Protecting the JVM from segfaults

[Maurizio Cimadamore]

On 07/04/2021 12:06, Maurizio Cimadamore wrote:
> he way the Foreign Memory Access API protects against SEGV is not by 
> wrapping them as regular Java exception (I'm not sure how that would 
> be possible in the general case)

To expand a bit more on that: we do have signal handlers installed by 
the VM whenever we are accessing off-heap memory _already_.

The main problem with this strategy is that when you access memory 
already freed, you are not _guaranteed_ to get a SEGV - if the memory 
has been reallocated by somebody else, you might just end up silently 
corrupting that memory. The same thing happens for access out of bounds 
(you might accidentally access a piece of memory which is actually 
accessible, but is not "for you to access").

In other words, I don't think the try/catch model works for these kind 
of situations.

Maurizio