On 2021-06-09 12:23:23, Jens Wiklander wrote:
> Uses the new simplified tee_shm_alloc_kernel_buf() function instead of
> the old deprecated tee_shm_alloc() function which required specific
> TEE_SHM-flags.
>
> Signed-off-by: Jens Wiklander <[email protected]>
Since this series is essentially a rewrite of the shm allocation logic,
it is worth pointing out that the rewrite still uses contiguous
allocations (from alloc_pages()). The tee_bnxt_fw driver is performing
an order-10 allocation which is the max, by default. I've only tested
tee_bnxt_fw when it was built-in to the kernel and tee_bnxt_fw_probe()
was called early in boot but I suspect that it might not succeed when
built as a module and loaded later after memory is segmented. I think
this driver would benefit from being able to request a non-contiguous
allocation.
Is this rewrite a good time to offer drivers a way to perform a
non-contiguous allocation?
Tyler
> ---
> drivers/firmware/broadcom/tee_bnxt_fw.c | 5 ++---
> 1 file changed, 2 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/firmware/broadcom/tee_bnxt_fw.c b/drivers/firmware/broadcom/tee_bnxt_fw.c
> index ed10da5313e8..56d00ddd4357 100644
> --- a/drivers/firmware/broadcom/tee_bnxt_fw.c
> +++ b/drivers/firmware/broadcom/tee_bnxt_fw.c
> @@ -212,10 +212,9 @@ static int tee_bnxt_fw_probe(struct device *dev)
>
> pvt_data.dev = dev;
>
> - fw_shm_pool = tee_shm_alloc(pvt_data.ctx, MAX_SHM_MEM_SZ,
> - TEE_SHM_MAPPED | TEE_SHM_DMA_BUF);
> + fw_shm_pool = tee_shm_alloc_kernel_buf(pvt_data.ctx, MAX_SHM_MEM_SZ);
> if (IS_ERR(fw_shm_pool)) {
> - dev_err(pvt_data.dev, "tee_shm_alloc failed\n");
> + dev_err(pvt_data.dev, "tee_shm_alloc_kernel_buf failed\n");
> err = PTR_ERR(fw_shm_pool);
> goto out_sess;
> }
> --
> 2.31.1
>
On Wed, Jun 9, 2021 at 4:58 PM Tyler Hicks <[email protected]> wrote:
>
> On 2021-06-09 12:23:23, Jens Wiklander wrote:
> > Uses the new simplified tee_shm_alloc_kernel_buf() function instead of
> > the old deprecated tee_shm_alloc() function which required specific
> > TEE_SHM-flags.
> >
> > Signed-off-by: Jens Wiklander <[email protected]>
>
> Since this series is essentially a rewrite of the shm allocation logic,
> it is worth pointing out that the rewrite still uses contiguous
> allocations (from alloc_pages()). The tee_bnxt_fw driver is performing
> an order-10 allocation which is the max, by default. I've only tested
> tee_bnxt_fw when it was built-in to the kernel and tee_bnxt_fw_probe()
> was called early in boot but I suspect that it might not succeed when
> built as a module and loaded later after memory is segmented. I think
> this driver would benefit from being able to request a non-contiguous
> allocation.
>
> Is this rewrite a good time to offer drivers a way to perform a
> non-contiguous allocation?
Good idea, I'll look into that. I'll add it as a separate patch if it works OK.
Cheers,
Jens