Skip to contentSkip to conversation

rtl8187: Fix driver to return TX retry info for RTL8187L

PATCH Superseded Larry Finger 4 participants 13 messages Jan 23, 2009 – Jan 30, 2009 Networking drivers (wireless)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 rtl8187: Fix driver to return TX retry info for RTL8187L

Current code for the RTL8187 is not returning valid retry information, = thus the
rate-setting mechanism is not functioning. As a further complication, t= his info
is only obtained by reading a register, which cannot be read while in i= nterrupt
context.

This patch implements the TX status return to mac80211 through the use = of a
work queue.

---

John,

This patch fixes a long-standing bug for the RTL8187 that has been pres= ent
since the driver was made to work with mac80211. Technically, it is not=
a
regression, but it would be nice if we could get it into 2.6.29 so that=
it
could be pushed to stable. I have not Cc'd stable as the patch will nee= d
rebasing to work with 2.6.28, or older.

Larry
---

Index: wireless-testing/drivers/net/wireless/rtl818x/rtl8187.h =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

drivers/net/wireless/rtl818x/rtl8187.h +3 -1
@@ -100,6 +100,8 @@ struct rtl8187_priv {
 	struct usb_device *udev;
 	u32 rx_conf;
 	struct usb_anchor anchored;
+	struct delayed_work work;
+	struct ieee80211_hw *dev;
 	u16 txpwr_base;
 	u8 asic_rev;
 	u8 is_rtl8187b;
@@ -117,7 +119,7 @@ struct rtl8187_priv {
 	struct {
 		__le64 buf;
 		struct sk_buff_head queue;
-	} b_tx_status;
+	} b_tx_status; /* This queue is used by both -b and non-b devices */
 };

=20
void rtl8187_write_phy(struct ieee80211_hw *dev, u8 addr, u32 data);
Index: wireless-testing/drivers/net/wireless/rtl818x/rtl8187_dev.c =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D

drivers/net/wireless/rtl818x/rtl8187_dev.c +0 -0
@@ -177,25 +177,33 @@ static void rtl8187_tx_cb(struct urb *ur
 					  sizeof(struct rtl8187_tx_hdr));
 	ieee80211_tx_info_clear_status(info);

