Description of problem: I get very slow connection speed (1Mbps) with Intel 5100 Wi-Fi adapter - Acer 6935G notebook Version-Release number of selected component (if applicable): How reproducible: All the time Steps to Reproduce: 1. Boot Fedora 2. Connect to Wi-Fi access point (standard 802.11G, 54Mbit 3. Connection is 1Mbps Actual results: Connection is 1Mbps Expected results: Connection should be 54Mbps Additional info: lspci -v for Wi-Fi adapter 05:00.0 Network controller: Intel Corporation WiFi Link 5100 Subsystem: Intel Corporation WiFi Link 5100 ABG Flags: bus master, fast devsel, latency 0, IRQ 46 Memory at d8100000 (64-bit, non-prefetchable) [size=8K] Capabilities: [c8] Power Management version 3 Capabilities: [d0] MSI: Enable+ Count=1/1 Maskable- 64bit+ Capabilities: [e0] Express Endpoint, MSI 00 Capabilities: [100] Advanced Error Reporting Capabilities: [140] Device Serial Number 00-16-ea-ff-ff-c6-7d-0c Kernel driver in use: iwlagn Kernel modules: iwlagn
NOTE: computer is located 1m from Access Point Signal quality is very good - see below: #nm-tool (wireless part) - Device: wlan0 [Auto dd-wrt] ------------------------------------------------- Type: 802.11 WiFi Driver: iwlagn State: connected Default: no HW Address: 00:16:EA:C6:7D:0C Capabilities: Speed: 1 Mb/s Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points (* = current AP) dima_wireless: Infra, 00:1E:58:BF:F9:CF, Freq 2437 MHz, Rate 54 Mb/s, Strength 35 WPA ZXDSL531BII-66852E: Infra, 00:16:E3:66:85:2E, Freq 2462 MHz, Rate 54 Mb/s, Strength 42 TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 44 WPA *dd-wrt: Infra, 00:1B:FC:91:83:4C, Freq 2412 MHz, Rate 54 Mb/s, Strength 100 WPA2 hpsetup: Ad-Hoc, 06:5D:57:D0:41:7D, Freq 2437 MHz, Rate 11 Mb/s, Strength 52 pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 45 WEP Thom_D005807: Infra, 00:24:D1:C8:69:07, Freq 2462 MHz, Rate 54 Mb/s, Strength 60 WPA WPA2 dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 70 WPA IPv4 Settings: Address: 192.168.1.61 Prefix: 24 (255.255.255.0) Gateway: 192.168.1.2 DNS: 192.168.1.2 Computers with other Wi-Fi adapters, for example based on Atheros chipset, can connect to AP at 54Mbit This compute ris also capable to connect at 54Mbit using Windows (Vista).
Hi, I see you opened other bugzillas for other hardware ..., but I think this is rather common software problem. I would told that the problem is in rate scaling algorithm, but iwlwifi use other rate-scaling code than other drivers, so there is most likely something other wrong in mac80211. Does loading iwlagn module with 11n_disable50=1 option helps? We can not also exclude this is Access Point problem, since lot's of drivers are affected. Did you tried to update/downgrade AP firmware, or check with other AP?
*** Bug 641250 has been marked as a duplicate of this bug. ***
*** Bug 640767 has been marked as a duplicate of this bug. ***
*** Bug 640779 has been marked as a duplicate of this bug. ***
Hi, This problem is specific to Wireless Client hardware (chipsets) I opened bugs for. Doesn't work ----------------- -Atheros 9170usb adapter (TP-Link WN821N) -Atheros 9271 USB adapter (TP-Link WN722N) -Atheros AR9285 Wireless Network Adapter (ath9k driver) Works for me: ------------------ -Ralink RT2800 -R73usb -Realtek RTL8179 Besides, all of those adapters work correctly (54Mbit, being located2m from AP) in Windows. Tested Windows Vista and Windows XP. What should I do to test "loading iwlagn module with 11n_disable50=1 option"? Do I need to unload module first and tan reload, using some extra arguments at CLI? Pls help/advise. P.S. I have good connection with enough bandwidth, and can test LiveCDs with nightly builds. All hardware remains available at my premises for further tests.
BTW: should adapters I listed work without compat-wireless package? Or may be it is still required, even with modern kernel?
(In reply to comment #6) > What should I do to test "loading iwlagn module with 11n_disable50=1 option"? > Do I need to unload module first and tan reload, using some extra arguments at > CLI? > Pls help/advise. You can do this by command line as root: $ rmmod iwlagn iwlcore $ modprobe iwlagn 11n_disable50=1 Or (also as root), to use option every time module load, i.e. system start: $echo "options iwlagn 11n_disable50=1" >> /etc/modprobe.d/iwlwifi.conf $ rmmod iwlagn iwlcore $ modprobe iwlagn > P.S. I have good connection with enough bandwidth, and can test LiveCDs with > nightly builds. No idea, maybe ask at some fedora mailing list. (In reply to comment #7) > BTW: should adapters I listed work without compat-wireless package? > Or may be it is still required, even with modern kernel? Hard to tell, some devices may requires news driver with compat-wireless to work.
executed # rmmod iwlagn iwlcore # modprobe iwlagn 11n_disable50=1 It seems this doesn't help # iwconfig lo no wireless extensions. eth0 no wireless extensions. wlan0 IEEE 802.11abg ESSID:"TP-L" Mode:Managed Frequency:2.427 GHz Access Point: D8:5D:4C:BF:25:B4 Bit Rate=2 Mb/s Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Encryption key:off Power Management:off Link Quality=70/70 Signal level=-15 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0 If I start pinging AP, than speed increases to about 18Mbps, and sometimes even to 48Mbps. Than it again drops to 1Mbps. What exactly does "11n_disable50=1" do? To which adapters it is applicable and, what adapters do not need it? I have several other Wi-Fi adapters, which suffer with current version of kernel. May be, problem is indeed common to all of them. But on th eother hand, there are some adapters which work ok.
(In reply to comment #9) > If I start pinging AP, than speed increases to about 18Mbps, and sometimes even > to 48Mbps. > Than it again drops to 1Mbps. After drop to 1Mbps, if you start doing a big data transfer through i.e. ftp or http, does rate stays constantly at 1Mbps? > What exactly does "11n_disable50=1" do? Disable 802.11n support for device. Devices which works for you (ralink,realtek) does not support 802.11n, so I thought that is the problem. > To which adapters it is applicable and, what adapters do not need it? Normally, option should be not need. This is just workaround, it can help find out where the problem is. It is only available for iwlagn driver.
