[cmaster-next] [PATCH] zebra-mpls: fix regression caused by wrong conflict resolution
Signed-off-by: Renato Westphal <renato@opensourcerouting.org> --- zebra/zebra_vty.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+) diff --git a/zebra/zebra_vty.c b/zebra/zebra_vty.c index f32b51f..bde9431 100644 --- a/zebra/zebra_vty.c +++ b/zebra/zebra_vty.c @@ -918,6 +918,15 @@ vty_show_ip_route_detail (struct vty *vty, struct route_node *rn, int mcast) default: break; } + + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + vty_out (vty, "%s", VTY_NEWLINE); } vty_out (vty, "%s", VTY_NEWLINE); @@ -1158,6 +1167,14 @@ vty_show_ip_route (struct vty *vty, struct route_node *rn, struct rib *rib, break; } + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_BLACKHOLE)) vty_out (vty, ", bh"); if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_REJECT)) -- 1.9.1
Guys, I couldn't identify which merge commit introduced this regression. I then came to realize that our git history is becoming unnecessarily way too hard to follow. Wouldn't it be better if we start avoiding merge commits? In other words, whenever we want to merge a branch, rebase it on top of the base branch before. Then we get a clean fast-forward merge and a commit history that is easy to follow. I think we could do this at least for small branches like -isis-simpl which had only two simple commits. A merge commit for that was unnecessary IMO. On Thu, Dec 8, 2016 at 10:04 PM, Renato Westphal <renato@opensourcerouting.org> wrote:
Signed-off-by: Renato Westphal <renato@opensourcerouting.org> --- zebra/zebra_vty.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
diff --git a/zebra/zebra_vty.c b/zebra/zebra_vty.c index f32b51f..bde9431 100644 --- a/zebra/zebra_vty.c +++ b/zebra/zebra_vty.c @@ -918,6 +918,15 @@ vty_show_ip_route_detail (struct vty *vty, struct route_node *rn, int mcast) default: break; } + + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + vty_out (vty, "%s", VTY_NEWLINE); } vty_out (vty, "%s", VTY_NEWLINE); @@ -1158,6 +1167,14 @@ vty_show_ip_route (struct vty *vty, struct route_node *rn, struct rib *rib, break; }
+ /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_BLACKHOLE)) vty_out (vty, ", bh"); if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_REJECT)) -- 1.9.1
-- Renato Westphal
If we remove the merge commit then when we want to merge the branch again you cannot. I am not sure I am convinced that I want to sacrifice that. donald On Thu, Dec 8, 2016 at 7:20 PM, Renato Westphal <renato@opensourcerouting.org> wrote:
Guys,
I couldn't identify which merge commit introduced this regression. I then came to realize that our git history is becoming unnecessarily way too hard to follow.
Wouldn't it be better if we start avoiding merge commits? In other words, whenever we want to merge a branch, rebase it on top of the base branch before. Then we get a clean fast-forward merge and a commit history that is easy to follow. I think we could do this at least for small branches like -isis-simpl which had only two simple commits. A merge commit for that was unnecessary IMO.
On Thu, Dec 8, 2016 at 10:04 PM, Renato Westphal <renato@opensourcerouting.org> wrote:
Signed-off-by: Renato Westphal <renato@opensourcerouting.org> --- zebra/zebra_vty.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
diff --git a/zebra/zebra_vty.c b/zebra/zebra_vty.c index f32b51f..bde9431 100644 --- a/zebra/zebra_vty.c +++ b/zebra/zebra_vty.c @@ -918,6 +918,15 @@ vty_show_ip_route_detail (struct vty *vty, struct route_node *rn, int mcast) default: break; } + + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + vty_out (vty, "%s", VTY_NEWLINE); } vty_out (vty, "%s", VTY_NEWLINE); @@ -1158,6 +1167,14 @@ vty_show_ip_route (struct vty *vty, struct route_node *rn, struct rib *rib, break; }
+ /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_BLACKHOLE)) vty_out (vty, ", bh"); if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_REJECT)) -- 1.9.1
-- Renato Westphal
_______________________________________________ cmaster-next mailing list cmaster-next@lists.nox.tf https://lists.nox.tf/listinfo/cmaster-next
82992fed8 => d64ff607e d64ff607e mis-resolved the conflict; 82992fe was then a "clean" merge because of the earlier mis-resolution. To find this, the "-m" option on "git log" is very useful. -David On Thu, Dec 08, 2016 at 10:20:44PM -0200, Renato Westphal wrote:
Guys,
I couldn't identify which merge commit introduced this regression. I then came to realize that our git history is becoming unnecessarily way too hard to follow.
Wouldn't it be better if we start avoiding merge commits? In other words, whenever we want to merge a branch, rebase it on top of the base branch before. Then we get a clean fast-forward merge and a commit history that is easy to follow. I think we could do this at least for small branches like -isis-simpl which had only two simple commits. A merge commit for that was unnecessary IMO.
On Thu, Dec 8, 2016 at 10:04 PM, Renato Westphal <renato@opensourcerouting.org> wrote:
Signed-off-by: Renato Westphal <renato@opensourcerouting.org> --- zebra/zebra_vty.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
diff --git a/zebra/zebra_vty.c b/zebra/zebra_vty.c index f32b51f..bde9431 100644 --- a/zebra/zebra_vty.c +++ b/zebra/zebra_vty.c @@ -918,6 +918,15 @@ vty_show_ip_route_detail (struct vty *vty, struct route_node *rn, int mcast) default: break; } + + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + vty_out (vty, "%s", VTY_NEWLINE); } vty_out (vty, "%s", VTY_NEWLINE); @@ -1158,6 +1167,14 @@ vty_show_ip_route (struct vty *vty, struct route_node *rn, struct rib *rib, break; }
+ /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_BLACKHOLE)) vty_out (vty, ", bh"); if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_REJECT)) -- 1.9.1
-- Renato Westphal
_______________________________________________ cmaster-next mailing list cmaster-next@lists.nox.tf https://lists.nox.tf/listinfo/cmaster-next
Applied, thanks! donald On Thu, Dec 8, 2016 at 7:04 PM, Renato Westphal <renato@opensourcerouting.org> wrote:
Signed-off-by: Renato Westphal <renato@opensourcerouting.org> --- zebra/zebra_vty.c | 17 +++++++++++++++++ 1 file changed, 17 insertions(+)
diff --git a/zebra/zebra_vty.c b/zebra/zebra_vty.c index f32b51f..bde9431 100644 --- a/zebra/zebra_vty.c +++ b/zebra/zebra_vty.c @@ -918,6 +918,15 @@ vty_show_ip_route_detail (struct vty *vty, struct route_node *rn, int mcast) default: break; } + + /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + vty_out (vty, "%s", VTY_NEWLINE); } vty_out (vty, "%s", VTY_NEWLINE); @@ -1158,6 +1167,14 @@ vty_show_ip_route (struct vty *vty, struct route_node *rn, struct rib *rib, break; }
+ /* Label information */ + if (nexthop->nh_label && nexthop->nh_label->num_labels) + { + vty_out (vty, " label %s", + mpls_label2str (nexthop->nh_label->num_labels, + nexthop->nh_label->label, buf, BUFSIZ)); + } + if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_BLACKHOLE)) vty_out (vty, ", bh"); if (CHECK_FLAG (rib->flags, ZEBRA_FLAG_REJECT)) -- 1.9.1
_______________________________________________ cmaster-next mailing list cmaster-next@lists.nox.tf https://lists.nox.tf/listinfo/cmaster-next
participants (3)
-
David Lamparter -
Donald Sharp -
Renato Westphal