CxOrg

Just a note that as of updates to Sabayon 01-02-2013 to userspace the modules file is no longer required. It in fact causes shutdown and wake problems using the "uswsusp" methods apparently.

This is likely to be the last update here as I've jumped ship and gone for a more up to date KDE distro. After a bit of research I've settled for Sabayon 10 which has KDE 4.9.3 and 3.5.0 kernel. Out of the box it is optimised for AMD 64 hardware with fglrx as standard. It is based on Gentoo which is a much more "hair shirt" basic experience. Sabayon on the other hand provides all the creature comforts of a good desktop distribution. I now have: Rock solid suspend/resume Generally faster/stable desktop Latest versions of most applications Small fixes were required for: Added the "modules" file with only the SLEEP_MODULE="uswsusp" to secure good suspend/resume Bluetooth demon not starting at login (rc-update add bluetooth default) USB modeswitch tweaks for the Huawei E353 3G dongle (as for previous Mandriva fix) Sound distortion problems with Skype/Pidgin/Thunderbird - Mostly good now except Skype microphone speech generates background static. OpenVPN Network Manager demon not starting at login - OpenVPN now works again from Networkmanager though the networking interface info is not shown.(/usr/libexec/nm-openvpn-service then rc-update add /usr/libexec/nm-openvpn-service default) It is best to back up everything from your home folder and allow the installer to reformat the disk. I restored key configurations such as Thunderbird mail, Mozilla Firefox, Pidgin and Skype after install to get back to my working environment more quickly. The learning process with Mandriva has certainly helped in this move. I considered moving to Mandriva Rosa 2012 but the release has been too slow. It may well run as well as Sabayon 10 but for now I'll stick with this flavour of Linux. The only real adjustment is loosing the Draktools collection and having to work with a different package manager. The standard KDE control panels are mostly adequate. Package management seems less user controlled. There is obviosly much magic under the hood of Rigo the GUI package tool, but to see what is going on I find it often easier to work with "equo" from the command line. Care has to be exercised as no warning is given that removing one package may rip out essential parts of KDE!! When I fully get to grips with the "Gentoo way" where many packages are installed from source and compiled, I may sing its praises. This is slower than pre-compiled .rpm or .deb packages. I will have to examine how the .ebuild system works. The advantage is probably that the app is built on the target machine.

Adding a pm-utils hook for bluetooth to recover correctly. Having tried a few methods it seems best to work within pm-utils to do this. The hook provided is /usr/lib64/pm-utils/sleep.d/49bluetooth this is best removed as it is actually obsolete and seems to cause the wakeup to fail in conjunction with a new hook which simply closes down bluetooth nicely and starts (re-starts) it on resume: I've called it 03wake-bt #!/bin/bash . "${PM_FUNCTIONS}" case $1 in hibernate|suspend) hciconfig hci0 down ;; thaw|resume) hciconfig hci0 down && hciconfig hci0 up ;; *) exit $NA ;; esac Since it is not part of the distribution I've put it in /etc/pm/sleep.d/ This seems to bring back the BT by stopping and starting it.

I experienced problems with plugged in USB/SD media failing to suspend and it seems causing a failure to resume. By using the SUSPEND_MODULES for ehci_hd and xhci_hcd this becomes more robust. These modules control USB2 and USB3 respectively. However it may still be necessary to eject all USB drives to achieve suspend. In addition I'm experimenting with using "kernel" as the suspend module. Keeping s2ram as a wrapper to pass video shutdown commands. The S2RAM_OPTS are absent for the moment but I will be experimenting if I have some failures. /etc/pm/config.d/modules is now SLEEP_MODULE="kernel" SUSPEND_MODULES="ehci_hd xhci_hcd" This seems a working solution and provides repeated suspend/resumes.

3.0 + Kernel update: I could not previously update the kernel to 3.0 or better as the Broadcom wireless kernel module would not load (link?) at boot time. I'm now using the Linus kernel 3.2.18. This may be relevant to the suspend/resume solution above. To get it installed also involves applying a patch to get the Broadcom wireless to work again. Go here Wireless BCM4312 3.2 kernel to find out how to patch and compile. Make sure you have the Broadcom sources and compile environment installed then it seemed to work without problems after that.

