Description of problem: attempting to configure broadcom BCM4311 wifi as it is not enabled by default gives the following outcome: [root@localhost ~]# urpmi dkms-broadcom-wl In order to satisfy the 'devel(libstdc++(64bit))' dependency, one of the following packages is needed: 1- libstdc++-devel-5.4.0-1.mga6.x86_64: Header files and libraries for C++ development (to install) 2- libstdc++5-devel-3.3.6-15.mga6.x86_64: Header files and libraries for C++ development (to install) What is your choice? (1-2) 1 To satisfy dependencies, the following packages are going to be installed: Package Version Release Arch (medium "Live Core") dkms 2.0.19 37.mga6 noarch gcc 5.4.0 1.mga6 x86_64 gcc-cpp 5.4.0 1.mga6 x86_64 glibc-devel 2.22 21.mga6 x86_64 isl 0.16.1 1.mga6 x86_64 kernel-desktop-devel-4.8.12-3> 1 1.mga6 x86_64 kernel-desktop-devel-latest 4.8.12 3.mga6 x86_64 (recommended) kernel-userspace-headers 4.8.12 1.mga6 x86_64 lib64isl15 0.16.1 1.mga6 x86_64 lib64mpc3 1.0.3 1.mga6 x86_64 lib64ncurses-devel 6.0 7.mga6 x86_64 libstdc++-devel 5.4.0 1.mga6 x86_64 make 4.1 5.mga6 x86_64 (medium "Live Nonfree") dkms-broadcom-wl 6.30.223.271 45.mga6.nonf> x86_64 143MB of additional disk space will be used. 34MB of packages will be retrieved. Proceed with the installation of the 14 packages? (Y/n) y installing lib64mpc3-1.0.3-1.mga6.x86_64.rpm libstdc++-devel-5.4.0-1.mga6.x86_64.rpm gcc-cpp-5.4.0-1.mga6.x86_64.rpm gcc-5.4.0-1.mga6.x86_64.rpm isl-0.16.1-1.mga6.x86_64.rpm lib64isl15-0.16.1-1.mga6.x86_64.rpm glibc-devel-2.22-21.mga6.x86_64.rpm kernel-userspace-headers-4.8.12-1.mga6.x86_64.rpm from /var/local/mga_rpms/core Preparing... ##################################################### 1/14: lib64mpc3 ##################################################### 2/14: kernel-userspace-headers ##################################################### error: Couldn't fork %prein(glibc-devel-6:2.22-21.mga6.x86_64): Cannot allocate memory ERROR: 'script' failed for glibc-devel-6:2.22-21.mga6.x86_64 error: glibc-devel-6:2.22-21.mga6.x86_64: install failed 3/14: isl ##################################################### 4/14: lib64isl15 ##################################################### 5/14: gcc-cpp ##################################################### error: Couldn't fork %post(gcc-cpp-5.4.0-1.mga6.x86_64): Cannot allocate memory ERROR: 'script' failed for gcc-cpp-5.4.0-1.mga6.x86_64 6/14: libstdc++-devel ##################################################### 7/14: gcc ##########################################Killed [root@localhost ~]# Version-Release number of selected component (if applicable): Mageia-6-sta2-LiveDVD-Plasma-x86_64-DVD.iso DATE.txt: Sat Dec 10 01:40:59 CET 2016 How reproducible:every time Steps to Reproduce: 1. 2. 3.
Created attachment 8769 [details] output of lspcidrake -v
Keywords: (none) => 6sta2
Component: RPM Packages => Release (media or process)Assignee: bugsquad => isobuildCC: (none) => marja11, sysadmin-bugs
I've managed to reproduce this in VirtualBox by reducing the system memory size to 1.25GB. What is the output of free -m on your system Ben? (when running the Live system)
CC: (none) => mageiaKeywords: (none) => NEEDINFO
(In reply to Martin Whitaker from comment #2) > I've managed to reproduce this in VirtualBox by reducing the system memory > size to 1.25GB. What is the output of > > free -m > > on your system Ben? (when running the Live system) [live@localhost ~]$ free -m total used free shared buff/cache available Mem: 1488 447 76 104 965 772 Swap: 0 0 0 [live@localhost ~]$ unit is a Compaq Presario C700 lappy, I was using it for the 5.1 testing
I've established the cutoff is somewhere between 1.5GB and 2GB (I've not tried to pin it down closer than that). The problem on a Live system is that the root file system is a union of the data stored on the Live media (DVD or USB) and a RAM disk. Any changes you make to the system, like installing packages, go onto the RAM disk, so use up some of your free memory. By default 6sta2 is using quite a bit more memory than 5.1, which is why you didn't see this before. I've found three big contributors to the increased memory usage. Disabling these will hopefully claw back enough to let you install the broadcom-wl driver. If you'd like to test this in advance of the next round of ISOs, do the following as root in a terminal: systemctl disable dnf-makecache.timer systemctl disable dnf-makecache.service systemctl disable cups.service systemctl disable cups-browsed.service then in the MCC remove all the remote software media sources, just leaving "Live Core" and "Live Nonfree" (which contain all the packages you need for installing the wl driver). Note that you need to disable dnf-makecache.timer reasonably quickly after boot, otherwise it kicks off a cache update. The rest you can take your time about.
