning function without returning a value
    #0 0x7003a708 in f() compiler-rt/test/ubsan/TestCases/Misc/Linux/diag-stacktrace.cpp:14:35
    #1 0x7003a708 in f() compiler-rt/test/ubsan/TestCases/Misc/Linux/diag-stacktrace.cpp:14:35
    #2 0x7003a714 in g() compiler-rt/test/ubsan/TestCases/Misc/Linux/diag-stacktrace.cpp:17:38
```
which isn't seen with `fast_unwind_on_fatal=0`.

This turns out to be another fallout from fixing
`__builtin_return_address`/`__builtin_extract_return_addr` on SPARC. In
`sanitizer_stacktrace_sparc.cpp` (`BufferedStackTrace::UnwindFast`) the
`pc` arg is the return address, while `pc1` from the stack frame
(`fr_savpc`) is the address of the `call` insn, leading to a double
entry for the innermost frame in `trace_buffer[]`.

This patch fixes this by moving the adjustment before all uses.

Tested on `sparc64-unknown-linux-gnu` and `sparcv9-sun-solaris2.11`
(with the `ubsan/TestCases/Misc/Linux` tests enabled).

Author: rorth

compiler-rt/lib/sanitizer_common/sanitizer_stacktrace_sparc.cpp

@@ -58,17 +58,16 @@ void BufferedStackTrace::UnwindFast(uptr pc, uptr bp, uptr stack_top,
   // Avoid infinite loop when frame == frame[0] by using frame > prev_frame.
   while (IsValidFrame(bp, stack_top, bottom) && IsAligned(bp, sizeof(uhwptr)) &&
          size < max_depth) {
-    uhwptr pc1 = ((uhwptr *)bp)[15];
+    // %o7 contains the address of the call instruction and not the
+    // return address, so we need to compensate.
+    uhwptr pc1 = GetNextInstructionPc(((uhwptr *)bp)[15]);
     // Let's assume that any pointer in the 0th page is invalid and
     // stop unwinding here.  If we're adding support for a platform
     // where this isn't true, we need to reconsider this check.
     if (pc1 < kPageSize)
       break;
-    if (pc1 != pc) {
-      // %o7 contains the address of the call instruction and not the
-      // return address, so we need to compensate.
-      trace_buffer[size++] = GetNextInstructionPc((uptr)pc1);
-    }
+    if (pc1 != pc)
+      trace_buffer[size++] = pc1;
     bottom = bp;
     bp = (uptr)((uhwptr *)bp)[14] + STACK_BIAS;
   }