A recent update of laptop-mode-tools seems to have made the difference allowing pm-utils to finally do its job properly. Check you have laptop-mode-tools 1.60 or better installed. I have removed the S2RAM line to allow it to decide how it is going to run. The specific module suspends are now handled correctly so we only need to tell it the SLEEP_MODULE to use. /etc/pm/config.d/modules is now: SLEEP_MODULE="uswsusp" The output of the successful pm-suspend.log file: Initial commandline parameters: Sat Nov 24 14:14:31 GMT 2012: Running hooks for suspend. Running hook /usr/lib64/pm-utils/sleep.d/00logging suspend suspend: Linux localhost.localdomain 3.2.18-1mdv #1 SMP Mon May 21 08:30:09 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux Module Size Used by nls_iso8859_1 4641 0 nls_cp437 6359 0 vfat 10615 0 fat 54040 1 vfat uas 9323 0 usb_storage 51541 0 fuse 73143 3 af_packet 27668 0 ipt_REJECT 2360 4 xt_comment 1040 18 xt_tcpudp 3732 11 ipt_LOG 8056 5 xt_multiport 2710 6 xt_addrtype 3137 4 xt_conntrack 3248 16 xt_mark 1323 1 rfcomm 66387 4 iptable_mangle 1814 1 iptable_nat 5518 0 nf_nat 22517 1 iptable_nat nf_conntrack_ipv4 15537 19 iptable_nat,nf_nat nf_conntrack 85009 4 xt_conntrack,iptable_nat,nf_nat,nf_conntrack_ipv4 nf_defrag_ipv4 1617 1 nf_conntrack_ipv4 bnep 15956 2 iptable_raw 1502 0 nfs 343808 0 lockd 81929 1 nfs fscache 36918 1 nfs auth_rpcgss 45540 1 nfs lib80211_crypt_tkip 8614 0 iptable_filter 1762 1 nfs_acl 2763 1 nfs sunrpc 237495 5 nfs,lockd,auth_rpcgss,nfs_acl wl 2569352 0 ip_tables 20029 4 iptable_mangle,iptable_nat,iptable_raw,iptable_filter x_tables 25078 13 ipt_REJECT,xt_comment,xt_tcpudp,ipt_LOG,xt_multiport,xt_addrtype,xt_conntrack,xt_mark,iptable_mangle,iptable_nat,iptable_raw,iptable_filter,ip_tables lib80211 5421 2 lib80211_crypt_tkip,wl binfmt_misc 8140 1 tun 16107 0 fglrx 3214160 388 snd_hda_codec_realtek 215562 1 uvcvideo 68252 0 videodev 97299 1 uvcvideo media 13809 2 uvcvideo,videodev v4l2_compat_ioctl32 9696 1 videodev snd_hda_codec_hdmi 26289 1 btusb 16400 2 usbhid 45398 0 hid 89679 1 usbhid snd_hda_intel 27893 2 eeepc_wmi 2682 0 asus_wmi 17052 1 eeepc_wmi sparse_keymap 4250 1 asus_wmi pci_hotplug 29674 1 asus_wmi snd_hda_codec 114341 3 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_intel snd_hwdep 6916 1 snd_hda_codec snd_seq_dummy 1758 0 joydev 10854 0 dm_mirror 14363 0 dm_region_hash 11054 1 dm_mirror dm_log 9956 2 dm_mirror,dm_region_hash bluetooth 252892 13 rfcomm,bnep,btusb dm_mod 82529 2 dm_mirror,dm_log snd_seq_oss 30953 0 snd_seq_midi_event 7243 1 snd_seq_oss snd_seq 56488 5 snd_seq_dummy,snd_seq_oss,snd_seq_midi_event snd_seq_device 6772 3 snd_seq_dummy,snd_seq_oss,snd_seq snd_pcm_oss 40974 0 sg 30776 0 evdev 10474 19 snd_pcm 90664 4 snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_pcm_oss xhci_hcd 134457 0 snd_timer 23439 2 snd_seq,snd_pcm snd_mixer_oss 16227 1 snd_pcm_oss rfkill 18735 4 asus_wmi,bluetooth snd 73218 16 snd_hda_codec_realtek,snd_hda_codec_hdmi,snd_hda_intel,snd_hda_codec,snd_hwdep,snd_seq_oss,snd_seq,snd_seq_device,snd_pcm_oss,snd_pcm,snd_timer,snd_mixer_oss i2c_piix4 11308 0 i2c_core 31014 2 videodev,i2c_piix4 battery 11825 0 ac 3576 0 video 13188 0 processor 31409 2 sp5100_tco 5583 0 wmi 10552 1 asus_wmi soundcore 7763 1 snd k10temp 3502 0 snd_page_alloc 8628 2 snd_hda_intel,snd_pcm button 5500 1 fglrx atl1c 37505 0 ipv6 337898 30 autofs4 33709 5 ext4 466971 3 jbd2 81795 1 ext4 crc16 1691 2 bluetooth,ext4 sd_mod 37910 5 crc_t10dif 1547 1 sd_mod ahci 21132 5 libahci 24085 1 ahci libata 207632 2 ahci,libahci ehci_hcd 57108 0 ohci_hcd 31909 0 ssb 58157 1 ohci_hcd usbcore 203527 9 uas,usb_storage,uvcvideo,btusb,usbhid,xhci_hcd,ehci_hcd,ohci_hcd scsi_mod 217729 5 uas,usb_storage,sg,sd_mod,libata mmc_core 103427 1 ssb pcmcia 56013 1 ssb usb_common 1181 1 usbcore pcmcia_core 19372 1 pcmcia total used free shared buffers cached Mem: 3646664 3144484 502180 0 166052 1501176 -/+ buffers/cache: 1477256 2169408 Swap: 4088504 0 4088504 /usr/lib64/pm-utils/sleep.d/00logging suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/00powersave suspend suspend: /usr/lib64/pm-utils/sleep.d/00powersave suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/01bootloader suspend suspend: /usr/lib64/pm-utils/sleep.d/01bootloader suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/06mysqld suspend suspend: [ OK ] Stopping mysqld (via systemctl): [ OK ] /usr/lib64/pm-utils/sleep.d/06mysqld suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/10network suspend suspend: Cannot find nfs service Configured devices: lo eth0 eth2 wlan0 Currently active devices: lo Stopping network (via systemctl): [ OK ] /usr/lib64/pm-utils/sleep.d/10network suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/15sound suspend suspend: /usr/lib64/pm-utils/sleep.d/15sound suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/30pcmcia suspend suspend: /usr/lib64/pm-utils/sleep.d/30pcmcia suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/40xlock suspend suspend: /usr/lib64/pm-utils/sleep.d/40xlock suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/49bluetooth suspend suspend: /usr/lib64/pm-utils/sleep.d/49bluetooth suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/55NetworkManager suspend suspend: Having NetworkManager put all interaces to sleep...Done. /usr/lib64/pm-utils/sleep.d/55NetworkManager suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/75modules suspend suspend: Unloading kernel module usbcore...Done. /usr/lib64/pm-utils/sleep.d/75modules suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/90clock suspend suspend: /usr/lib64/pm-utils/sleep.d/90clock suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/91laptop-mode suspend suspend: /usr/lib64/pm-utils/sleep.d/91laptop-mode suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/92disk suspend suspend: /usr/lib64/pm-utils/sleep.d/92disk suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/94cpufreq suspend suspend: /usr/lib64/pm-utils/sleep.d/94cpufreq suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/95led suspend suspend: /usr/lib64/pm-utils/sleep.d/95led suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/98video-quirk-db-handler suspend suspend: ATI Catalyst driver detected, not using quirks. /usr/lib64/pm-utils/sleep.d/98video-quirk-db-handler suspend suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/99video suspend suspend: /usr/lib64/pm-utils/sleep.d/99video suspend suspend: success. Sat Nov 24 14:14:35 GMT 2012: performing suspend Switching from vt8 to vt1 fbcon fb0 state 1 fbcon fb0 state 0 switching back to vt8 Sat Nov 24 14:14:56 GMT 2012: Awake. Sat Nov 24 14:14:56 GMT 2012: Running hooks for resume Running hook /usr/lib64/pm-utils/sleep.d/99video resume suspend: /usr/lib64/pm-utils/sleep.d/99video resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/98video-quirk-db-handler resume suspend: /usr/lib64/pm-utils/sleep.d/98video-quirk-db-handler resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/95led resume suspend: /usr/lib64/pm-utils/sleep.d/95led resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/94cpufreq resume suspend: /usr/lib64/pm-utils/sleep.d/94cpufreq resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/92disk resume suspend: /usr/lib64/pm-utils/sleep.d/92disk resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/91laptop-mode resume suspend: Reloading laptop-mode configuration (via systemctl): [ OK ] /usr/lib64/pm-utils/sleep.d/91laptop-mode resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/90clock resume suspend: /usr/lib64/pm-utils/sleep.d/90clock resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/75modules resume suspend: Reloaded unloaded modules. /usr/lib64/pm-utils/sleep.d/75modules resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/55NetworkManager resume suspend: Having NetworkManager wake interfaces back up...Done. /usr/lib64/pm-utils/sleep.d/55NetworkManager resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/49bluetooth resume suspend: /usr/lib64/pm-utils/sleep.d/49bluetooth resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/40xlock resume suspend: /usr/lib64/pm-utils/sleep.d/40xlock resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/30pcmcia resume suspend: /usr/lib64/pm-utils/sleep.d/30pcmcia resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/15sound resume suspend: /usr/lib64/pm-utils/sleep.d/15sound resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/10network resume suspend: Starting network (via systemctl): [ OK ] Starting network-up (via systemctl): [ OK ] /usr/lib64/pm-utils/sleep.d/10network resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/06mysqld resume suspend: Starting mysqld (via systemctl): [ OK ] /usr/lib64/pm-utils/sleep.d/06mysqld resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/01bootloader resume suspend: /usr/lib64/pm-utils/sleep.d/01bootloader resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/00powersave resume suspend: /usr/lib64/pm-utils/sleep.d/00powersave resume suspend: success. Running hook /usr/lib64/pm-utils/sleep.d/00logging resume suspend: /usr/lib64/pm-utils/sleep.d/00logging resume suspend: success. Sat Nov 24 14:15:08 GMT 2012: Finished. It is looking very good now with many suspend/resumes working smoothly. I use the "lid close" more often though I should road test Fn+Zz as well. Finally suspend/resume on the Asus 1215b is now working really reliably. Lets hope that is the end of this issue!! Post here if your 1215b EeePC or Laptop with AMD E450 is helped by this last configuration?

