On Mon, May 20, 2019 at 11:00 AM Jakub Jelinek <ja...@redhat.com> wrote: > > On Mon, May 20, 2019 at 10:16:26AM +0200, Richard Biener wrote: > > There's not many details there but I always thought that the > > "interesting" parts of the GCC history (official branches and trunk) > > do _not_ contain merge commits so the GCC history forms a tree > > rather than a DAG. (I'm not sure if there's an easy way to verify > > whether there are any merge commits on a specific branch) > > Some people do merge commits to release branches, or even to trunk. > I see svn:mergeinfo property e.g. on the following branches: > gcc-4.4 /trunk:155835,156042 > gcc-4.7 > /trunk:185428,186391,186587,187608,187999,190407,190800,191014,197180,197284,206124-206126 > gcc-4.8 /:197284 > /trunk:197180,197284,197677,197951,202766,204547,206123-206126,206395,206628,207387,210668,211733,215049 > gcc-4.9 /branches/gcc-5-branch:230092 > /trunk:210615,210668,211733,214798,215049,215136,215176 > gcc-5 /trunk:222186,230397,231455,232818,241491,243528,243962,244001 > gcc-6 /trunk:239173,239656,241491,243962,244001,249272 > gcc-7 /branches/gcc-8-branch:262746-262747 > /trunk:249272,249411,250097,253187,253451,254099,254866,262104,262442,262744 > gcc-8 /trunk:260603,262104,262221,262442,262744 > Current trunk doesn't have svn:mergeinfo, but in the past I've removed it > several times. I see e.g. r250835, r249115, r249050, r236795, r210113, ... > > Would be nice if during the conversion we could say we want to ignore > svn:mergeinfo on some branches altogether.
I think the best would be to convert to linear history - the thing we want to enforce for later commits anyway. IIRC svn:mergeinfo isn't always created (as you said we removed that property multiple times) so I wonder how the conversion can pick up "merges" from SVN data anyways? Richard. > > Jakub
