Re: kern/57683: uvm_pglistalloc_s_ps() et al integer overflow

[Tobias Nygren <tnn%nygren.pp.se@localhost>]

The following reply was made to PR kern/57683; it has been noted by GNATS.

From: Tobias Nygren <tnn%nygren.pp.se@localhost>
To: matthew green <mrg%netbsd.org@localhost>
Cc: gnats-bugs%netbsd.org@localhost
Subject: Re: kern/57683: uvm_pglistalloc_s_ps() et al integer overflow
Date: Mon, 6 Nov 2023 09:15:01 +0100

 On Mon, 06 Nov 2023 13:25:18 +1100
 matthew green <mrg%netbsd.org@localhost> wrote:
 
 > this is not a small problem AFAICT.  getting to 16T by using unsigned
 > page count would help, but it's not nearly enough for the future.
 
 The problem can be divided into supporting large physical
 address space and actually mapping (or avoiding to map) large areas
 with normal sized pages. Word-addressable CXL SSDs that map into
 physical address space are already announced products from the vendors.
 Those will need to use special apertures with large pages to keep the
 page count sane.