I have 3G data working fine using a ZTE MF112. My mobile operator 3 UK decided I should have an upgrade and so sent me a Huawei E353 which they claim connects faster and supports the newer fastest speeds. The Huawei E353 did not work straight out of the box and is a different kind of device in that it operates more like a router. It seems to provide a local network and NAT to the operator's network. This is a better security model than the simple modem style direct WAN connection. Install usb_modeswitch from the above repos but refer to the developer's site for more information at www.draisberghof.de. should be installed to get most USB 3G devices to work. It gets round the installer (CD) drive which most 3G dongles have for Windows use. A config file will have to be added for this device so create a file /usr/share/usb_modeswitch/12d1:1f01 with the contents: ####################################################### # Huawei E353 (3.se) # # Contributor: Ulf Eklund DefaultVendor= 0x12d1 DefaultProduct=0x1f01 TargetVendor= 0x12d1 TargetProduct= 0x14db MessageContent="55534243123456780000000000000a11062000000000000100000000000000" # Driver is cdc_ether NoDriverLoading=1 The contents of this come from the usb_modeswitch site and the reference file device_reference.txt. Make sure the permissions match the other files then with the dongle inserted use: usb_modeswitch --config-file /usr/share/usb_modeswitch/12d1:1f01 The connection should now appear in Networkmanager. If not just create one under "wired" connections with DHCP and on eth2. You can name this eth2 network connection so it is easy to identify. You may well now be connected otherwise click on the connection to connect. Now to automate this we need to add an entry in /lib/udev/rules.d/40-usb_modeswitch.rules Add at the end of the Huawei devices another entry: # Huawei E353 ATTRS{idVendor}=="12d1", ATTRS{idProduct}=="1f01", RUN+="usb_modeswitch '%b/%k'" Now usb_modeswitch will set up the device when it is plugged in. And it does connect quickly and seems to get good data speeds. Great!!

