Date:   Tue, 7 Dec 2021 17:26:56 +0530
From:   Vinod Koul <vkoul@kernel.org>
To:     Georgi Djakov <djakov@kernel.org>
Cc:     Rob Herring <robh+dt@kernel.org>, linux-arm-msm@vger.kernel.org,
        Bjorn Andersson <bjorn.andersson@linaro.org>,
        Andy Gross <agross@kernel.org>,
        Odelu Kukatla <okukatla@codeaurora.org>,
        linux-pm@vger.kernel.org, devicetree@vger.kernel.org,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] interconnect: qcom: Add SM8450 interconnect provider
 driver
Message-ID: <Ya9MCCYHIAjaW6GM@matsya>
References: <20211201072557.3968915-1-vkoul@kernel.org>
 <20211201072557.3968915-3-vkoul@kernel.org>
 <b4c43541-ffcd-d43c-3405-86d770905dd2@kernel.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <b4c43541-ffcd-d43c-3405-86d770905dd2@kernel.org>
Precedence: bulk

On 01-12-21, 18:46, Georgi Djakov wrote:
> Hi Vinod,
> 
> Thanks for working on this!
> 
> On 1.12.21 9:25, Vinod Koul wrote:
> > Add driver for the Qualcomm interconnect buses found in SM8450 based
> > platforms. The topology consists of several NoCs that are controlled by
> > a remote processor that collects the aggregated bandwidth for each
> > master-slave pairs.
> > 
> > This is based on the downstream driver by
> > Vivek Aknurwar <viveka@codeaurora.org>
> 
> Maybe CC him too.

Will do

> > Signed-off-by: Vinod Koul <vkoul@kernel.org>
> > ---
> >   drivers/interconnect/qcom/Kconfig  |    9 +
> >   drivers/interconnect/qcom/Makefile |    2 +
> >   drivers/interconnect/qcom/sm8450.c | 1988 ++++++++++++++++++++++++++++
> >   drivers/interconnect/qcom/sm8450.h |  169 +++
> >   4 files changed, 2168 insertions(+)
> >   create mode 100644 drivers/interconnect/qcom/sm8450.c
> >   create mode 100644 drivers/interconnect/qcom/sm8450.h
> > 
> [..]
> > +static struct platform_driver qnoc_driver = {
> > +	.probe = qnoc_probe,
> > +	.remove = qnoc_remove,
> > +	.driver = {
> > +		.name = "qnoc-sm8450",
> > +		.of_match_table = qnoc_of_match,
> > +		.sync_state = icc_sync_state,
> 
> We should set this callback only after we enable the client drivers to
> request bandwidth. Otherwise some path might get disabled because of no
> users. I would suggest to add sync_state after we describe the paths in
> DT.

Okay thanks for the suggestion. yeah we have only few paths described. I
will keep this as a separate patch and send few months down the line :)

-- 
~Vinod