skia2/fuzz/fuzz.cpp
scroggo 19b91531e9 Add support for multiple frames in SkCodec
Add an interface to decode frames beyond the first in SkCodec, and
add an implementation for SkGifCodec.

Add getFrameData to SkCodec. This method reads ahead in the stream
to return a vector containing meta data about each frame in the image.
This is not required in order to decode frames beyond the first, but
it allows a client to learn extra information:
- how long the frame should be displayed
- whether a frame should be blended with a prior frame, allowing the
  client to provide the prior frame to speed up decoding

Add a new fields to SkCodec::Options:
- fFrameIndex
- fHasPriorFrame

The API is designed so that SkCodec never caches frames. If a
client wants a frame beyond the first, they specify the frame in
Options.fFrameIndex. If the client does not have the
frame's required frame (the frame that this frame must be blended on
top of) cached, they pass false for
Options.fHasPriorFrame. Unless the frame is
independent, the codec will then recursively decode all frames
necessary to decode fFrameIndex. If the client has the required frame
cached, they can put it in the dst they pass to the codec, and the
codec will only draw fFrameIndex onto it.

Replace SkGifCodec's scanline decoding support with progressive
decoding, and update the tests accordingly.

Implement new APIs in SkGifCodec. Instead of using gif_lib, use
GIFImageReader, imported from Chromium (along with its copyright
headers) with the following changes:
- SkGifCodec is now the client
- Replace blink types
- Combine GIFColorMap::buildTable and ::getTable into a method that
  creates and returns an SkColorTable
- Input comes from an SkStream, instead of a SegmentReader. Add
  SkStreamBuffer, which buffers the (potentially partial) stream in
  order to decode progressively.
  (FIXME: This requires copying data that previously was read directly
  from the SegmentReader. Does this hurt performance? If so, can we
  fix it?)
- Remove UMA code
- Instead of reporting screen width and height to the client, allow the
  client to query for it
- Fail earlier if the first frame AND screen have size of zero
- Compute required previous frame when adding a new one
- Move GIFParseQuery from GIFImageDecoder to GIFImageReader
- Allow parsing up to a specific frame (to skip parsing the rest of the
  stream if a client only wants the first frame)
- Compute whether the first frame has alpha and supports index 8, to
  create the SkImageInfo. This happens before reporting that the size
  has been decoded.

Add GIFImageDecoder::haveDecodedRow to SkGifCodec, imported from
Chromium (along with its copyright header), with the following changes:
- Add support for sampling
- Use the swizzler
- Keep track of the rows decoded
- Do *not* keep track of whether we've seen alpha

Remove SkCodec::kOutOfOrder_SkScanlineOrder, which was only used by GIF
scanline decoding.

Call onRewind even if there is no stream (SkGifCodec needs to clear its
decoded state so it will decode from the beginning).

Add a method to SkSwizzler to access the offset into the dst, taking
subsetting into account.

Add a GM that animates a GIF.
Add tests for the new APIs.

*** Behavior changes:
* Previously, we reported that an image with a subset frame and no transparent
index was opaque and used the background index (if present) to fill the
background. This is necessary in order to support index 8, but it does not
match viewers/browsers I have seen. Examples:
- Chromium and Gimp render the background transparent
- Firefox, Safari, Linux Image Viewer, Safari Preview clip to the frame (for
  a single frame image)
This CL matches Chromium's behavior and renders the background transparent.
This allows us to have consistent behavior across products and simplifies
the code (relative to what we would have to do to continue the old behavior
on Android). It also means that we will no longer support index 8 for some
GIFs.
* Stop checking for GIFSTAMP - all GIFs should be either 89a or 87a.
This matches Chromium. I suspect that bugs would have been reported if valid
GIFs started with "GIFVER" instead of "GIF89a" or "GIF87a" (but did not decode
in Chromium).