As a result of some messing with the install of some plasma desktop items I damaged KDE which resulted in no borders or controls to application windows. I hoped my backups using Clonezilla would get me back to my last backup state. Unfortunately the result the partitions were recovered but it failed to recreate the GRUB boot configuration. I could have persisted with this but decided to simply re-install from my USB thumb-drive ISO (64 bit 2011). By having a partition configuration of / /usr/share and /home I don't have to format the /home partition and so keep most of my desktop configuration. Make sure you use the same login name and all should be well. I now have a better install with wine32 working properly as well as kdropbox and Skype. One discovery was to use the repos as shown in MIB from here and choose 64 bit if you are using it. The given enabled repos will let you install most things. Because I had a problem with wine's wine32 install I installed this first before any updates. Updating may take some time. You may prefer to disable the MIB repos initially until all the official versions are installed. I installed everything from these repos except a new kernel 3.1.10 from MIB. This kernel proved quite reliable for suspend/resume especially with my latest config for suspend/resume as above.

Suspend/Resume Latest: (The best so far as of 23/10/2012) Make sure s2ram (suspend-s2ram) is installed! I had thought that pm-utils would be taking care of suspend/resume it seems s2ram does the job nicely. /etc/pm/config.d/modules is now: SLEEP_MODULE="uswsusp" S2RAM_OPTS="-f -a 1 -s" SUSPEND_MODULES="usbcore fglrx ext4" SUSPEND_MODULES: usbcore suspends the whole USB subsystem which helps with problems if USB devices are plugged in including the internal Bluetooth, fglrx unloads fglrx (video) and ext4 which must deal with disk access. The order of module unloading/loading seems to be important here though the reasoning is less clear! Using lsmod displays all the module names and their dependencies, which at least explains the relationships. The suspend is now much more reliable. Including ext4 seems top have been the key to repeatable wakeup. No failures yet with this configuration using lid-close/Fn+Zz/Menu controls and space bar to wake. That should do for now. I hope these guides help?

