linux-kernel - Re: [PATCH] remoteproc: qcom_q6v5_mss: Re-order writes to the IMEM region

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: <CAD=FV=VnQCO+y_wy=KQhK3wGwHGfO0+MQntgoPh78ZygcgNiig@mail.gmail.com>
Date: Mon, 19 Aug 2024 16:40:07 -0700
From: Doug Anderson <dianders@...omium.org>
To: Sibi Sankar <quic_sibis@...cinc.com>
Cc: andersson@...nel.org, mathieu.poirier@...aro.org, 
	linux-arm-msm@...r.kernel.org, linux-remoteproc@...r.kernel.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] remoteproc: qcom_q6v5_mss: Re-order writes to the IMEM region

Hi,

On Mon, Aug 19, 2024 at 12:30 AM Sibi Sankar <quic_sibis@...cinc.com> wrote:
>
> Any write access to the IMEM region when the Q6 is setting up XPU
> protection on it will result in a XPU violation. Fix this by ensuring
> IMEM writes related to the MBA post-mortem logs happen before the Q6
> is brought out of reset.
>
> Fixes: 318130cc9362 ("remoteproc: qcom_q6v5_mss: Add MBA log extraction support")
> Signed-off-by: Sibi Sankar <quic_sibis@...cinc.com>
> ---
>  drivers/remoteproc/qcom_q6v5_mss.c | 6 +++---
>  1 file changed, 3 insertions(+), 3 deletions(-)

As discussed offlist, this isn't a perfect fix since writes to this
IMEM could happen by other drivers and those could still cause things
to go boom if they run in parallel with this driver. That being said:
* It seems like a more proper fix needs a coordinated effort between a
device's built-in firmware and the modem firmware. This is difficult /
near impossible to get done properly.
* Even if we do a more proper fix, making this change won't hurt.
* This change will immediately improve things by avoiding the XPU
violation in the most common case.

I've confirmed that the test case I had where things were going boom
is fixed. Thus:

Reviewed-by: Douglas Anderson <dianders@...omium.org>
Tested-by: Douglas Anderson <dianders@...omium.org>