[PATCH v2 1/1] usb: Assimilate usb_get_descriptor() to linux

Before this commit, usb_get_descriptor() failed for some flakey USB devices. We hereby adopt the more robust linux implementation [1].
For instance, for the "Alcor Micro Corp. Flash Drive" (VID 0x058f, PID 0x6387), the following behavior occurs from time to time:
=> usb start starting USB... Bus xhci_pci: Register 10000840 NbrPorts 16 Starting the controller USB XHCI 1.20 scanning bus xhci_pci for devices... usb_new_device: Cannot read configuration, skipping device 058f:6387
Signed-off-by: Philip Oberfichtner pro@denx.de
[1] From a38297e3fb012 (Linux 6.9), see https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/driver... ---
Notes: Changes in V2: Adapt commit message to - state kernel version - state which exact USB device is being fixed - include error log
common/usb.c | 37 ++++++++++++++++++++++++++++++------- 1 file changed, 30 insertions(+), 7 deletions(-)
diff --git a/common/usb.c b/common/usb.c index 99e6b857c7..661ec0a9c4 100644 --- a/common/usb.c +++ b/common/usb.c @@ -215,8 +215,9 @@ int usb_int_msg(struct usb_device *dev, unsigned long pipe, * clear keyboards LEDs). For data transfers, (storage transfers) we don't * allow control messages with 0 timeout, by previousely resetting the flag * asynch_allowed (usb_disable_asynch(1)). - * returns the transferred length if OK or -1 if error. The transferred length - * and the current status are stored in the dev->act_len and dev->status. + * returns the transferred length if OK, otherwise a negative error code. The + * transferred length and the current status are stored in the dev->act_len and + * dev->status. */ int usb_control_msg(struct usb_device *dev, unsigned int pipe, unsigned char request, unsigned char requesttype, @@ -258,11 +259,14 @@ int usb_control_msg(struct usb_device *dev, unsigned int pipe, break; mdelay(1); } + + if (timeout == 0) + return -ETIMEDOUT; + if (dev->status) return -1;
return dev->act_len; - }
/*------------------------------------------------------------------- @@ -563,10 +567,29 @@ int usb_clear_halt(struct usb_device *dev, int pipe) static int usb_get_descriptor(struct usb_device *dev, unsigned char type, unsigned char index, void *buf, int size) { - return usb_control_msg(dev, usb_rcvctrlpipe(dev, 0), - USB_REQ_GET_DESCRIPTOR, USB_DIR_IN, - (type << 8) + index, 0, buf, size, - USB_CNTL_TIMEOUT); + int i; + int result; + + if (size <= 0) /* No point in asking for no data */ + return -EINVAL; + + memset(buf, 0, size); /* Make sure we parse really received data */ + + for (i = 0; i < 3; ++i) { + /* retry on length 0 or error; some devices are flakey */ + result = usb_control_msg(dev, usb_rcvctrlpipe(dev, 0), + USB_REQ_GET_DESCRIPTOR, USB_DIR_IN, + (type << 8) + index, 0, buf, size, + USB_CNTL_TIMEOUT); + if (result <= 0 && result != -ETIMEDOUT) + continue; + if (result > 1 && ((u8 *)buf)[1] != type) { + result = -ENODATA; + continue; + } + break; + } + return result; }
/**********************************************************************

On 6/4/24 12:18 PM, Philip Oberfichtner wrote:
Before this commit, usb_get_descriptor() failed for some flakey USB devices. We hereby adopt the more robust linux implementation [1].
For instance, for the "Alcor Micro Corp. Flash Drive" (VID 0x058f, PID 0x6387), the following behavior occurs from time to time:
=> usb start starting USB... Bus xhci_pci: Register 10000840 NbrPorts 16 Starting the controller USB XHCI 1.20 scanning bus xhci_pci for devices... usb_new_device: Cannot read configuration, skipping device 058f:6387
Signed-off-by: Philip Oberfichtner pro@denx.de
[1] From a38297e3fb012 (Linux 6.9), see https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/tree/driver...
Reviewed-by: Marek Vasut marex@denx.de
Thanks
participants (2)
-
Marek Vasut
-
Philip Oberfichtner