Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <11e03e8d-7711-330d-e0d4-808ef9acec3a@igalia.com>
Date:   Sat, 8 Oct 2022 11:14:11 -0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.2.2
Subject: Re: [PATCH 5/8] pstore: Fix long-term implicit conversions in the
 compression routines
Content-Language: en-US
To:     Kees Cook <keescook@chromium.org>, Ard Biesheuvel <ardb@kernel.org>
Cc:     linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
        kernel-dev@igalia.com, kernel@gpiccoli.net, anton@enomsg.org,
        ccross@android.com, tony.luck@intel.com
References: <20221006224212.569555-1-gpiccoli@igalia.com>
 <20221006224212.569555-6-gpiccoli@igalia.com>
 <202210061634.758D083D5@keescook>
 <CAMj1kXF27wZYzXm1u3kKSBtbG=tcK7wOwq6YTwpFg+Z7ic4siQ@mail.gmail.com>
 <202210071234.D289C8C@keescook>
From:   "Guilherme G. Piccoli" <gpiccoli@igalia.com>
In-Reply-To: <202210071234.D289C8C@keescook>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 07/10/2022 16:37, Kees Cook wrote:
> On Fri, Oct 07, 2022 at 10:47:01AM +0200, Ard Biesheuvel wrote:
>> [...]
>> Isn't this the stuff we want to move into the crypto API?
> 
> It is, yes. Guilherme, for background:
> https://lore.kernel.org/linux-crypto/20180802215118.17752-1-keescook@chromium.org/
> 

Thanks a bunch Kees / Ard for pointing me that!

But I'm confused with regards to the state of this conversion: the
patches seem to be quite mature and work fine, but at the same time,
they focus in what Herbert consider a deprecated/old API, so they were
never merged right?

The proposal from Ard (to move to crypto scomp/acomp) allow to rework
the zbufsize worst case thing and wire it on such new API, correct? Do
you intend to do that Kees?

At the same time, I feel it is still valid to avoid these bunch of
implicit conversions on pstore, as this patch proposes - what do you all
think?

I could rework this one on top of Ard's acomp migration while we don't
have an official zbufsize API for on crypto scomp - and once we have,
it'd be just a matter of removing the zbufsize functions of pstore and
make use of the new API, which shouldn't be affected by this implicit
conversion fix.

Cheers,


Guilherme