summaryrefslogtreecommitdiff
path: root/core/fxcodec
AgeCommit message (Collapse)Author
2017-03-14Replace FX_CHAR and FX_WCHAR with underlying types.Dan Sinclair
Change-Id: I96e0a20d66b9184d22f64d8e4ce0dadd5a78c1e8 Reviewed-on: https://pdfium-review.googlesource.com/2967 Commit-Queue: dsinclair <dsinclair@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org>
2017-03-13Remove openjpeg write support.thestig
Review-Url: https://codereview.chromium.org/2071693002
2017-03-13Fix boundary value negation in bmp_read_headerchromium/3041Nicolas Pena
When the value read is equal to -INT_MIN, we cannot negate it since it will be out of bounds, so return error in this case. BUG=chromium:628559 Change-Id: I7e47a71ef0d35cfb2d1fddc0ba644f9aac79ec3f Reviewed-on: https://pdfium-review.googlesource.com/2965 Reviewed-by: Tom Sepez <tsepez@chromium.org> Commit-Queue: Nicolás Peña <npm@chromium.org>
2017-03-13Fix some nits in fx_codec_faxNicolas Pena
Change-Id: I16acba7b3eda26b4edfbfe0cb356267bd536fee8 Reviewed-on: https://pdfium-review.googlesource.com/2964 Reviewed-by: dsinclair <dsinclair@chromium.org> Commit-Queue: Nicolás Peña <npm@chromium.org>
2017-03-13Check run lengths in FaxG4GetRowNicolas Pena
The spec says a1 is to the right of a0, a2 to the right of a1. I think that means that the run lengths have to be positive, but that certainly means that they cannot be negative. BUG=chromium:699340 Change-Id: Ic07a272e63610f7a66c5073179cdb2768f80e2b8 Reviewed-on: https://pdfium-review.googlesource.com/2963 Reviewed-by: Tom Sepez <tsepez@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org> Commit-Queue: Nicolás Peña <npm@chromium.org>
2017-03-07fx_codec_flate code cleanup.chromium/3035Dan Sinclair
This CL cleans up the namespacing along with some type confusion in the flate codec class. Change-Id: I4772ee571bcd78f1c839b5cefc370860aca06bb3 Reviewed-on: https://pdfium-review.googlesource.com/2917 Commit-Queue: dsinclair <dsinclair@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org>
2017-03-07Limit BMP width to avoid overflowsNicolas Pena
BMP_WIDTHBYTES starts with: (width * bitCount) + 31. Since bitCount can be as large as 32, to avoid this overflowing we need width <= 67108863. BUG=chromium:628559 Change-Id: I4fd33b65da76225c8200a22380f2bfc4523c5c8d Reviewed-on: https://pdfium-review.googlesource.com/2934 Commit-Queue: Nicolás Peña <npm@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org>
2017-03-06Check size before writtingchromium/3033dan sinclair
Before writting to the stream buffer make sure that we won't walk off the end of the allocated size. In this specific case the dest_size of the buffer is 0, so we're basically just looping over to free the temp results. BUG=chromium:697847 Change-Id: I229eea96179692216cb2685facbb7d5379c501c7 Reviewed-on: https://pdfium-review.googlesource.com/2903 Commit-Queue: dsinclair <dsinclair@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org>
2017-03-01Create virtual codec APIs so chrome/fuzzers can link separatelyTom Sepez
The one step to make an actual concrete class is conditionalized in fpdfview and is unconditional in the fuzzer. Also replace the clumsy C-style callbacks with a delegate interface as long as we are making new interfaces. Change-Id: I733a437483ce5e0c34211cfbbda05105336f55b5 Reviewed-on: https://pdfium-review.googlesource.com/2887 Commit-Queue: Tom Sepez <tsepez@chromium.org> Commit-Queue: dsinclair <dsinclair@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org>
2017-02-27Fix nits from commit db764708.Lei Zhang
- Constify test data - ASSERT_EQ() buffer size - EXPECT_EQ() arguments are reversed - Fix typos in comments Change-Id: I846bfcc29bbf43d0a8b333250d1cce1aae5a8d59 Reviewed-on: https://pdfium-review.googlesource.com/2841 Reviewed-by: Tom Sepez <tsepez@chromium.org> Commit-Queue: Lei Zhang <thestig@chromium.org>
2017-02-27Allow building XFA without additional codecs.Tom Sepez
