linux-pci.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
To: Johan Hovold <johan@kernel.org>
Cc: Andy Gross <agross@kernel.org>,
	Bjorn Andersson <bjorn.andersson@linaro.org>,
	Rob Herring <robh+dt@kernel.org>,
	Krzysztof Kozlowski <krzk@kernel.org>,
	Jingoo Han <jingoohan1@gmail.com>,
	Gustavo Pimentel <gustavo.pimentel@synopsys.com>,
	Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>,
	Bjorn Helgaas <bhelgaas@google.com>,
	Stanimir Varbanov <svarbanov@mm-sol.com>,
	Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>,
	Vinod Koul <vkoul@kernel.org>,
	linux-arm-msm@vger.kernel.org, linux-pci@vger.kernel.org,
	devicetree@vger.kernel.org
Subject: Re: [PATCH v11 3/7] PCI: dwc: Handle MSIs routed to multiple GIC interrupts
Date: Mon, 23 May 2022 16:39:56 +0300	[thread overview]
Message-ID: <8ce50a9f-241d-c37a-15e9-1a97d410f61e@linaro.org> (raw)
In-Reply-To: <Yos9fkgxAN1jJ4jO@hovoldconsulting.com>

On 23/05/2022 10:53, Johan Hovold wrote:
> On Fri, May 20, 2022 at 09:31:10PM +0300, Dmitry Baryshkov wrote:
>> On some of Qualcomm platforms each group of 32 MSI vectors is routed to the
>> separate GIC interrupt. Implement support for such configurations by
>> parsing "msi0" ... "msiN" interrupts and attaching them to the chained
>> handler.
>>
>> Note, that if DT doesn't list an array of MSI interrupts and uses single
>> "msi" IRQ, the driver will limit the amount of supported MSI vectors
>> accordingly (to 32).
>>
>> Signed-off-by: Dmitry Baryshkov <dmitry.baryshkov@linaro.org>
>> ---
>>   .../pci/controller/dwc/pcie-designware-host.c | 58 +++++++++++++++++--
>>   1 file changed, 54 insertions(+), 4 deletions(-)
>>
>> diff --git a/drivers/pci/controller/dwc/pcie-designware-host.c b/drivers/pci/controller/dwc/pcie-designware-host.c
>> index a076abe6611c..381bc24d5715 100644
>> --- a/drivers/pci/controller/dwc/pcie-designware-host.c
>> +++ b/drivers/pci/controller/dwc/pcie-designware-host.c
>> @@ -288,6 +288,43 @@ static void dw_pcie_msi_init(struct pcie_port *pp)
>>   	dw_pcie_writel_dbi(pci, PCIE_MSI_ADDR_HI, upper_32_bits(msi_target));
>>   }
>>   
>> +static const char * const split_msi_names[] = {
>> +	"msi0", "msi1", "msi2", "msi3",
>> +	"msi4", "msi5", "msi6", "msi7",
>> +};
>> +
>> +static int dw_pcie_parse_split_msi_irq(struct pcie_port *pp)
>> +{
>> +	struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
>> +	struct device *dev = pci->dev;
>> +	struct platform_device *pdev = to_platform_device(dev);
>> +	int irq;
>> +	u32 ctrl;
>> +
>> +	irq = platform_get_irq_byname_optional(pdev, split_msi_names[0]);
>> +	if (irq == -ENXIO)
>> +		return -ENXIO;
> 
> You still need to check for other errors and -EPROBE_DEFER here.

I think even the if (irq < 0) return irq; will work here.

> 
>> +
>> +	pp->msi_irq[0] = irq;
>> +
>> +	/* Parse as many IRQs as described in the DTS. */
> 
> s/DTS/devicetree/
> 
>> +	for (ctrl = 1; ctrl < MAX_MSI_CTRLS; ctrl++) {
>> +		irq = platform_get_irq_byname_optional(pdev, split_msi_names[ctrl]);
>> +		if (irq == -ENXIO)
>> +			break;
>> +		if (irq < 0)
>> +			return dev_err_probe(dev, irq,
>> +					     "Failed to parse MSI IRQ '%s'\n",
>> +					     split_msi_names[ctrl]);
>> +
>> +		pp->msi_irq[ctrl] = irq;
>> +	}
>> +
>> +	pp->num_vectors = ctrl * MAX_MSI_IRQS_PER_CTRL;
>> +
>> +	return 0;
>> +}
>> +
>>   static int dw_pcie_msi_host_init(struct pcie_port *pp)
>>   {
>>   	struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
>> @@ -295,22 +332,34 @@ static int dw_pcie_msi_host_init(struct pcie_port *pp)
>>   	struct platform_device *pdev = to_platform_device(dev);
>>   	int ret;
>>   	u32 ctrl, num_ctrls;
>> +	bool has_split_msi_irq = false;
> 
> This one should go in the follow-on patch that starts using it.
> 
>> -	num_ctrls = pp->num_vectors / MAX_MSI_IRQS_PER_CTRL;
>> -	for (ctrl = 0; ctrl < num_ctrls; ctrl++)
>> +	for (ctrl = 0; ctrl < MAX_MSI_CTRLS; ctrl++)
>>   		pp->irq_mask[ctrl] = ~0;
>>   
>> +	if (!pp->msi_irq[0]) {
>> +		ret = dw_pcie_parse_split_msi_irq(pp);
>> +		if (ret < 0 && ret != -ENXIO)
>> +			return ret;
>> +	}
>> +
>> +	if (!pp->num_vectors)
>> +		pp->num_vectors = MSI_DEF_NUM_VECTORS;
> 
> This works, but now you override num_vectors unconditionally when using
> split msis (and not just when num_vectors is set to zero) >
> Is it work allowing to use num_vectors as a maximum as in previous
> versions (if only for consistency)?