Created attachment 454926 [details] NetworkManager Connection Information - 54Mbps iwlagn driver, Connection Information - 54Mbps Access Point: TP-Link TL-WR841ND Wireless Router Settings for this AP/Router: see http://vplessky.blogspot.com/2010/10/tp-link-tl-wr841nd-wireless-router.html after reboot, I have 54Mbit connection. But not 300Mbit or 150Mbit, as you can expect with 11n adapter and 11n AP. Than ... NM-Connection Information shows again 1Mbps See next screenshot
Created attachment 454928 [details] NetworkManager - Connection Information 1Mbps I had 54Mbps connection speed after reboot - according to Connection Information.. After some time (probably 10 sec. to 1 min.) it shows 1Mbps - see screenshot. Started pinging AP from Linux desktop. Speed in COnnection Information is again 54Mbps. $ nm-tool NetworkManager Tool State: connected - Device: eth0 ----------------------------------------------------------------- Type: Wired Driver: ATL1E State: unavailable Default: no HW Address: 00:A0:D1:A9:D2:AF Capabilities: Carrier Detect: yes Wired Properties Carrier: off - Device: wlan0 [Auto TP-L] --------------------------------------------------- Type: 802.11 WiFi Driver: iwlagn State: connected Default: yes HW Address: 00:16:EA:C6:7D:0C Capabilities: Speed: 54 Mb/s Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points (* = current AP) homenet: Infra, 00:26:18:74:54:EE, Freq 2412 MHz, Rate 54 Mb/s, Strength 38 WPA2 TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 48 WPA *TP-L: Infra, D8:5D:4C:BF:25:B4, Freq 2427 MHz, Rate 54 Mb/s, Strength 100 WPA pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 65 WEP dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 67 WPA IPv4 Settings: Address: 192.168.1.202 Prefix: 24 (255.255.255.0) Gateway: 192.168.1.2 DNS: 192.168.1.2 Response to ping is also very fast: [vadim@localhost ~]$ ping 192.168.1.1 PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data. 64 bytes from 192.168.1.1: icmp_req=1 ttl=64 time=1.69 ms 64 bytes from 192.168.1.1: icmp_req=2 ttl=64 time=0.800 ms 64 bytes from 192.168.1.1: icmp_req=3 ttl=64 time=0.745 ms 64 bytes from 192.168.1.1: icmp_req=4 ttl=64 time=0.730 ms 64 bytes from 192.168.1.1: icmp_req=5 ttl=64 time=0.688 ms ^C --- 192.168.1.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4003ms rtt min/avg/max/mdev = 0.688/0.931/1.694/0.383 ms $ iwconfig lo no wireless extensions. eth0 no wireless extensions. wlan0 IEEE 802.11abg ESSID:"TP-L" Mode:Managed Frequency:2.427 GHz Access Point: D8:5D:4C:BF:25:B4 Bit Rate=54 Mb/s Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Power Management:off Link Quality=70/70 Signal level=-13 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0
Created attachment 454935 [details] ar9170usb - NetworkManager Connection Information - 1Mbps Wi-Fi USB adapter information: --------------------------------------- Bus 002 Device 003: ID 0cf3:1002 Atheros Communications, Inc. TP-Link TL-WN821N v2 [Atheros AR9001U-(2)NG] Device Descriptor: bLength 18 bDescriptorType 1 bcdUSB 2.00 bDeviceClass 255 Vendor Specific Class bDeviceSubClass 255 Vendor Specific Subclass bDeviceProtocol 255 Vendor Specific Protocol bMaxPacketSize0 64 idVendor 0x0cf3 Atheros Communications, Inc. idProduct 0x1002 TP-Link TL-WN821N v2 [Atheros AR9001U-(2)NG] bcdDevice 1.06 iManufacturer 16 iProduct 32 iSerial 48 bNumConfigurations 1 Configuration Descriptor: bLength 9 bDescriptorType 2 wTotalLength 46 bNumInterfaces 1 bConfigurationValue 1 iConfiguration 0 bmAttributes 0x80 (Bus Powered) MaxPower 500mA Interface Descriptor: bLength 9 bDescriptorType 4 bInterfaceNumber 0 bAlternateSetting 0 bNumEndpoints 4 bInterfaceClass 255 Vendor Specific Class bInterfaceSubClass 0 bInterfaceProtocol 0 iInterface 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x01 EP 1 OUT bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x82 EP 2 IN bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x83 EP 3 IN bmAttributes 3 Transfer Type Interrupt Synch Type None Usage Type Data wMaxPacketSize 0x0040 1x 64 bytes bInterval 1 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x04 EP 4 OUT bmAttributes 3 Transfer Type Interrupt Synch Type None Usage Type Data wMaxPacketSize 0x0040 1x 64 bytes bInterval 1
Created attachment 454936 [details] ar9170usb - NetworkManager Connection Information - 48Mbps When I start pinging, using ar9170usb as the only connection to AP, connection speed goes up to 48Mbps. (which is not bad, but MUCH-MUCH below 300Mbps adaper capable of) ping goes fine: [vadim@localhost ~]$ ping 192.168.1.1 PING 192.168.1.1 (192.168.1.1) 56(84) bytes of data. 64 bytes from 192.168.1.1: icmp_req=1 ttl=64 time=0.875 ms 64 bytes from 192.168.1.1: icmp_req=2 ttl=64 time=1.15 ms 64 bytes from 192.168.1.1: icmp_req=3 ttl=64 time=1.14 ms 64 bytes from 192.168.1.1: icmp_req=4 ttl=64 time=1.53 ms 64 bytes from 192.168.1.1: icmp_req=5 ttl=64 time=0.978 ms ^C --- 192.168.1.1 ping statistics --- 5 packets transmitted, 5 received, 0% packet loss, time 4005ms rtt min/avg/max/mdev = 0.875/1.138/1.539/0.228 ms $ iwconfig lo no wireless extensions. eth0 no wireless extensions. wlan0 IEEE 802.11abg Mode:Managed Frequency:2.427 GHz Access Point: Not-Associated Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Power Management:off wlan1 IEEE 802.11bgn ESSID:"TP-L" Mode:Managed Frequency:2.427 GHz Access Point: D8:5D:4C:BF:25:B4 Bit Rate=54 Mb/s Tx-Power=20 dBm Retry long limit:7 RTS thr:off Fragment thr:off Power Management:off Link Quality=70/70 Signal level=-21 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0