*** Future work not included in this CL:
* Move some checks out of haveDecodedRow, since they are the same for the
  entire frame e.g.
- intersecting the frameRect with the full image size
- whether there is a color table
* Change when we write transparent pixels
- In some cases, Chromium deemed this unnecessary, but I suspect it is slower
  than the fallback case. There will continue to be cases where we should
  *not* write them, but for e.g. the first pass where we have already
  cleared to transparent (which we may also be able to skip) writing the
  transparent pixels will not make anything incorrect.
* Report color type and alpha type per frame
- Depending on alpha values, disposal methods, frame rects, etc, subsequent
  frames may have different properties than the first.
* Skip copies of the encoded data
- We copy the encoded data in case the stream is one that cannot be rewound,
  so we can parse and then decode (possibly not immediately). For some input
  streams, this is unnecessary.
  - I was concerned this cause a performance regression, but on average the
    new code is faster than the old for the images I tested [1].
  - It may cause a performance regression for Chromium, though, where we can
    always move back in the stream, so this should be addressed.

Design doc:
https://docs.google.com/a/google.com/document/d/12Qhf9T92MWfdWujQwCIjhCO3sw6pTJB5pJBwDM1T7Kc/

[1] https://docs.google.com/a/google.com/spreadsheets/d/19V-t9BfbFw5eiwBTKA1qOBkZbchjlTC5EIz6HFy-6RI/

GOLD_TRYBOT_URL= https://gold.skia.org/search2?unt=true&query=source_type%3Dgm&master=false&issue=2045293002

Review-Url: https://codereview.chromium.org/2045293002
2016-10-24 09:03:26 -07:00