Keywords: NEEDINFO => (none)Summary: [6sta2] Broadcom BCM4311 wireless is not enabled in "live" mode => [6sta2] Broadcom BCM4311 wireless is not enabled in "live" mode on system with < 2GB RAM
commit 7fbfd70bd4fdd12e06e336c814d0720682e198ff Author: Martin Whitaker <mageia@...> Date: Sun Dec 18 09:14:18 2016 +0000 live.cfg: reduce memory footprint to allow dkms drivers to build (mga#19943). - don't add online media, that significantly increases urpmi memory use - disable dnf-makecache service, the cache is big --- Commit Link: http://gitweb.mageia.org/software/build-system/draklive-config/commit/?id=7fbfd70bd4fdd12e06e336c814d0720682e198ff
Summary: [6sta2] Broadcom BCM4311 wireless is not enabled in "live" mode on system with < 2GB RAM => Broadcom BCM4311 wireless is not enabled in "live" mode on system with < 2GB RAM
Ben, am I right in thinking this is not a problem when using the Xfce Lives? If so, I think we will declare that to be the fix.
For Mageia 7 we also have the option of enabling persistence, which will reduce the RAM usage (and have the advantage you don't need to rebuild the driver every time).
Another thought. From reading around, the BCM4311 should be supported by the free b43 driver. I wonder if the problem is just a missing firmware file. If you boot the Live ISO without building the wl driver, and, as root, execute the commands modprobe ssb modprobe b43 what (if any) messages appear in the system log.
from xfce Live Mga8. I hope this is what you were looking for: dmesg after running those commands [ 22.622362] ssb: Found chip with id 0x4311, rev 0x02 and package 0x00 [ 22.677288] b43-pci-bridge 0000:01:00.0: Sonics Silicon Backplane found on PCI device 0000:01:00.0 [ 23.153734] input: AlpsPS/2 ALPS GlidePoint as /devices/platform/i8042/serio1/input/input8 [ 23.679109] iTCO_vendor_support: vendor-support=0 [ 23.761749] iTCO_wdt: Intel TCO WatchDog Timer Driver v1.11 [ 23.761825] iTCO_wdt: Found a ICH8M TCO device (Version=2, TCOBASE=0x0460) [ 23.778559] iTCO_wdt: initialized. heartbeat=30 sec (nowayout=1) [ 23.839171] gpio_ich gpio_ich.2.auto: GPIO from 446 to 495 [ 24.312138] 8139too 0000:02:01.0 enp2s1: renamed from eth0 [ 24.945828] cfg80211: Loading compiled-in X.509 certificates for regulatory database [ 24.964522] cfg80211: Loaded X.509 cert 'sforshee: 00b28ddf47aef9cea7' [ 25.267976] b43-phy0: Broadcom 4311 WLAN found (core revision 13) [ 25.285282] b43-phy0: Found PHY: Analog 4, Type 2 (G), Revision 9 [ 25.285304] b43-phy0: Found Radio: Manuf 0x17F, ID 0x2050, Revision 2, Version 0 [ 25.293568] Broadcom 43xx driver loaded [ Features: PNLS ] [ 25.294881] b43 ssb0:0: Direct firmware load for b43/ucode13.fw failed with error -2 [ 25.294916] b43 ssb0:0: Direct firmware load for b43/ucode13.fw failed with error -2 [ 25.295038] b43 ssb0:0: Direct firmware load for b43-open/ucode13.fw failed with error -2 [ 25.295091] b43 ssb0:0: Direct firmware load for b43-open/ucode13.fw failed with error -2 [ 25.295095] b43-phy0 ERROR: Firmware file "b43/ucode13.fw" not found [ 25.295099] b43-phy0 ERROR: Firmware file "b43-open/ucode13.fw" not found [ 25.295103] b43-phy0 ERROR: You must go to http://wireless.kernel.org/en/users/Drivers/b43#devicefirmware and download the correct firmware for this driver version. Please carefully read all instructions on this website.
Yes, that was what I was looking for. Unfortunately, reading up on this indicates you need a working internet connection to get that file - we can't redistribute it. I tried sticking an old Broadcom WiFi adapter into my laptop to see if I could investigate why the wl driver wouldn't work without a reboot, but it sprang into life as soon as I installed the dkms-broadcom-wl package. It is a newer Broadcom device than yours though.
ok, marking as wontfix as it is beyond Mageia control
Status: NEW => RESOLVEDResolution: (none) => WONTFIX