Skip to contentSkip to conversation

ip: promote missed packets to the -s row

PATCH Applied Jakub Kicinski 3 participants 7 messages Sep 16, 2020 – Sep 23, 2020

Subsystems (2)
lore
Share & tools
lore

Jakub Kicinski

author maintainer
#0 lore raw

PATCH 1/1 ip: promote missed packets to the -s row

missed_packet_errors are much more commonly reported:

linux$ git grep -c '[.>]rx_missed_errors ' -- drivers/ | wc -l 64
linux$ git grep -c '[.>]rx_over_errors ' -- drivers/ | wc -l 37

Plus those drivers are generally more modern than those using rx_over_errors.

Since recently merged kernel documentation makes this preference official, let's make ip -s output more informative and let rx_missed_errors take the place of rx_over_errors.

Before:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

---
 ip/ipaddress.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
ip/ipaddress.c +4 -4
index ccf67d1dd55c..0822a8d2d792 100644
@@ -744,7 +744,7 @@ static void __print_link_stats(FILE *fp, struct rtattr *tb[])
 		close_json_object();
 	} else {
 		/* RX stats */
-		fprintf(fp, "    RX: bytes  packets  errors  dropped overrun mcast   %s%s",
+		fprintf(fp, "    RX: bytes  packets  errors  dropped missed  mcast   %s%s",
 			s->rx_compressed ? "compressed" : "", _SL_);
 
 		fprintf(fp, "    ");
@@ -752,7 +752,7 @@ static void __print_link_stats(FILE *fp, struct rtattr *tb[])
 		print_num(fp, 8, s->rx_packets);
 		print_num(fp, 7, s->rx_errors);
 		print_num(fp, 7, s->rx_dropped);
-		print_num(fp, 7, s->rx_over_errors);
+		print_num(fp, 7, s->rx_missed_errors);
 		print_num(fp, 7, s->multicast);
 		if (s->rx_compressed)
 			print_num(fp, 7, s->rx_compressed);
@@ -760,14 +760,14 @@ static void __print_link_stats(FILE *fp, struct rtattr *tb[])
 		/* RX error stats */
 		if (show_stats > 1) {
 			fprintf(fp, "%s", _SL_);
-			fprintf(fp, "    RX errors: length   crc     frame   fifo    missed%s%s",
+			fprintf(fp, "    RX errors: length   crc     frame   fifo    overrun%s%s",
 				s->rx_nohandler ? "   nohandler" : "", _SL_);
 			fprintf(fp, "               ");
 			print_num(fp, 8, s->rx_length_errors);
 			print_num(fp, 7, s->rx_crc_errors);
 			print_num(fp, 7, s->rx_frame_errors);
 			print_num(fp, 7, s->rx_fifo_errors);
-			print_num(fp, 7, s->rx_missed_errors);
+			print_num(fp, 7, s->rx_over_errors);
 			if (s->rx_nohandler)
 				print_num(fp, 7, s->rx_nohandler);
 		}
Signature · 1 line
2.26.2
#1 lore raw
Quoted 29 lines · Jakub Kicinski

missed_packet_errors are much more commonly reported:

linux$ git grep -c '[.>]rx_missed_errors ' -- drivers/ | wc -l 64
linux$ git grep -c '[.>]rx_over_errors ' -- drivers/ | wc -l 37

Plus those drivers are generally more modern than those using rx_over_errors.

Since recently merged kernel documentation makes this preference official, let's make ip -s output more informative and let rx_missed_errors take the place of rx_over_errors.

Before:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

Quoted 10 lines

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

changes to ip output are usually not allowed.

author maintainer
#2 lore raw
Quoted 25 lines · On Fri, 18 Sep 2020 09:44:35 -0600 David Ahern wrote:
Quoted 13 lines · Jakub Kicinski

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

Quoted 10 lines

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

Quoted 1 line

changes to ip output are usually not allowed.

