On Fri. 22 Apr. 2022 at 22:38, Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> wrote: > On Thu, Mar 17, 2022 at 12:55:05PM +0900, Vincent Mailhol wrote: > > This series rework usb_maxpacket() to change its prototype from: > > | static inline __u16 > > | usb_maxpacket(struct usb_device *udev, int pipe, int is_out) > > > > into: > > | static inline u16 usb_maxpacket(struct usb_device *udev, int pipe) > > > > and rewrite the function using usb_pipe_endpoint(). > > > > Rationale: > > > > * The third argument of usb_maxpacket(): is_out can be derived from > > its second one: pipe using usb_pipeout(pipe). Furthermore, > > usb_pipeout(pipe) is being called within usb_maxpacket() > > regardless to confirm the input. > > > > * This function is not exposed to the UAPI so return type should be > > u16, not __u16. > > > > * usb_pipe_endpoint() (defined a couple of lines before) does most > > of the work. Use it instead of reimplementing the same thing > > again. > > > > In order to do so, usb_maxpacket() is temporarily changed into a variadic > > function to ignore its third argument. > > > > The series goes as follow: > > > > * Patch 1: make usb_maxpacket() variadic to accommodate during the > > migration > > > > * Patch 2 to 7: migrate all the users of usb_maxpacket() (one > > patch per tree) > > > > * Patch 8: remove the third argument of usb_maxpacket() now that > > everyone is migrated > > > > * Patch 9: rework usb_maxpacket()'s body using usb_pipe_endpoint() > > Now queued up, thanks for sticking with this! Thanks to you for your patience and explanations! > Can you send the follow-up patch that removes the vararg? I'll keep > that in my tree to try to catch any remaining things that come in and > get it merged when it's ok to. It is this patch: https://lore.kernel.org/linux-usb/20220317035514.6378-9-mailhol.vincent@xxxxxxxxxx/ Or do you prefer me to resend it on the mailing list? Yours sincerely, Vincent Mailhol
