Re: [PATCH v3 8/9] mailbox: mediatek: Add CMDQ secure mailbox driver

From: Jason-JH Lin (林睿祥)
Date: Wed Jan 03 2024 - 01:54:07 EST


On Thu, 2023-12-28 at 01:38 +0000, CK Hu (胡俊光) wrote:
> On Wed, 2023-12-27 at 07:25 +0000, Jason-JH Lin (林睿祥) wrote:
> > Hi CK,
> >
> > Thanks for the reviews.
> >
> > On Tue, 2023-12-26 at 08:58 +0000, CK Hu (胡俊光) wrote:
> > > Hi, Jason:
> > >
> > > On Fri, 2023-12-22 at 12:52 +0800, Jason-JH.Lin wrote:
> > > > To support secure video path feature, GCE have to read/write
> > > > registgers
> > > > in the secure world. GCE will enable the secure access
> > > > permission
> > > > to
> > > > the
> > > > HW who wants to access the secure content buffer.
> > > >
> > > > Add CMDQ secure mailbox driver to make CMDQ client user is able
> > > > to
> > > > sending their HW settings to the secure world. So that GCE can
> > > > execute
> > > > all instructions to configure HW in the secure world.
> > > >
> > > > Signed-off-by: Jason-JH.Lin <jason-jh.lin@xxxxxxxxxxxx>
> > > > ---
> > >
> > > [snip]
> > >
> > > > +
> > > > +static int cmdq_sec_fill_iwc_msg(struct cmdq_sec_context
> > > > *context,
> > > > + struct cmdq_sec_task
> > > > *sec_task, u32
> > > > thrd_idx)
> > > > +{
> > > > + struct iwc_cmdq_message_t *iwc_msg = NULL;
> > > > + struct cmdq_sec_data *data = (struct cmdq_sec_data
> > > > *)sec_task-
> > > > > task.pkt->sec_data;
> > > >
> > > > + u32 size = 0, offset = 0, *instr;
> > > > +
> > > > + iwc_msg = (struct iwc_cmdq_message_t *)context-
> > > > >iwc_msg;
> > > > +
> > > > + if (sec_task->task.pkt->cmd_buf_size + 4 *
> > > > CMDQ_INST_SIZE >
> > > > CMDQ_TZ_CMD_BLOCK_SIZE) {
> > > > + pr_err("sec_task:%p size:%zu > %u",
> > > > + sec_task, sec_task->task.pkt-
> > > > >cmd_buf_size,
> > > > CMDQ_TZ_CMD_BLOCK_SIZE);
> > > > + return -EFAULT;
> > > > + }
> > > > +
> > > > + if (thrd_idx == CMDQ_INVALID_THREAD) {
> > > > + iwc_msg->command.cmd_size = 0;
> > > > + iwc_msg->command.metadata.addr_list_length = 0;
> > > > + return -EINVAL;
> > > > + }
> > > > +
> > > > + iwc_msg->command.thread = thrd_idx;
> > > > + iwc_msg->command.scenario = sec_task->scenario;
> > > > + iwc_msg->command.eng_flag = sec_task->engine_flag;
> > > > + size = sec_task->task.pkt->cmd_buf_size;
> > > > + memcpy(iwc_msg->command.va_base + offset, sec_task-
> > > > >task.pkt-
> > > > > va_base, size);
> > >
> > > I think it's not necessary to copy from normal command buffer to
> > > normal
> > > command buffer. Just
> > >
> > > iwc_msg->command.va_base = sec_task->task.pkt->va_base;
> > >
> > > You have a secure command buffer for secure gce to execute
> > > command.
> > > If
> > > you want to modify command buffer, copy from normal command
> > > buffer
> > > to
> > > secure command buffer, and modify the secure command buffer.
> > >
> > > Regards,
> > > CK
> > >
> > >
> >
> > I think secure world won't recognize normal world's va, so we nee
> > to
> > copy to iwc_msg->command.va_base, which is a world share memory
> > first.
>
> Referring to cmdq_sec_allocate_wsm(), any kernel buffer could use
> tee_shm_register_kernel_buf() to become share buffer between normal
> world and secure world. In client driver, use
> tee_client_open_context()
> to create tee context and register normal command buffer to be share
> buffer. This would reduce the redundant memory copy.
>
> Regards,
> CK
>

OK, I'll try to register normal cmd buffer to share memory.

Regards,
Jason-JH.Lin

> >
> > Regards,
> > Jason-JH.Lin