You are not logged in.
@kultex
if you have intel it's i915 : lsmod | grep i915
maybe rebuilding xf86-xorg-video-vesa without the patch you find would ignore loaded kernel modules bound to a graphic card.
Offline
@kultex
i tried to build xf86-xorg-video-vesa reversing the patch you found, X starts but the output translate upwards and you loose the top panel, probably i915 interferes and there was a reason why vesa's developers introduced the patch.
please, could you try my patched version of spacefm here
http://forum.slitaz.org/topic/tazbox-su-problems/page/2
?
it calls tazbox as default su app, you can check if it works opening a root window (menu file)
to install my version you should remove spacefm, rm -r /home/ $USER/.config/spacefm then logout.
this is required because there is a spacefm --desktop in the background so new package would not be loaded.
then login again and install my package, now it should works.
please, let me know what happens, thanks
Offline
@Aleksej sorry - only the first line was for you, all the rest relating the xorg bug
@ernia
your patch of spacefm works perfect - thanks a lot
I just found in Xorg.0.log, whats blocking the video card:
[ 675.662] (II) Loader magic: 0x81df690
[ 675.662] (II) Module ABI versions:
[ 675.662] X.Org ANSI C Emulation: 0.4
[ 675.662] X.Org Video Driver: 12.0
[ 675.662] X.Org XInput driver : 16.0
[ 675.662] X.Org Server Extension : 6.0
[ 675.664] (--) PCI:*(0:0:2:0) 8086:2592:1043:82d9 rev 4, Mem @ 0xf7f00000/524288, 0xd0000000/268435456, 0xf7ec0000/262144, I/O @ 0x0000ec00/8
[ 675.664] (--) PCI: (0:0:2:1) 8086:2792:1043:82d9 rev 4, Mem @ 0xf7f80000/524288
by the way, its quite stupid, that dmesg.log stopps, as soon as X starts....
Offline
but this should not be the problem, because this exists in 4.0 also
I do not find real differences until:
[ 136.253] (WW) xf86OpenConsole: setpgid failed: Operation not permitted
[ 136.259] vesa: Ignoring device with a bound kernel driver
these 2 Xorg.0.log are from the Thinkpad X31 one from 4.0, one from rolling - that one earlier was fromm an EEE-PC
Offline
I just remembered, that the old rolling from 04. 09. booted with vesa on my Portege R100 - and still it is doing so
so next I will try the rolling from now - but takes 30 min to boot
EDIT:
the same with actual rolling
There is a funny thing with the R100 - I have to type 2 x startx until it starts....
Offline
so thats the important part after dmesg.log stopps on the Portege R100 - I have to find a way to get tis out of the X31
input: Power Button as /devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input4
ACPI: Power Button [PWRB]
input: Lid Switch as /devices/LNXSYSTM:00/device:00/PNP0C0D:00/input/input5
ACPI: Lid Switch [LID]
input: Power Button as /devices/LNXSYSTM:00/LNXPWRBN:00/input/input6
ACPI: Power Button [PWRF]
Linux agpgart interface v0.103
agpgart-intel 0000:00:00.0: Intel 855PM Chipset
[Firmware Bug]: ACPI: No _BQC method, cannot determine initial brightness
acpi device:1a: registered as cooling_device2
input: Video Bus as /devices/LNXSYSTM:00/device:00/PNP0A03:00/device:19/LNXVIDEO:00/input/input7
ACPI: Video Device [VGA] (multi-head: yes rom: yes post: no)
agpgart-intel 0000:00:00.0: AGP aperture is 256M @ 0xc0000000
Offline
here we are - dmesg > test.txt did the trick
on the Portege R100 after agpgart-intel was nothing x related
and on the X31 there is aq lot of dri/drm stuff and this is blocking the video card
Offline
yes but it shows only, that drm/dri is the bastard and that it boots on the Portegee R100 with the trident card (there is nothing mesa/dri/drm stuff for trident in SliTaz)
dmesg on the thinkpad X31 with the ati card has nearly no difference between 4.0 (att. in this post) and rolling (att. in last post)
also the package list from 4.0 and rolling regarding mesa/dri/drm is the same
so for me the only difference that just remains this vesa commit
http://cgit.freedesktop.org/xorg/driver/xf86-video-vesa/commit/?id=b1f7f190f9d4f2ab63d3e9ade3e7e04bb4b1f89f
I think, we are having quite a big trouble with vesa
here I am finshed - I have no more idea and I also will have very little time until the beginning of february.