Skype: If you want Skype you can now install it. Here you must temporarily enable the 32bit main Mandriva repository. Do not leave it enabled! as there is a risk of trashing your installation in subsequent updates. This should go OK. I use skype4pidgin (64 bit). If you start Skype first then Pidgin can use it for messaging. The version installed via the MIB repositories or direct from Skype seem to be largely 32bit so the skype4pidgin link causes a problem and very slow plasma/kde response. Looking round for a more 64bit version I found a Skype compilation for Fedora, this still requires some 32bit libraries (8) but works much better. Remember to disable the 32bit repos after installation!!

Some time has passed since my last update: In that time I've re-installed Mandriva 2011 but this time the 64bit version. This is mostly straightforward loading from a USB drive. The effort has been worth it as I can now use all the 4Gb of RAM and the general operation is smoother and probably quicker since full use is made of the processor and graphics. On install select custom partition configuration to select the root /, /home, and if you have it a /usr or /usr/share partition. Avoid re-formatting your /home partition so most of your desktop, data and application settings will be preserved. I use /usr/share as a partition since it holds most of the applications you may install. You will have to re-install all the applications except Windows apps which run under Wine. We are now in a 64 bit environment! It is important to add the standard repositories and PLF repos to start with. There will be a lot of updates. persist with this till everything is installed. Do the libraries last and proceed carefully to install the correct release version. If there are difficulties you may have to enable the backports repos to resolve some installs. Only enable these to install a particular package then disable. It is much safer! Install the given version of FGLRX (DKMS module) for ATI video. This will get updated later but at least it is installed. Configure the desktop effects to use OpenGL if you want the full cube desktop transitions. Add the MIB repositories (Ignore the Russian but choose MIB). Again update until done. This will include the latest FGLRX (video). NSS security may give you problems. Here you may be obliged to re-install nss security (3 packages) and a number of applications which depend on them. Note down which as they should also be re-installed. I think it was Firefox, Pidgin, Thunderbird and knetwork manager (networkmanager) and it's plasma panel. Getting the latest NSS security will allow installation of the latest Firefox and Thunderbird. Next we can upgrade the kernel I've used the netbook kernel in MIB's online repos. This is the latest which will work with the current wireless and sound driver support. It works very nicely with suspend/resume also working with the config files below. (well most of the time!!)

Waiting for a working version of the newest kernel 3.2.13 http://mib.pianetalinux.org/forum/viewtopic.php?f=38&t=3226'>http://mib.pianetalinux.org/forum/viewtopic.php?f=38&t=3226 The current blog build status at http://mib.pianetalinux.org/forum/viewtopic.php?f=38&t=3226 tells us to wait till they sort out some module problems. Trying one of these builds fails to add fglrx (video), the sound driver module and networking so it is not usable for now.

Some other issues: Bluetooth: For some reason in Mandriva 2011 it was decided to replace Bluez with Bluedevil as the Bluetooth stack and application. I had been used to bluez working very well with Blueman as the user interface. Bluedevil certainly did not work well with the 1215B bluetooth when I first tried it on installing Mandriva 2011 as it could not scan and detect any devices. Removing Bluedevil and then installing the bluez software stack followed by blueman solved this and Blueman provides most features anyone might need. I have been using a bluetooth mouse which is convenient but the wakeup from sleep can be slow to start and requires switching off and on to get it going. It may well be the mouse which is at fault as it is often the same with XP. Using a 2.4 Ghz band mouse and USB dongle works much better as it literally appears to the PC as a USB mouse. This means it does not need to maintain the comms stack and radio connection in software as with Bluetooth. As an experiment I now tried removing Bluez again and replacing with the Bluedevil BT stack. This seems to work well so some bugs must have been fixed. I also works happily under Blueman which still provides the best user interface and control features. Removing the last library of bluez I encountered problems with many dependencies with plasma elements. Using: rpm -e libbluez3 --nodeps Allows removal without upheaval in the Plasma desktop. Then add the essential part of Bluedevil, Suspend and resume seems good as well with the mouse recovering more reliably. Bluedevil uses some of Bluez so it is not a completely new development. Touchpad: The touchpad showed some erratic behaviour particularly when on battery which made it unusable making an external mouse essential. This is in fact a hardware defect caused by a ground connection to the touchpad backplane. Since we don't have any driver update which XP/7 users report fixes this the backplane should be isolated. There are a number of guides online to do this. A good guide to start with is http://www.insidemylaptop.com/disassemble-asus-eee-pc-1215b-laptop/ then identify the conductive foam pad or wire link and disconnect/isolate it with something insulating. Then all normal touchpad features and additional gesture features work OK.

