1. Forum
  2. FidoNet
  3. SYNC SYSOPS

  • sexyz sometimes crashes with a 'buffer overflow' message

    From Eric Oulashin@1:103/705 to GitLab issue in main/sbbs on Monday, September 04, 2023 16:52:34
    open https://gitlab.synchro.net/main/sbbs/-/issues/614

    I saw this with sexyz for Linux built on September 2nd 10:14AM). Upon uploading, it looks like sexyz sometimes crashes with a 'buffer overflow' error.I made a core dump available here (this is for the release build though):http://www.digitaldistortionbbs.com:81/synchronet_stuff/core.sexyz.1000.58c6d2c0dc9145e69c895496d017a1eb.406883.1693871056000000.zstFrom syslog:Sep 4 16:44:16 EO-BBS-NIX synchronet: term Node 2 <Nightfox> sexyz: Reading /home/erico/BBS/sbbs/exec/sexyz.iniSep 4 16:44:16 EO-BBS-NIX synchronet: term Node 2 <Nightfox> sexyz: *** buffer overflow detected ***: terminatedSep 4 16:44:16 EO-BBS-NIX systemd[1]: Started Process Core Dump (PID 406885/UID 0).Sep 4 16:44:16 EO-BBS-NIX systemd-coredump[406886]: Process 406883 (sexyz) of user 1000 dumped core.#012#012Found module /home/erico/BBS/sbbs/exec/sexyz with build-id: 5c4b0c867a5cd63e01670b5b61ce8476606c3d6c#012Found module linux-vdso.so.1 with build-id: 5a18f0aeef268b5653f8a845ea6994eaa99c909e#012Found module ld-linux-x86-64.so.2 with build-id: 61ef896a699bb1c2e4e231642b2e1688b2f1a61e#012Found module libc.so.6 with build-id: 69389d485a9793dbe873f0ea2c93e02efaa9aa3d#012Stack trace of thread 406883:#012#0 0x00007f98a43a1a7c __pthread_kill_implementation (libc.so.6 + 0x96a7c)#012#1 0x00007f98a434d476 __GI_raise (libc.so.6 + 0x42476)#012#2 0x00007f98a43337f3 __GI_abort (libc.so.6 + 0x287f3)#012#3 0x00007f98a43946f6 __libc_message (libc.so.6 + 0x896f6)#012#4 0x00007f98a444176a __GI___fortify_fail (libc.so.6 + 0x13676a)#012#5 0x00007f98a44400c6 __GI___chk_fail (libc.so.6 + 0x1350c6)#012#6 0x00007f98a44416ab __fdelt_chk (libc.so.6 + 0x1366ab)#012#7 0x000055b014b39fb5 n/a (/home/erico/BBS/sbbs/exec/sexyz + 0x8fb5)
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Nigel Reed@1:103/705 to GitLab note in main/sbbs on Monday, September 04, 2023 23:00:10
    https://gitlab.synchro.net/main/sbbs/-/issues/614#note_3891

    Did you try a cleanall.sh and then rebuild? I've seen peripheral programs fail because they didn't recompile for some reason.Also, what client was your user using? I just uploaded to my BBS 10 times with no core dumps.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Eric Oulashin@1:103/705 to GitLab note in main/sbbs on Tuesday, September 05, 2023 10:48:23
    https://gitlab.synchro.net/main/sbbs/-/issues/614#note_3893

    I hadn't tried cleannall and rebuild yet.I don't know what client my user was using, but I tried it myself with SyncTerm and saw it coredump. I don't think the user's client would matter, as sexyz is crashing on the Synchronet side before the upload even starts.
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)
  • From Rob Swindell@1:103/705 to GitLab note in main/sbbs on Tuesday, September 05, 2023 11:25:56
    https://gitlab.synchro.net/main/sbbs/-/issues/614#note_3896

    What are the contents of the sexyz.ini file being read?Can you provide a gdb-backtrace from the coredump?Have you enabled any special build options (e.g. SANITIZE or FORTIFY)?
    --- SBBSecho 3.20-Linux
    * Origin: Vertrauen - [vert/cvs/bbs].synchro.net (1:103/705)