linux-kernel - Re: [PATCH v4] bus: mhi: host: pci_generic: Add support for QDU100 device

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <540fa0c1-2e84-0c11-87e9-09198d6f270f@quicinc.com>
Date: Wed, 11 Dec 2024 08:24:18 -0700
From: Jeffrey Hugo <quic_jhugo@...cinc.com>
To: Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>,
        Vivek Pernamitta
	<quic_vpernami@...cinc.com>
CC: <quic_krichai@...cinc.com>, <mhi@...ts.linux.dev>,
        <linux-arm-msm@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v4] bus: mhi: host: pci_generic: Add support for QDU100
 device

On 12/10/2024 11:41 PM, Manivannan Sadhasivam wrote:
> On Mon, Dec 09, 2024 at 11:09:50AM +0530, Vivek Pernamitta wrote:
>> Add MHI controller configuration for QDU100 device.
>>
>> The Qualcomm X100 5G RAN Accelerator card is designed to enhance Open vRAN
>> servers by offloading CPUs from intensive 5G baseband functions.
>>
>> Link: https://docs.qualcomm.com/bundle/publicresource/87-79371-1_REV_A_Qualcomm_X100_5G_RAN_Accelerator_Card_Product_Brief.pdf
>>
>> Signed-off-by: Vivek Pernamitta <quic_vpernami@...cinc.com>
>> ---
>> changes from V3:
>> - Removed IP_SW1 and IP_SW2 support currently, will add once
>>    SW_IP1/2 channels support are added to mhi-net driver.
>> - Removed qdu100 edl images, as EDL images are not needed.
>> - Added space inbetween kernel_ulong_t and mhi_qcom_qdu100_info.
>> - QDU100 XBL FW images were pushed to linux-firmware git repo
>>    https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/qcom/qdu100
>>
>> changes from V2:
>> - updated commit text.
>>
>> changes from V1:
>> - Changing naming convention from modem_qcom_qdu100*
>>    to mhi_qcom_qdu100*.
>> - Updated commit text.
>> - Fixed and corrected by passing mhi_pci_dev_info struct
>>    instead of mhi_controller_config.
>> ---
>>
>> ---
>>   drivers/bus/mhi/host/pci_generic.c | 55 ++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 55 insertions(+)
>>
>> diff --git a/drivers/bus/mhi/host/pci_generic.c b/drivers/bus/mhi/host/pci_generic.c
>> index 07645ce2119a71c9277356e962252b840379cd81..dee9fa9e7ae441fbc9a86e53694568c0ba192002 100644
>> --- a/drivers/bus/mhi/host/pci_generic.c
>> +++ b/drivers/bus/mhi/host/pci_generic.c
>> @@ -245,6 +245,58 @@ struct mhi_pci_dev_info {
>>   		.channel = ch_num,		\
>>   	}
>>   
>> +static const struct mhi_channel_config mhi_qcom_qdu100_channels[] = {
>> +	MHI_CHANNEL_CONFIG_UL(0, "LOOPBACK", 32, 2),
>> +	MHI_CHANNEL_CONFIG_DL(1, "LOOPBACK", 32, 2),
>> +	MHI_CHANNEL_CONFIG_UL_SBL(2, "SAHARA", 128, 1),
>> +	MHI_CHANNEL_CONFIG_DL_SBL(3, "SAHARA", 128, 1),
>> +	MHI_CHANNEL_CONFIG_UL(4, "DIAG", 64, 3),
>> +	MHI_CHANNEL_CONFIG_DL(5, "DIAG", 64, 3),
>> +	MHI_CHANNEL_CONFIG_UL(9, "QDSS", 64, 3),
>> +	MHI_CHANNEL_CONFIG_UL(14, "NMEA", 32, 4),
>> +	MHI_CHANNEL_CONFIG_DL(15, "NMEA", 32, 4),
>> +	MHI_CHANNEL_CONFIG_UL(16, "CSM_CTRL", 32, 4),
>> +	MHI_CHANNEL_CONFIG_DL(17, "CSM_CTRL", 32, 4),
>> +	MHI_CHANNEL_CONFIG_UL(40, "MHI_PHC", 32, 4),
>> +	MHI_CHANNEL_CONFIG_DL(41, "MHI_PHC", 32, 4),
>> +	MHI_CHANNEL_CONFIG_UL(46, "IP_SW0", 256, 5),
>> +	MHI_CHANNEL_CONFIG_DL(47, "IP_SW0", 256, 5),
>> +};
>> +
>> +static struct mhi_event_config mhi_qcom_qdu100_events[] = {
>> +	/* first ring is control+data ring */
>> +	MHI_EVENT_CONFIG_CTRL(0, 64),
>> +	/* SAHARA dedicated event ring */
>> +	MHI_EVENT_CONFIG_SW_DATA(1, 256),
>> +	/* Software channels dedicated event ring */
>> +	MHI_EVENT_CONFIG_SW_DATA(2, 64),
>> +	MHI_EVENT_CONFIG_SW_DATA(3, 256),
>> +	MHI_EVENT_CONFIG_SW_DATA(4, 256),
>> +	/* Software IP channels dedicated event ring */
>> +	MHI_EVENT_CONFIG_SW_DATA(5, 512),
>> +	MHI_EVENT_CONFIG_SW_DATA(6, 512),
>> +	MHI_EVENT_CONFIG_SW_DATA(7, 512),
>> +};
>> +
>> +static const struct mhi_controller_config mhi_qcom_qdu100_config = {
>> +	.max_channels = 128,
>> +	.timeout_ms = 120000,
>> +	.num_channels = ARRAY_SIZE(mhi_qcom_qdu100_channels),
>> +	.ch_cfg = mhi_qcom_qdu100_channels,
>> +	.num_events = ARRAY_SIZE(mhi_qcom_qdu100_events),
>> +	.event_cfg = mhi_qcom_qdu100_events,
>> +};
>> +
>> +static const struct mhi_pci_dev_info mhi_qcom_qdu100_info = {
>> +	.name = "qcom-qdu100",
>> +	.fw = "qcom/qdu100/xbl_s.melf",
> 
> Can you first push this fw to linux-firmware? I haven't enforced this
> requirement earlier, but that was my mistake.

Mani, this is already done.  There is a handy link in the change log. 
Did you miss that, or is there something still missing?

> 
>> +	.edl_trigger = true,
>> +	.config = &mhi_qcom_qdu100_config,
>> +	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
>> +	.dma_data_width = 32,
>> +	.sideband_wake = false,
>> +};
>> +
>>   static const struct mhi_channel_config modem_qcom_v1_mhi_channels[] = {
>>   	MHI_CHANNEL_CONFIG_UL(4, "DIAG", 16, 1),
>>   	MHI_CHANNEL_CONFIG_DL(5, "DIAG", 16, 1),
>> @@ -822,6 +874,9 @@ static const struct pci_device_id mhi_pci_id_table[] = {
>>   	/* NETPRISMA FCUN69 (SDX6X) */
>>   	{ PCI_DEVICE(PCI_VENDOR_ID_NETPRISMA, 0x1001),
>>   		.driver_data = (kernel_ulong_t) &mhi_netprisma_fcun69_info },
>> +	/* QDU100, x100-DU */
>> +	{ PCI_DEVICE(PCI_VENDOR_ID_QCOM, 0x0601),
>> +		.driver_data = (kernel_ulong_t) &mhi_qcom_qdu100_info },
> 
> Did you read the comment on top of mhi_pci_id_table[]?
> 
> - Mani
>