It looks like all this can be done in the pm-utils file which is now simplified and the only file in the config.d folder. I removed 2 others which contained suspend_modules commands for "ndiswrapper" and "rt73usb' neither of which feature in my modules list (lsmod). File "modules" now with contents: SLEEP_MODULE="uswsusp" S2RAM_OPTS="-f -a 2" SUSPEND_MODULES="rfkill xhci_hcd button bluetooth" xhci_hcd controls USB 3.0 and button controls fglrx the video driver. These modules when suspended should suspend the state of USB 3.0 and fglrx video then restart them on resume. I've added rfkill and bluetooth which seems to improve the closedown of wireless "bluetooth" seems to improve the recovery of bluetooth and reconnection of a BT mouse. Testing will prove if it is 100% reliable. I'm sticking to Fn +Zz to suspend and the power button to resume for now to be consistent, though lid close seems to work OK to suspend and the space bar wakes most of the time. One thing I've noticed is the centre of the base of the laptop remains warm when suspended and kept in a laptop sleeve. This must show some battery drain, probably to RAM which must not be in its lowest power mode so suspend would not last too long. I was used to suspend working for days with my old Samsung NC20. This may well be something to investigate. It may well be some other module which is at fault? Please contribute your experiences to make this work 100%.

Looks like there is more to this story! Most of my testing so far has been at my desk where I tend to use an Ethernet connection. Sleeping with wireless connected causes more failures which of course is the likely scenario for a laptop. Disabling networking in the Networkmanager applet before suspending is a work around for this. I'll check the shutdown scenario scripts for lid close Fn+Zz key etc to see if we can add a network shutdown line and then add a re-enable on wake up.

Current contents of my /etc/pm/config.d/modules SLEEP_MODULE="uswsusp" S2RAM_OPTS="-f -a 2" SUSPEND_MODULES="button uhci_hcd" Now sleeping with Fn + Zz keys and lid close seems to work. Waking with short press of power button seems the most reliable. Again testing in real use will prove if there is more work to do or not.

Most things work nicely on this little laptop with the AMD brazos e450 processor/GPU. The thing which has always proved troublesome is suspend to ram. Suspend works but it would never wake up requiring the battery to be removed to get it going again. Suspend to disk always worked fine. Hoping that updates from AMD/ATI for the video driver or new kernels would eventually solve it. I now have a MIB sourced kernel http://mib.pianetalinux.org/MIB/2011.0/32/projects/kernels/3.1.6/ and the latest ATI driver http://www2.ati.com/drivers/linux/amd-driver-installer-12-2-x86.x86_64.run which work really well making for a pretty reasonable video and desktop performance with all the KDE plasma effects and animations. However suspend/resume still did not want to play? I tried re-compiling the kernel which for a brief moment seemed to allow suspend/resume having read this solved someone's resume problem from one source. Finding a comprehensive guide for Suse linux which explains all you need to know about pm-utils and s2ram - http://en.opensuse.org/SDB:Suspend_to_RAM and here http://old-en.opensuse.org/Pm-utils Testing to find the best parameters to use with the Asus 1215B and ATI video driver I settled for s2ram -f -a 2 To make it work with the keyboard and power saving system I added this file /etc/pm/config.d/modules containing a single config line: S2RAM_OPTS="-f -a 2" set the permissions to 755 So far this works OK. I'll have to test against some other kernel(s) and use suspend to ram daily to see if it breaks?

It is a while since I added anything new here this is due to my upgrade to an EeePC 1215b with the wonderful AMD Fusion E-350 APU. A very efficient dual core and graphic processing unit. I'll post a new thread on this soon with the experience of upgrading to Mandriva 2011. This issue affects both 2010.2 and 2011 when using kvpnc to manage openvpn connections. Openvpn is now 2.2.1 and introduces script security which is not handled by kvpnc. It is a simple command line parameter as in $ openvpn --script-security 3 myopenvpn.config. There is no way to add additional command line parameters to kvpnc so we get error messages in the log display like: WARNING: External program may not be called unless '--script-security 2' or higher is enabled. Use '--script-security 3 system' for backward compatibility with 2.1_rc8 and earlier The solution until this is fixed in kvpnc is to create a shell script which calls openvpn and adds the parameter. create and put this script openvpn1 in /usr/sbin and set the permissions ( sudo chmod 755 openvpn1 ) #!/bin/bash openvpn --script-security 3 $@ exit 1 Then change in the daemons section of kvpnc configuration the file called for openvpn to /usr/sbin/openvpn1 This takes care of starting any connection but breaks starting any new connection as openvpn remains loaded as a process and we cannot start a new one. I am sure there is a way to do this through the script but here is the quick and dirty solution. In the "after disconnect" section of your kvpnc profile add "killall openvpn" (no quotes). This will remove the remaining openvpn process and a new connection starts from scratch as if it were the 1st connection. A bit brutal but it works. Lets hope for a fix from the kvpnc team!