=20
- if (!urb->status &&
- !(info->flags & IEEE80211_TX_CTL_NO_ACK) &&
- priv->is_rtl8187b) {
- skb_queue_tail(&priv->b_tx_status.queue, skb);
-
- /* queue is "full", discard last items */
- while (skb_queue_len(&priv->b_tx_status.queue) > 5) {
- struct sk_buff *old_skb;
-
- dev_dbg(&priv->udev->dev,
- "transmit status queue full\n");
-
- old_skb =3D skb_dequeue(&priv->b_tx_status.queue);
- ieee80211_tx_status_irqsafe(hw, old_skb);
- }
- } else {
- if (!(info->flags & IEEE80211_TX_CTL_NO_ACK) && !urb->status)
+ if (!(urb->status) && !(info->flags & IEEE80211_TX_CTL_NO_ACK)) {
+ if (priv->is_rtl8187b) {
+ skb_queue_tail(&priv->b_tx_status.queue, skb);
+
+ /* queue is "full", discard last items */
+ while (skb_queue_len(&priv->b_tx_status.queue) > 5) {
+ struct sk_buff *old_skb;
+
+ dev_dbg(&priv->udev->dev,
+ "transmit status queue full\n");
+
+ old_skb =3D skb_dequeue(&priv->b_tx_status.queue);
+ ieee80211_tx_status_irqsafe(hw, old_skb);
+ }
+ return;
+ } else {
info->flags |=3D IEEE80211_TX_STAT_ACK;
+ }
+ }
+ if (priv->is_rtl8187b)
ieee80211_tx_status_irqsafe(hw, skb);
+ else {
+ /* Retry information for the RTI8187 is only available by
+ * reading a register in the device. We are in interrupt mode
+ * here, thus queue the skb and finish on a work queue. */
+ skb_queue_tail(&priv->b_tx_status.queue, skb);
+ queue_delayed_work(hw->workqueue, &priv->work, 0);
}
}
=20
@@ -854,6 +862,34 @@ static int rtl8187b_init_hw(struct ieee8
return 0;
}
=20
+static void rtl8187_work(struct work_struct *work)
+{
+ /* The RTL8187 returns the retry count through register 0xFFFA. In
+ * addition, it appears to be a cumulative retry count, not the
+ * value for the current TX packet. When multiple TX entries are
+ * queued, the retry count will be valid for the last one in the queu=
e.
+ * The "error" should not matter for purposes of rate setting. */
+ struct rtl8187_priv *priv =3D container_of(work, struct rtl8187_priv,
+ work.work);
+ struct ieee80211_tx_info *info;
+ struct ieee80211_hw *dev =3D priv->dev;
+ static u16 retry;
+ u16 tmp;
+
+ mutex_lock(&priv->conf_mutex);
+ tmp =3D rtl818x_ioread16(priv, (__le16 *)0xFFFA);
+ while (skb_queue_len(&priv->b_tx_status.queue) > 0) {
+ struct sk_buff *old_skb;
+
+ old_skb =3D skb_dequeue(&priv->b_tx_status.queue);
+ info =3D IEEE80211_SKB_CB(old_skb);
+ info->status.rates[0].count =3D tmp - retry + 1;
+ ieee80211_tx_status_irqsafe(dev, old_skb);
+ }
+ retry =3D tmp;
+ mutex_unlock(&priv->conf_mutex);
+}
+
static int rtl8187_start(struct ieee80211_hw *dev)
{
struct rtl8187_priv *priv =3D dev->priv;
@@ -868,6 +904,7 @@ static int rtl8187_start(struct ieee8021
mutex_lock(&priv->conf_mutex);
=20
init_usb_anchor(&priv->anchored);
+ priv->dev =3D dev;
=20
if (priv->is_rtl8187b) {
reg =3D RTL818X_RX_CONF_MGMT |
@@ -935,6 +972,7 @@ static int rtl8187_start(struct ieee8021
reg |=3D RTL818X_CMD_TX_ENABLE;
reg |=3D RTL818X_CMD_RX_ENABLE; rtl818x_iowrite8(priv, &priv->map->CMD, reg);
+ INIT_DELAYED_WORK(&priv->work, rtl8187_work);
mutex_unlock(&priv->conf_mutex);
=20
return 0;
@@ -965,6 +1003,8 @@ static void rtl8187_stop(struct ieee8021
dev_kfree_skb_any(skb);
=20
usb_kill_anchored_urbs(&priv->anchored);
+ if (!priv->is_rtl8187b)
+ cancel_delayed_work_sync(&priv->work);
mutex_unlock(&priv->conf_mutex);
}
=20
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#1 lore raw
Quoted 5 lines · Larry Finger

Current code for the RTL8187 is not returning valid retry information, thus the rate-setting mechanism is not functioning. As a further complication, this info is only obtained by reading a register, which cannot be read while in interrupt context.

Quoted 7 lines

This patch implements the TX status return to mac80211 through the use of a work queue.

I have tested this patch in compat-wireless-2009-01-23 (kernel 2.6.27, ubuntu) with my rtl8187 card (build on Asus P5B Delux Wifi, 0bda:8187 Realtek Semiconductor Corp. RTL8187 Wireless Adapter) and for me it does not work.

With this patch I get a connection to the AP, but the bandwidth is very low. After a reboot it was connected to the AP, but the connection was unusable for me, no traffic went through the connection at all.

Hauke

#2 lore raw
Quoted 9 lines · Hauke Mehrtens

I have tested this patch in compat-wireless-2009-01-23 (kernel 2.6.27, ubuntu) with my rtl8187 card (build on Asus P5B Delux Wifi, 0bda:8187 Realtek Semiconductor Corp. RTL8187 Wireless Adapter) and for me it does not work.

With this patch I get a connection to the AP, but the bandwidth is very low. After a reboot it was connected to the AP, but the connection was unusable for me, no traffic went through the connection at all.