This is something we'd like to try for initial XFA launches adding in codecs as justified by results in the wild. Adding statistics for the unsupported cases is a follow-up exercise once this builds correctly. We always build all the additional libraries, to allow fuzzers to link against them even if we are not shipping them. The linker will sort it out for the actual code. Rename some files to match the classes contained within. That the existing tests seem to pass with the codecs disabled warrants further investigation. Change-Id: Iad269db91289f12dc9f5dda8f48121d27a0c4367 Reviewed-on: https://pdfium-review.googlesource.com/2836 Commit-Queue: Tom Sepez <tsepez@chromium.org> Reviewed-by: Lei Zhang <thestig@chromium.org>
2017-02-14Prevent heap-buffer-overflow in CCodec_ProgressiveDecoderNicolas Pena
In CCodec_ProgressiveDecoder::GifInputRecordPositionBufCallback, m_pSrcPalette can be allocated size pal_num. So if pal_index >= pal_num, then bail out. BUG=691278 Change-Id: Ib0157cf51cbf52ecd5d60b027e5fc32898a906ed Reviewed-on: https://pdfium-review.googlesource.com/2699 Commit-Queue: Nicolás Peña <npm@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org>
2017-01-25Calculate code_store safely in CGifLZWDecoder::DecodeNicolas Pena
BUG=682628 Change-Id: I8e88cc0c8392b078afb73f9549ea4dea9a5717fd Reviewed-on: https://pdfium-review.googlesource.com/2390 Commit-Queue: Nicolás Peña <npm@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org>
2017-01-23Fix bug found by fuzzer in A85 encoder (bad shift).rbpotter
A85 encoder is not called yet; fix before adding calls to this code in Chromium. BUG=682550 Review-Url: https://codereview.chromium.org/2649263002
2017-01-16Return early when gif_img_row_bytes is 0chromium/2984Nicolas Pena
We can get into a loop when gif_img_row_bytes is 0 since Decode will return 3 when the second parameter is 0, and there is a while(ret!=0). BUG=681170 Change-Id: I63502a8487c07030fce2373f74cec6b4f0c98297 Reviewed-on: https://pdfium-review.googlesource.com/2211 Reviewed-by: dsinclair <dsinclair@chromium.org> Commit-Queue: dsinclair <dsinclair@chromium.org>
2017-01-16Check blue,green,red bit count in bmp_decode_rgbNicolas Pena
If the values are going to overflow, return error code, which seems to be 2. BUG=668822 Change-Id: I89b3fcf277e98d65b8c3438e6d9bb84fe62a8de9 Reviewed-on: https://pdfium-review.googlesource.com/2213 Commit-Queue: Nicolás Peña <npm@chromium.org> Commit-Queue: dsinclair <dsinclair@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org>
2017-01-16Return error when bits_left is too bigNicolas Pena
BUG=680062 Change-Id: I9e2fd960915cd1de5e0cc15aeabf9ccf27e5a795 Reviewed-on: https://pdfium-review.googlesource.com/2212 Commit-Queue: Nicolás Peña <npm@chromium.org> Reviewed-by: dsinclair <dsinclair@chromium.org>
2017-01-16Use FX_Alloc() properly in fxcodec.tsepez
It already has a sizeof() and a cast built into the macro, so we don't need to do sizeof() and cast on each usage. Review-Url: https://codereview.chromium.org/2625133009
2017-01-12Add postscript pathrbpotter
This patch adds the additional functions required to make postscript printing functional. The most significant additions are are two added compression functions and a new API for setting the postscript level. Not currently called from Chromium, Chromium patch to come. BUG= Review-Url: https://codereview.chromium.org/2612243005
2017-01-11Make tiff_read return actual length readNicolas Pena
The return value is used to determine whether TIFFReadFile fails. If we return just the length, libtiff will try reading uninitilized values afterwards, on corrupted files. BUG=679230, 670928 Change-Id: I579adc9d8a00e8cafab45dbdb728f1cb702da051 Reviewed-on: https://pdfium-review.googlesource.com/2172 Commit-Queue: Nicolás Peña <npm@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org>
2017-01-10Restrict code size in CGifLZWDecoderchromium/2978Nicolas Pena