Having done the above work on the DHCP and wlan the general performance is again a bit better. Improved network performance helps most applications get on with what they should be doing without waiting. Thunderbird and Firefox work much better. I also removed 2 of the 4 virtual desktops which may reduce memory use. 2 desktops is normally enough for me!

Wireless connections have always been a bit slow to negotiate and a bit embarrassing compared to Windows and OSX which are generally quick and easy. I tried using NDIS wrapper to the Atheros 5001 XP driver thinking the Linux driver was the cause, but that did not work well at all. I noticed the problem occurs on DHCP negotiation so I tried another DHCP client and installed dhcpcd . Configuring a known wireless connection to use this DHCP daemon (in advanced right at the bottom!) and the connection was then swift as we know it should be. Fantastic!! I'll have to find out how to make this the default DHCP client, possibly by removing the default dhclient. It turns out that once dhcpcd is selected for the wireless interface it remains the choice so it is not necessary to remove the ISC DHCP client. In fact trying this it also removes network-monitor so I have kept it installed. This may well be a solution for other Atheros wifi module users with other Linux distributions where DHCP negotiation seems to be the problem. PS: After road testing for some days of using this both at home and on my work I have tried most combinations and even dhcpxd as the client daemon. I'm finding dhcpcd is the most reliable client ahead of dhcpxd and dhclient for wlan0 connections. Wired ethernet seems to work best with the default dhclient. It seems they can co-exist though changing between clients for a specific connection seems to be problematic requiring at least several connection attempts if not a re-boot. Anyway wireless connections seem to be more consistent with dhcpcd.

An update kernel from the MIB repositories to 2.26.36.4 seems to give a helpful boost to application speed. I recommend adding the MIB repositories. These can be installed at http://http://urpmi.mandriva.ru/?language=en No change to the BT module initialisation but the suspend/resume is perhaps a bit quicker. The improvement may well be due to the improved threading and scheduling I read about being implemented in the standard kernel. Having this improvement I would also recommend clearing the caches in Firefox as my Firefox startup had really slowed. After clearing the startup returned to less than 10 seconds and more comparable to Chrome's swift startup. I prefer Firefox still due to the many useful extensions I use. I am sure that the startup would be quicker without them.

Yes in fact you are right. It makes no difference removing gtk-qt-engine. The system config GTK control panel remains together with the selection of qtcurve as the selected theme. Using gtk-qt-engine's "use my KDE style in GTK applications" results in some unexpected half mapped styles and sometimes freezes of KDE so I do not actually use it.

On discovery of the excellent kde4-style-qtcurve to style application windows and features, I became more interested in getting all applications to use this same styling. I had added gtk-qt-engine to get GTK+ applications to take up the main styling but this is problematic and unnecessary when using qtcurve. This installation does add the configuration panel "GTK Styles and fonts". Removing gtk-qt-engine again leaves this control panel in place so that qtcurve can be selected as the GTK style. The principal application I use is Pidgin which is a GTK+ app so by setting the GTK styles and fonts control panel in system settings qtcurve is available to GTK apps. If you use Skype it has its own selector for themes in the configuration where qtcurve can be selected, The last thing was to persuade the KDE/Mandriva "Configure your computer" and "Network manager" to play also with the same theme. Some of the default themes when selected seem to be used. Nothing seemed to work for qtcurve until I realised that these applications are looking in the /root account for the style information. The modifications made to qtcurve are stored in /home/youraccount/.config/qtcurve/stylerc . These don't seem to be applied to the root account /root/.config/qtcurve/stylerc . So all that needs to be done is to copy this file from the home account to the root account. I use mucommander (http://www.mucommander.com) started as root though Dolphin started as root or copy in a terminal session (su or sudo) will of course work. As a starting point you could use my stylerc: [settings] borderMenuitems=true comboBtn=selected crColor=selected customAlphas=0 customMenuNormTextColor=#202020 customMenuSelTextColor=#4A50BF doubleGtkComboArrow=false dwtAppearance=dullglass dwtSettings=23 flatSbarButtons=false grooveAppearance=flat handles=dots hideShortcutUnderline=true inactiveTitlebarAppearance=gradient lighterPopupMenuBgnd=42 lvAppearance=flat lvLines=true menuBgndAppearance=soft menuStripe=#F4F4F4 menubarAppearance=gradient menuitemAppearance=dullglass progressAppearance=agua progressGrooveAppearance=flat round=full roundMbTopOnly=false sbarBgndAppearance=soft selectionAppearance=soft shadeSliders=origselected sliderAppearance=agua sliderStyle=circular sliderThumbs=none sliderWidth=13 sortedLv=selected square=256 stdSidebarButtons=true thin=7 thinSbarGroove=false titlebarAppearance=gradient titlebarButtonAppearance=harsh toolbarAppearance=soft tooltipAppearance=dullglass useHighlightForMenu=true vArrows=false version=1.8.4 windowDrag=1 Now most things look consistent in most applications. Wine/Windows apps need to depend on Wine's own basic windows sourced themes. I do not know if this is a bug in the qtcurve control panel which does not also apply the style changes to the root account. It would prompt for a password to do this as the change of login theme does. The KDE/Mandriva panel has one annoying issue that the left hand selector menu uses a white roll over in this menu and white text so the text disappears on mouse over. If anyone knows how to style this please let me know?

