Restrict maximum size for seed corpus in fuzzer archives.

R=mmoroz@chromium.org,metzman@chromium.org
TBR=maxmorin@chromium.org

Change-Id: I3929bc2a22537cc0f83aec69693e99b8ae10223a
Bug: 952036
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1565241
Commit-Queue: Abhishek Arya <inferno@chromium.org>
Reviewed-by: Max Moroz <mmoroz@chromium.org>
Cr-Commit-Position: refs/heads/master@{#650389}

testing/libfuzzer/archive_corpus.py

@@ -16,6 +16,8 @@ import sys
 import warnings
 import zipfile
 
+SEED_CORPUS_LIMIT_MB = 100
+
 
 def main():
   parser = argparse.ArgumentParser(description="Generate fuzzer config.")
@@ -26,6 +28,7 @@ def main():
   args = parser.parse_args()
 
   corpus_files = []
+  seed_corpus_path = args.output
 
   for directory in args.corpus_directories:
     if not os.path.exists(directory):
@@ -36,7 +39,7 @@ def main():
         full_filename = os.path.join(dirpath, filename)
         corpus_files.append(full_filename)
 
-  with zipfile.ZipFile(args.output, 'w') as z:
+  with zipfile.ZipFile(seed_corpus_path, 'w') as z:
     # Turn warnings into errors to interrupt the build: crbug.com/653920.
     with warnings.catch_warnings():
       warnings.simplefilter("error")
@@ -45,6 +48,10 @@ def main():
         arcname = '%016d' % i
         z.write(corpus_file, arcname)
 
+  if os.path.getsize(seed_corpus_path) > SEED_CORPUS_LIMIT_MB * 1024 * 1024:
+    print('Seed corpus %s exceeds maximum allowed size (%d MB).' %
+          (seed_corpus_path, SEED_CORPUS_LIMIT_MB))
+    sys.exit(-1)
 
 if __name__ == '__main__':
   main()

third_party/pffft/generate_seed_corpus.py

@@ -9,6 +9,7 @@ from __future__ import print_function
 from array import array
 import os
 import random
+import shutil
 import sys
 
 MAX_INPUT_SIZE = 5000  # < 1 MB so we don't blow up fuzzer build sizes.
@@ -42,9 +43,10 @@ def main():
     sys.exit(1)
 
   output_path = sys.argv[1]
-  # Create output directory if missing.
-  if not os.path.exists(output_path):
-    os.makedirs(output_path)
+  # Start with a clean output directory.
+  if os.path.exists(output_path):
+    shutil.rmtree(output_path)
+  os.makedirs(output_path)
 
   # List of valid input sizes.
   N = [n for n in range(MAX_INPUT_SIZE) if IsValidSize(n)]