What is your status without this patch? Is your connection stable at 1 Mb/s?

With the patch, what rate did iwconfig report, and which rate-setting algorithm are you using?

Larry

#3 lore raw
In reply to Hauke Mehrtens
Quoted 2 lines · Hauke Mehrtens
Quoted 1 line · Larry Finger

Current code for the RTL8187 is not returning valid retry informati=

on, thus the

Quoted 1 line
Quoted 1 line

rate-setting mechanism is not functioning. As a further complicatio=

n, this info

Quoted 1 line
Quoted 1 line

is only obtained by reading a register, which cannot be read while =

in interrupt

Quoted 3 lines
Quoted 3 lines

context.
=20
This patch implements the TX status return to mac80211 through the =

use of a

Quoted 8 lines
Quoted 6 lines

work queue.
=20

=20
I have tested this patch in compat-wireless-2009-01-23 (kernel 2.6.27=

,

Quoted 2 lines

ubuntu) with my rtl8187 card (build on Asus P5B Delux Wifi, 0bda:8187 Realtek Semiconductor Corp. RTL8187 Wireless Adapter) and for me it d=

oes

Quoted 3 lines

not work.
=20
With this patch I get a connection to the AP, but the bandwidth is ve=

ry

Quoted 1 line

low. After a reboot it was connected to the AP, but the connection wa=

s

Quoted 1 line

unusable for me, no traffic went through the connection at all.

Just curious if you tried limiting the rate to 11Mb as Larry said someone else had done? Did that work any better?

John
--=20
John W. Linville Someday the world will need a hero, and you
[email protected] might be all we have. Be ready.
--
To unsubscribe from this list: send the line "unsubscribe linux-wireles= s" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#4 lore raw
Quoted 4 lines · John W. Linville

Just curious if you tried limiting the rate to 11Mb as Larry said someone else had done? Did that work any better?

John

I did some more testing with and without the patch. With the patch it works when the rate is limited to 11M with pid and with minstrel algorithm. Without this limitation it works only with pid.

Without this patch it also works with the limiting the rate to 11M with both rate control algorithms, but with minstrel it also works without limiting the rate.

Without the patch, without limiting the rate and with minstrel iwconfig reports a rate of 1Mb/s, but I get 3Mb/s with wget, properly an other bug. With pid the rate goes up to 54Mb/s and it does not work.

With the patch, without limiting the rate and with minstrel iwconfig reports a rate of 24Mb/s and it does not work. With pid it starts with 1Mb/s and the connection is stable. The rate is changing between 1mb/s, 2Mb/s, 5.5Mb/s and 9Mb/s.

I will do more testing with pid and with this patch.

I have tested with compat-wireless-2009-01-24 and reversed the patch.

Hauke

#5 lore raw
Quoted 3 lines · Hauke Mehrtens

I did some more testing with and without the patch. With the patch it works when the rate is limited to 11M with pid and with minstrel algorithm. Without this limitation it works only with pid.

Of course, when you set a manual rate, the rate-setting method is immaterial.

Quoted 3 lines

Without this patch it also works with the limiting the rate to 11M with both rate control algorithms, but with minstrel it also works without limiting the rate.

Without the patch, minstrel is internally limited to 1 Mb/s TX rate.

Quoted 3 lines

Without the patch, without limiting the rate and with minstrel iwconfig reports a rate of 1Mb/s, but I get 3Mb/s with wget, properly an other bug. With pid the rate goes up to 54Mb/s and it does not work.

The rate reported by iwconfig is for TX. The throughput from wget is dependent on the TX rate of the AP. This info is useful, but the rates from wput, or any other transmitting program, are more instructive to find TX problems.

Quoted 4 lines

With the patch, without limiting the rate and with minstrel iwconfig reports a rate of 24Mb/s and it does not work. With pid it starts with 1Mb/s and the connection is stable. The rate is changing between 1mb/s, 2Mb/s, 5.5Mb/s and 9Mb/s.

Please submit your rc_stats data for the minstrel case.