nm-tool - connection with ar9170usb $ nm-tool NetworkManager Tool State: connected - Device: eth0 ----------------------------------------------------------------- Type: Wired Driver: ATL1E State: unavailable Default: no HW Address: 00:A0:D1:A9:D2:AF Capabilities: Carrier Detect: yes Wired Properties Carrier: off - Device: wlan0 ---------------------------------------------------------------- Type: 802.11 WiFi Driver: iwlagn State: disconnected Default: no HW Address: 00:16:EA:C6:7D:0C Capabilities: Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points homenet: Infra, 00:26:18:74:54:EE, Freq 2412 MHz, Rate 54 Mb/s, Strength 38 WPA2 TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 50 WPA TP-L: Infra, D8:5D:4C:BF:25:B4, Freq 2427 MHz, Rate 54 Mb/s, Strength 48 WPA pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 64 WEP dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 68 WPA - Device: wlan1 [Auto TP-L] --------------------------------------------------- Type: 802.11 WiFi Driver: ar9170usb State: connected Default: yes HW Address: D8:5D:4C:8D:4B:72 Capabilities: Speed: 54 Mb/s Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points (* = current AP) ZXDSL531BII-66852E: Infra, 00:16:E3:66:85:2E, Freq 2462 MHz, Rate 54 Mb/s, Strength 41 *TP-L: Infra, D8:5D:4C:BF:25:B4, Freq 2427 MHz, Rate 54 Mb/s, Strength 100 WPA pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 84 WEP TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 58 WPA dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 71 WPA IPv4 Settings: Address: 192.168.1.200 Prefix: 24 (255.255.255.0) Gateway: 192.168.1.2 DNS: 192.168.1.2 $ uname -a Linux localhost.localdomain 2.6.35.6-43.fc14.i686 #1 SMP Wed Oct 13 21:36:03 UTC 2010 i686 i686 i386 GNU/Linux
Created attachment 454940 [details] Atheros - TP-Link WN721N USB - NM Connection Information - 1Mbps TP-Link TL-WN722N Wireless USB adapter (Atheros chipset - 'ath9k') Same problem - Network Manager Connection Information shows 1Mbps And driver name is strange: 'usb' According to Linux-Wireless page http://linuxwireless.org/en/users/Drivers/ath9k driver should be 'ath9k' http://wireless.kernel.org/en/users/Drivers/ath9k_htc/devices Vendor Product Chipset USB Vendor USB Product MaxPower TP-Link TL-WN721N AR9271 0x0cf3 0x9271 500mA TP-Link TL-WN722N AR9271 0x0cf3 0x9271 TP-Link TL-WN422G v2 AR9271 0x0cf3 0x1006 Atheros AR7010 0x0cf3 0x7010 Atheros AR9271 0x0cf3 0x7015 Netgear WNA1100 AR9271 0x0846 0x9030 Netgear WNDA3200 AR9271 0x0846 0x9018 D-Link 150 AR9271 0x07d1 0x3a10 Azurewave AR9271 0x13d3 0x3327 Azurewave AR9271 0x13d3 0x3328 LiteOn AR9271 0x04ca 0x4605 SMC Networks AR9271 0x083a 0xa704 Device information ------------------------ $lsusb -v ... Bus 002 Device 004: ID 0cf3:9271 Atheros Communications, Inc. AR9271 802.11n Device Descriptor: bLength 18 bDescriptorType 1 bcdUSB 2.00 bDeviceClass 255 Vendor Specific Class bDeviceSubClass 255 Vendor Specific Subclass bDeviceProtocol 255 Vendor Specific Protocol bMaxPacketSize0 64 idVendor 0x0cf3 Atheros Communications, Inc. idProduct 0x9271 AR9271 802.11n bcdDevice 1.08 iManufacturer 16 iProduct 32 iSerial 48 bNumConfigurations 1 Configuration Descriptor: bLength 9 bDescriptorType 2 wTotalLength 60 bNumInterfaces 1 bConfigurationValue 1 iConfiguration 0 bmAttributes 0x80 (Bus Powered) MaxPower 500mA Interface Descriptor: bLength 9 bDescriptorType 4 bInterfaceNumber 0 bAlternateSetting 0 bNumEndpoints 6 bInterfaceClass 255 Vendor Specific Class bInterfaceSubClass 0 bInterfaceProtocol 0 iInterface 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x01 EP 1 OUT bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x82 EP 2 IN bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x83 EP 3 IN bmAttributes 3 Transfer Type Interrupt Synch Type None Usage Type Data wMaxPacketSize 0x0040 1x 64 bytes bInterval 1 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x04 EP 4 OUT bmAttributes 3 Transfer Type Interrupt Synch Type None Usage Type Data wMaxPacketSize 0x0040 1x 64 bytes bInterval 1 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x05 EP 5 OUT bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0 Endpoint Descriptor: bLength 7 bDescriptorType 5 bEndpointAddress 0x06 EP 6 OUT bmAttributes 2 Transfer Type Bulk Synch Type None Usage Type Data wMaxPacketSize 0x0200 1x 512 bytes bInterval 0
TP-Link TL-WN722N (150Mbps 802.11n adapter) ---------------------- Speed: 1 Mb/s $ nm-tool NetworkManager Tool State: connected - Device: eth0 ----------------------------------------------------------------- Type: Wired Driver: ATL1E State: unavailable Default: no HW Address: 00:A0:D1:A9:D2:AF Capabilities: Carrier Detect: yes Wired Properties Carrier: off - Device: wlan0 ---------------------------------------------------------------- Type: 802.11 WiFi Driver: iwlagn State: disconnected Default: no HW Address: 00:16:EA:C6:7D:0C Capabilities: Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points ZXDSL531BII-66852E: Infra, 00:16:E3:66:85:2E, Freq 2462 MHz, Rate 54 Mb/s, Strength 38 TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 48 WPA homenet: Infra, 00:26:18:74:54:EE, Freq 2412 MHz, Rate 54 Mb/s, Strength 41 WPA2 TP-L: Infra, D8:5D:4C:BF:25:B4, Freq 2427 MHz, Rate 54 Mb/s, Strength 50 WPA pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 71 WEP dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 67 WPA - Device: wlan2 [Auto TP-L] --------------------------------------------------- Type: 802.11 WiFi Driver: usb State: connected Default: yes HW Address: D8:5D:4C:8E:D7:D2 Capabilities: Speed: 1 Mb/s Wireless Properties WEP Encryption: yes WPA Encryption: yes WPA2 Encryption: yes Wireless Access Points (* = current AP) TRENDnet: Infra, 00:14:D1:57:F8:11, Freq 2412 MHz, Rate 54 Mb/s, Strength 38 WPA *TP-L: Infra, D8:5D:4C:BF:25:B4, Freq 2427 MHz, Rate 54 Mb/s, Strength 100 WPA ZXDSL531BII-66852E: Infra, 00:16:E3:66:85:2E, Freq 2462 MHz, Rate 54 Mb/s, Strength 31 pantherx: Infra, 00:1C:C5:D8:34:8C, Freq 2462 MHz, Rate 54 Mb/s, Strength 70 WEP dlink: Infra, 00:26:5A:32:B7:39, Freq 2437 MHz, Rate 54 Mb/s, Strength 100 WPA IPv4 Settings: Address: 192.168.1.71 Prefix: 24 (255.255.255.0) Gateway: 192.168.1.2 DNS: 192.168.1.2
