On Sat, 2010-05-08 at 16:07 -0500, Jonathan Nieder wrote: > Without this change, is_alias is likely to happen to be nonzero, > resulting in "fatal: invalid --pretty format" when the fake alias > cannot be resolved. > > Use memset instead of initializing the members one by one to make it > easier to expand the struct in the future if needed. > > t4205 (log --pretty) does not pass for me without this fix. > > Cc: Will Palmer <wmpalmer@xxxxxxxxx> > Signed-off-by: Jonathan Nieder <jrnieder@xxxxxxxxx> > --- > Sorry I missed this before. Sane? > > Jonathan Ah, looks right. I think previous versions of my patch were building in a local variable (which /was/ initialized), then copied the whole thing into the newly-allocated space. When this was changed to building in-place, the initialization was lost. Good catch, thanks. Not sure if Signed-off-by or Reviewed-by is the appropriate tag to mention here, but one of those, I assume. -- -- Will -- 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