The code_size variable is the number of bits. We should make sure that the size is at most 31 to avoid having undefined shifts etc. BUG=620661 Change-Id: Ia533386d01de93a55048cfd63d63989b2731a210 Reviewed-on: https://pdfium-review.googlesource.com/2161 Reviewed-by: dsinclair <dsinclair@chromium.org> Reviewed-by: Tom Sepez <tsepez@chromium.org> Commit-Queue: Nicolás Peña <npm@chromium.org>
2017-01-09Check validity of width and height in CCodec_TiffContext::LoadFrameInfoNicolas Pena
We are using pdfium::base::checked_cast to get the width and height, but we may overflow and abort. Therefore, we should instead early return if the obtained width and height are not valid int32_t's. BUG=655056 Change-Id: Ic0c6b88a16dc3d547fe82736bb14ed3122cd356a Reviewed-on: https://pdfium-review.googlesource.com/2160 Reviewed-by: Tom Sepez <tsepez@chromium.org> Commit-Queue: Nicolás Peña <npm@chromium.org>
2017-01-09Remove CFX_ArrayTemplate from fpdftext and fxcodec.tsepez
Remove unused m_Segments. Review-Url: https://codereview.chromium.org/2618863004
2017-01-06Revert postscript code removal.rbpotter
Revert CL http://crrev.com/2608663003 in preparation for adding postscript generation to Pdfium. Note postscript generation code will not be called until additional patches land. These patches will also include modifications needed to make this code functional (currently missing a few compression functions). BUG= Review-Url: https://codereview.chromium.org/2615703002
2017-01-03Check for overflow in JBig2 Huffman decoderDan Sinclair
This CL updates the Huffman decoder in the JBig2 codex to check the low field does not overflow. BUG=chromium:675236 Change-Id: I7f5f6fe8329df4ece6f317fac521fe2373686479 Reviewed-on: https://pdfium-review.googlesource.com/2131 Reviewed-by: Tom Sepez <tsepez@chromium.org> Commit-Queue: dsinclair <dsinclair@chromium.org>
2016-12-14Remove CJBig2_List in favor of std::vector<std::unique_ptr<>>tsepez
Review-Url: https://codereview.chromium.org/2578663002
2016-12-14Avoid the ptr.reset(new XXX()) anti-patterntsepez
Be suspicious of |new|. This removes some of the easy cases. Review-Url: https://codereview.chromium.org/2571913002
2016-12-13Return unique_ptr<>s from fxcodec/chromium/2951tsepez
Review-Url: https://codereview.chromium.org/2572843002
2016-12-12Fix JBig2 decoding logicweili
While decoding among instances, variable "FIRSTS" should hold its value, not be reset. This was accidently changed by earlier refactoring. BUG=chromium:625848,pdfium:636 Review-Url: https://codereview.chromium.org/2569023002
2016-12-07Refcount all the IFX_ stream classes all the time.tsepez
We can remove a lot of "bOwnsStream" logic in the process. Always pass these by const reference, in case the called method wants to hang on to the stream (one exception is where we stick a raw pointer into a void* slot in a context from another layer). Review-Url: https://codereview.chromium.org/2451493002
2016-12-06Return unique_ptrs from CFX_DIBitmap::Clone().tsepez
Because that's what clone does. Perform immediate release in some spots to avoid disrupting too much at once. Review-Url: https://codereview.chromium.org/2534953004
2016-12-01Improve early return in CGifLZWDecoder::Decodenpm
|code_size_cur| could be larger than |code_size|, so |code| could be larger than |code_end|. If this happens, early return, since the Decode has failed. BUG=659417 Review-Url: https://codereview.chromium.org/2542673004
2016-11-28Replace ASSERT with error return in bmp_read_headernpm
The width and compress flag are read, so returning with error is more appropriate than having an ASSERT. BUG=659497 Review-Url: https://codereview.chromium.org/2535863002
2016-11-22Multiply safely in CCodec_TiffContext::Decodenpm
BUG=667074 Review-Url: https://codereview.chromium.org/2520253003
2016-11-22pdfium: Fix inconsistent number of color components of ICC profilekcwu
