On Fri, 2016-04-15 at 18:25 +0700, Duy Nguyen wrote:
> On Thu, Apr 14, 2016 at 1:47 AM, David Turner <
> dturner@xxxxxxxxxxxxxxxx> wrote:
> > > > + fd = unix_stream_connect(socket_path);
> > > > + if (refresh_cache) {
> > > > + ret = write_in_full(fd, "refresh", 8) != 8;
> > >
> > > Since we've moved to unix socket and had bidirectional
> > > communication,
> > > it's probably a good idea to read an "ok" back, giving index
> > > -helper
> > > time to prepare the cache. As I recall the last discussion with
> > > Johannes, missing a cache here when the index is around 300MB
> > > could
> > > hurt more than wait patiently once and have it ready next time.
> >
> > It is somewhat slower to wait for the daemon (which requires a disk
> > load + a memcpy) than it is to just load it ourselves (which is
> > just a
> > disk load).
>
> You forgot the most costly part, SHA-1 verification. For very large
> index, I assume the index-helper is already in the middle of hashing
> the index content. If you ignore index-helper, you need to go hash
> the
> whole thing again. The index-helper can hand it to you if you wait
> just a bit more. This wait time should be shorter because index
> -helper
> is already in the middle of hashing (and in optimistic case, very
> close to finishing it).
You're right -- I did forget that part.
In "index-helper: use watchman to avoid refreshing index with lstat()",
we switch from just poking to poking and waiting for a reply. Then in
"read-cache: config for waiting for index-helper", we make that waiting
optional. So what if I just remove that patch? Does that solve it?
--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html