linux-kernel - Re: [PATCH 14/21] binder: do not add pages to LRU in release path

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

Message-ID: <ZWmIJ4vmjBOICFDJ@google.com>
Date:   Fri, 1 Dec 2023 07:15:51 +0000
From:   Carlos Llamas <cmllamas@...gle.com>
To:     Alice Ryhl <aliceryhl@...gle.com>
Cc:     Arve Hjønnevåg <arve@...roid.com>,
        Christian Brauner <brauner@...nel.org>,
        Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Joel Fernandes <joel@...lfernandes.org>,
        kernel-team@...roid.com, linux-kernel@...r.kernel.org,
        Martijn Coenen <maco@...roid.com>,
        Suren Baghdasaryan <surenb@...gle.com>,
        Todd Kjos <tkjos@...roid.com>
Subject: Re: [PATCH 14/21] binder: do not add pages to LRU in release path

On Tue, Nov 07, 2023 at 09:08:29AM +0000, Alice Ryhl wrote:
> Carlos Llamas <cmllamas@...gle.com> writes:
> > In binder_alloc_deferred_release() pages are added to the LRU list via
> > binder_free_buf_locked(). However, this is pointless because these pages
> > are kfree'd immediately afterwards. Add an option to skip the LRU list.
> 
> They aren't freed with kfree, buf with __free_page.
>  
> > Signed-off-by: Carlos Llamas <cmllamas@...gle.com>
> 
> The change itself looks correct, so I'll give my tag for this:
> 
> Reviewed-by: Alice Ryhl <aliceryhl@...gle.com>
> 
> But I'm wondering whether process cleanup really is so performance
> intensive to justify the added complexity of this?

So, this was needed on an earlier version of the patchset and I was
hoping that it would also help with an issue reported here:
https://lore.kernel.org/all/ZSHmtLqtNZRAtaZ0@google.com/

However, I do agree that it is unecessary at this stage so I've decided
to drop it from v2.

Thanks,
--
Carlos Llamas