Created attachment 454944 [details] System monitor - TP-Link WN722N - Network transfer speed (Wi-Fi) is 2.0 MiB/s to 3.5 MiB/s TP-Link WN722N (Atheros) System monitor in GNOME shows that - Network transfer speed (Wi-Fi) is 2.0 MiB/s to 3.5 MiB/s This doesn't correspond with 1Mb/s in Network Manager Connection Information
Created attachment 455114 [details] Windows - TP-Link WN821N - connected at 300Mbps Access Point is the same as for Linux test - WR841ND Wireless Router. (802.11n, 300Mbps)
Created attachment 455116 [details] Windows - TP-Link WN722N - connected at 150Mbps Access Point is the same as for Linux test - WR841ND Wireless Router. (802.11n, 300Mbps) This adapter supports 150Mbps connectivity (802.11N Light) As you can see, both WN821N and WN722N can connect to AP at 11n speed.
Created attachment 455126 [details] Windows - Intel WiFi Link 5100 - connected at 54Mbps Final piece of this puzzle - Intel Wi-Fi Adapter 5100, connected to the same Access Point (TP-Link WR841ND router) at 54Mbps According to sticker on computer (Acer 6935G notebook), it is 802.11a/b/g WLAN adapter. 802.11n is not supported (despite driver in Linux is 'iwlagn')
Thank you for detailed informations. Currently in linux we show only legacy (non 802.11n) rates, I think this should be fixed. Also we show rate of last transmitted frame, perhaps we should show also maximum capable rate by device. Both of these fixes need some upstream work, I'm going to ask about that. If you compare in practice windows and linux, does they differ? You don't need to provide detailed information, just something like: "on linux 100MB ftp file transfer between two local machines takes 50 seconds whereas on windows it was 10 seconds on the same environment (same devices, same accommodation, same AP)".
I need to check transfer time on Linux and Windows, for same environment. But looking at response time (ping) - Intel 5100 was somewhat unstable and slow, Atheros 9170 USB and other Atheros USB stick (TP-Link WN722N, TP-Link WN821) are also slower than in Windows. Transfer of 650MB ISO image between two computers (one connected via Ethernet to router, another - via Wi-Fi 11n, Windows) was very fast, about 1.5-2 min. Transfer rate was 5.5MB-7MB/sec. (megabytes per second), no difference to wired transfer. I guess this was also kind of limit of wired Ethernet interface, and host (Server) itself - it was Windows Vista x86 (not 64 bit). With 54Mbit adapter, transfer rate was about 1.8MB-2MB/sec. and it took 6-7 minutes to transfer ISO image. You would be surprised, but with Ralink RT73usb adapter (TP-Link TL-WN321, 11g), which is "cheapest and most simple" Wi-Fi adapter in TP-Link range (about $13 here!) was delivering same 54MBit/sec. (1.6MB-1.9MB/sec.) in Linux as in Windows. And as far as I recall, NetworkManager Connection Manager was displaying 54Mbit. (not 1Mbit)
(In reply to comment #22) > Thank you for detailed informations. > > Currently in linux we show only legacy (non 802.11n) rates, I think this should > be fixed. Also we show rate of last transmitted frame, perhaps we should show > also maximum capable rate by device. Both of these fixes need some upstream > work, I'm going to ask about that. > Re: perhaps we should show also maximum capable rate by device The best is to have Status monitor. You will see in a graphical representation what current uplink speed, and if it is changing comparing to what it was few seconds/minute ago. Tx/RX transmit speed should be possible to display in System Monitor like on this picture: https://bugzilla.redhat.com/attachment.cgi?id=454944 TP-Link Wireless Utility shown on https://bugzilla.redhat.com/show_bug.cgi?id=640612#c19 displays Current Receive Rate and Current Transmit Rate measured in Mbps. And those values are changed dynamically, sometimes for 11n connection it goes down to 270Mbps, than up. And - THIS IS IMPORTANT - if 2nd channel for full-speed 802.11n 300MBps connection is busy/noisy (See also Control Channel and Extension Channel), connetcion speed would be just 150Mbps, with same hardware. In order to resolve this, I was changing settings on router (from 2nd Channel Up - to 2nd Channel Down)
Found this interesting page: http://en.wikipedia.org/wiki/Comparison_of_open_source_wireless_drivers *** Non-free firmware required: Yes *** Does Fedora14 install "non-free" firmware by default? I guess compat-wireless package is needed, which provides firmware and scripts for adapters requiring "non-free" firmware.
re: comment 21 -- not all IWL5100 devices support 802.11n. It would seem that yours is one that does not. Given that you seem to be reporting transfer speeds as high as 54 Mbps, is there any actual problem here? re: comment 22 -- for the record, 802.11n speeds are reported ever since commit 254416aae70ab2e6b57fd79782c8a67196234d02 ("wireless: report reasonable bitrate for MCS rates through wext"). re: comment 25 -- some of the information at that link is suspect, out-of-date, or just wrong. In particular, the firmware for the iwlagn devices is provided as part of the standard Fedora install.
