Bug#1104930: glib2.0: CVE-2025-4373

[Simon McVittie]

On Thu, 08 May 2025 at 21:28:38 +0200, Salvatore Bonaccorso wrote:
>The following vulnerability was published for glib2.0.
>
>CVE-2025-4373[0]:
>| A flaw was found in GLib, which is vulnerable to an integer overflow
>| in the g_string_insert_unichar() function. When the position at
>| which to insert the character is large, the position will overflow,
>| leading to a buffer underwrite.

If I'm reading correctly, this can only happen if the string length is 
around 2GB in 32-bit processes, or extremely large (half the address 
space) in 64-bit processes, such that a gssize (same type as ssize_t) 
would reach a signed overflow and wrap around?

If that assessment is correct, then this doesn't seem like an urgent 
problem to fix, or particularly likely to be exploited in practice.

     smcv