linux-kernel - Re: [PATCH] KVM: guest_memfd: Refactor kvm_gmem into inode->i

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <ZRMxfTd65Ijn3RAj@google.com>
Date:   Tue, 26 Sep 2023 12:31:09 -0700
From:   Sean Christopherson <seanjc@...gle.com>
To:     isaku.yamahata@...el.com
Cc:     kvm@...r.kernel.org, linux-kernel@...r.kernel.org,
        isaku.yamahata@...il.com, Michael Roth <michael.roth@....com>,
        Paolo Bonzini <pbonzini@...hat.com>,
        linux-coco@...ts.linux.dev, Chao Peng <chao.p.peng@...ux.intel.com>
Subject: Re: [PATCH] KVM: guest_memfd: Refactor kvm_gmem into inode->i_private

On Tue, Sep 26, 2023, isaku.yamahata@...el.com wrote:
> From: Isaku Yamahata <isaku.yamahata@...el.com>
> 
> Refactor guest_memfd to use inode->i_private to store info about kvm_gmem.

Why!?!?!?  Sadly, I don't have telepathic superpowers.  This changelog only
explains *what* the code is doing, what I need to know is *why* you want to make
this change.

The below kinda sorta suggests that this is to simplify the code, but it's not
at all obvious that that's the actual motivation, or the only motiviation.

> Currently it is stored in the following way.
> - flags in inode->i_private
> - struct kvm_gmem in file->private_data
> - struct kvm_gmem in linked linst in inode->i_mapping->private_list
>   And this list has single entry.
> 
> The relationship between struct file, struct inode and struct kvm_gmem is
> 1:1, not 1:many. 

No, it's not.  Or at least it won't be in the future.  As I explained before[1]:

 : The code is structured to allow for multiple gmem instances per inode.  This isn't
 : actually possible in the initial code base, but it's on the horizon[*].  I included
 : the list-based infrastructure in this initial series to ensure that guest_memfd
 : can actually support multiple files per inode, and to minimize the churn when the
 : "link" support comes along.

 : [*] https://lore.kernel.org/all/cover.1691446946.git.ackerleytng@google.com

[1] https://lore.kernel.org/all/ZQsAiGuw%2F38jIOV7@google.com