In my search for solutions I discovered the latest KDE incremental release KDE 4.5.5 . Not being able to resist an upgrade I configured the repositories from the MIB site. These commands will configure the repos for you. (copied from the MIB site) urpmi.addmedia --update MIB-KDE455_release http://mib.pianetalinux.org/MIB/2010.2/others/kde/4.5.5/32/release urpmi.addmedia --update MIB-KDE455_others http://mib.pianetalinux.org/MIB/2010.2/others/kde/4.5.5/32/others urpmi.addmedia --update MIB-KDE455_updates http://mib.pianetalinux.org/MIB/2010.2/others/kde/4.5.5/32/updates urpmi.addmedia --update MIB-KDE455_extras http://mib.pianetalinux.org/MIB/2010.2/others/kde/45X-extras/32/ urpmi --auto-select --auto-update urpmi --auto-select --auto urpmi --auto-select --auto urpmi --auto-select --auto # bye, bye from mikala, NicCo and all MIB Team These commands remove all the currently configured repos so I would suggest re-adding these using the online site: http://urpmi.mandriva.ru/ . Add the Official Medias and the PLF Medias. The version should be 1010.2 and i586 on the selectors. This upgrade can take some time and make sure you have some space on your partition where /usr is located. I removed some development applications which I do not use and deleted the application folders which should be in /usr/share. It seemed to need up to 4Gb during download and installation which reduced as the KDE code and applications replaced the old ones. I use the "update system" Draktools panel to perform the update as we get the nice progress bars. KdiskFree gives a nice view on disk space. Install this if you don't have it. Verify that all updates are completed in the "Install and Remove Applications" panel. Again I had to remove some old applications which were blocking the install of some new packages. It is wise to verify that the boot system will default to the newly installed upgrade though it seems to use the same 2.6.36 kernel as before. Rebooting offers no great surprises except a much more fluid performance for handling applications and windows. Bluetooth works straight away on boot using a plug in BT dongle however the internal BT device still requires the previous fix to get it going but once launched it does not seem to fail and need re-initialising even after suspend to ram. Suspend to ram works nicely and seems to be a bit quicker to go to sleep. The conclusion is that this KDE 4.5.5 version is much improved with probably many fixes which contribute to the stability and speed. The internal bluetooth device (Broadcom BCM92045NHD) has been flagged as causing similar problems on other laptops - Acer etc so since it works (after a delay) with the /etc/rc.d/rc.local fix. I can live with that.

Problems with bluetooth not starting at boot time with Mandriva 2010.1 or .2 An update to this entry: I found that the fix above was unreliable and often becoming "most of the time" BT would not start. I have done some more snooping about to find a better fix. I should have remembered this from my days struggling with Xandros on the EeePC. Simply use /etc/rc.d/rc.local which runs after all other startup scripts have run. It is in fact bluetoothd which does not run on the initialisation of the internal Broadcom BT device. This should be run according to the udev control processes. See /etc/udev/rules.d/97-bluetooth.rules # Run helper every time a Bluetooth device appears # On remove actions, bluetoothd should go away by itself ACTION=="add", SUBSYSTEM=="bluetooth", RUN+="/usr/local/sbin/bluetoothd --udev" ACTION=="change", SUBSYSTEM=="bluetooth", RUN+="/usr/local/sbin/bluetoothd --udev" So using the udev command to run bluetoothd but in rc.local makes sure it does start. So updating it so: /usr/local/sbin/bluetoothd --udev touch /var/lock/subsys/local Probably there will be a fix in the real udev method in these rules. I notice that my aging EeePC900 has no such problem with the BT mouse BT works even in the KDE boot splash screen. If a fix comes through it will show up with the mouse working in this splash screen.