[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <0b019df0-f874-4218-a7fc-222b0e35727a@oss.qualcomm.com>
Date: Fri, 14 Nov 2025 08:53:59 +0530
From: Krishna Chaitanya Chundru <krishna.chundru@....qualcomm.com>
To: Bjorn Andersson <andersson@...nel.org>
Cc: cros-qcom-dts-watchers@...omium.org,
Konrad Dybcio <konradybcio@...nel.org>, Rob Herring <robh@...nel.org>,
Krzysztof Kozlowski <krzk+dt@...nel.org>,
Conor Dooley
<conor+dt@...nel.org>,
Lorenzo Pieralisi <lpieralisi@...nel.org>,
Manivannan Sadhasivam <mani@...nel.org>,
Krzysztof Wilczyński <kwilczynski@...nel.org>,
Bjorn Helgaas <bhelgaas@...gle.com>, Jingoo Han <jingoohan1@...il.com>,
linux-arm-msm@...r.kernel.org, devicetree@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-pci@...r.kernel.org,
quic_vbadigan@...cinc.com, quic_mrana@...cinc.com,
quic_vpernami@...cinc.com, mmareddy@...cinc.com,
Konrad Dybcio <konrad.dybcio@....qualcomm.com>
Subject: Re: (subset) [PATCH v8 0/5] PCI: dwc: Add ECAM support with iATU
configuration
On 11/13/2025 11:03 PM, Bjorn Andersson wrote:
> On Thu, Nov 13, 2025 at 09:27:36AM +0530, Krishna Chaitanya Chundru wrote:
>> On 11/10/2025 11:51 PM, Bjorn Andersson wrote:
>>> On Tue, Oct 28, 2025 at 11:12:23PM +0530, Krishna Chaitanya Chundru wrote:
>>>> On 10/28/2025 4:07 AM, Bjorn Andersson wrote:
>>>>> On Thu, 28 Aug 2025 13:04:21 +0530, Krishna Chaitanya Chundru wrote:
>>>>>> The current implementation requires iATU for every configuration
>>>>>> space access which increases latency & cpu utilization.
>>>>>>
>>>>>> Designware databook 5.20a, section 3.10.10.3 says about CFG Shift Feature,
>>>>>> which shifts/maps the BDF (bits [31:16] of the third header DWORD, which
>>>>>> would be matched against the Base and Limit addresses) of the incoming
>>>>>> CfgRd0/CfgWr0 down to bits[27:12]of the translated address.
>>>>>>
>>>>>> [...]
>>>>> Applied, thanks!
>>>>>
>>>>> [1/5] arm64: dts: qcom: sc7280: Increase config size to 256MB for ECAM feature
>>>>> commit: 03e928442d469f7d8dafc549638730647202d9ce
>>>> Hi Bjorn,
>>>>
>>>> Can you revert this change, this is regression due to this series due to
>>>> that we have change the logic,
>>> How is that possible? This is patch 1 in the series, by definition it
>>> doesn't have any outstanding dependencies.
>> The regression is due to the drivers changes on non qcom platforms.
>>
> Please be specific when you're answering, this way of saying "go fish"
> isn't helpful.
>
> By investing a little bit more time and writing a single sentence to
> share what you know, you could have enlightened me and other readers of
> this email list.
Sorry, for not being clear in the previous reply.
The driver is implemented in a assumption that config space start
address will
be the dbi address. The driver is using config space start address as
dbi address
and ignoring the dbi address provided in the device tree. This
assumption is
breaking other vendor driver i.e SiFive driver. To fix this regression
we have removed
this assumption keeping dbi and config space as separate entity.
Due to these driver changes, we have to move DBI to different location
(controller driver
has ability to move the dbi space) for this reason we need to revert this.
- Krishna Chaitanya.
> Regards,
> Bjorn
>
>> - Krishna Chaitanya.
>>> I've reverted the change.
>>>
>>> Regards,
>>> Bjorn
>>>
>>>> we need to update the dtsi accordingly, I will send a separate for all
>>>> controllers to enable this ECAM feature.
>>>>
>>>> - Krishna Chaitanya.
>>>>
>>>>
>>>>> Best regards,
Powered by blists - more mailing lists