Re: [dpdk-dev] [PATCH v8 8/9] bus/pci: support Windows with bifurcated drivers

[Dmitry Kozlyuk <dmitry.kozliuk@gmail.com>]

[snip]
> +		/* kernel driver type is unsupported */
> +		RTE_LOG(DEBUG, EAL,
> +			"kernel driver type for PCI device " PCI_PRI_FMT ","
> +			" is unsupported",

Nit: log messages usually start with a capital.

> +			dev->addr.domain, dev->addr.bus,
> +			dev->addr.devid, dev->addr.function);
> +		return -1;
> +	}
> +
> +	return ERROR_SUCCESS;
> +}
[snip]
> +static int
> +pci_scan_one(HDEVINFO dev_info, PSP_DEVINFO_DATA device_info_data)
> +{
> +	struct rte_pci_device *dev;
> +	int ret = -1;
> +
> +	dev = malloc(sizeof(*dev));
> +	if (dev == NULL) {
> +		goto end;
> +	}

Braces not needed here. Checkpatch should've complained.

> +
> +	memset(dev, 0, sizeof(*dev));
> +
> +	char  pci_device_info[PATH_MAX];
> +	BOOL  res;
> +	struct rte_pci_addr addr;
> +	struct rte_pci_id pci_id;
> +
> +	/* Retrieve PCI device IDs */
> +	res = SetupDiGetDeviceRegistryPropertyA(dev_info, device_info_data,
> +			SPDRP_HARDWAREID, NULL, (BYTE *)&pci_device_info,
> +			sizeof(pci_device_info), NULL);
> +	if (!res) {
> +		RTE_LOG_WIN32_ERR(
> +			"SetupDiGetDeviceRegistryPropertyA(SPDRP_HARDWAREID)");
> +		goto end;
> +	}
> +
> +	ret = get_pci_hardware_info((const char *)&pci_device_info, &pci_id);

What do you think of calling SetupDiGetDeviceRegistryPropertyA from
get_pci_hardware_info (like other functions in this file work) and renaming
the latter to something like get_device_pci_id (for what it exactly does)?
Current get_pci_hardware_info may become parse_pci_id static helper.

[snip]
> @@ -165,5 +366,44 @@ pci_uio_remap_resource(struct rte_pci_device *dev __rte_unused)
>  int
>  rte_pci_scan(void)
>  {
> +	int   ret = -1;
> +	DWORD device_index = 0, found_device = 0;
> +	HDEVINFO dev_info;
> +	SP_DEVINFO_DATA device_info_data;
> +
> +	/* for debug purposes, PCI can be disabled */
> +	if (!rte_eal_has_pci())
> +		return 0;
> +
> +	dev_info = SetupDiGetClassDevs(&GUID_DEVCLASS_NET, TEXT("PCI"), NULL,
> +				DIGCF_PRESENT);
> +	if (dev_info == INVALID_HANDLE_VALUE) {
> +		RTE_LOG_WIN32_ERR("SetupDiGetClassDevs(pci_scan)");
> +		RTE_LOG(ERR, EAL, "Unable to enumerate PCI devices.\n");
> +		goto end;
> +	}
> +
> +	device_info_data.cbSize = sizeof(SP_DEVINFO_DATA);
> +	device_index = 0;
> +
> +	while (SetupDiEnumDeviceInfo(dev_info, device_index,
> +	    &device_info_data)) {
> +		device_index++;
> +		ret = pci_scan_one(dev_info, &device_info_data);
> +		if (ret == ERROR_SUCCESS)
> +			found_device++;
> +		else if (ret != ERROR_CONTINUE)
> +			goto end;
> +
> +		memset(&device_info_data, 0, sizeof(SP_DEVINFO_DATA));
> +		device_info_data.cbSize = sizeof(SP_DEVINFO_DATA);
> +	}
> +
> +	RTE_LOG(DEBUG, EAL, "PCI scan found %lu devices\n", found_device);
>  	return 0;

"dev_info" leaks here.

> +end:
> +	if (dev_info != INVALID_HANDLE_VALUE)
> +		SetupDiDestroyDeviceInfoList(dev_info);
> +
> +	return ret;
>  }
[snip]

-- 
Dmitry Kozlyuk