Segfault with certain language packs on certain images

[cchadowitz]

Current Behavior

When running the tesseract binary with specific language packs enabled, the binary segfaults (different places for different combinations).

1. Using the chi_sim and the mrz language packs together for this input image:

   
   root@78ce6cbf2cc5:/opt/tesseract/bin# gdb ./tesseract 
   GNU gdb (Ubuntu 14.0.50.20230907-0ubuntu1) 14.0.50.20230907-git
   Copyright (C) 2023 Free Software Foundation, Inc.
   License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
   This is free software: you are free to change and redistribute it.
   There is NO WARRANTY, to the extent permitted by law.
   Type "show copying" and "show warranty" for details.
   This GDB was configured as "x86_64-linux-gnu".
   Type "show configuration" for configuration details.
   For bug reporting instructions, please see:
   <https://www.gnu.org/software/gdb/bugs/>.
   Find the GDB manual and other documentation resources online at:
   <http://www.gnu.org/software/gdb/documentation/>.

For help, type "help". Type "apropos word" to search for commands related to "word"... Reading symbols from ./tesseract... (No debugging symbols found in ./tesseract) (gdb) run sample_012631.jpg - --tessdata-dir /tessdata/ -l chi_sim+mrz Starting program: /opt/tesseract/bin/tesseract sample_012631.jpg - --tessdata-dir /tessdata/ -l chi_sim+mrz warning: Error disabling address space randomization: Operation not permitted [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Estimating resolution as 548 Detected 108 diacritics

Program received signal SIGSEGV, Segmentation fault. 0x00005619906db4e8 in tesseract::ELIST_ITERATOR::forward() () (gdb) bt

0 0x00005619906db4e8 in tesseract::ELIST_ITERATOR::forward() ()

1 0x00005619906bc1ff in tesseract::PAGE_RES_IT::ReplaceCurrentWord(tesseract::PointerVector*) ()

2 0x000056199074ea84 in tesseract::Tesseract::classify_word_and_language(int, tesseract::PAGE_RES_IT, tesseract::WordData) ()

3 0x0000561990752adc in tesseract::Tesseract::RecogAllWordsPassN(int, tesseract::ETEXT_DESC, tesseract::PAGE_RES_IT, std::vector<tesseract::WordData, std::allocator >*) ()

4 0x00005619907537ba in tesseract::Tesseract::recog_all_words(tesseract::PAGE_RES, tesseract::ETEXT_DESC, tesseract::TBOX const, char const, int) ()

5 0x0000561990719b46 in tesseract::TessBaseAPI::Recognize(tesseract::ETEXT_DESC*) ()

6 0x000056199071a3d2 in tesseract::TessBaseAPI::ProcessPage(Pix, int, char const, char const, int, tesseract::TessResultRenderer) ()

7 0x000056199071b4ea in tesseract::TessBaseAPI::ProcessPagesInternal(char const, char const, int, tesseract::TessResultRenderer*) ()

8 0x000056199071bb92 in tesseract::TessBaseAPI::ProcessPages(char const, char const, int, tesseract::TessResultRenderer*) ()

9 0x00005619906a6331 in main ()

2. Using the `eng`, `ara`, and `ocrb_int` language packs together for this input image:
![eng-ara-orcb_int-segfault](https://github.com/tesseract-ocr/tesseract/assets/24324908/d7dd2880-2725-4f7f-af67-297bec764501)

root@78ce6cbf2cc5:/opt/tesseract/bin# gdb ./tesseract GNU gdb (Ubuntu 14.0.50.20230907-0ubuntu1) 14.0.50.20230907-git Copyright (C) 2023 Free Software Foundation, Inc. License GPLv3+: GNU GPL version 3 or later http://gnu.org/licenses/gpl.html This is free software: you are free to change and redistribute it. There is NO WARRANTY, to the extent permitted by law. Type "show copying" and "show warranty" for details. This GDB was configured as "x86_64-linux-gnu". Type "show configuration" for configuration details. For bug reporting instructions, please see: https://www.gnu.org/software/gdb/bugs/. Find the GDB manual and other documentation resources online at: http://www.gnu.org/software/gdb/documentation/.

For help, type "help". Type "apropos word" to search for commands related to "word"... Reading symbols from ./tesseract... (No debugging symbols found in ./tesseract) (gdb) run sample_000111.jpg - --tessdata-dir /tessdata/ -l eng+ara+ocrb_int Starting program: /opt/tesseract/bin/tesseract sample_000111.jpg - --tessdata-dir /tessdata/ -l eng+ara+ocrb_int warning: Error disabling address space randomization: Operation not permitted [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". Estimating resolution as 975

Program received signal SIGSEGV, Segmentation fault. 0x000055e53cf153c6 in tesseract::PAGE_RES_IT::DeleteCurrentWord() () (gdb) bt

0 0x000055e53cf153c6 in tesseract::PAGE_RES_IT::DeleteCurrentWord() ()

