scripts: Added -w/--word-bits to bound dbgleb128/dbgle32 parsing

This is limited to dbgle32.py, dbgleb128.py, and dbgtag.py for now.

This more closely matches how littlefs behaves, in that we read a
bounded number of bytes before leb128 decoding. This minimizes bugs
related to leb128 overflow and avoids reading inherently undecodable
data.

The previous unbounded behavior is still available with -w0.

Note this gives dbgle32.py much more flexibility in that it can now
decode other integer widths. Uh, ignore the name for now. At least it's
self documenting that the default is 32-bits...

---

Also fixed a bug in fromleb128 where size was reported incorrectly on
offset + truncated leb128.

scripts/dbgbmap.py

@@ -213,7 +213,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgbmapd3.py

@@ -243,7 +243,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgbtree.py

@@ -151,7 +151,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgle32.py

@@ -5,6 +5,7 @@ if __name__ == "__main__":
     __import__('sys').path.pop(0)
 
 import io
+import math as mt
 import os
 import struct
 import sys
@@ -21,18 +22,27 @@ def openio(path, mode='r', buffering=-1):
     else:
         return open(path, mode, buffering)
 
-def fromle32(data, j=0):
+def dbg_le32s(data, *,
-    return struct.unpack('<I', data[j:j+4].ljust(4, b'\0'))[0]
+        word_bits=32):
+    # figure out le32 size in bytes
+    if word_bits != 0:
+        n = mt.ceil(word_bits / 8)
 
-def dbg_le32s(data):
+    # parse le32s, or le<whatevers>
     lines = []
     j = 0
     while j < len(data):
-        word = fromle32(data, j)
+        word = 0
+        d = 0
+        while (j+d < len(data)
+                and (d < n if word_bits != 0 else True)):
+            word |= data[j+d] << d
+            d += 1
+
         lines.append((
-                ' '.join('%02x' % b for b in data[j:j+4]),
+                ' '.join('%02x' % b for b in data[j:j+d]),
                 word))
-        j += 4
+        j += d
 
     # figure out widths
     w = [0]
@@ -47,18 +57,21 @@ def dbg_le32s(data):
 
 def main(le32s, *,
         hex=False,
-        input=None):
+        input=None,
+        word_bits=32):
     hex_ = hex; del hex
 
     # interpret as a sequence of hex bytes
     if hex_:
         bytes_ = [b for le32 in le32s for b in le32.split()]
-        dbg_le32s(bytes(int(b, 16) for b in bytes_))
+        dbg_le32s(bytes(int(b, 16) for b in bytes_),
+                word_bits=word_bits)
 
     # parse le32s in a file
     elif input:
         with openio(input, 'rb') as f:
-            dbg_le32s(f.read())
+            dbg_le32s(f.read(),
+                    word_bits=word_bits)
 
     # we don't currently have a default interpretation
     else:
@@ -84,6 +97,12 @@ if __name__ == "__main__":
     parser.add_argument(
             '-i', '--input',
             help="Read le32s from this file. Can use - for stdin.")
+    parser.add_argument(
+            '-w', '--word-bits',
+            nargs='?',
+            type=lambda x: int(x, 0),
+            const=0,
+            help="Word size in bits. 0 is unbounded. Defaults to 32.")
     sys.exit(main(**{k: v
             for k, v in vars(parser.parse_intermixed_args()).items()
             if v is not None}))

scripts/dbgleb128.py

@@ -5,6 +5,7 @@ if __name__ == "__main__":
     __import__('sys').path.pop(0)
 
 import io
+import math as mt
 import os
 import struct
 import sys
@@ -31,13 +32,25 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
-def dbg_leb128s(data):
+def dbg_leb128s(data, *,
+        word_bits=32):
+    # figure out leb128 size in bytes
+    if word_bits != 0:
+        n = mt.ceil(word_bits / 7)
+
+    # parse leb128s
     lines = []
     j = 0
     while j < len(data):
+        # bounded leb128s?
+        if word_bits != 0:
+            word, d = fromleb128(data[j:j+n])
+        # unbounded?
+        else:
             word, d = fromleb128(data, j)
+
         lines.append((
                 ' '.join('%02x' % b for b in data[j:j+d]),
                 word))
@@ -56,18 +69,21 @@ def dbg_leb128s(data):
 
 def main(leb128s, *,
         hex=False,
-        input=None):
+        input=None,
+        word_bits=32):
     hex_ = hex; del hex
 
