Hi, Baolin Wang <baolin.wang@xxxxxxxxxx> writes: > On 26 July 2016 at 07:15, Felipe F. Tonello <eu@xxxxxxxxxxxxxxxxx> wrote: >> USB spec specifies wMaxPacketSize to be little endian (as other properties), >> so when using this variable in the driver we should convert to the current >> CPU endianness if necessary. >> >> Signed-off-by: Felipe F. Tonello <eu@xxxxxxxxxxxxxxxxx> >> --- >> drivers/usb/gadget/function/f_midi.c | 2 +- >> 1 file changed, 1 insertion(+), 1 deletion(-) >> >> diff --git a/drivers/usb/gadget/function/f_midi.c b/drivers/usb/gadget/function/f_midi.c >> index 58fc199a18ec..a83d852b1da5 100644 >> --- a/drivers/usb/gadget/function/f_midi.c >> +++ b/drivers/usb/gadget/function/f_midi.c >> @@ -362,7 +362,7 @@ static int f_midi_set_alt(struct usb_function *f, unsigned intf, unsigned alt) >> struct usb_request *req = >> midi_alloc_ep_req(midi->out_ep, >> max_t(unsigned, midi->buflen, >> - bulk_out_desc.wMaxPacketSize)); >> + le16_to_cpu(bulk_out_desc.wMaxPacketSize))); > > I think here we should use usb_ep_align_maybe() function instead of > max_t() to handle 'quirk_ep_out_aligned_size' quirk, please see the > patch I've send out: https://lkml.org/lkml/2016/7/12/106 agree, if usb_ep_align_maybe() has a bug with endianness, let's fix it since there are other gadgets using usb_ep_align_maybe() -- balbi
Attachment:
signature.asc
Description: PGP signature