(In reply to comment #26) > re: comment 21 -- not all IWL5100 devices support 802.11n. It would seem that > yours is one that does not. Given that you seem to be reporting transfer > speeds as high as 54 Mbps, is there any actual problem here? According to sticker on computer (Acer 6935G notebook), it is 802.11a/b/g WLAN adapter. So it has been sold as 54Mbit - I do not expect from this adapter 11N connectivity. As I may need 11N speed- I bought recently couple of Wi-Fi USB sticks. They work in Windows at 300Mbps and 150Mbps Results in Linux are mixed (see above) > > re: comment 22 -- for the record, 802.11n speeds are reported ever since commit > 254416aae70ab2e6b57fd79782c8a67196234d02 ("wireless: report reasonable bitrate > for MCS rates through wext"). I was not able to get such a speed. See above. > > re: comment 25 -- some of the information at that link is suspect, out-of-date, > or just wrong. In particular, the firmware for the iwlagn devices is provided > as part of the standard Fedora install. Wikipedia pages indeed can be outdated. But sometimes they provide very valuable information.
(In reply to comment #22) > Thank you for detailed informations. > > Currently in linux we show only legacy (non 802.11n) rates, I think this should > be fixed. Also we show rate of last transmitted frame, perhaps we should show > also maximum capable rate by device. Both of these fixes need some upstream > work, I'm going to ask about that. > > If you compare in practice windows and linux, does they differ? You don't need > to provide detailed information, just something like: "on linux 100MB ftp file > transfer between two local machines takes 50 seconds whereas on windows it was > 10 seconds on the same environment (same devices, same accommodation, same > AP)". I did "real-life" test - transfer of ~690MB ISO file from FTP server (Filezilla running on Windows Vista) to Windows and Fedora 14. Results are quite interesting. I attach screenshots from each transfer below. Fedora 14 -Intel 5100 (11g 54Mbit) ~ 2.3 MB/sec. transfer speed -TP-Link WN821N (ar9170usb, 11n 300Mbit) ~ 2.0 MB/sec. This is lower than Intel 5100, and much lower comparing to Windows. !! should be faster than Intel 5100/54Mbit! ftp client: Nautilus 2.32.0 Windows Vista - Intel 5100 (11g 54Mbit) ~ 2.3 MB/sec. transfer speed - TP-Link WN821N (ar9170usb, 11n 300Mbit) ~ 3.0 MB/sec. This is faster than Intel 5100, both in Linux and Windows ftp client: Filezilla client Network utilization was about 40% for Intel 5100 (54Mbit) and 12% for TP-Link WN821N (300Mbit) So it seems 3MB/sec is kind of physical limitation for ftp server (Filezilla on Windows Vista) or Vista networking stack. And for both adapters, it seems Network Manager Connection Information (Fedora 14) returns not correct info. With 1Mbit link real transfer speed would be much lower than what I had.
Created attachment 456079 [details] File transfer (ftp) speed - Intel5100 54Mbit - Fedora 14
Created attachment 456080 [details] File transfer (ftp) speed - TP-Link WN821N 300Mbit - Fedora 14
Created attachment 456081 [details] File transfer (ftp) speed - Intel5100 54Mbit - Windows Vista
Created attachment 456082 [details] File transfer (ftp) speed - TP-Link WN821N (ar9170usb) 300Mbit - Windows Vista
Hello, Is there anything I can do to help here ? I am using ath9k and I am having speed issues with my wifi connection. This is a regression compared to Fedora 13 where everything was working great regarding wireless. Do not hesitate to ask for any information or test case. Thank you very much for your help !
(In reply to comment #33) > Is there anything I can do to help here ? Ideally if you attach patch with fix :-) We have a few bug reports about performance problem after update to F-14. However these problems are AP specific, I can not reproduce performance degradation with APs that I have, even if I have the same wifi devices as bug reporters. So what is your AP, maybe one of my colleague the same, so I can borrow it. Also are you willing to install upstream kernel from git repository and perform bisection to find commit that cause problems? If so I can give you some advices instructions.
I have following Access Points/Routers: - ASUS WL-500G Premium v1 Wireless Router (Broadcom chipset) - flashed with DD-Wrt build 14896 http://vplessky.blogspot.com/2010/11/dd-wrt-in-client-mode-status.html (screenshot is for Client Mode setup, but AP mode works as well, it was default setup) - TP-Link WR1043ND Wireless Router with factory firmware (Atheros chipset) http://vplessky.blogspot.com/2010/11/tp-link-tl-wr1043nd-wireless-router.html - TP-Link WR841ND Wireless Router with OpenWrt 10.03.1-RC4 (Atheros chipset) http://vplessky.blogspot.com/2010/10/openwrt-running-on-tp-link-wr841nd.html - TP-Link WA801ND Access Point - factory firmware (Atheros chipset) http://vplessky.blogspot.com/2010/11/tp-link-wa801nd-in-repeater-mode-status.html (SS for Repeater mode, works in AP mode just fine) And here is SS with Signal Strength from my routers/APs. From router: http://vplessky.blogspot.com/2010/11/tp-link-wa801nd-in-repeater-mode-scan.html From Windows XP (RaUI utility, TP-Link Wireless Utility): http://vplessky.blogspot.com/2010/11/wa801nd-repeater-to-openwrt-scan.html I have not noticed any difference when connecting with same Wi-Fi adapter to different devices. In MS Windows: any adapter (with exception of Ralink RT2500, which can't connect over 2 concrete walls) connects to those routers/APs without any problem, at full speed (54Mbit, 150Mbit, 300Mbit - depending on adapter and radio in AP) So I think problem is caused by driver or middleware software (like NetworkManager) in Linux.
(In reply to comment #34) > (In reply to comment #33) > > Is there anything I can do to help here ? > Ideally if you attach patch with fix :-) I knew I would get such answer :-) > So what is your AP, maybe one of my colleague the same, so I can borrow it. > I am in France and my AP is the one from my french ISP (Free - Freebox v5). I think it has 802.11n capabilities. I am not sure we will be able to test it together ... unless you come to Nice, France :-) I am 99% convinced that the issue comes from the wifi stack (core kernel or module I do not know) since I have no issue watching a streaming video (using flash player square preview on x86_64 and Firefox) with network cable but issues after a couple of minutes through the wireless connection. Also I tested my connection (speedtest.net, pingtest.net) and there are no issue in this area. > Also are you willing to install upstream kernel from git repository and perform > bisection to find commit that cause problems? If so I can give you some advices > instructions. This is totally an area where I can help. I am quiet used to bisecting (with mercurial), I am curious to see how it will go with git and the kernel. Also I have a simple test case to reproduce the issue. Unfortunatly it requires a manual testing (watch a video through a streaming platform, it will fail after 1 or 2 minutes). Here are a few questions that would help me : - Which branch should I clone from ? a git url would be really great, please :-) - Which changeset should I flag as good (not sure which changeset corresponds to latest F13 kernel) ? - Which changeset should I flag as bad (not sure which changeset corresponds to current F14 kernel) ? - Do I have to recompile the whole kernel and reboot with it at each iteration or I can boot one kernel and recompile/reload some modules only ? - It is a long time since I compiled the kernel by hand. When I have to do it I usually use rpmbuild/mock and this is all simple. Do I have a simple way to not mess up my current installation with new kernel files and headers, I mean will I be able to easily remove the test kernel+headers ? - I have no idea about kernel development tools. Is there any tools beside bisect to test that issue quickly or the only test possible is build/install/reboot/test ? Thank you very much for your time and your help here. I will try to be as a good tester as possible !
