On Fri, Mar 13, 2015 at 9:20 PM, Paul Bolle <[email protected]> wrote: > On Wed, 2015-03-11 at 16:04 +0100, Valentin Rothberg wrote: >> Paul, how long does your monster run? Maybe I just call it wrong or >> mess up with caches. > > Even longer, I presume. Because an update for just a new linux next > release can take over a minute on my fastest machine (a ThinkPad X220). > > (Recall that my monster runs daily over just the blobs added to the tree > for the latest linux-next tag and stores an intermediate parse as a git > note to each of those new blobs. It then does a second parse of all the > git notes relevant for that tag and stores the final result as a git > note to that tag. > > And I do all this to make the daily update run at a decent speed. A > downside of this approach is that the very first run, which has to > parse, say, 50.000 new blobs, takes ages.)
Thank you for the explanation. I was really surprised how long (compared to reset) it takes. > My suggestion won't do here. So let me just say that messing with the > state of peoples repository might make you the target of a flame or two. > Are you sure you want to go down that route? In case running the script would delete one's day of work or worse ... no ... I don't want to take responsibility for that. The patch at the current state is unacceptable, but I see two options to solve the issue while being fast: (1) Test if the current tree is dirty, warn the user and ask if she wants to continue or not. (2) Abort if the tree is dirty. Personally, I prefer option 2. The script would still be fast, and there is no way that it deletes data by accident. Kind regards, Valentin -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
