You are not logged in.
Pages: 1
Hello, I'm a rather new Arch user, and so far I haven't had many issues outside of this particular one. I have this laptop which I use to dualboot Win10 and Arch, and the wifi speed in Arch in particular is really, really bad. My connection isn't that great, most of my devices (including this very same laptop within the Win10 partition) usually average around 17 mbps for download, but my laptop gets around 1.5 mbps at best, and normally just stays at 0.5 mbps during the day. With an ethernet connection it works great, I get ~30mbps. So I'm pretty sure the connection or the hardware aren't the roots of the problem.
I've been having this problem basically since I installed arch, and right now I'm running the kernel 5.5.7-arch1-1 x86_64. For other specifics, here are some outputs:
lspci -nn -s 01:00 | grep -i net (For the specific network adapter model)
01:00.0 Network controller [0280]: Broadcom Inc. and subsidiaries BCM4313 802.11bgn Wireless Network Adapter [14e4:4727] (rev 01)As for the configuration, I have the following:
sudo lshw -C network
*-network
description: Wireless interface
product: BCM4313 802.11bgn Wireless Network Adapter
vendor: Broadcom Inc. and subsidiaries
physical id: 0
bus info: pci@0000:01:00.0
logical name: wlp1s0b1
version: 01
serial: 84:4b:f5:8e:5e:d9
width: 64 bits
clock: 33MHz
capabilities: pm msi pciexpress bus_master cap_list ethernet physical wireless
configuration: broadcast=yrd driver=brcmsmac driverversion=5.5.7-arch1-1 firmware=610.812 ip=192.168.0.34 latency=0 link=yes multicast=yes wireless=IEEE 802.11
resources: irq:16 memory:c250000-c2503ffinxi -Fs (Network)
Device-1: Broadcom and subsidiaries BCM4313 802.11bgn Wireless Network Adapter driver: bcma-pci-bridgeI'm posting both because they both list different drivers. According to the answer to this question (which is actually about Ubuntu), this particular adapter does use two drivers, both of which are found within the linux-firmware package and which seem to be indeed the ones that the commands above list. So I think the drivers aren't the issue here either.
A few things I tried after browsing for a while where configuring the regdomain (which only resulted in getting slightly higher ping) and setting net.ipv4.tcp_ecn=0, both of which didn't really help.
I'm not sure what other information could be useful so if I'm missing something important I'll post it soon.
Last edited by ghsrkm (2020-03-07 06:43:22)
Offline
which I use to dualboot Win10 and Arch
First things first, ensure https://wiki.archlinux.org/index.php/Du … ibernation is disabled. You will have to reboot at least windows and pot. linux (ie. shutdown linux, boot windows, disable fast boot, reboot windows, boot linux, in doubt reboot linux) to ensure the HW is in a defined state.
If this doesn't help please post the ouputs of https://wiki.archlinux.org/index.php/Ne … _interface
Offline
First things first, ensure https://wiki.archlinux.org/index.php/Du … ibernation is disabled
Thanks, both Hibernation and Fast Boot were already disabled, just logged in to check the status and sure enough neither of them is set on.
Here's the info about the wireless network interface:
ghsrkm% ip link show wlp1s0b1
3: wlp1s0b1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP mode DORMANT group default qlen 1000
link/ether 84:4b:f5:8e:5e:d9 brd ff:ff:ff:ff:ff:ffghsrkm% iw dev wlp1s0b1 link
Connected to 48:5b:39:21:94:7d (on wlp1s0b1)
SSID: castillo 2
freq: 2462
RX: 3327384 bytes (37224 packets)
TX: 104722 bytes (886 packets)
signal: -35 dBm
rx bitrate: 36.0 MBits/s
tx bitrate: 6.0 MBits/s
bss flags: short-slot-time
dtim period: 1
beacon int: 100Offline
That seems ok.
Check your journal ("sudo journalctl -b").
Do you get frequent disconnects, or kernel module crashes (in doubt post the journal)?
In case you're using NM, try to disable https://wiki.archlinux.org/index.php/Ne … domization
Offline
My connection is actually fairly stable, it's really slow but I can't say I've had many disconnects at all.
Here's the journal output, I can't see an error or crash that would seem to be related to the issue here so far.
Disabling MAC randomization doesn't seem to have changed much sadly, I'm getting about the same test results
Offline
The logged session seems rather short - did you try to download sth. over the WiFi connection during those 3 minutes?
If so (or afterwards):
iwconfig wlp1s0b1(iwconfig from wireless_tools package, has some more details about netto ./. brutto RX)
Offline
Let's take a peek at the transport layer.
What is the output of cat /proc/net/wireless ?
Nevermind. Somehow I missed most of the thread.