Needed tools can be installed by: # yum-builddep kernel # yum install ncurses-devel You can clone linus' tree. $ git clone git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git First step would be compile this kernel and check if problem is fixed there. If it's fixed, we will need to find what commit fix the problem. If not, we will need find first commit that broke things. Kernel compilation can be challenge by itself. First you have to copy config from /boot and configure kernel based on that, for example: $ cp /boot/config-2.6.35.6-48.fc14.i686 .config $ make oldconfig Since version are different "make oldconfig" will ask you to enable/or disable some new option, you can simply press enter many times to accept defaults. This configure kernel more or less the same as fedora kernel. However compilation of such configured kernel may take very log, i.e few hours if your machine is slow. Since you will be compiling kernel on every bisection step is good to make it compile fast, remove all things that you not needed (i.e infiniband, 10Gbit ethernet, GFS, etc ...). To do so use below command: $ make menuconfig Configuration need some knowledge about kernel itself and hardware you have. There are lots of things in "Networking support", "Device Drivers", "File systems" you can remove. But be careful, removing too much may cause kernel does not even boot, i.e. if you remove support for file system that is on your root partition. Now you can compile kernel by $ make -j N+1 where N is number of cpu cores you have, this make the compilation faster. After compilation finish install kernel: # make modules_install # make install This install kernel in /boot, modules in /lib/modules and add entry in /etc/grub.conf (check if that file exist and is link to /boot/grub/grub.conf). There is no uninstall, you have remove things by yourself (but be careful to remove proper version). Now you can boot the kernel. If it not boot check if you do not remove something vital from configuration. If not try with some older version $ git checkout -b b2.6.36 v2.6.36 This will switch to 2.6.36, now the same procedure make oldconfig, make, make modules_install, make install. This have to be done at every bisection step. Regarding booting, there is also one problem in fedora, the X window driver may not be compatible with upstream kernel (this mostly case with nvidia novou driver). So to make thinks boot, you need to boot into runlevel 3. Change to "id:3:initdefault:" in /etc/inittab. If this will be your problem you will also need find steps to reproduce from console (you can use cnetworkmanager to connect and some tools like wget to generate network traffic). How to perform bisection is described here: http://www.kernel.org/pub/software/scm/git/docs/git-bisect.html IIRC things works in 2.6.34 kernel and does not work in 2.6.35, if so to start bisection do: git bisect start -- drivers/net/wireless/ net/wireless/ net/mac80211/ git bisect bad v2.6.35 git bisect good v2.6.34 It thinks works in upstream, we want to know what commit do the fix to backport it to fedora. In git bad commit must be allays newer then good one. So now bad kernel is now the one that works, remember about that at bisection step, to start: git bisect start -- drivers/net/wireless/ net/wireless/ net/mac80211/ git bisect bad HEAD git bisect good v2.6.35 Hope this answer your questins, happy bisecting :-)
Thank you very much Stanislaw for all those very valuable information. I have all the information I need. Also do not worry too much about me : this will not be my first kernel compilation. However since it is a long time I did it from scratch (I usually run rpmbuild/mock) I was very happy to have all the details, in particular the git one since I am more a mercurial user in my daily life. I guess this will take some time for me (unless I am lucky with the bisection) but I will report here as soon as I have any valuable information, in particular whether the bug is upstream or not. Again, a big thank you for your time and all the very clear information :-) See you all !
I am also here and looking forward to do testing. It would be very convenient to have common "test base" TP-Link WR841ND wireless router (which I also mentioned above) http://www.tp-link.com/products/productDetails.asp?pmodel=TL-WR841ND has price just about $40 here (less than Euro 30) Form my experience- and I tested a lot of Wi-Fi devices - it provides very good wireless signal and stable connection. This is due to the fact (in my opinion) that it uses Atheros AR71xx chipset (1), and Linux kernel 2.6.15 in factory firmware (2). Major disadvantage of factory firmware is that you can't connect to router via telnet or ssh. This can be fixed by flashing OpenWrt firmware. 1) http://wiki.openwrt.org/toh/tp-link/tl-wr841nd 2) TP-Link GPL Code Center http://www.tp-link.com/support/gpl.asp OpenWrt 10.03.1-RC4 uses kernel 2.6.32 you can get all required information about wireless part and network configuration - via #ifconfig, iwconfig (telnet/ssh).
I'd like to add that I also reported (on 2010-11-08) this bug at Intel Linux Wireless (http://www.intellinuxwireless.org/) - http://bugzilla.intellinuxwireless.org/show_bug.cgi?id=2268 Bug is assigned to: wey-yi.w.guy QA contact: jiajia.zheng But unfortunately there is no update for this bug since beginning of Nov., 2010
Created attachment 467714 [details] 0001-iwlagn-remove-low-ack-count-check.patch Before bisecting can You test this patch on top of fedora kernel (or upstream if it not work). I can prepare kernel packages with that patch if wish.
Both of you have problem on intel 5xxx, right?
I have problem on Intel 5xxx, and: -Atheros 9170usb adapter (TP-Link WN821N) -Atheros 9271 USB adapter (TP-Link WN722N) -Atheros AR9285 Wireless Network Adapter (ath9k driver) You closed these bugs as duplicates of Bug 640612: https://bugzilla.redhat.com/show_bug.cgi?id=641250 https://bugzilla.redhat.com/show_bug.cgi?id=640767 https://bugzilla.redhat.com/show_bug.cgi?id=640779
(In reply to comment #42) > Both of you have problem on intel 5xxx, right? Nope, see comment #33 : I am using ath9k. And yes I can test whatever patch you want me to test :-)
Hello, Here are some good news from my testing so far : The issue is fixed in Linus's tree for kernel 2.6.36 and 2.6.37-rc5. I have not tested yet the 2.6.35 from Linus's tree so I cannot tell whether this is a regression in 2.6.35 from upstream or from Fedora. I used the exact same config on all Kernels. I just copied the one from my Fedora's 2.6.35 config in /boot. Next steps for me during the weekend : test upstream 2.6.35 and see if the issue is there two. test Fedora's 2.6.36 in koji and see if the issue is fixed there too. See you !
