> This makes more than one patch. I count: > > - using buf size as a compile time constant > > - changing the index scheme > > - a few gratuitous changes like > unsigned len; > ... > len = ...; > > becoming > unsigned len = ...; Really, I was just doing #2. #1 I did because it was less work than any alternative scheme for guaranteeing that that buf_size was always a power of two. I suppose I could split it up, but doing that will reduce the patch by 2 deleted lines and 4 changed ones, a total of 10 lines of delta. I'll happily do it if you like, but I plead honest belief that it was too small a difference to be worth dividing. It would have been extra effort to create an intermediate version and I was trying to create a simple patch, not a patch series. As for #3, if you look, I added a line of the form + unsigned put_pos = gb->buf_put % WRITE_BUF_SIZE; And I adpated the "len =" assignment to match. It wasn't entirely gratuitous. My personal habit is to use initializers for short simple expressions. I wrote my additions that way, then noticed the inconsistency was awkward, so one or the other had to change. I thought making them both initializers was cleanest. Is there a style preference against initializers in the USB code? > - maybe more I also added a clarifying comment to gs_buf_free. Again, it is my habit to add a comment when I spend more than a few seconds wondering "why did they do that?", and in that case, the reason for setting gb->buf_buf to NULL was not completely obvious. Anyway, the following series of 6 patches is the smallest units I can break it down to. Number 3 is the big one, but number 2 is a prerequisite. Number 6 is particularly unimportant. -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