It is obvious that the RTL8187L does not do as well as other wireless chips in low S/N conditions. It is also possible that some flavors of the chips are better than others. In my Netgear WG111v2 with the patch and using the minstrel algorithm, I get a TX rate between 21 and 22 Mb/s as measured with tcpperf at a distance of 2 m from my AP. The rate reported by iwconfig is 54 Mb/s. When I move to ~15 m from the AP, the rate falls off to ~12 Mb/s with iwconfig reporting 24-48 Mb/s. At 20 m, the connection is essentially unusable - sounds like yours the iwconfig rate is 24 Mb/s.

We are clearly flying blind here as we do not have a data sheet for the RTL8187L. I got some help from the Realtek engineer regarding the retry info, but that is the limit of my info. I do have what I think is the latest version of the vendor driver, and I will compare the power setting code with theirs.

The patch definitely helps my device as long as the reported signal level is above -60 dBm. With lower signals, then it doesn't work very well. Your device must be in that range.

Larry

#7 lore raw

[RFT] rtl8187: Fix OFDM power settings for RTL8187L

In reply to Hauke Mehrtens

Hauke,

Please apply this patch on top of current wireless-testing, or compat-wireless.

Larry
====================

After reports of poor performance, a review of the latest vendor driver (rtl8187_linux_26.1025.0328.2007) for RTL8187L devices was undertaken.

A difference was found in the code used to index the OFDM power tables. When the Linux driver was changed, my unit works at a much greater range than before.

---

drivers/net/wireless/rtl818x/rtl8187_rtl8225.c +4 -1
Index: wireless-testing/drivers/net/wireless/rtl818x/rtl8187_rtl8225.c
===================================================================
@@ -536,7 +536,10 @@ static void rtl8225z2_rf_set_tx_power(st
 	cck_power += priv->txpwr_base & 0xF;
 	cck_power = min(cck_power, (u8)35);
 
-	ofdm_power = min(ofdm_power, (u8)15);
+	if (ofdm_power > (u8)15)
+		ofdm_power = 25;
+	else
+		ofdm_power += 10;
 	ofdm_power += priv->txpwr_base >> 4;
 	ofdm_power = min(ofdm_power, (u8)35);
 
#8 lore raw

PATCH v2 1/1 rtl8187: Fix OFDM power settings for RTL8187L

In reply to Hauke Mehrtens

Hauke,

Please apply this patch on top of current wireless-testing, or compat-wireless.

V1 only did one of the two kinds of radios.

Larry
====================

After reports of poor performance, a review of the latest vendor driver (rtl8187_linux_26.1025.0328.2007) for RTL8187L devices was undertaken.

A difference was found in the code used to index the OFDM power tables. When the Linux driver was changed, my unit works at a much greater range than before.

---

drivers/net/wireless/rtl818x/rtl8187_rtl8225.c +8 -1
Index: wireless-testing/drivers/net/wireless/rtl818x/rtl8187_rtl8225.c
===================================================================
@@ -285,6 +285,10 @@ static void rtl8225_rf_set_tx_power(stru
 	ofdm_power = priv->channels[channel - 1].hw_value >> 4;
 
 	cck_power = min(cck_power, (u8)11);
+	if (ofdm_power > (u8)15)
+		ofdm_power = 25;
+	else
+		ofdm_power += 10;
 	ofdm_power = min(ofdm_power, (u8)35);
 
 	rtl818x_iowrite8(priv, &priv->map->TX_GAIN_CCK,
@@ -536,7 +540,10 @@ static void rtl8225z2_rf_set_tx_power(st
 	cck_power += priv->txpwr_base & 0xF;
 	cck_power = min(cck_power, (u8)35);
 
-	ofdm_power = min(ofdm_power, (u8)15);
+	if (ofdm_power > (u8)15)
+		ofdm_power = 25;
+	else
+		ofdm_power += 10;
 	ofdm_power += priv->txpwr_base >> 4;
 	ofdm_power = min(ofdm_power, (u8)35);
 
#9 lore raw

Re: [RFT V2] rtl8187: Fix OFDM power settings for RTL8187L

Quoted 8 lines · Larry Finger

Hauke,

Please apply this patch on top of current wireless-testing, or compat-wireless.