470 lines
18 KiB
C++

/*
* Copyright 2016 Google Inc.
*
* Use of this source code is governed by a BSD-style license that can be
* found in the LICENSE file.
*/
#include "Fuzz.h"
#include "SkCanvas.h"
#include "SkCodec.h"
#include "SkCommandLineFlags.h"
#include "SkData.h"
#include "SkImage.h"
#include "SkImageEncoder.h"
#include "SkMallocPixelRef.h"
#include "SkPicture.h"
#include "SkPicture.h"
#include "SkPicture.h"
#include "SkSLCompiler.h"
#include "SkStream.h"
#include <cmath>
#include <signal.h>
#include <stdlib.h>
DEFINE_string2(bytes, b, "", "A path to a file. This can be the fuzz bytes or a binary to parse.");
DEFINE_string2(name, n, "", "If --type is 'api', fuzz the API with this name.");
DEFINE_string2(type, t, "api", "How to interpret --bytes, either 'image_scale', 'image_mode', 'skp', 'icc', or 'api'.");
DEFINE_string2(dump, d, "", "If not empty, dump 'image*' or 'skp' types as a PNG with this name.");
static int printUsage(const char* name) {
SkDebugf("Usage: %s -t <type> -b <path/to/file> [-n api-to-fuzz]\n", name);
return 1;
}
static uint8_t calculate_option(SkData*);
static int fuzz_api(sk_sp<SkData>);
static int fuzz_img(sk_sp<SkData>, uint8_t, uint8_t);
static int fuzz_skp(sk_sp<SkData>);
static int fuzz_icc(sk_sp<SkData>);
static int fuzz_color_deserialize(sk_sp<SkData>);
static int fuzz_sksl2glsl(sk_sp<SkData>);
int main(int argc, char** argv) {
SkCommandLineFlags::Parse(argc, argv);
const char* path = FLAGS_bytes.isEmpty() ? argv[0] : FLAGS_bytes[0];
sk_sp<SkData> bytes(SkData::MakeFromFileName(path));
if (!bytes) {
SkDebugf("Could not read %s\n", path);
return 2;
}
uint8_t option = calculate_option(bytes.get());
if (!FLAGS_type.isEmpty()) {
if (0 == strcmp("api", FLAGS_type[0])) {
return fuzz_api(bytes);
}
if (0 == strcmp("color_deserialize", FLAGS_type[0])) {
return fuzz_color_deserialize(bytes);
}
if (0 == strcmp("icc", FLAGS_type[0])) {
return fuzz_icc(bytes);
}
if (0 == strcmp("image_scale", FLAGS_type[0])) {
return fuzz_img(bytes, option, 0);
}
if (0 == strcmp("image_mode", FLAGS_type[0])) {
return fuzz_img(bytes, 0, option);
}
if (0 == strcmp("skp", FLAGS_type[0])) {
return fuzz_skp(bytes);
}
if (0 == strcmp("sksl2glsl", FLAGS_type[0])) {
return fuzz_sksl2glsl(bytes);
}
}
return printUsage(argv[0]);
}
// This adds up the first 1024 bytes and returns it as an 8 bit integer. This allows afl-fuzz to
// deterministically excercise different paths, or *options* (such as different scaling sizes or
// different image modes) without needing to introduce a parameter. This way we don't need a
// image_scale1, image_scale2, image_scale4, etc fuzzer, we can just have a image_scale fuzzer.
// Clients are expected to transform this number into a different range, e.g. with modulo (%).
static uint8_t calculate_option(SkData* bytes) {
uint8_t total = 0;
const uint8_t* data = bytes->bytes();
for (size_t i = 0; i < 1024 && i < bytes->size(); i++) {
total += data[i];
}
return total;
}
int fuzz_api(sk_sp<SkData> bytes) {
const char* name = FLAGS_name.isEmpty() ? "" : FLAGS_name[0];
for (auto r = SkTRegistry<Fuzzable>::Head(); r; r = r->next()) {
auto fuzzable = r->factory();
if (0 == strcmp(name, fuzzable.name)) {
SkDebugf("Fuzzing %s...\n", fuzzable.name);
Fuzz fuzz(bytes);
fuzzable.fn(&fuzz);
SkDebugf("[terminated] Success!\n");
return 0;
}
}
SkDebugf("When using --type api, please choose an API to fuzz with --name/-n:\n");
for (auto r = SkTRegistry<Fuzzable>::Head(); r; r = r->next()) {
auto fuzzable = r->factory();
SkDebugf("\t%s\n", fuzzable.name);