Thomas
Offline
Xorg.0.log.txt => xorg-xf86-video-intel, xorg driver module, is not installed, so Xorg tried to load vesa but vesa refuses to start because the intel kernel module is loaded - nothing wrong, as i say you in my previous post i patched back xorg-xf86-video-intel to start with intel kernel module loaded and it worked in a strange way.
patch:
[c]--- xorg-xf86-video-vesa-2.3.1/src/vesa.c
+++ ../stuff/src/vesa.c
@@ -454,14 +454,7 @@
pScrn = xf86ConfigPciEntity(NULL, 0, entity_num, NULL,
NULL, NULL, NULL, NULL, NULL);
if (pScrn != NULL) {
- VESAPtr pVesa;
-
- if (pci_device_has_kernel_driver(dev)) {
- ErrorF("vesa: Ignoring device with a bound kernel driver\n");
- return FALSE;
- }
-
- pVesa = VESAGetRec(pScrn);
+ VESAPtr pVesa = VESAGetRec(pScrn);
VESAInitScrn(pScrn);
pVesa->pciInfo = dev;
}[/c]
receipt
[c]# SliTaz package receipt.
PACKAGE="xorg-xf86-video-vesa"
VERSION="2.3.1"
CATEGORY="x-window"
SHORT_DESC="Xorg Vesa video driver."
MAINTAINER="pankso@slitaz.org"
SOURCE="xf86-video-vesa"
TARBALL="$SOURCE-$VERSION.tar.bz2"
WEB_SITE="http://www.x.org/"
WGET_URL="$XORG_MIRROR/driver/$TARBALL"
# We can use xorg-server-light with evdev and vesa driver.
#DEPENDS="xorg-server"
BUILD_DEPENDS="xorg-server-dev"
MODS='/usr/lib/X11/modules'
# Rules to configure and make the package.
compile_rules()
{
cd $src
echo $src
patch -Np1 -i $stuff/dontcareaboutboundedkerneldriver.diff
./configure \
--sysconfdir=/etc \
--localstatedir=/var \
--with-xorg-module-dir=$MODS \
$CONFIGURE_ARGS &&
make && make install
}
# Rules to gen a SliTaz package suitable for Tazpkg.
genpkg_rules()
{
mkdir -p ${fs}${MODS}/drivers
cp -a ${install}${MODS}/drivers/*.so ${fs}${MODS}/drivers
}[/c]
package for rolling (but i'm not responsable of your hardware)
http://www.filedropper.com/xorg-xf86-video-vesa-231
Xorg.0.log-4.0.txt => there is a vga=normal kernel option, what does it do? maybe avoid video kernel modules loading? vesa starts ok.
Xorg.0.log-rolling.txt => vesa refuses to start because the intel kernel module is loaded - nothing wrong,
Xorg.0.log-portege-roll-09.txt => vesa has not found a kernel driver bound to the grapgic card so starts well
i can't see nothin wrong, the problem is not in vesa but in how kernel modules are loaded - unloaded
Offline
Another CRAZY idea:
Switch to single (actually, dual) user system. 95% of all home systems, regardless of the OS, use no more than two users: admin/root + main user. Why not do this with SliTaz? By doing this we can simplify everything tremendously - no need to mess with multi-user stuff. Everything is done by one main user, except system daemons launched under root, no need to think about something complicated.
Offline
@devl547
Puppy(rus) linux already exists! (:
What is overhead of multy-user stuff? Size of packages (containing adduser, deluser, addgroup, etc.), or a number of customized non-standard scripts (plus /etc/skel)? Sorry, I'm not in theme.