V1 only did one of the two kinds of radios.

Larry

Hi Larry,

With this patch included and pid the rate goes up to 24Mb/s, but the connection is stable. Without this patch is is only up to 11Mb/s. The connection is still unusable with minstrel, it goes up too fare. It starts with 24Mb/s and the connection drops. After some time it goes up to 54Mb/s, but then it goes down again and the connection is working again for a minute till the procedure begins again.

What do you mean with rc_stats ?

Complete iwconfig output with minstrel (No connection):
wlan0 IEEE 802.11bg ESSID:"mehrtenswlan"
Mode:Managed Frequency:2.412 GHz Access Point:
00:1A:92:EA:73:12
Bit Rate=48 Mb/s Tx-Power=27 dBm
Retry min limit:7 RTS thr:off Fragment thr=2352 B
Power Management:off
Link Quality=57/100 Signal level:-52 dBm
Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:0 Invalid misc:0 Missed beacon:0

dmesg |fgrep rtl shows the following:
phy0: hwaddr ffff8801399ac080, RTL8187vB (default) V1 + rtl8225z2
usbcore: registered new interface driver rtl8187

#10 lore raw

Re: [RFT V2] rtl8187: Fix OFDM power settings for RTL8187L

Quoted 14 lines · Hauke Mehrtens

Hi Larry,

With this patch included and pid the rate goes up to 24Mb/s, but the
connection is stable. Without this patch is is only up to 11Mb/s. The
connection is still unusable with minstrel, it goes up too fare. It
starts with 24Mb/s and the connection drops. After some time it goes up
to 54Mb/s, but then it goes down again and the connection is working
again for a minute till the procedure begins again.

Quoted 1 line

What do you mean with rc_stats ?

if you have mac80211 debug switched on,

cat /sys/kernel/debug/ieee80211/phyXX/stations/<MAC ADDR>/rc_stats

shows some rc_stats info... where XX and <MAC_ADDRESS> differs depending on your config/setup.

#11 lore raw

Re: [RFT V2] rtl8187: Fix OFDM power settings for RTL8187L

In reply to Hauke Mehrtens
Quoted 8 lines · Hauke Mehrtens

With this patch included and pid the rate goes up to 24Mb/s, but the connection is stable. Without this patch is is only up to 11Mb/s. The connection is still unusable with minstrel, it goes up too fare. It starts with 24Mb/s and the connection drops. After some time it goes up to 54Mb/s, but then it goes down again and the connection is working again for a minute till the procedure begins again.

What do you mean with rc_stats ?

The statistics for the mistrel algorithn are presented in the /sys tree in a file named rc_stats. The easy way to see that data is to do the following:

STATS=$(find /sys -name rc_stats)
sudo cat $STATS

You will get output that looks like

rate throughput ewma prob this prob this succ/attempt
success attempts

     1         0.9       99.9      100.0          0(  0)        855
      873
     2         1.8       95.7      100.0          0(  0)         11
       11
  P  5.5       5.0       99.9      100.0          0(  0)         77
       77
    11         9.1       95.9      100.0          0(  0)         14
       16
     6         5.4       95.6      100.0          0(  0)         16
       19
     9         8.0       95.7      100.0          0(  0)         11
       11
    12        10.6       95.7      100.0          0(  0)         11
       11
    18        15.5       95.7      100.0          0(  0)         11
       11
    24        20.4       96.0      100.0          0(  0)        276
      279
    36         5.9       96.0      100.0          0(  0)         76
       78
 t  48         6.5       76.5      100.0          0(  0)       2958
     3034
T   54        20.0       99.9      100.0          1(  1)       6220
     6400

Total packet count:: ideal 9216 lookaround 485

These data are for my RTL8187, which is running at 54 Mb/s (the line with the T) and the success rate at this speed is 6220/6400 (97%).

Let the mistrel method go through several of those cycles that you mention and then dump and post this table. Perhaps the experts will see what is wrong.