}
return 1;
}
static void dump_png(SkBitmap bitmap) {
if (!FLAGS_dump.isEmpty()) {
SkImageEncoder::EncodeFile(FLAGS_dump[0], bitmap, SkImageEncoder::kPNG_Type, 100);
SkDebugf("Dumped to %s\n", FLAGS_dump[0]);
}
}
int fuzz_img(sk_sp<SkData> bytes, uint8_t scale, uint8_t mode) {
// We can scale 1x, 2x, 4x, 8x, 16x
scale = scale % 5;
float fscale = (float)pow(2.0f, scale);
SkDebugf("Scaling factor: %f\n", fscale);
// We have 4 different modes of decoding, just like DM.
mode = mode % 4;
SkDebugf("Mode: %d\n", mode);
// This is mostly copied from DMSrcSink's CodecSrc::draw method.
SkDebugf("Decoding\n");
SkAutoTDelete<SkCodec> codec(SkCodec::NewFromData(bytes));
if (nullptr == codec.get()) {
SkDebugf("[terminated] Couldn't create codec.\n");
return 3;
}
SkImageInfo decodeInfo = codec->getInfo();
SkISize size = codec->getScaledDimensions(fscale);
decodeInfo = decodeInfo.makeWH(size.width(), size.height());
// Construct a color table for the decode if necessary
SkAutoTUnref<SkColorTable> colorTable(nullptr);
SkPMColor* colorPtr = nullptr;
int* colorCountPtr = nullptr;
int maxColors = 256;
if (kIndex_8_SkColorType == decodeInfo.colorType()) {
SkPMColor colors[256];
colorTable.reset(new SkColorTable(colors, maxColors));
colorPtr = const_cast<SkPMColor*>(colorTable->readColors());
colorCountPtr = &maxColors;
}
SkBitmap bitmap;
SkMallocPixelRef::ZeroedPRFactory zeroFactory;
SkCodec::Options options;
options.fZeroInitialized = SkCodec::kYes_ZeroInitialized;
if (!bitmap.tryAllocPixels(decodeInfo, &zeroFactory, colorTable.get())) {
SkDebugf("[terminated] Could not allocate memory. Image might be too large (%d x %d)",
decodeInfo.width(), decodeInfo.height());
return 4;
}
switch (mode) {
case 0: {//kCodecZeroInit_Mode, kCodec_Mode
switch (codec->getPixels(decodeInfo, bitmap.getPixels(), bitmap.rowBytes(), &options,
colorPtr, colorCountPtr)) {
case SkCodec::kSuccess:
SkDebugf("[terminated] Success!\n");
break;
case SkCodec::kIncompleteInput:
SkDebugf("[terminated] Partial Success\n");
break;
case SkCodec::kInvalidConversion:
SkDebugf("Incompatible colortype conversion\n");
// Crash to allow afl-fuzz to know this was a bug.
raise(SIGSEGV);
default:
SkDebugf("[terminated] Couldn't getPixels.\n");
return 6;
}
break;
}
case 1: {//kScanline_Mode
if (SkCodec::kSuccess != codec->startScanlineDecode(decodeInfo, NULL, colorPtr,
colorCountPtr)) {
SkDebugf("[terminated] Could not start scanline decoder\n");
return 7;
}
void* dst = bitmap.getAddr(0, 0);
size_t rowBytes = bitmap.rowBytes();
uint32_t height = decodeInfo.height();
switch (codec->getScanlineOrder()) {
case SkCodec::kTopDown_SkScanlineOrder:
case SkCodec::kBottomUp_SkScanlineOrder:
// We do not need to check the return value. On an incomplete
// image, memory will be filled with a default value.
codec->getScanlines(dst, height, rowBytes);
break;
}
SkDebugf("[terminated] Success!\n");
break;
}
case 2: { //kStripe_Mode
const int height = decodeInfo.height();
// This value is chosen arbitrarily. We exercise more cases by choosing a value that
// does not align with image blocks.
const int stripeHeight = 37;
const int numStripes = (height + stripeHeight - 1) / stripeHeight;
// Decode odd stripes
if (SkCodec::kSuccess != codec->startScanlineDecode(decodeInfo, NULL, colorPtr,
colorCountPtr)
|| SkCodec::kTopDown_SkScanlineOrder != codec->getScanlineOrder()) {
// This mode was designed to test the new skip scanlines API in libjpeg-turbo.
// Jpegs have kTopDown_SkScanlineOrder, and at this time, it is not interesting