I did testing with TP-Link WR1043ND (Atheros AR71xx), flashed with OpenWrt 10.03.1-RC4 Wireless config on router root@OpenWrt:~# cat /etc/config/wireless config 'wifi-device' 'radio0' option 'type' 'mac80211' option 'macaddr' 'd8:5d:4c:b9:d9:3c' option 'hwmode' '11ng' option 'htmode' 'HT20' list 'ht_capab' 'SHORT-GI-40' list 'ht_capab' 'DSSS_CCK-40' option 'disabled' '0' option 'channel' '5' config 'wifi-iface' option 'device' 'radio0' option 'network' 'lan' option 'mode' 'ap' option 'encryption' 'psk' option 'key' 'mykey' option 'ssid' 'WR1043ND-OWrt' Connected from two computers - one is Windows Vista (Atheros chipset). Windows Network Center says that I connected at 130Mbps it's ok for htmode HT20, single channel) Linux (Network Manager) - 1Mbps Let's check what is visible from router root@OpenWrt:~# iw wlan0 station dump Station 00:13:d3:84:fe:9d (on wlan0) - Ralink RT2500pci (Linux) inactive time: 15790 ms rx bytes: 9257 rx packets: 114 tx bytes: 10151 tx packets: 59 signal: -35 dBm tx bitrate: 1.0 MBit/s Station 00:22:43:30:99:0d (on wlan0) - Atheros AR928x (Windows) inactive time: 5260 ms rx bytes: 124388 rx packets: 1012 tx bytes: 47818 tx packets: 300 signal: -20 dBm tx bitrate: 1.0 MBit/s Doesn't look good... root@OpenWrt:~# iw wlan0 info Interface wlan0 ifindex 7 type AP Than I started ping from 00:22:43:30:99:0d to router root@OpenWrt:~# iw wlan0 station dump Station 00:13:d3:84:fe:9d (on wlan0) inactive time: 520 ms rx bytes: 9929 rx packets: 130 tx bytes: 11901 tx packets: 69 signal: -38 dBm tx bitrate: 1.0 MBit/s Station 00:22:43:30:99:0d (on wlan0) inactive time: 900 ms rx bytes: 132280 rx packets: 1151 tx bytes: 58552 tx packets: 408 signal: -25 dBm tx bitrate: 58.5 MBit/s MCS 6 continous ping (ping -t) from Windows command line root@OpenWrt:~# iw wlan0 station dump Station 00:13:d3:84:fe:9d (on wlan0) inactive time: 4440 ms rx bytes: 10060 rx packets: 131 tx bytes: 12056 tx packets: 70 signal: -40 dBm tx bitrate: 12.0 MBit/s Station 00:22:43:30:99:0d (on wlan0) inactive time: 170 ms rx bytes: 134877 rx packets: 1204 tx bytes: 62050 tx packets: 437 signal: -26 dBm tx bitrate: 58.5 MBit/s MCS 6 root@OpenWrt:~# iw wlan0 station dump Station 00:13:d3:84:fe:9d (on wlan0) inactive time: 22490 ms rx bytes: 10060 rx packets: 131 tx bytes: 12056 tx packets: 70 signal: -40 dBm tx bitrate: 12.0 MBit/s Station 00:22:43:30:99:0d (on wlan0) inactive time: 160 ms rx bytes: 136631 rx packets: 1223 tx bytes: 63964 tx packets: 464 signal: -25 dBm tx bitrate: 78.0 MBit/s MCS 12 All this looks really strange to me. How is 'tx bitrate' calculated? And what is MCS 6, MCS 12?.. Both computers are located in 1m from router, signal strength is 100% Than I start pinging router form Linux PC. Rate (reported by router) increases to 48Mbit/sec., but not higher If you try command #iw wlan0 station dump several times, you can get for Windows PC speed up to 130Mbit/sec., and than it can go down again to 1Mbps. Looks like driver makes some sampling,and than returns "appx. speed"
Changed on wireless router option 'htmode' 'HT20' to option 'htmode' 'HT40-' (two channels, second channel down) Windows Vista client (AR928x) connects now at 300Mbps See screenshot: http://vplessky.blogspot.com/2010/12/openwrt-on-wr1043nd-client-ar928x.html Station dump on router --------------------------------------- root@OpenWrt:~# iw wlan0 station dump Station 00:22:43:30:99:0d (on wlan0) inactive time: 10 ms rx bytes: 43196513 rx packets: 51376 tx bytes: 3108619 tx packets: 26106 signal: -25 dBm tx bitrate: 270.0 MBit/s MCS 14 40Mhz short GI Station 00:13:d3:84:fe:9d (on wlan0) inactive time: 20 ms rx bytes: 193712 rx packets: 1744 tx bytes: 258599 tx packets: 1493 signal: -39 dBm tx bitrate: 54.0 MBit/s NOTE: both Windows and Linux have continuous ping to router ---------------------------------------------- root@OpenWrt:~# cat /etc/config/wireless config 'wifi-device' 'radio0' option 'type' 'mac80211' option 'macaddr' 'd8:5d:4c:b9:d9:3c' option 'hwmode' '11ng' option 'htmode' 'HT40-' list 'ht_capab' 'SHORT-GI-40' list 'ht_capab' 'DSSS_CCK-40' option 'disabled' '0' option 'channel' '5' config 'wifi-iface' option 'device' 'radio0' option 'network' 'lan' option 'mode' 'ap' option 'encryption' 'psk' option 'key' 'mykey' option 'ssid' 'WR1043ND-OWrt'
I use TP-LINK TL-WN722N USB wireless interface (ID 0cf3:9271 Atheros Communications, Inc. AR9271 802.11n). It work, but iwconfig reported only 1 Mb/s. Real data transfer rate >10 Mb/s. I experienced this with kernel 2.6.35.10-77.fc14.x86_64. I compiled rawhide 2.6.37-2.fc15 kernel to Fedora 14, and I experienced same problem. If I cange transfer rate manually, I get an error message: iwconfig wlan2 rate 54M Error for wireless request "Set Bit Rate" (8B20) : SET failed on device wlan2 ; Operation not supported.
with Intel 5100 Wi-Fi adapter, Fedora14 - kernel-2.6.35.10-74.fc14.i686 (23 Dec.2010): [vadim@Acer6935 ~]$ iw wlan0 station dump Station d8:5d:4c:b9:d9:3c (on wlan0) inactive time: 0 ms rx bytes: 230463719 rx packets: 273301 tx bytes: 6044469 tx packets: 69713 signal: -54 dBm tx bitrate: 48.0 MBit/s Real transfer rate is ~ 2.0-2.1MB/sec. (ftp transfer of 700MB ISO image from NAS) This is rather low, for 54Mbps transfer rate should be 3.0-3.5MB/sec.