Last edited by ewaller (2020-03-06 17:58:02)
Nothing is too wonderful to be true, if it be consistent with the laws of nature -- Michael Faraday
The shortest way to ruin a country is to give power to demagogues.— Dionysius of Halicarnassus
---
How to Ask Questions the Smart Way
Offline
Hi, I see you are using a Broadcom chip, BC43xx series. Do you have the corresponding firmware installed? The package is https://aur.archlinux.org/packages/b43-firmware/
Your kernel dmesg should show something like the following:
[ 18.544177] b43-phy0: Loading firmware version 784.2 (2012-08-15 21:35:19)
[ 18.894172] b43-phy0: Loading firmware version 784.2 (2012-08-15 21:35:19)Can you please post the output of the following?
dmesg|grep b43Last edited by topcat01 (2020-03-06 18:39:20)
Offline
When everything is fine, the command:
dmesg|grep 'b43\|Broadcom'should produce output similar to the following:
[ 14.483616] b43-phy0: Broadcom 43224 WLAN found (core revision 23)
[ 14.484102] b43-phy0: Found PHY: Analog 8, Type 4 (N), Revision 6
[ 14.484151] b43-phy0: Found Radio: Manuf 0x17F, ID 0x2056, Revision 11, Version 0
[ 14.484931] Broadcom 43xx driver loaded [ Features: PNLS ]
[ 14.926343] b43 bcma0:1 wlp3s0b1: renamed from wlan0
[ 18.544177] b43-phy0: Loading firmware version 784.2 (2012-08-15 21:35:19)
[ 18.894172] b43-phy0: Loading firmware version 784.2 (2012-08-15 21:35:19)
[ 28.877489] b43-phy0: Loading firmware version 784.2 (2012-08-15 21:35:19)I see in your log posting that the bcma0 step returned an error. This suggests you do not have the firmware?
Offline
The logged session seems rather short - did you try to download sth. over the WiFi connection during those 3 minutes?
If so (or afterwards):
I used the command shortly after booting up the computer and connecting to WiFi so that's probably why it was pretty short. Now I just tried a few downloads and here's what I got:
ghsrkm% iwconfig wlp1s0b1
wlp1s0b1 IEEE 802.11 ESSID:"castillo 2"
Mode:Managed Frequency:2.462 GHz Access Point: 48:5B:39:21:94:7D
Bit Rate=36 Mb/s Tx-Power=19 dBm
Retry short limit:7 RTS thr:off Fragment thr:off
Power Management:off
Link Quality=70/70 Signal level=-22 dBm
Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:567 Invalid misc:1760 Missed beacon:0
ghsrkm% iwconfig wlp1s0b1
wlp1s0b1 IEEE 802.11 ESSID:"castillo 2"
Mode:Managed Frequency:2.462 GHz Access Point: 48:5B:39:21:94:7D
Bit Rate=36 Mb/s Tx-Power=19 dBm
Retry short limit:7 RTS thr:off Fragment thr:off
Power Management:off
Link Quality=70/70 Signal level=-19 dBm
Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0
Tx excessive retries:640 Invalid misc:1957 Missed beacon:0I ran the command twice, the first one is while starting a download and the second one after that download is over. For comparison's sake, it was a ~12 mb file, and here I had an average download speed of like 70 kb/s, while on a different computer connected to the same network through wifi the download speed had an average of 1.3 mb/s (with the same file of course). The increase of Tx excessive retries seems meaningful, considering it's remained fairly stable after the download was over
Offline
Can you please post the output of the following?
dmesg|grep b43
Yes, I get the following:
ghsrkm% dmesg|grep b43
[ 11.503537] Support for cores revisions 0x17 and 0x18 disabled by module param allhwsupport=0. Try b43.allhwsupport=1
[ 11.503542] b43: probe of bcma0:1 failed with error -524So that seems to be an issue, I did not have the firmware installed. However, I just tried installing it from the AUR, and even after the installation finishes seemingly successfully, I'm still getting the same output, even after rebooting.
When everything is fine, the command:
dmesg|grep 'b43\|Broadcom'should produce output similar to the following:
After reinstalling, I'm getting this output
ghsrkm% dmesg|grep 'b43\|Broadcom'
[ 1.850571] usb 1-1.3: Manufacturer: Broadcom Corp
[ 11.503537] Support for cores revisions 0x17 and 0x18 disabled by module param allhwsupport=0. Try b43.allhwsupport=1
[ 11.503542] b43: probe of bcma0:1 failed with error -524
[ 11.503554] Broadcom 43xx driver loaded [ Features: PNLS ]The same error is showing up, so I don't know if it's an installation issue or something else, but I'll probably try what the error message suggests first.
Offline
How did that go? Btw, do you have the linux-firmware package installed as well?
Last edited by topcat01 (2020-03-06 22:28:37)
Offline
BCM4313 isn't supported by the b43 module, https://wireless.wiki.kernel.org/en/users/Drivers/b43
Tx excessive retries are scary, https://bbs.archlinux.org/viewtopic.php?id=211702 suggests the prop. wl driver works better (or at least used to 3 years ago…)
Offline
Right I missed that, thanks! Hope wl works!
Offline
https://bbs.archlinux.org/viewtopic.php?id=211702 suggests the prop. wl driver works better (or at least used to 3 years ago…)
Well, turns out it still works better to this day. The problem has been solved by installing that module, even if the process was a bit trickier than expected, but now I'm getting speeds on par with the rest of my devices. Thanks for your help and patience.
Offline
Pages: 1