Larry

Larry

#12 lore raw

Re: [RFT V2] rtl8187: Fix OFDM power settings for RTL8187L

Quoted 48 lines · Larry Finger
Quoted 8 lines · Hauke Mehrtens

With this patch included and pid the rate goes up to 24Mb/s, but the connection is stable. Without this patch is is only up to 11Mb/s. The connection is still unusable with minstrel, it goes up too fare. It starts with 24Mb/s and the connection drops. After some time it goes up to 54Mb/s, but then it goes down again and the connection is working again for a minute till the procedure begins again.

What do you mean with rc_stats ?

The statistics for the mistrel algorithn are presented in the /sys tree in a file named rc_stats. The easy way to see that data is to do the following:

STATS=$(find /sys -name rc_stats)
sudo cat $STATS

You will get output that looks like

rate throughput ewma prob this prob this succ/attempt
success attempts

     1         0.9       99.9      100.0          0(  0)        855
      873
     2         1.8       95.7      100.0          0(  0)         11
       11
  P  5.5       5.0       99.9      100.0          0(  0)         77
       77
    11         9.1       95.9      100.0          0(  0)         14
       16
     6         5.4       95.6      100.0          0(  0)         16
       19
     9         8.0       95.7      100.0          0(  0)         11
       11
    12        10.6       95.7      100.0          0(  0)         11
       11
    18        15.5       95.7      100.0          0(  0)         11
       11
    24        20.4       96.0      100.0          0(  0)        276
      279
    36         5.9       96.0      100.0          0(  0)         76
       78
 t  48         6.5       76.5      100.0          0(  0)       2958
     3034
T   54        20.0       99.9      100.0          1(  1)       6220
     6400

Total packet count:: ideal 9216 lookaround 485

Quoted 8 lines

These data are for my RTL8187, which is running at 54 Mb/s (the line with the T) and the success rate at this speed is 6220/6400 (97%).

Let the mistrel method go through several of those cycles that you mention and then dump and post this table. Perhaps the experts will see what is wrong.

Larry

Hi Larry,

I captured the rc_stats for my rtl8187 device. These stats are created with kernel 2.6.29-rc3-wl (master-2009-01-28). I tried to get traffic through the link, but I was almost dead. Most of the time it showed me 54Mb/s in iwconfig.

Hauke

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.2       25.0      100.0          0(  0)          1           1
     2         0.0        3.1       12.5          0(  0)          1           8
     5.5       1.2       25.0      100.0          0(  0)          1           1
  P 11         5.5       57.8       12.5          0(  0)       3924        4064
     6         1.4       25.0      100.0          0(  0)          1           1
     9         2.1       25.0      100.0          0(  0)          1           1
    12         2.7       25.0      100.0          0(  0)          1           1
    18         2.0       12.5       50.0          0(  0)          1           2
    24         5.3       25.0      100.0          0(  0)          1           1
 t  36         6.8       22.3       12.5          0(  0)          6          34
T   48         9.1       23.3       12.5          1(  8)         27         190
    54         1.3        3.1       12.5          0(  0)          1           8

Total packet count:: ideal 212 lookaround 11

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.2       25.0      100.0          0(  0)          1           1
     2         0.0        3.1       12.5          0(  0)          1           8
     5.5       1.2       25.0      100.0          0(  0)          1           1
  P 11         5.5       57.8       12.5          0(  0)       3924        4064
     6         1.4       25.0      100.0          0(  0)          1           1
     9         2.1       25.0      100.0          0(  0)          1           1
    12         2.7       25.0      100.0          0(  0)          1           1
    18         2.0       12.5       50.0          0(  0)          1           2
    24         5.3       25.0      100.0          0(  0)          1           1
 t  36         5.8       19.2       10.0          0(  0)          7          44
T   48         6.0       15.4       12.5          1(  8)         36         273
    54         1.3        3.1       12.5          0(  0)          1           8