     # interpret as a sequence of hex bytes
     if hex_:
         bytes_ = [b for leb128 in leb128s for b in leb128.split()]
-        dbg_leb128s(bytes(int(b, 16) for b in bytes_))
+        dbg_leb128s(bytes(int(b, 16) for b in bytes_),
+                word_bits=word_bits)
 
     # parse leb128s in a file
     elif input:
         with openio(input, 'rb') as f:
-            dbg_leb128s(f.read())
+            dbg_leb128s(f.read(),
+                    word_bits=word_bits)
 
     # we don't currently have a default interpretation
     else:
@@ -93,6 +109,12 @@ if __name__ == "__main__":
     parser.add_argument(
             '-i', '--input',
             help="Read leb128s from this file. Can use - for stdin.")
+    parser.add_argument(
+            '-w', '--word-bits',
+            nargs='?',
+            type=lambda x: int(x, 0),
+            const=0,
+            help="Word size in bits. 0 is unbounded. Defaults to 32.")
     sys.exit(main(**{k: v
             for k, v in vars(parser.parse_intermixed_args()).items()
             if v is not None}))

scripts/dbglfs.py

@@ -170,7 +170,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgmtree.py

@@ -151,7 +151,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgrbyd.py

@@ -161,7 +161,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0

scripts/dbgtag.py

@@ -5,6 +5,7 @@ if __name__ == "__main__":
     __import__('sys').path.pop(0)
 
 import io
+import math as mt
 import os
 import struct
 import sys
@@ -74,7 +75,7 @@ def fromleb128(data, j=0):
         if not b & 0x80:
             return word, d+1
         d += 1
-    return word, len(data)
+    return word, d
 
 def fromtag(data, j=0):
     d = 0
@@ -234,7 +235,12 @@ def list_tags():
                 w[0], 'LFSR_'+n,
                 c))
 
-def dbg_tags(data):
+def dbg_tags(data, *,
+        word_bits=32):
+    # figure out tag size in bytes
+    if word_bits != 0:
+        n = 2 + 2*mt.ceil(word_bits / 7)
+
     lines = []
     # interpret as ints?
     if not isinstance(data, bytes):
@@ -247,7 +253,13 @@ def dbg_tags(data):
     else:
         j = 0
         while j < len(data):
+            # bounded tags?
+            if word_bits != 0:
+                v, tag, w, size, d = fromtag(data[j:j+n])
+            # unbounded?
+            else:
                 v, tag, w, size, d = fromtag(data, j)
+
             lines.append((
                     ' '.join('%02x' % b for b in data[j:j+d]),
                     tagrepr(tag, w, size)))
@@ -271,7 +283,8 @@ def dbg_tags(data):
 def main(tags, *,
         list=False,
         hex=False,
-        input=None):
+        input=None,
+        word_bits=32):
     list_ = list; del list
     hex_ = hex; del hex
 
@@ -282,16 +295,19 @@ def main(tags, *,
     # interpret as a sequence of hex bytes
     elif hex_:
         bytes_ = [b for tag in tags for b in tag.split()]
-        dbg_tags(bytes(int(b, 16) for b in bytes_))
+        dbg_tags(bytes(int(b, 16) for b in bytes_),
+                word_bits=word_bits)
 
     # parse tags in a file
     elif input:
         with openio(input, 'rb') as f:
-            dbg_tags(f.read())
+            dbg_tags(f.read(),
+                    word_bits=word_bits)
 
     # default to interpreting as ints
     else:
-        dbg_tags(int(tag, 0) for tag in tags)
+        dbg_tags((int(tag, 0) for tag in tags),
+                word_bits=word_bits)
 
 
 if __name__ == "__main__":
@@ -315,6 +331,12 @@ if __name__ == "__main__":
     parser.add_argument(
             '-i', '--input',
             help="Read tags from this file. Can use - for stdin.")
+    parser.add_argument(
+            '-w', '--word-bits',
+            nargs='?',
+            type=lambda x: int(x, 0),
+            const=0,
+            help="Word size in bits. 0 is unbounded. Defaults to 32.")
     sys.exit(main(**{k: v
             for k, v in vars(parser.parse_intermixed_args()).items()
             if v is not None}))