[Debian-med-packaging] Bug#923433: EncoderStrategy::Flush (this=0x5555557452f0) at ./src/encoderstrategy.h:156

[Mathieu Malaterre]

Package: dcmtk
Version: 3.6.4-2

It seems there is a regression in the JPEG-LS/CharLS encoder at least
for some files. dcmcrle / dcmcjpeg seems to be ok. GDCM is able to
compress it:

$ gdcmconv --jpegls /tmp/foo.dcm /tmp/o.dcm

Steps:

$ cd /tmp
$ curl https://raw.githubusercontent.com/neurolabusc/dcm_qa/master/In/Orientation/ax/axasc35/MR.1.3.12.2.1107.5.2.32.35131.2014031012493950715786673
> foo.dcm
$ gdb dcmcjpls
(gdb) r /tmp/foo.dcm /tmp/o.dcm
...
Starting program: /usr/bin/dcmcjpls /tmp/foo.dcm /tmp/o.dcm
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".

Program received signal SIGSEGV, Segmentation fault.
0x00007ffff76dd3b3 in EncoderStrategy::Flush (this=0x5555557452f0) at
./src/encoderstrategy.h:156
156                 *_position = static_cast<uint8_t>(_bitBuffer >> 24);
(gdb) bt full
#0  0x00007ffff76dd3b3 in EncoderStrategy::Flush (this=0x5555557452f0)
at ./src/encoderstrategy.h:156
        i = 0
        i = <optimized out>
#1  0x00007ffff76dd555 in EncoderStrategy::AppendToBitStream
(this=0x5555557452f0, bits=93, bitCount=<optimized out>) at
./src/encoderstrategy.h:87
        mask = <optimized out>
        __PRETTY_FUNCTION__ = "void
EncoderStrategy::AppendToBitStream(int32_t, int32_t)"
#2  0x00007ffff76edfb8 in JlsCodec<LosslessTraitsT<unsigned short,
16>, EncoderStrategy>::EncodeMappedValue (limit=64,
mappedError=<optimized out>,
    k=<optimized out>, this=0x5555557452f0) at ./src/losslesstraits.h:112
        highbits = <optimized out>
        highbits = <optimized out>
#3  JlsCodec<LosslessTraitsT<unsigned short, 16>,
EncoderStrategy>::DoRegular (pred=<optimized out>, x=112,
Qs=<optimized out>, this=0x5555557452f0)
    at ./src/scan.h:271
        ctx = @0x555555745ad4: {A = 1573, B = -7, C = -1, N = 8}
        k = <optimized out>
        Px = 159
        ErrVal = -47
        sign = <optimized out>
        __PRETTY_FUNCTION__ = "typename TRAITS::SAMPLE
JlsCodec<TRAITS, STRATEGY>::DoRegular(int32_t, int32_t, int32_t,
EncoderStrategy*) [with TRAITS = LosslessTraitsT<short unsigned int,
16>; STRATEGY = EncoderStrategy; typename "...
        sign = <optimized out>
        ctx = <optimized out>
        k = <optimized out>
        Px = <optimized out>
        ErrVal = <optimized out>
#4  JlsCodec<LosslessTraitsT<unsigned short, 16>,
EncoderStrategy>::DoLine (this=this at entry=0x5555557452f0) at
./src/scan.h:660
        Ra = <optimized out>
        Rc = <optimized out>
        Qs = <optimized out>
        index = 162
        Rb = 59
        Rd = 79
#5  0x00007ffff76ee16a in JlsCodec<LosslessTraitsT<unsigned short,
16>, EncoderStrategy>::DoScan (this=this at entry=0x5555557452f0)
    at /usr/include/c++/8/bits/stl_vector.h:930
        component = <optimized out>
        line = <optimized out>
        pixelstride = 388
        components = 1
        vectmp = std::vector of length 776, capacity 776 = {0, 0, 25,
21, 18, 23, 23, 22, 25, 23, 27, 41, 69, 73, 93, 78, 72, 43, 231, 216,
92, 88, 128, 87,
          85, 54, 50, 22, 15, 26, 37, 32, 31, 33, 43, 60, 16, 36, 33,
53, 24, 43, 49, 88, 47, 49, 47, 121, 100, 124, 174, 100, 90, 47, 53,
39, 15, 20, 20,
          20, 20, 20, 25, 15, 20, 0, 24, 23, 25, 29, 22, 16, 33, 30,
24, 19, 54, 44, 126, 171, 191, 83, 87, 47, 30, 113, 51, 37, 131, 232,
53, 39, 33, 26,
          52, 43, 26, 18, 24, 43, 28, 30, 22, 32, 29, 65, 57, 55, 85,
49, 68, 80, 39, 97, 162, 92, 72, 85, 57, 39, 21, 19, 20, 15, 30, 24,
32, 22, 16, 0,