[Olsr-dev] Strange shift behaviour (ipcalc.c)

[Hannes Gredler]

On Mon, Dec 01, 2008 at 01:12:40PM +0100, Bernd Petrovitsch wrote:
| On Mon, 2008-12-01 at 11:36 +0100, Hannes Gredler wrote:
| > On Mon, Dec 01, 2008 at 12:06:56AM +0100, Bernd Petrovitsch wrote:
| > | On Son, 2008-11-30 at 22:42 +0100, Bernd Petrovitsch wrote:
| > | [...]
| > | > I think I will make a sane inline function for this in src/ipcalc.h and
| > | > use that everywhere.
| > | 
| > | see http://gredler.at/hg/olsrd/rev/15bb00f25929
| > | The commit message should sday it all:
| > | ----  snip  ----
| > | Fixed an embarrassing bug if we want to generate a netmask from a prefix == 0.
| > | Added an inline function to avoid someone else making the same mistake
| > | 
| > | All credits belong to Markus Kittenberger <(spam-protected)> for finding,
| > | reporting and hunting it down to the root cause.
| > | ----  snip  ----
| > | Ok, the English could be better....
| > 
| > sorry i lost track a bit .. does the bug also apply to 0.5.6 ?
| 
| http://gredler.at/hg/olsrd-0.5.6/?
| Yes AFAICS.
| 
| > if so it might be good to add before cutting 0.5.6-r3 ...
| 
| Hmm, I don't know how you generate a new release?
| transplanting worthwile patches or just taking
| http://gredler.at/hg/olsrd or ?????

agreed procedure is to backport "sensible" (the ipcalc one certainly does apply)
patches to the stable branch (http://gredler.at/hg/olsrd-0.5.6/ ) and then
simply tag this tree with the R3 tag and publish it.
so transplant is the way to go ...

/hannes