[opensuse] openSuSE 10.3 - Madwifi/Atheros Card Signal Strength Cut in Half
Listmates: With 10.3, my AR5212/5213 wireless card using the madwifi driver and wpa_supplicant I get only 50% of the signal strength that it did with 10.2. This Signal Strength Change Occurred Between openSuSE 10.2 and 10.3. Since the system is a dual boot system, I can confirm that signal strength is still 100% normal on the dark side of my laptop. With openSuSE 10.2 using madwifi and wpa_supplicant, signal strength was always 100% no matter where I was in my house or office. (Both have linksys WRT54G access points). The distances involved are 15 meters or less. The signal strength loss occurred when I upgraded my laptop from 10.2 to 10.3. The loss is present even if I put my laptop (Toshiba P35) right in front of the AP. (I have had the same wireless setup for over 2 years) Whatever changed in the latest 1 or 2 releases of madwifi or wpa has definitively caused changes to the ability of the driver to get good signal strength. I don't know what is causing it, but I can CONFIRM the problem with madwifi under openSuSE 10.3 with the Atheros AR5212/5213 driver on my Toshiba P35. athstats now shows crc errors that were not there before: [root Rankin-P35a/home/david] # athstats 379 tx management frames 88 long on-chip tx retries 345 tx frames with no ack marked 10 tx frames with an alternate rate 299 rx failed due to bad CRC 28 PHY errors 28 CCK restart 1312 periodic calibrations 1 rfgain value change rssi of last rcv: 28 79 switched default/rx antenna Antenna profile: [1] tx 1188 rx 24964 [2] tx 176 rx 4211 [root Rankin-P35a/home/david] # athstats ath0 input output altrate short long xretry crcerr crypt phyerr rssi rate 30621 1592 10 0 88 0 312 0 28 30 0M Any thoughts? Anybody else seeing this in 10.3? Any help would be appreciated. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
David C. Rankin J.D. P.E. wrote:
Listmates:
With 10.3, my AR5212/5213 wireless card using the madwifi driver and wpa_supplicant I get only 50% of the signal strength that it did with 10.2. This Signal Strength Change Occurred Between openSuSE 10.2 and 10.3. Since the system is a dual boot system, I can confirm that signal strength is still 100% normal on the dark side of my laptop.
With openSuSE 10.2 using madwifi and wpa_supplicant, signal strength was always 100% no matter where I was in my house or office. (Both have linksys WRT54G access points). The distances involved are 15 meters or less.
The signal strength loss occurred when I upgraded my laptop from 10.2 to 10.3. The loss is present even if I put my laptop (Toshiba P35) right in front of the AP. (I have had the same wireless setup for over 2 years)
Whatever changed in the latest 1 or 2 releases of madwifi or wpa has definitively caused changes to the ability of the driver to get good signal strength. I don't know what is causing it, but I can CONFIRM the problem with madwifi under openSuSE 10.3 with the Atheros AR5212/5213 driver on my Toshiba P35. athstats now shows crc errors that were not there before:
[root Rankin-P35a/home/david] # athstats
379 tx management frames 88 long on-chip tx retries 345 tx frames with no ack marked 10 tx frames with an alternate rate 299 rx failed due to bad CRC 28 PHY errors 28 CCK restart 1312 periodic calibrations 1 rfgain value change rssi of last rcv: 28 79 switched default/rx antenna Antenna profile: [1] tx 1188 rx 24964 [2] tx 176 rx 4211
[root Rankin-P35a/home/david] # athstats ath0
input output altrate short long xretry crcerr crypt phyerr rssi rate
30621 1592 10 0 88 0 312 0 28 30 0M
Any thoughts? Anybody else seeing this in 10.3? Any help would be appreciated.
Here is a bit more information. It looks like there have been some kernel changes. It looks like the ath_pci module is now under the "weak-updates" directory. Please tell the kernel guys that we don't want "weak" signal, we want "strong" signals. ;-) /lib/modules/2.6.22.9-0.4-default/weak-updates/ath/ath_pci.ko /lib/modules/2.6.22.9-0.4-default/weak-updates/ath_rate/onoe/ath_rate_onoe.ko /lib/modules/2.6.22.9-0.4-default/weak-updates/ath_rate/amrr/ath_rate_amrr.ko /lib/modules/2.6.22.9-0.4-default/weak-updates/ath_rate/sample/ath_rate_sample.ko /lib/modules/2.6.22.9-0.4-default/weak-updates/ath_hal/ath_hal.ko /lib/modules/2.6.22.9-0.4-default/kernel/drivers/md/multipath.ko /lib/modules/2.6.22.9-0.4-default/kernel/drivers/md/dm-multipath.ko -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Fri, 2007-10-19 at 14:35 -0500, David C. Rankin J.D. P.E. wrote:
Listmates:
With 10.3, my AR5212/5213 wireless card using the madwifi driver and wpa_supplicant I get only 50% of the signal strength that it did with 10.2. This Signal Strength Change Occurred Between openSuSE 10.2 and 10.3. Since the system is a dual boot system, I can confirm that signal strength is still 100% normal on the dark side of my laptop.
With openSuSE 10.2 using madwifi and wpa_supplicant, signal strength was always 100% no matter where I was in my house or office. (Both have linksys WRT54G access points). The distances involved are 15 meters or less.
The signal strength loss occurred when I upgraded my laptop from 10.2 to 10.3. The loss is present even if I put my laptop (Toshiba P35) right in front of the AP. (I have had the same wireless setup for over 2 years)
Whatever changed in the latest 1 or 2 releases of madwifi or wpa has definitively caused changes to the ability of the driver to get good signal strength. I don't know what is causing it, but I can CONFIRM the problem with madwifi under openSuSE 10.3 with the Atheros AR5212/5213 driver on my Toshiba P35. athstats now shows crc errors that were not there before:
The fact that it was always 100% was wrong in NetworkManager (we had a
workaround in earlier suse versions but it broke with the newer madwifi
drivers):
https://bugzilla.novell.com/show_bug.cgi?id=293556
We fixed this in a recent 10.3 update because it was extremely
deceptive. The problem is then either in the madwifi driver and how it
outputs signal strength or NM and how it calculates signal strength.