Total packet count:: ideal 222 lookaround 11

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.2       25.0      100.0          0(  0)          1           1
     2         0.1        5.4       12.5          0(  0)          2          16
     5.5       1.2       25.0      100.0          0(  0)          1           1
  P 11         4.4       46.4       12.5          0(  0)       3925        4072
     6         1.4       25.0      100.0          0(  0)          1           1
     9         2.1       25.0      100.0          0(  0)          1           1
    12         2.7       25.0      100.0          0(  0)          1           1
    18         2.0       12.5       12.5          0(  0)          2          10
    24         4.6       21.8       12.5          0(  0)          2           9
 t  36         4.6       15.3       12.5          0(  0)         11          76
T   48         6.4       16.3       12.5          1(  8)         83         649
    54         1.3        3.1       12.5          0(  0)          1           8

Total packet count:: ideal 274 lookaround 14

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.1       17.7       12.5          0(  0)          4          25
     2         0.3       16.6       50.0          0(  0)          3          18
     5.5       1.9       39.6      100.0          0(  0)          4          18
    11         3.0       31.6       12.5          0(  0)       3927        4088
     6         0.9       17.4       12.5          0(  0)          4          27
     9         1.4       17.7       12.5          0(  0)          4          25
    12         4.6       41.9      100.0          0(  0)          4          14
  P 18         7.9       48.8       12.5          0(  0)        120         206
    24         7.4       34.7       12.5          0(  0)        133         418
T   36        11.6       38.1       12.5          1(  8)         21         135
    48         5.3       13.6       12.5          0(  0)        217        1684
 t  54         8.3       19.2       12.5          0(  0)        139         963

Total packet count:: ideal 793 lookaround 41

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.1       17.7       12.5          0(  0)          4          25
     2         0.3       16.6       50.0          0(  0)          3          18
     5.5       1.9       39.6      100.0          0(  0)          4          18
    11         3.0       31.6       12.5          0(  0)       3927        4088
     6         0.9       17.4       12.5          0(  0)          4          27
     9         1.4       17.7       12.5          0(  0)          4          25
    12         4.6       41.9      100.0          0(  0)          4          14
  P 18         7.9       48.8       12.5          0(  0)        120         206
    24         7.4       34.7       12.5          0(  0)        133         418
T   36        11.6       38.1       12.5          1(  8)         21         135
    48         5.3       13.6       12.5          0(  0)        217        1684
 t  54         8.3       19.2       12.5          0(  0)        139         963

Total packet count:: ideal 793 lookaround 41

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
     1         0.1       17.7       12.5          0(  0)          4          25
     2         0.7       37.4      100.0          0(  0)          4          19
     5.5       1.9       39.6      100.0          0(  0)          4          18
    11         2.5       26.8       12.5          0(  0)       3928        4096
     6         0.9       17.4       12.5          0(  0)          4          27
     9         1.4       17.7       12.5          0(  0)          4          25
  P 12         4.6       41.9      100.0          0(  0)          4          14
    18         5.5       33.8       12.5          0(  0)        124         229
    24         5.3       25.0       12.5          0(  0)        136         442
 t  36         5.8       19.3       12.5          0(  0)         33         224
    48         5.3       13.6       12.5          0(  0)        217        1684
T   54         6.7       15.5       12.5          1(  8)        153        1061

Total packet count:: ideal 826 lookaround 43

hauke@hauke-desktop:~/kamikaze/openwrt5$ sudo cat $STATS
rate     throughput  ewma prob   this prob  this succ/attempt   success    attempts
 tP  1         0.7       81.8       50.0          0(  0)         21          23
     2         0.0        0.0        0.0          0(  0)          0           0
     5.5       0.0        0.0        0.0          0(  0)          0           0
    11         0.0        0.0        0.0          0(  0)          0           0
     6         0.0        0.0        0.0          0(  0)          0           0
     9         0.0        0.0        0.0          0(  0)          0           0
    12         0.0        0.0        0.0          0(  0)          0           0
    18         0.0        0.0        0.0          0(  0)          0           0
    24         0.0        0.0        0.0          0(  0)          0           0
T   36         4.5       14.8       12.5          1(  8)         15         106
    48         0.0        0.0        0.0          0(  0)          0           0
    54         0.0        0.0        0.0          0(  0)          0           0

Total packet count:: ideal 36 lookaround 1