I always need: regular user (1 pcs), admin user 'root' (1 pcs), and system need several users like 'www' (>1 pcs). Where we can drop multy-userness? Btw, I don't know why, but I don't like PuppyRus because of it root-orientation.
Offline
Another (not so) crazy idea.
Multiple distributions in one filesystem.
Try http://tank.slitaz.org/~bellard/demo-subroot.sh
and see http://doc.slitaz.org/en:guides:uncommoninst#subdirectory-install-in-a-posix-filesystem
Offline
@Aleksej
Well, we dont have to work under root.
Also, you cannot log in as a www/dbus/etc user.
I just suggest optimizing the system for a single user experience.
>Size of packages (containing adduser, deluser, addgroup, etc.), or a number of customized non-standard scripts (plus /etc/skel)?
Well, for example we can drop slim and use getty autologin (keeping console-based login on tty1/tty2/etc). All root tasks in Xorg can be done with su/sudo.
We can tweak user permissions to allow everything like storage mount, cd recording and etc and dont have to mess with groups any more.
The reason I am talking of this is simple - why not make a system, that is like "install-and-it-just-works", like MacOS, while keeping oldschool Linux simplicity and speed.
Offline
@devl547
as a special flavour I think it could be possible, but not as default, because it necessary for business use.
I for example run SliTAz in an cultural office with 5 computers for 10 users - and every user has access to his account from every computer
Offline
Just so you know, but made you a Mod again because I'm working on a better solution for file uploads using File Thingie.
Offline
>I for example run SliTAz in an cultural office with 5 computers for 10 users
Install slim and everything gonna be ok)
//That was just an idea, dont take it so seriously.
Offline
Found one more issue while cooking:
mount: mounting /home/slitaz/src on /dev/shm/aufsmnt2911root/home/slitaz/src failed: No such file or directory
...
umount: can't umount /dev/shm/aufsmnt2911root/home/slitaz/src: No such file or directory
That's on freshly installed rolling.
Offline
I have tested today rolling from 21. 1. and its totally broken???
sorry had no time to search for details...
Offline
ok tried again the rolling from 30.1.
the CD is not bootable anymore - it hangs on searching the soundcard
earlier cookings had Start SliTaz on the first boot screen, form where it was bootable - this line disappeard now and it hangs no matter what language I choose
and I cannot get the iso from the usb stick - does anybody know, how to start rolling from the stick?
it would be really cool, if we get the iso to boot - otherwhise nobody can test
Offline
hi there,
i tried the last rolling-iso's on my lifebook.
several times it hangs on searching the sound card. when i switch to german lang, it boots.
after boot i couldn't get access neither to my harddisk nor to my usb stick. on pcmanfm the hd and sd doesn't appear and i couldn't find the mountbox tool witch i have on the 4.0-iso.
for testing i need access to the internet, but the ipw2200 drivers i have on my hd or sd i can not copy to the system to setup the wireless access.
Offline
funny - I tried now really 10 times on my eee-PC it hangs in German and all other languages
Offline
Just downloaded the iso which is the same size than the one i downloaded jan 17th but seems to have a different md5sum.... is this a new rolling version uploaded recently ?
Offline
do you have a pc with a reset button?
does boot hangs on restart too?
sometimes i have the 5% problem in hd install too, but never when i restart from other OS, only when i boot from cold pc.
Offline
@ christophe yes its new - from 30. 01
@ ernia - this eee-pc does not have a restet button, but its the same if I boot first to another OS and go then to restart
Offline
thanks kultex, i have no time (and probably not enough skills) to fix it, but it does affect hd install to, not at every bot but sometime it happens.
putting some sleep 15 inside scripts to have time to read messages makes problem disappear, has the old root delay trick already been tried?
Offline
[ Generated in 0.017 seconds, 7 queries executed - Memory usage: 1.59 MiB (Peak: 1.77 MiB) ]