Does that mean "no" or "you need to be more convincing"? :)

JSON output is not changed. I don't think we care about screen scrapers. If we cared about people how interpret values based on their position in the output we would break that with every release, no?

#3 lore raw
Quoted 26 lines · Jakub Kicinski
Quoted 25 lines · On Fri, 18 Sep 2020 09:44:35 -0600 David Ahern wrote:
Quoted 23 lines · Jakub Kicinski

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

Quoted 9 lines
Quoted 1 line

changes to ip output are usually not allowed.

Does that mean "no" or "you need to be more convincing"? :)

JSON output is not changed. I don't think we care about screen scrapers. If we cared about people how interpret values based on their position in the output we would break that with every release, no?

In this case you are not adding or inserting a new column, you are changing the meaning of an existing column.

It's an 'error' stat so probably not as sensitive. I do not have a strong religion on it since it seems to be making the error stat more up to date.

#4 lore raw
Quoted 37 lines · On Fri, 18 Sep 2020 22:52:56 -0600 David Ahern <[email protected]> wrote:
Quoted 26 lines · Jakub Kicinski
Quoted 25 lines · On Fri, 18 Sep 2020 09:44:35 -0600 David Ahern wrote:
Quoted 23 lines · Jakub Kicinski

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

Quoted 9 lines
Quoted 1 line

changes to ip output are usually not allowed.

Does that mean "no" or "you need to be more convincing"? :)

JSON output is not changed. I don't think we care about screen scrapers. If we cared about people how interpret values based on their position in the output we would break that with every release, no?

Quoted 6 lines

In this case you are not adding or inserting a new column, you are changing the meaning of an existing column.

It's an 'error' stat so probably not as sensitive. I do not have a strong religion on it since it seems to be making the error stat more up to date.

Is there any way to see the old error column at all?

author maintainer
#5 lore raw
Quoted 7 lines · On Sat, 19 Sep 2020 09:06:58 -0700 Stephen Hemminger wrote:
Quoted 7 lines
Quoted 6 lines

Does that mean "no" or "you need to be more convincing"? :)

JSON output is not changed. I don't think we care about screen scrapers. If we cared about people how interpret values based on their position in the output we would break that with every release, no?

Quoted 6 lines
Quoted 6 lines

In this case you are not adding or inserting a new column, you are changing the meaning of an existing column.

It's an 'error' stat so probably not as sensitive. I do not have a strong religion on it since it seems to be making the error stat more up to date.

I hope this change would encourage vendors to use the standard stats. Because..

Quoted 1 line

Is there any way to see the old error column at all?

I had no idea how to even show error stats. I read the code multiple times, only when I sat down to actually implement displaying did I realize that -s -s (repeat the option) shows all the stats.

Applied
#6 lore raw
In reply to Jakub Kicinski
Quoted 40 lines · Jakub Kicinski

missed_packet_errors are much more commonly reported:

linux$ git grep -c '[.>]rx_missed_errors ' -- drivers/ | wc -l 64
linux$ git grep -c '[.>]rx_over_errors ' -- drivers/ | wc -l 37

Plus those drivers are generally more modern than those using rx_over_errors.

Since recently merged kernel documentation makes this preference official, let's make ip -s output more informative and let rx_missed_errors take the place of rx_over_errors.

Before:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped overrun mcast
6.04T 4.67G 0 0 0 67.7M
RX errors: length crc frame fifo missed
0 0 0 0 7
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

After:

2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DEFAULT group default qlen 1000
link/ether 00:0a:f7:c1:4d:38 brd ff:ff:ff:ff:ff:ff
RX: bytes packets errors dropped missed mcast
6.04T 4.67G 0 0 7 67.7M
RX errors: length crc frame fifo overrun
0 0 0 0 0
TX: bytes packets errors dropped carrier collsns
3.13T 2.76G 0 0 0 0
TX errors: aborted fifo window heartbeat transns
0 0 0 0 6

Quoted 5 lines
---
 ip/ipaddress.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

applied to iproute2-next