fx_codec_icc.cpp specify default number of color components as 3 for unknown profiles. However, lcms may know such profile with different number of components. The inconsistency may lead to array access violation. This CL uses cmsChannelsOf() from lcms to ensure consistency. And rejects unexpected number according to PDF spec. BUG=chromium:667694 Review-Url: https://codereview.chromium.org/2522933002
2016-11-21Check dimensions and content of bitmaps in EmbedderTests.chromium/2927thestig
Review-Url: https://codereview.chromium.org/2514173002
2016-11-21Fixup lint flags.Dan Sinclair
The -build/include setting was masking out build/include_what_you_use. This CL restores them, fixes any build errors, and adds NOLINT as needed. As well, the runtime/explicit and runtime/printf flags are aslo enabled and NOLINT'd. lint cleanups Change-Id: Ib013b3eb29c8d0e48cad74c5df9028684130719f Reviewed-on: https://pdfium-review.googlesource.com/2030 Reviewed-by: Tom Sepez <tsepez@chromium.org>
2016-11-11Relax colorspace checks in CPDF_DIBSource::CreateDecoder().chromium/2917thestig
Previously the log just made sure the colorspace and the image have exact matches for the number of colorspace components. Now, for some colorspace types, check the type and make sure the number of components meets or exceeds what is required by the spec. Also do some refactoring. BUG=chromium:650230 Review-Url: https://codereview.chromium.org/2486123002
2016-11-09Add early returns in CJBig2_TRDProc::decode_Arith when decode fails.npm
None of the decodes in the method are currently being checked. This is causing pdfium to take a long time rendering corrupted files. Thus, I added a couple of early returns to help prevent this from happening. BUG=450971 Review-Url: https://codereview.chromium.org/2493633002
2016-11-09Revert of Clean up fx_codec_icc.cpp (patchset #1 id:1 of ↵kcwu
https://codereview.chromium.org/2482663002/ ) Reason for revert: Max cmsChannelsOf() is 15, which is larger than expectation of existing code and cause crashes (at least the fuzzer). BUG=chromium:663240 Original issue's description: > Clean up fx_codec_icc.cpp > > Committed: https://pdfium.googlesource.com/pdfium/+/a94fc11866adb1b9ca4a4e1afb4fb574ed472e07 TBR=dsinclair@chromium.org # Not skipping CQ checks because original CL landed more than 1 days ago. Review-Url: https://codereview.chromium.org/2485363002
2016-11-07Clean up fx_codec_icc.cppkcwu
Review-Url: https://codereview.chromium.org/2482663002
2016-11-02Remove FX_BOOL from coretsepez
Review-Url: https://codereview.chromium.org/2477443002
2016-11-01Fix founding difference in pdfium_test on AdobeCMYK_to_sRGBbrucedawson
An optimization to speed up float-to-int rounding caused a different result for one input value. This tweaks the conversion constant so that the results are identical across the entire valid range, and adds a test that checks the part of the range that is most sensitive to errors. BUG=pdfium:624 Review-Url: https://codereview.chromium.org/2466203002
2016-10-28Change some ints to bools in JBIG2 code.thestig
Review-Url: https://codereview.chromium.org/2450393004
2016-10-28Fix more FX_BOOL / int noise in core/fpdfsdktsepez
Review-Url: https://codereview.chromium.org/2454263002
2016-10-27Fix FX_BOOL / int noise in JBig2, pass 3tsepez
Review-Url: https://codereview.chromium.org/2457943002
2016-10-27Fix FX_BOOL / int noise in JBig2, pass 2tsepez
Review-Url: https://codereview.chromium.org/2461543002
2016-10-27Fix FX_BOOL / int noise in JBig2_Grd*Proctsepez
Replace most of these with ints since the are used in integer operations. If it walks like a duck, and quacks like a duck ... despite what the hungarian notation might say. Review-Url: https://codereview.chromium.org/2455523005
2016-10-27Fix some FX_BOOL / int noise in core/fx*chromium/2903tsepez
Review-Url: https://codereview.chromium.org/2459673002
2016-10-26Add more checks to tiff_read() and tiff_seek().thestig
BUG=chromium:659519 Review-Url: https://codereview.chromium.org/2456553002