1 0x000055e53cfaf5a8 in tesseract::Tesseract::recog_all_words(tesseract::PAGE_RES, tesseract::ETEXT_DESC, tesseract::TBOX const, char const, int) ()

2 0x000055e53cf75b46 in tesseract::TessBaseAPI::Recognize(tesseract::ETEXT_DESC*) ()

3 0x000055e53cf763d2 in tesseract::TessBaseAPI::ProcessPage(Pix, int, char const, char const, int, tesseract::TessResultRenderer) ()

4 0x000055e53cf774ea in tesseract::TessBaseAPI::ProcessPagesInternal(char const, char const, int, tesseract::TessResultRenderer*) ()

5 0x000055e53cf77b92 in tesseract::TessBaseAPI::ProcessPages(char const, char const, int, tesseract::TessResultRenderer*) ()

6 0x000055e53cf02331 in main ()

### Expected Behavior

No segfaults.

### Suggested Fix

_No response_

### tesseract -v

root@78ce6cbf2cc5:/opt/tesseract/bin# ./tesseract -v tesseract 5.3.3-1-gdc22 leptonica-1.82.0 libgif 5.2.1 : libjpeg 8d (libjpeg-turbo 2.1.2) : libpng 1.6.40 : libtiff 4.5.1 : zlib 1.2.13 : libwebp 1.2.4 : libopenjp2 2.5.0 Found AVX2 Found AVX Found FMA Found SSE4.1 Found libarchive 3.6.2 zlib/1.2.13 liblzma/5.4.0 bz2lib/1.0.8 liblz4/1.9.4 libzstd/1.5.2 Found libcurl/8.2.1 OpenSSL/3.0.10 zlib/1.2.13 brotli/1.0.9 zstd/1.5.5 libidn2/2.3.4 libpsl/0.21.2 (+libidn2/2.3.3) libssh/0.10.5/openssl/zlib nghttp2/1.55.1 librtmp/2.3 OpenLDAP/2.6.6

### Operating System

_No response_

### Other Operating System

Ubuntu 23.10

### uname -a

root@78ce6cbf2cc5:/opt/tesseract/bin# uname -a Linux 78ce6cbf2cc5 5.4.0-150-generic #167~18.04.1-Ubuntu SMP Wed May 24 00:51:42 UTC 2023 x86_64 x86_64 x86_64 GNU/Linux

### Compiler

GCC 13.2.0

### CPU

Intel Xeon CPU E5-2609 v4 @ 1.70GHz x16

### Virtualization / Containers

I'm building and running `tesseract` inside a docker (v24.0.2) container, where the container is running the OS and compiler versions listed above.

### Other Information

_No response_

[zdenop]

Similar to #4148 (but crash on the different place):

1. You should be able to demonstrate a problem with official data only, and it looks like you use custom-generated languages (mrz, ocrb_int).

[stweil]

@cchadowitz, please provide sample_012631.jpg, mrz.traineddata and ocrb_int.traineddata. Those files are needed to reproduce the issue.

[zdenop]

@stweil : seems like ocrb_int is @Shreeshrii training from https://github.com/Shreeshrii/tessdata_ocrb and mrz is from https://github.com/DoubangoTelecom/tesseractMRZ

[stweil]

I can reproduce one of the segmentation faults:

tesseract sample_000111.jpg - -l tessdata_fast/eng+tessdata_fast/ara+ocrb_int
Estimating resolution as 975
Speicherzugriffsfehler

The segmentation fault does not occur with eng and ara from tessdata_best or tessdata.

But I got another issue with a non-existing 2nd language model:

tesseract sample_000111.jpg - -l ara+orcb_int
Error: LSTM requested, but not present!! Loading tesseract.
mgr->GetComponent(TESSDATA_INTTEMP, &fp):Error:Assert failed:in file ../../../src/classify/adaptmatch.cpp, line 539
Abgebrochen

[cchadowitz]

Thanks for taking a look!

@stweil looks like you were able to reproduce, but just fyi the images I referenced are the same as the attached ones (Github just changed the filenames). I shared the example run after the image for each example, so the first image is the referenced sample_012631.jpg and the second is sample_000111.jpg. And @zdenop was correct with the sources of the language packs (originally I found them from https://tesseract-ocr.github.io/tessdoc/Data-Files-Contributions.html)

Anyways, let me know if I can provide anything else in the meantime.

[farfromrefug]

I am getting a lot of reports of crashes in recog_all_words (android app). I am trying to get better reports with exact call-stack .

[marcreichman-pfi]

@stweil are you able to reproduce the segfaults from this and from #4148 against 5.3.4 or current code? We tested against the Ubuntu 24.04-distributed 5.3.4 and it doesn't seem like we can anymore. However, I cannot seem to figure out what in the branch diffs would have changed this. If you have a moment to over it I'd love to know your thoughts.