[Sanitizers] Fix lint failure with Python 3.6

[nemanjai]

Summary

There are some files in compiler-rt that use UTF-8 characters in some of the comments. This causes lint failures with some versions of Python. This patch just makes the encoding implicit in the call to open.

[nemanjai]

Sample of the build bot failure this resolves:
http://lab.llvm.org:8011/builders/clang-ppc64le-rhel/builds/6122/steps/ninja%20check%201/logs/stdio

[NeHuang]

LGTM.

[nemanjai]

I forgot to add the differential revision to the commit message.
Committed in https://reviews.llvm.org/rGacb3b8dce1cd

compiler-rt/lib/sanitizer_common/scripts/litlint.py

def LintFile(p):
 
   Args:
     p: str, valid path to a file
 
   Returns:
     The number of errors detected.
   """
   errs = 0
-  with open(p, 'r') as f:
+  with open(p, 'r', encoding='utf-8') as f:
     for i, s in enumerate(f.readlines(), start=1):
       msg, col = LintLine(s)
       if msg != None:
         errs += 1
         errorMsg = 'litlint: {}:{}:{}: error: {}.\n{}{}\n'
         arrow = (col-1) * ' ' + '^'
         sys.stderr.write(errorMsg.format(p, i, col, msg, s, arrow))
   return errs