On 2021-06-24 03:45 PM, Loic Poulain wrote:
Hi Bjorn,
On Thu, 24 Jun 2021 at 22:30, Bjorn Andersson
<bjorn.andersson@xxxxxxxxxx> wrote:
On Thu 24 Jun 15:28 CDT 2021, Loic Poulain wrote:
> Currently, the MHI controller driver defines which channels should
> have their inbound buffers allocated and queued. But ideally, this is
> something that should be decided by the MHI device driver instead,
> which actually deals with that buffers.
>
> Add a flag parameter to mhi_prepare_for_transfer allowing to specify
> if buffers have to be allocated and queued by the MHI stack.
>
> Keep auto_queue flag for now, but should be removed at some point.
>
> Signed-off-by: Loic Poulain <loic.poulain@xxxxxxxxxx>
> Tested-by: Bhaumik Bhatt <bbhatt@xxxxxxxxxxxxxx>
> Reviewed-by: Bhaumik Bhatt <bbhatt@xxxxxxxxxxxxxx>
> Reviewed-by: Hemant Kumar <hemantk@xxxxxxxxxxxxxx>
> Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@xxxxxxxxxx>
> Acked-by: Jakub Kicinski <kuba@xxxxxxxxxx>
> Link: https://lore.kernel.org/r/1621603519-16773-1-git-send-email-loic.poulain@xxxxxxxxxx
> Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@xxxxxxxxxx>
What's the intention with this patch? Why is Mani the last S-o-b when
you're the one sending it and why is it sent only to linux-arm-msm@?
Actually the previous version of that patch has already been applied
to mhi-next, but has been nacked as part of Mani's PR, so it's a quick
follow-up fix to address the issue.
And completely separate of the practical matters, is it true that qrtr
is the only client that use this pre-allocation feature?
yes.
Would it be a net gain to simplify the core and add buffer allocation
to qrtr instead?
Yes, I 100% agree, but I however would prefer to keep that for a
follow-up series since this patch fixes a real issue for MHI/PCI
modems (no inbound QRTR buffers allocated).
Regards,
Loic
I had discussed this with Chris Lew a while ago and he also agreed qrtr
can
queue inbound buffers.
This patch is good and we can continue to have flexibility for clients
since
this allows MHI controller to not have to bother about some of the
unnecessary
plugins/booleans we maintain on behalf of clients.
They can have more control this way and less headache for us.
Thanks,
Bhaumik
---
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
Forum,
a Linux Foundation Collaborative Project
