On Sun, Mar 10, 2019 at 04:46:08PM +0200, Yuri Benditovich wrote:
> When processing list of USB devices, avoid allocating memory
> for devices which later will be skipped. Use existing libusb
> struct first to check whether the device shall be excluded.
I probably would reword the short log
"win-usb-dev: do not allocate memory for hub devices"
-> "win-usb-dev: skip hub devices early in g_udev_client_list_devices"
> -static gboolean g_udev_skip_search(GUdevDevice *udev)
> +static gboolean g_udev_skip_search(libusb_device *dev)
> {
> - GUdevDeviceInfo* udevinfo;
> gboolean skip;
> + uint8_t addr = libusb_get_device_address(dev);
> + struct libusb_device_descriptor desc;
>
> - g_return_val_if_fail(G_UDEV_DEVICE(udev), FALSE);
> -
> - udevinfo = udev->priv->udevinfo;
> - g_return_val_if_fail(udevinfo != NULL, FALSE);
> + libusb_get_device_descriptor(dev, &desc);
Do you know if these
libusb_get_device_descriptor/libusb_get_device_address are going to be
expensive compared to using cached values?
>
> - skip = ((udevinfo->addr == 0xff) || /* root hub (HCD) */
> + skip = ((addr == 0xff) || /* root hub (HCD) */
> #if defined(LIBUSBX_API_VERSION) && (LIBUSBX_API_VERSION >= 0x010000FF)
> - (udevinfo->addr == 1) || /* root hub addr for libusbx >= 1.0.13 */
> + (addr == 1) || /* root hub addr for libusbx >= 1.0.13 */
Note for later: I'm not sure this still needs to be conditional as we
depend on libusb 1.0.16 now.
Looks good otherwise,
Christophe
Attachment:
signature.asc
Description: PGP signature
_______________________________________________ Spice-devel mailing list Spice-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/spice-devel