Miklos Vajna <vmiklos@xxxxxxxxxxxxxx> writes: > On Sun, Jul 13, 2008 at 07:46:59PM +0200, Miklos Vajna <vmiklos@xxxxxxxxxxxxxx> wrote: >> So far what I see is that the input for the reduce_heads() function is >> (c1, c0, c2, c0, c1). The expected output would be (c1, c2), but the >> actual output is c2. So I suspect the bug is not in builtin-merge.c >> itself but in reduce_heads(). > > This fixes the problem for me. Junio, does the fix looks correct to you > as well? You are correct, the "item"s are the highlander (i.e. "there can be only one") objects but commit-list elements that hold pointers to them are not, so we need to dereference and compare. Thanks. -- 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