Let me take a look.

> 
>> +	num_ctrls = pp->num_vectors / MAX_MSI_IRQS_PER_CTRL;
>> +
>>   	if (!pp->msi_irq[0]) {
>>   		int irq = platform_get_irq_byname_optional(pdev, "msi");
>>   
>>   		if (irq < 0) {
>>   			irq = platform_get_irq(pdev, 0);
>>   			if (irq < 0)
>> -				return irq;
>> +				return dev_err_probe(dev, irq, "Failed to parse MSI irq\n");
>>   		}
>>   		pp->msi_irq[0] = irq;
>>   	}
>>   
>> +	dev_dbg(dev, "Using %d MSI vectors\n", pp->num_vectors);
>> +
>>   	pp->msi_irq_chip = &dw_pci_msi_bottom_irq_chip;
>>   
>>   	ret = dw_pcie_allocate_domains(pp);
>> @@ -407,7 +456,8 @@ int dw_pcie_host_init(struct pcie_port *pp)
>>   				     of_property_read_bool(np, "msi-parent") ||
>>   				     of_property_read_bool(np, "msi-map"));
>>   
>> -		if (!pp->num_vectors) {
>> +		/* for the has_msi_ctrl the default assignment is handled inside dw_pcie_msi_host_init() */
>> +		if (!pp->has_msi_ctrl && !pp->num_vectors) {
>>   			pp->num_vectors = MSI_DEF_NUM_VECTORS;
>>   		} else if (pp->num_vectors > MAX_MSI_IRQS) {
>>   			dev_err(dev, "Invalid number of vectors\n");
> 
> Johan


-- 
With best wishes
Dmitry

  reply	other threads:[~2022-05-23 13:40 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-20 18:31 [PATCH v11 0/7] PCI: qcom: Fix higher MSI vectors handling Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 1/7] PCI: dwc: Convert msi_irq to the array Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 2/7] PCI: dwc: split MSI IRQ parsing/allocation to a separate function Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 3/7] PCI: dwc: Handle MSIs routed to multiple GIC interrupts Dmitry Baryshkov
2022-05-23  7:53   ` Johan Hovold
2022-05-23 13:39     ` Dmitry Baryshkov [this message]
2022-05-23 14:02       ` Johan Hovold
2022-05-23 15:17         ` Dmitry Baryshkov
2022-05-23 15:32           ` Johan Hovold
2022-05-23 15:36             ` Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 4/7] PCI: dwc: Implement special ISR handler for split MSI IRQ setup Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 5/7] dt-bindings: PCI: qcom: Support additional MSI interrupts Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 6/7] arm64: dts: qcom: sm8250: provide " Dmitry Baryshkov
2022-05-20 18:31 ` [PATCH v11 7/7] dt-bindings: mfd: qcom,qca639x: add binding for QCA639x defvice Dmitry Baryshkov
2022-05-23  7:42 ` [PATCH v11 0/7] PCI: qcom: Fix higher MSI vectors handling Johan Hovold
2022-05-23 13:03   ` Dmitry Baryshkov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=8ce50a9f-241d-c37a-15e9-1a97d410f61e@linaro.org \
    --to=dmitry.baryshkov@linaro.org \
    --cc=agross@kernel.org \
    --cc=bhelgaas@google.com \
    --cc=bjorn.andersson@linaro.org \
    --cc=devicetree@vger.kernel.org \
    --cc=gustavo.pimentel@synopsys.com \
    --cc=jingoohan1@gmail.com \
    --cc=johan@kernel.org \
    --cc=krzk@kernel.org \
    --cc=linux-arm-msm@vger.kernel.org \
    --cc=linux-pci@vger.kernel.org \
    --cc=lorenzo.pieralisi@arm.com \
    --cc=manivannan.sadhasivam@linaro.org \
    --cc=robh+dt@kernel.org \
    --cc=svarbanov@mm-sol.com \
    --cc=vkoul@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).