And here is what I get on router, for Intel 5100 adapter connected to this router: root@OpenWrt:~# iw wlan0 station dump Station 00:16:ea:c6:7d:0c (on wlan0) inactive time: 10 ms rx bytes: 21143 rx packets: 143 tx bytes: 21880 tx packets: 94 signal: -53 dBm tx bitrate: 18.0 MBit/s root@OpenWrt:~# iw wlan0 station dump Station 00:16:ea:c6:7d:0c (on wlan0) inactive time: 10 ms rx bytes: 25627 rx packets: 195 tx bytes: 23800 tx packets: 104 signal: -51 dBm tx bitrate: 48.0 MBit/s OpenWrt version is Backfire 10.03.1-RC4 Hardware: Atheros AR71xx/AR9132 (TP-Link WR1043ND) root@OpenWrt:~# iw --version iw version 0.9.21
For TP-Link TL-WN821N (802.11N) USB adapter: On router: ------------------------------- root@OpenWrt:~# iw wlan0 station dump Station d8:5d:4c:8d:4b:72 (on wlan0) inactive time: 10 ms rx bytes: 5057829 rx packets: 56865 tx bytes: 140530607 tx packets: 92136 signal: -62 dBm tx bitrate: 120.0 MBit/s MCS 11 40Mhz short GI on Acer 6935G notebook (client), Fedora 14, kernel-2.6.35.10-74.fc14.i686 ---------------------------------------------- [vadim@Acer6935 ~]$ iw dev wlan1 station dump Station d8:5d:4c:b9:d9:3c (on wlan1) inactive time: 1 ms rx bytes: 121352713 rx packets: 86785 tx bytes: 4286146 tx packets: 48190 signal: -57 dBm tx bitrate: 54.0 MBit/s
And finally with TP-Link WN722N USB Wi-Fi adapter (with 1x external 4dBi antenna, 150Mbps) on router: --------------------------------------------- root@OpenWrt:~# iw wlan0 station dump Station d8:5d:4c:8e:d7:d2 (on wlan0) inactive time: 10 ms rx bytes: 48044 rx packets: 287 tx bytes: 135529 tx packets: 246 signal: -50 dBm tx bitrate: 150.0 MBit/s MCS 7 40Mhz short GI on Acer 6935G notebook (client), Fedora 14, kernel-2.6.35.10-74.fc14.i686 --------------------------------------------- [vadim@Acer6935 ~]$ iw dev wlan2 station dump Station d8:5d:4c:b9:d9:3c (on wlan2) inactive time: 2206 ms rx bytes: 487948 rx packets: 1856 tx bytes: 72857 tx packets: 424 signal: -55 dBm tx bitrate: 1.0 MBit/s Transfer rate for 700MB ISO file is 2.5MB-3MB/sec.
I unlocked 11N functionality using iwleeprom (http://code.google.com/p/iwleeprom/) on this card. But inconsistency in Fedora 14 and what is reported by router (WR1043ND-OpenWrt) remains. OpenWrt: ------------- root@OpenWrt:~# iw wlan0 station dump Station 00:16:ea:c6:7d:0c (on wlan0) inactive time: 0 ms rx bytes: 26675447 rx packets: 308568 tx bytes: 1805938318 tx packets: 1177534 signal: -27 dBm tx bitrate: 300.0 MBit/s MCS 15 40Mhz short GI computer / Fedora 14: -------------------- [vadim@Acer6935 ~]$ iw wlan0 station dump Station d8:5d:4c:b9:d9:3c (on wlan0) inactive time: 4 ms rx bytes: 1775953476 rx packets: 1162110 tx bytes: 26400527 tx packets: 305145 signal: -16 dBm tx bitrate: 150.0 MBit/s MCS 7 40Mhz short GI ** --> 300 MBit/s MCS 15 tx bitrate on router, and tx bitrate: 150.0 MBit/s MCS 7 on computer. Both are using 40Mhz channel with short GI. ---------------------------------------------------------------- [vadim@Acer6935 ~]$ ifconfig wlan0 wlan0 Link encap:Ethernet HWaddr 00:16:EA:C6:7D:0C inet addr:10.10.20.61 Bcast:10.10.20.255 Mask:255.255.255.0 inet6 addr: fe80::216:eaff:fec6:7d0c/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:830387 errors:0 dropped:0 overruns:0 frame:0 TX packets:238683 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:1256686923 (1.1 GiB) TX bytes:20651445 (19.6 MiB) [vadim@Acer6935 ~]$ iwconfig wlan0 wlan0 IEEE 802.11abgn ESSID:"WR1043ND-OWrt" Mode:Managed Frequency:2.417 GHz Access Point: D8:5D:4C:B9:D9:3C Bit Rate=150 Mb/s Tx-Power=15 dBm Retry long limit:7 RTS thr:off Fragment thr:off Power Management:off Link Quality=70/70 Signal level=-9 dBm Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 Tx excessive retries:0 Invalid misc:0 Missed beacon:0 Signal level is very good - and that is reported by both wireless router and wi-fi adapter. Network Manager -> Connection Information - reports 1Mbit/sec. when there are no transfers - "jumps" to 150 Mbit/sec. if I start FTP transfer - never seen 300 Mbit/sec. in Network Manager (but it was 300MBit/sec. on router)
Is this by a chance fixed on F-15 or F-16 ?
If this is fixed- than only "by chance" To my best knowledge, upstream bug (at Intel site) even not touched by Intel engineers.
There was various fixed applied (and some new bug too) in the iwlwifi driver since last year. Can you test on the latest fedora kernel 3.2.2 or 3.3-rc1 ?
This bug will not be fixed on F-14, if issue is still present on new releases, reopen this bug report.