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: <CAN5uoS_0crJm+ywd6Jg3NxtzO5gyREp35TTfbvFrpsuUyG6MLw@mail.gmail.com>
Date:   Fri, 26 Nov 2021 08:59:45 +0100
From:   Etienne Carriere <etienne.carriere@...aro.org>
To:     Cristian Marussi <cristian.marussi@....com>
Cc:     Sudeep Holla <sudeep.holla@....com>,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
        Vincent Guittot <vincent.guittot@...aro.org>
Subject: Re: [PATCH] firmware: arm_scmi: optee: Drop the support for the OPTEE
 shared dynamic buffer

Hello Sudeep,

On Thu, 25 Nov 2021 at 19:25, Cristian Marussi <cristian.marussi@....com> wrote:
>
> On Thu, Nov 25, 2021 at 03:07:30PM +0000, Sudeep Holla wrote:
> > The shared memory buffer allocated by the optee driver is normal cached
> > memory and can't be used with IOMEM APIs used in shmem_*.
> >
> > We currently support only IO memory for shared memory and supporting
> > normal cached memory needs more changes and needs to be thought through
> > properly. So for now, let us drop the support for this OPTEE shared buffer.
> >
> > Cc: Cristian Marussi <cristian.marussi@....com>
> > Cc: Etienne Carriere <etienne.carriere@...aro.org>
> > Signed-off-by: Sudeep Holla <sudeep.holla@....com>
> > ---
>
> Hi,
>
>
> >  drivers/firmware/arm_scmi/optee.c | 19 +------------------
> >  1 file changed, 1 insertion(+), 18 deletions(-)
> >
> > diff --git a/drivers/firmware/arm_scmi/optee.c b/drivers/firmware/arm_scmi/optee.c
> > index 901737c9f5f8..175b39bcd470 100644
> > --- a/drivers/firmware/arm_scmi/optee.c
> > +++ b/drivers/firmware/arm_scmi/optee.c
> > @@ -282,23 +282,6 @@ static void scmi_optee_clear_channel(struct scmi_chan_info *cinfo)
> >       shmem_clear_channel(channel->shmem);
> >  }
> >
> > -static int setup_dynamic_shmem(struct device *dev, struct scmi_optee_channel *channel)
> > -{
> > -     const size_t msg_size = SCMI_OPTEE_MAX_MSG_SIZE;
> > -
> > -     channel->tee_shm = tee_shm_alloc_kernel_buf(scmi_optee_private->tee_ctx, msg_size);
> > -     if (IS_ERR(channel->tee_shm)) {
> > -             dev_err(channel->cinfo->dev, "shmem allocation failed\n");
> > -             return -ENOMEM;
> > -     }
> > -
> > -     channel->shmem = (void *)tee_shm_get_va(channel->tee_shm, 0);
> > -     memset(channel->shmem, 0, msg_size);
> > -     shmem_clear_channel(channel->shmem);
> > -
> > -     return 0;
> > -}
> > -
> >  static int setup_static_shmem(struct device *dev, struct scmi_chan_info *cinfo,
> >                             struct scmi_optee_channel *channel)
> >  {
> > @@ -342,7 +325,7 @@ static int setup_shmem(struct device *dev, struct scmi_chan_info *cinfo,
> >       if (of_find_property(cinfo->dev->of_node, "shmem", NULL))
> >               return setup_static_shmem(dev, cinfo, channel);
> >       else
> > -             return setup_dynamic_shmem(dev, channel);
> > +             return -ENOMEM;
> >  }
> >

I would rather find an alternate way to support tee shared memory.
I think OP-TEE could use msg.c format when handling tee memory. Linux
and OP-TEE Scmi transport discovery negotiate the channel type and
support for msg format could allow OP-TEE to use its shm management. I
will prepare an implementation but if you prefer the current remove
support and later introduce back tee shm support, I'm fine.

Best regards,
Etienne

>
> LGTM.
>
> Reviewed-by: Cristian Marussi <cristian.marussi@....com>
>
> Thanks,
> Cristian

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