https://bugs.openldap.org/show_bug.cgi?id=10404

--- Comment #17 from Ondřej Kuzník <[email protected]> ---
On Wed, Feb 18, 2026 at 03:55:54PM +0000, [email protected] wrote:
> This trace is suspicious. There is no such function "slap_sasl_getconfpath" in
> OpenLDAP slapd.
> 
> ==1559117== 31 bytes in 1 blocks are definitely lost in loss record 542 of
> 1,022
> ==1559117==    at 0x4844818: malloc (vg_replace_malloc.c:446)
> ==1559117==    by 0x1A7031: slap_sasl_getconfpath (sasl.c:1248)
> ==1559117==    by 0x48E6044: sasl_server_init (in 
> /usr/lib/x86_64-linux-gnu/libsasl2.so.2.0.25)
> ==1559117==    by 0x1A7B6C: slap_sasl_init (sasl.c:1291)
> ==1559117==    by 0x17EEF4: slap_init (init.c:152)
> ==1559117==    by 0x130AAB: main (main.c:651)
> 
> It seems you're running code with 3rd party patches. We can't help you with
> that.

This code would appear to be coming from a Debian-specific patch:
https://salsa.debian.org/openldap-team/openldap/-/blob/master/debian/patches/sasl-default-path

SASL_CB_GETCONFPATH callback contract is undocumented but it does handle
global strings so the assumption is that it's the callback's
responsibility to manage the lifetime of the returned string *somehow*.

As such this is per-server-init allocation, so not really a meaningful
leak.

-- 
You are receiving this mail because:
You are on the CC list for the issue.

Reply via email to