[PATCH] revision: Don't queue uninteresting commits

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Currently all given commits are added to the topo_queue during
init_topo_walk(). Later on in get_revision_1() the uninteresting ones
are skipped because simplify_commit() tells it to.

Let's not add them to the topo_queue in the first place.

Signed-off-by: Øystein Walle <oystwa@xxxxxxxxx>
---

I noticed this while trying to understand the generation based algorithm
introduced in b45424181e (revision.c: generation-based topo-order
algorithm, 2018-11-01) in an attempt to write a similar one for
gitoxide. Comparing my solution to git's output I fixed a mismatch by
essentially doing this, and it turns out it works in git too. I am not
extremely confident, but all the tests pass...

For fun I also tried removing the UNINTERESTING check from
get_commit_action() altogether but then a lot of tests fail. I expected
that because both the flag and function predate the new algorithm.

 revision.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/revision.c b/revision.c
index 2f4c53ea20..deeab813c7 100644
--- a/revision.c
+++ b/revision.c
@@ -3681,7 +3681,8 @@ static void init_topo_walk(struct rev_info *revs)
 	for (list = revs->commits; list; list = list->next) {
 		struct commit *c = list->item;
 
-		if (*(indegree_slab_at(&info->indegree, c)) == 1)
+		if (*(indegree_slab_at(&info->indegree, c)) == 1 &&
+		    !(c->object.flags & UNINTERESTING))
 			prio_queue_put(&info->topo_queue, c);
 	}
 
-- 
2.34.1




[Index of Archives]     [Linux Kernel Development]     [Gcc Help]     [IETF Annouce]     [DCCP]     [Netdev]     [Networking]     [Security]     [V4L]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [Fedora Users]

  Powered by Linux