#205 buffer overflow core

[Chris]

so both release src 1.6 and 20250720 do this.
gcc.linux.x64.exe.debug]$ ./syncterm
*** buffer overflow detected ***: terminated
Aborted (core dumped)

I am assuming thats syncterm's test, because it way shouldn't happen with arch/manjaro.
Linux version 6.12.37-1-MANJARO

[Anonymous]

Same issue here on CachyOS (arch-based) running kernel 6.15.7-3-cachyos and KDE plasma on Wayland.

[Anonymous]

Same on CachyOS (arch-based)

[Anonymous]

Same happens with 1.5. Seems to be an issue with newer kernels or maybe even libc (seen tools with similar issues).

[Stephen James Hurd]

Sorry, been doing summer things. Could you attach a properly sanitized syncterm.ini and syncterm.lst file (ie: remove all user/password lines, etc)?

[Stephen James Hurd]

Actually, installing CachyOS in a VM now... will update if I can't repro.

[Anonymous]

Yes, it worked.

|)/`. /\/\/-|) a ()xy()zb()urne venture

On Sun, Aug 10, 2025 at 3:24 PM Stephen James Hurd
deuce@users.sourceforge.net wrote:

status: open --> pending
assigned_to: Stephen James Hurd
Comment:

Ok, should be fixed in git now, and I've kicked off a new test build to update the source/binaries on the website.

The issue appears to be a mismatch between GCC v15 and Cryptlib with regard to what sizeof returns for the size of a struct that ends in a flex array. It appears that GCC is standard compliant in this behaviour, so this was a bug in Cryptlib.

Please comment/close this ticket if the issue is resolved for you.

---

[tickets:#205] buffer overflow core

Status: pending
1.7: 1.7
Created: Mon Jul 21, 2025 02:35 AM UTC by Chris
Last Updated: Sun Aug 10, 2025 06:50 PM UTC
Owner: Stephen James Hurd

so both release src 1.6 and 20250720 do this.
gcc.linux.x64.exe.debug]$ ./syncterm
*** buffer overflow detected ***: terminated
Aborted (core dumped)

I am assuming thats syncterm's test, because it way shouldn't happen with arch/manjaro.
Linux version 6.12.37-1-MANJARO

---

Sent from sourceforge.net because you indicated interest in https://sourceforge.net/p/syncterm/tickets/205/

To unsubscribe from further messages, please visit https://sourceforge.net/auth/subscriptions/

[Stephen James Hurd]

Ok, should be fixed in git now, and I've kicked off a new test build to update the source/binaries on the website.

The issue appears to be a mismatch between GCC v15 and Cryptlib with regard to what sizeof returns for the size of a struct that ends in a flex array. It appears that GCC is standard compliant in this behaviour, so this was a bug in Cryptlib.

Please comment/close this ticket if the issue is resolved for you.

[Stephen James Hurd]

Fix confirmed.