// to run this test for image types that do not have this scanline ordering.
SkDebugf("[terminated] Could not start top-down scanline decoder\n");
return 8;
}
for (int i = 0; i < numStripes; i += 2) {
// Skip a stripe
const int linesToSkip = SkTMin(stripeHeight, height - i * stripeHeight);
codec->skipScanlines(linesToSkip);
// Read a stripe
const int startY = (i + 1) * stripeHeight;
const int linesToRead = SkTMin(stripeHeight, height - startY);
if (linesToRead > 0) {
codec->getScanlines(bitmap.getAddr(0, startY), linesToRead, bitmap.rowBytes());
}
}
// Decode even stripes
const SkCodec::Result startResult = codec->startScanlineDecode(decodeInfo, nullptr,
colorPtr, colorCountPtr);
if (SkCodec::kSuccess != startResult) {
SkDebugf("[terminated] Failed to restart scanline decoder with same parameters.\n");
return 9;
}
for (int i = 0; i < numStripes; i += 2) {
// Read a stripe
const int startY = i * stripeHeight;
const int linesToRead = SkTMin(stripeHeight, height - startY);
codec->getScanlines(bitmap.getAddr(0, startY), linesToRead, bitmap.rowBytes());
// Skip a stripe
const int linesToSkip = SkTMin(stripeHeight, height - (i + 1) * stripeHeight);
if (linesToSkip > 0) {
codec->skipScanlines(linesToSkip);
}
}
SkDebugf("[terminated] Success!\n");
break;
}
case 3: { //kSubset_Mode
// Arbitrarily choose a divisor.
int divisor = 2;
// Total width/height of the image.
const int W = codec->getInfo().width();
const int H = codec->getInfo().height();
if (divisor > W || divisor > H) {
SkDebugf("[terminated] Cannot codec subset: divisor %d is too big "
"with dimensions (%d x %d)\n", divisor, W, H);
return 10;
}
// subset dimensions
// SkWebpCodec, the only one that supports subsets, requires even top/left boundaries.
const int w = SkAlign2(W / divisor);
const int h = SkAlign2(H / divisor);
SkIRect subset;
SkCodec::Options opts;
opts.fSubset = &subset;
SkBitmap subsetBm;
// We will reuse pixel memory from bitmap.
void* pixels = bitmap.getPixels();
// Keep track of left and top (for drawing subsetBm into canvas). We could use
// fscale * x and fscale * y, but we want integers such that the next subset will start
// where the last one ended. So we'll add decodeInfo.width() and height().
int left = 0;
for (int x = 0; x < W; x += w) {
int top = 0;
for (int y = 0; y < H; y+= h) {
// Do not make the subset go off the edge of the image.
const int preScaleW = SkTMin(w, W - x);
const int preScaleH = SkTMin(h, H - y);
subset.setXYWH(x, y, preScaleW, preScaleH);
// And fscale
// FIXME: Should we have a version of getScaledDimensions that takes a subset
// into account?
decodeInfo = decodeInfo.makeWH(
SkTMax(1, SkScalarRoundToInt(preScaleW * fscale)),
SkTMax(1, SkScalarRoundToInt(preScaleH * fscale)));
size_t rowBytes = decodeInfo.minRowBytes();
if (!subsetBm.installPixels(decodeInfo, pixels, rowBytes, colorTable.get(),
nullptr, nullptr)) {
SkDebugf("[terminated] Could not install pixels.\n");
return 11;
}
const SkCodec::Result result = codec->getPixels(decodeInfo, pixels, rowBytes,
&opts, colorPtr, colorCountPtr);
switch (result) {
case SkCodec::kSuccess:
case SkCodec::kIncompleteInput:
SkDebugf("okay\n");
break;
case SkCodec::kInvalidConversion:
if (0 == (x|y)) {
// First subset is okay to return unimplemented.
SkDebugf("[terminated] Incompatible colortype conversion\n");
return 12;
}
// If the first subset succeeded, a later one should not fail.
// fall through to failure
case SkCodec::kUnimplemented:
if (0 == (x|y)) {
// First subset is okay to return unimplemented.
