minimize.MaskGen Class Reference

Inheritance diagram for minimize.MaskGen:

Public Member Functions
	__init__ (self, nbits, mask)
	__iter__ (self)
	__next__ (self)
	update_tested (self, new_mask)
	update_mask (self, new_mask)
	count_bits (self)
	inc_var (self)
	stop (self)
	next (self)

Public Attributes
	nbits
	mask
	convolution_idx
	max
	nbitsVar
	filter_mask
	tested
	gen_new_mask

Detailed Description

Generates bitpatterns of <nbits> bits which indicate which
statements or expressions should be masked when passed as the
--smask or --emask parameter to dartfuzz.dart.
The patterns are generated in the following manner:
11111111111111111111111111111111
11111111111111110000000000000000
00000000000000001111111111111111
11111111111111111111111100000000
11111111111111110000000011111111
11111111000000001111111111111111
00000000111111111111111111111111
...
If the error persists with a given a pattern it is stored in the
parameter <mask>.

Definition at line 22 of file minimize.py.

Constructor & Destructor Documentation

__init__()

minimize.MaskGen.__init__	(		self,
			nbits,
			mask
	)

Definition at line 40 of file minimize.py.

    def __init__(self, nbits, mask):
        # Mask bitlength.
        self.nbits = nbits
        # Mask determining which statements or expressions to skip.
        self.mask = mask
        # Current mask modification start index.
        self.convolution_idx = 0
        # Max mask with all bits set to 1.
        self.max = (1 << (nbits + 1)) - 1
        # Current number of bits set in the mask modification mask.
        self.nbitsVar = int(nbits)
        # Current mask modification mask.
        # This will be folded over the current value of self.mask
        # starting from 0.
        self.filter_mask = (1 << self.nbitsVar) - 1
        # All the masks already tested.
        self.tested = set()
        self.gen_new_mask = False
 

Member Function Documentation

__iter__()

minimize.MaskGen.__iter__

(

self

)

Definition at line 59 of file minimize.py.

    def __iter__(self):
        return self
 

__next__()

minimize.MaskGen.__next__

(

self

)

Definition at line 62 of file minimize.py.

    def __next__(self):
        return self.next()
 

count_bits()

minimize.MaskGen.count_bits

(

self

)

Definition at line 83 of file minimize.py.

    def count_bits(self):
        return bin(self.mask).count('1')
 

inc_var()

minimize.MaskGen.inc_var

(

self

)

Definition at line 86 of file minimize.py.

    def inc_var(self):
        self.convolution_idx = 0
        self.nbitsVar = self.nbitsVar >> 1
        self.filter_mask = (1 << self.nbitsVar) - 1
        return self.filter_mask == 0
 

next()

minimize.MaskGen.next

(

self

)

Definition at line 105 of file minimize.py.

    def next(self):
        while not self.stop():
            # Generate a new mask according to the mask generation algorithm.
            new_mask = (self.mask |
                        (self.filter_mask << self.convolution_idx)) & self.max
            # Update the counter variable for the mask generation algorithm.
            self.convolution_idx += self.nbitsVar
            # If the new_mask has new bits and has not been tested yet.
            # We need to check both because only masks that produce an
            # error are merged into self.mask.
            if new_mask != self.mask and \
                new_mask not in self.tested:
                # Add new mask to the set of tested masks.
                self.tested.add(new_mask)
                # Mark that we generated a new mask on this run.
                self.gen_new_mask = True
                # Return the new mask to be tested.
                return new_mask
        raise StopIteration()
 
 
# Generate a Dart program for the given statement (smask) and
# expression (emask) masks. Dartfuzz parameters for seed, ffi and
# fp are static and therefore contained in the dartfuzz_cmd parameter.

stop()

minimize.MaskGen.stop

(

self

)

Definition at line 92 of file minimize.py.

    def stop(self):
        if self.convolution_idx >= self.nbits and self.inc_var() > 0:
            # Check if the last run generated a new mask.
            if self.gen_new_mask:
                # Restart mask generation from beginning.
                self.nbitsVar = self.nbits >> 1
                self.inc_var()
                self.gen_new_mask = False
                return False
            print("STOP")
            return True
        return False
 

update_mask()

minimize.MaskGen.update_mask	(		self,
			new_mask
	)

Definition at line 76 of file minimize.py.

    def update_mask(self, new_mask):
        new_mask = (self.mask | new_mask) & self.max
        self.mask = new_mask
 

update_tested()

minimize.MaskGen.update_tested	(		self,
			new_mask
	)

Definition at line 67 of file minimize.py.

    def update_tested(self, new_mask):
        self.tested.add(new_mask)
 

Member Data Documentation

convolution_idx

minimize.MaskGen.convolution_idx

Definition at line 46 of file minimize.py.

filter_mask

minimize.MaskGen.filter_mask

Definition at line 54 of file minimize.py.

gen_new_mask

minimize.MaskGen.gen_new_mask

Definition at line 57 of file minimize.py.

mask

minimize.MaskGen.mask

Definition at line 44 of file minimize.py.

max

minimize.MaskGen.max

Definition at line 48 of file minimize.py.

nbits

minimize.MaskGen.nbits

Definition at line 42 of file minimize.py.

nbitsVar

minimize.MaskGen.nbitsVar

Definition at line 50 of file minimize.py.

tested

minimize.MaskGen.tested

Definition at line 56 of file minimize.py.

---

The documentation for this class was generated from the following file:

• third_party/dart-lang/sdk/runtime/tools/dartfuzz/minimize.py