I am curious why the range of the wide vtable is not validated.

[류대호]

Hello Mr. maintatiners My name is daepo, and i'm studying cybersecurity in korea university. I am curious why _IO_WIDE_JUMPS_FUNC does not validate wide vtables, whereas _IO_JUMPS_FUNC validates the vtable range. I couldn't find the answer through my searches. I am asking because I believe the libc-maintainers might know the reason or be able to provide a reasonable explanation. My own reasoning is: It is ineffective. Since FSOP can be performed using other techniques even if this is blocked, there is no need to go through the trouble of patching it. It causes significant performance degradation. I couldn't find anything through searching, so GPT was the only person I could ask. I explained my own speculative reason, but they said that the path for processing wide characters has changed, which could cause a much larger drop in performance. However, since they claim this isn't the official stance, I cannot trust it.I am curious why _IO_WIDE_JUMPS_FUNC does not validate wide vtables, whereas _IO_JUMPS_FUNC validates the vtable range. I couldn't find the answer through my searches. thank you for reading this mail.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://sourceware.org/pipermail/libc-alpha/attachments/20260322/90c7c9e9/attachment-0001.htm>