-JP
--
JP Rosevear
JP Rosevear wrote:
On Fri, 2007-10-19 at 14:35 -0500, David C. Rankin J.D. P.E. wrote:
Listmates:
With 10.3, my AR5212/5213 wireless card using the madwifi driver and wpa_supplicant I get only 50% of the signal strength that it did with 10.2. This Signal Strength Change Occurred Between openSuSE 10.2 and 10.3. Since the system is a dual boot system, I can confirm that signal strength is still 100% normal on the dark side of my laptop.
With openSuSE 10.2 using madwifi and wpa_supplicant, signal strength was always 100% no matter where I was in my house or office. (Both have linksys WRT54G access points). The distances involved are 15 meters or less.
The signal strength loss occurred when I upgraded my laptop from 10.2 to 10.3. The loss is present even if I put my laptop (Toshiba P35) right in front of the AP. (I have had the same wireless setup for over 2 years)
Whatever changed in the latest 1 or 2 releases of madwifi or wpa has definitively caused changes to the ability of the driver to get good signal strength. I don't know what is causing it, but I can CONFIRM the problem with madwifi under openSuSE 10.3 with the Atheros AR5212/5213 driver on my Toshiba P35. athstats now shows crc errors that were not there before:
The fact that it was always 100% was wrong in NetworkManager (we had a workaround in earlier suse versions but it broke with the newer madwifi drivers): https://bugzilla.novell.com/show_bug.cgi?id=293556
We fixed this in a recent 10.3 update because it was extremely deceptive. The problem is then either in the madwifi driver and how it outputs signal strength or NM and how it calculates signal strength.
-JP
Thanks JP, This is truly a strange problem. What is very very strange is that I think this is a display change or a calculation change in the way the signal strength is computed in one of the apps that feeds the kinternet signal strength display. Whatever the change, the resulting signal strength display is now incorrect. We have gone from an extreme high to an extreme low signal strength display. How do I know the display is FUBAR? I put my laptop right next to the Linksys AP ( within 1 foot ) and signal strength never exceeded 58%. More testing also reveals something is amiss. I have also conducted a test with *two separate* 10.3 installs on the same laptop. I have 2 laptop hard drives, the original 4200 rpm Toshiba and a 5400 rpm WD. Both have 10.3 loaded. The 4200 rpm drive was loaded with 10.3 when it was released and has not been updated. The 5400 rpm drive is the daily use drive (because it is faster) with all current updates via Yast -> software management -> all packages -> update if newer. The 4200 rpm drive install shows signal strength is 100% The 5400 rpm drive install will not show signal strength over 58% no matter what! Both 10.3 installs have the same kernel, madwifi version and modules installed: Rankin-P35a:/home/david # uname -r 2.6.22.9-0.4-default david@Rankin-P35a:~> rpm -qa | grep wifi madwifi-kmp-default-0.9.3.2_2.6.22.5_31-0.1 madwifi-devel-0.9.3.99-36 madwifi-0.9.3.99-36 The 4200 rpm drive shows 100% signal strength even though it shows LESS actual signal with iwconfig than the updated install: iwconfig ath0 IEEE 802.11g ESSID:"skyline" Nickname:"" Mode:Managed Frequency:2.437 GHz Access Point: 00:0F:66:52:A3:A1 Bit Rate:18 Mb/s Tx-Power:16 dBm Sensitivity=1/1 Retry:off RTS thr:off Fragment thr:off Encryption key:5FD4-3FF1-5F40-387A-0CF1-4990-AA11-D349 Security mode:restricted Power Management:off Link Quality=30/70 Signal level=-65 dBm Noise level=-95 dBm ^^^^^^^^^^^^^^^^^^^^ Rx invalid nwid:1757 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0 The 5400 rpm drive shows 45% signal strength even though it has MORE actual signal: iwconfig ath0 IEEE 802.11g ESSID:"skyline" Nickname:"" Mode:Managed Frequency:2.437 GHz Access Point: 00:0F:66:52:A3:A1 Bit Rate:54 Mb/s Tx-Power:16 dBm Sensitivity=1/1 Retry:off RTS thr:off Fragment thr:off Encryption key:3DA2-0064-7787-5868-648E-AE4E-1AEB-D774 Security mode:restricted Power Management:off Link Quality=32/70 Signal level=-62 dBm Noise level=-94 dBm ^^^^^^^^^^^^^^^^^^^^ Rx invalid nwid:10572 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0 So the bottom line is something has changed and it is not quite right yet. Signal strength cannot be a maximum of 58% 1 foot from the AP. I understand the deceptive overstatement of the link always showing 100%, but I think there is an equal deceptive understatement when the link shows 58% when the laptop is literally on top of the AP. I hope the additional information helps narrow down where the problem is and what needs to be fixed. From a throughput standpoint, everything is fine. Data passes back and forth between the laptop and the AP just fine regardless of whether the signal strength reads 40% or 100%, but I guarantee you that there will be a whole lot of people that freak over the dramatic drop on the meter. -- David C. Rankin, J.D., P.E. Rankin Law Firm, PLLC 510 Ochiltree Street Nacogdoches, Texas 75961 Telephone: (936) 715-9333 Facsimile: (936) 715-9339 www.rankinlawfirm.com -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
participants (2)
-
David C. Rankin J.D. P.E. -
JP Rosevear