On Wed, Jun 29, 2011 at 02:51:42PM -0400, Alan Stern wrote: > On Wed, 29 Jun 2011, Kirill Smelkov wrote: > > > Yes, but still it would be good to always keep the invariant > > > > allocated <= uframe_periodic_max > > > > and that debug is there to catch when this breaks. > > Then perhaps it should print out the maximum number of microseconds > already allocated for any uframe, instead of stopping as soon as it > finds something above the new limit. Yes, this makes sense, thanks. I've adjusted the patch accordingly. > > > Can you make that check conditional on DEBUG being set? > > > > Yes I can, but it seems to me we are starting to complicate the code. > > > > What's the problem with returning error on setting uframe_periodic_max < > > already allocated usb bandwith? > > No problem, really. > > > The checking is not a priority for me here, so if you think it's better not > > to check or do it under #ifdef - let's do it. Though of course we all > > have our preferences :) > > Yes, it's just a matter of taste. I prefer to add as little code as > possible for a feature that won't be used much. Look: we have 3 choices: to check or not to check, and if we are checking whether it is under #ifdef DEBUG. Not checking at all makes the code inconsistent with the check in periodic_usecs(). Then, if we are checking, putting it under #ifdef DEBUG goes against minimum complexity / "to add as little code as possible ...". It looks the good balance is to leave the checking as is module you above remark. And with the above-mentioned correction and comments inspired by Sarah's concern I'm resending the whole series as v3. Can we do it this way, please? Thanks, Kirill -- 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
