Giter Club home page Giter Club logo

Comments (6)

prashantpandey avatar prashantpandey commented on August 20, 2024

Hi @koadman ,

Could you please share the code to reproduce the bug?

Thanks,
Prashant

from cqf.

koadman avatar koadman commented on August 20, 2024

Hi Prashant thanks for the quick reply. I've had some difficulty isolating the exact same conditions from the rest of the codebase I'm working on but have now found two small-ish test cases that reproduce what look like a very similar or possibly the same issue.
The first is given in cqf_debug.cpp, which I am building as:

g++ cqf_debug.cpp  -o cqf_debug -O0 -g3 -I. /home/koadman/software/cqf/obj/*.o -std=c++11 -I/home/koadman/software/cqf/include/ -pthread -lz -lssl -lcrypto -lm

with gcc version 5.4.0 20160609 on intel x64.
When I run it I see:

gdb ./cqf_debug
...
Reading symbols from ./cqf_debug...done.
(gdb) r keys_1.txt keys_2.txt 
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
Resizing the CQF.
Resizing the CQF.
Resizing the CQF.
Resizing the CQF.

Program received signal SIGSEGV, Segmentation fault.
0x00000000004036a1 in shift_remainders (qf=0x7fffffffd8d0, start_index=7692, empty_index=7680) at src/gqf.c:711
711			*REMAINDER_WORD(qf, last_word) = shift_into_b(*REMAINDER_WORD(qf, last_word-1),
(gdb) bt
#0  0x00000000004036a1 in shift_remainders (qf=0x7fffffffd8d0, start_index=7692, empty_index=7680) at src/gqf.c:711
#1  0x0000000000405717 in insert1 (qf=0x7fffffffd8d0, hash=493509468, runtime_lock=4 '\004') at src/gqf.c:1372
#2  0x000000000040746b in qf_insert (qf=0x7fffffffd8d0, key=493509468, value=0, count=1, flags=4 '\004') at src/gqf.c:1922
#3  0x00000000004017f4 in main (argc=3, argv=0x7fffffffde08) at cqf_debug.cpp:44

The interesting thing is that if I comment out the code block that creates and loads keys to the first QF, then the 2nd QF does not segfault. Possibly there is some global or file-scoped variable that's creating an unexpected dependency between the two QFs?
cqf_debug.cpp.gz
keys_2.txt.gz
keys_1.txt.gz

from cqf.

koadman avatar koadman commented on August 20, 2024

The second, possibly related problem I have isolated into a small test example happens when merging two QFs. As above, compiling with:

g++ cqf_debug2.cpp  -o cqf_debug2 -O0 -g3 -I. /home/koadman/software/cqf/obj/*.o -std=c++11 -I/home/koadman/software/cqf/include/ -pthread -lz -lssl -lcrypto -lm

and running:

gdb ./cqf_debug2
...
Reading symbols from ./cqf_debug2...done.
(gdb) r
...

Program received signal SIGSEGV, Segmentation fault.
0x0000000000403c1b in run_end (qf=0x7fffffffdc60, hash_bucket_index=80) at src/gqf.c:601
601					runend_block_offset = bitselectv(get_block(qf,
(gdb) bt
#0  0x0000000000403c1b in run_end (qf=0x7fffffffdc60, hash_bucket_index=80) at src/gqf.c:601
#1  0x0000000000406544 in insert (qf=0x7fffffffdc60, hash=330765, count=16, runtime_lock=9 '\t') at src/gqf.c:1444
#2  0x0000000000407eb0 in qf_insert (qf=0x7fffffffdc60, key=330765, value=0, count=16, flags=9 '\t') at src/gqf.c:1924
#3  0x000000000040a0a6 in qf_multi_merge (qf_arr=0x621c20, nqf=2, qfr=0x7fffffffdc60) at src/gqf.c:2498
#4  0x00000000004018ba in merger (cqf1="cqf_2.1", cqf2="cqf_2.2", cqf_out="cqf_2.merged") at cqf_debug2.cpp:27
#5  0x0000000000401af7 in main (argc=1, argv=0x7fffffffde28) at cqf_debug2.cpp:36
(gdb)

cqf_debug2.cpp.gz
cqf_1.1.gz
cqf_1.2.gz
cqf_2.1.gz
cqf_2.2.gz
(the above files need to be unzipped first of course)

interestingly, I see a similar behavior as before, where if I comment out the merge of the first two QFs then the 2nd merge is able to complete successfully, otherwise it segfaults.

from cqf.

koadman avatar koadman commented on August 20, 2024

quick update: i still have these problems. I have been able to work around them by doing two things:

  1. only insert to a single CQF per program invocation
  2. avoid CQF auto-resizing

of course a solution that eliminates the segfault would be preferable!

from cqf.

kylebd99 avatar kylebd99 commented on August 20, 2024

Hey,
I had this same issue, and I think it derives from some reliance on the buffer's state. I changed the malloc call for the filter's buffer to a calloc call and this issue disappeared.

from cqf.

koadman avatar koadman commented on August 20, 2024

cool! maybe send a pull request?

from cqf.

Related Issues (8)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.