SkDebugf("[terminated] subset codec not supported\n");
return 13;
}
// If the first subset succeeded, why would a later one fail?
// fall through to failure
default:
SkDebugf("[terminated] subset codec failed to decode (%d, %d, %d, %d) "
"with dimensions (%d x %d)\t error %d\n",
x, y, decodeInfo.width(), decodeInfo.height(),
W, H, result);
return 14;
}
// translate by the scaled height.
top += decodeInfo.height();
}
// translate by the scaled width.
left += decodeInfo.width();
}
SkDebugf("[terminated] Success!\n");
break;
}
default:
SkDebugf("[terminated] Mode not implemented yet\n");
}
dump_png(bitmap);
return 0;
}
int fuzz_skp(sk_sp<SkData> bytes) {
SkMemoryStream stream(bytes);
SkDebugf("Decoding\n");
sk_sp<SkPicture> pic(SkPicture::MakeFromStream(&stream));
if (!pic) {
SkDebugf("[terminated] Couldn't decode as a picture.\n");
return 3;
}
SkDebugf("Rendering\n");
SkBitmap bitmap;
if (!FLAGS_dump.isEmpty()) {
SkIRect size = pic->cullRect().roundOut();
bitmap.allocN32Pixels(size.width(), size.height());
}
SkCanvas canvas(bitmap);
canvas.drawPicture(pic);
SkDebugf("[terminated] Success! Decoded and rendered an SkPicture!\n");
dump_png(bitmap);
return 0;
}
int fuzz_icc(sk_sp<SkData> bytes) {
sk_sp<SkColorSpace> space(SkColorSpace::MakeICC(bytes->data(), bytes->size()));
if (!space) {
SkDebugf("[terminated] Couldn't decode ICC.\n");
return 1;
}
SkDebugf("[terminated] Success! Decoded ICC.\n");
return 0;
}
int fuzz_color_deserialize(sk_sp<SkData> bytes) {
sk_sp<SkColorSpace> space(SkColorSpace::Deserialize(bytes->data(), bytes->size()));
if (!space) {
SkDebugf("[terminated] Couldn't deserialize Colorspace.\n");
return 1;
}
SkDebugf("[terminated] Success! deserialized Colorspace.\n");
return 0;
}
int fuzz_sksl2glsl(sk_sp<SkData> bytes) {
SkSL::Compiler compiler;
std::string output;
bool result = compiler.toGLSL(SkSL::Program::kFragment_Kind,
(const char*)bytes->data(), SkSL::GLCaps(), &output);
if (!result) {
SkDebugf("[terminated] Couldn't compile input.\n");
return 1;
}
SkDebugf("[terminated] Success! Compiled input.\n");
return 0;
}
Fuzz::Fuzz(sk_sp<SkData> bytes) : fBytes(bytes), fNextByte(0) {}
void Fuzz::signalBug () { SkDebugf("Signal bug\n"); raise(SIGSEGV); }
void Fuzz::signalBoring() { SkDebugf("Signal boring\n"); exit(0); }
size_t Fuzz::size() { return fBytes->size(); }
size_t Fuzz::remaining() {
return fBytes->size() - fNextByte;
}
template <typename T>
T Fuzz::nextT() {
if (fNextByte + sizeof(T) > fBytes->size()) {
this->signalBoring();
}
T val;
memcpy(&val, fBytes->bytes() + fNextByte, sizeof(T));
fNextByte += sizeof(T);
return val;
}
uint8_t Fuzz::nextB() { return this->nextT<uint8_t >(); }
bool Fuzz::nextBool() { return nextB()&1; }
uint32_t Fuzz::nextU() { return this->nextT<uint32_t>(); }
float Fuzz::nextF() { return this->nextT<float >(); }
float Fuzz::nextF1() {
// This is the same code as is in SkRandom's nextF()
unsigned int floatint = 0x3f800000 | (this->nextU() >> 9);
float f = SkBits2Float(floatint) - 1.0f;
return f;
}
uint32_t Fuzz::nextRangeU(uint32_t min, uint32_t max) {
if (min > max) {
SkDebugf("Check mins and maxes (%d, %d)\n", min, max);
this->signalBoring();
}
uint32_t range = max - min + 1;
if (0 == range) {
return this->nextU();
} else {
return min + this->nextU() % range;
}
}
float Fuzz::nextRangeF(float min, float max) {
if (min > max) {
SkDebugf("Check mins and maxes (%f, %f)\n", min, max);
this->signalBoring();
}
float f = std::abs(this->nextF());
if (!std::isnormal(f) && f != 0.0) {
this->signalBoring();
}
return min + fmod(f, (max - min + 1));
}