Mandrake 9.1 install locks on Sager 5670

freezing while it's uncompressing the kernel is bad, and errors listing stack traces like that are bad, too. at that point, it hasn't even gotten to loading the firewire driver yet, so that's not it. did you download and burn the iso? if so, it might have been corrupted during download; try redownloading and re-burning it, maybe even with CRC checking or MD5 hashing.

... and if you bought it... well... then there's something quite wrong. i'll try and see what i can find about it.

you want help getting the touchpad to work with 2.6.0-test1? ok:

first of all, the touchpad works a lot differently now - i don't know if you saw it, but the touchpad was detected and set up during the boot into 2.6. we still need the x driver, of course, but now it works through the event interface, as opposed to the ps2 interface. so here is what you need:

first of all, you need event interface support, under the 'input device support' of the kernel config. when you reboot, and go into /dev/input (if you're using devfs, which almost everyone uses, and you should be using it), you'll see a few event interfaces - event0, event1, etc. one of these is the touchpad - cat each one, and use the touchpad, and see which of them responds with weird symbols.

now, we need the synaptics driver. go to:

http://w1.894.telia.com/~u89404340/touchpad/

and get the driver there, synaptics-0.11.3p6. it's an updated version of that synaptics driver at tuxmobil.org. it's got event interface support added, which is what we need. plus, we have the bonus of making that scroll button in the middle work...

so set the mouse up with that driver in the xfree config. i'll assume you know how to do that; if not, i can help you out. but use that driver, with the correct event interface device as the device (mine is /dev/input/event0), and the protocol is "event", as opposed to ps/2.

that should do it. all the other edge settings and whatnot are the same as the other synaptics driver (the ones in the readme), and you can do a ZAxisMapping "4 5" to get that center scroll up/down button working, if you have one.

let me know if you have trouble.

i must have glanced over it, but i forgot to add:

software suspend (swsusp), or suspend to disk, is a kernel patch that is completely independent of any power management like apm or acpi. so you really don't need acpi to suspend. but acpi adds those other nice things like throttling, power switching between battery and a/c, hyperthreading if you have it, etc, so you probably want it anyway.

i really can't tell you what's up with your 2.4.21... it might be something in your configuration, because i've never had any problems with 2.4.21. it's a lot harder to troubleshoot if it's hanging, as opposed to giving error messages.

with the event interfaces:

the input core and event interface support, are they compiled directly into the kernel or as modules? also, do you have serial i/o support enabled?

but if it's detected and on isa0060/serio4, then that's a good thing.

on the 8139too driver: i've found that it's always better to compile the network card directly into the kernel. the 8139too module depends on the mii module, which is a common library for accessing some pci network cards. however, when loading modules, there are many opportunities for things to go wrong; your error, for example. there was another thread about the mii module not loading, and so the 8139too wouldn't load. but if it's compiled directly in, there is really no chance for dependent modules to get lost or something else - the module and its dependencies are all in there, and are all already loaded when the kernel is booted. so give that a shot.

i'll see what else i can come up with.

the protocol should be "event". in the lines from your x log, it says the auto-dev protocol is changing the device it uses - from /dev/input/event1 to /dev/input/js0 event1. and that device doesn't exist. i'm almost certain that changing 'auto-dev' to 'event' will fix it.

for the sound card driver, the /sbin/modprobe line means it probably has something to do with the module loading system, presumably because, like you said, you don't have module-init-tools (the 2.5/2.6 module tools) installed. is the alsa driver and its features compiled directly into the kernel?

you could build a giant kernel - i compile a lot of things in to avoid the hassle of loading a module. however, that is going to cause problems - direct rendering, and consequently 3d acceleration through drm, will _not_ work at all if the radeon drm module is not a driver, for example. i strongly recommend getting them working. what problems did you have building the tools?

hi Sez!

i don't know if i can be of any help but it looks like you are linking to a library that is not there, which in this case is the c library. the -l means a link to something and after the -l there is a c, which is the c library.

do you have libc-dev installed? if not you will be to install it (or it may need a version of libc like libc6-dev).

regards

i can't really help you out if something is missing from slackware since i've never used it, sorry. maybe you can find a module-init-tools package for slackware?

as to two mouses:

is your ServerLayout section correct? i.e. do you have two separate mouse sections with separate identifiers, with one InputDevice set as the CorePointer, and the other one set with SendCoreEvents? for example, mine looks like:

InputDevice "touchpad" "CorePointer"
InputDevice "usbmouse" "SendCoreEvents"

i set touchpad as the corepointer since it's (naturally) always connected, while my usb mouse could be connected/disconnected at times.

hmm. well actually, i know very little about the packages in mandrake, too; i had bad results back in the day with the older versions and never really went back...

i suppose i could compile a package for you, if you wanted. i would need to know your computer's specs for the correct CFLAGS, though.

i use the intel8x0 alsa driver without problems. in the kernel config, i chose alsa, sequencer support, oss api emulation, oss mixer api, oss pcm (digital audio) api, oss sequencer api, rtc timer support, and intel i8x0 under pci devices - all directly in the kernel, no modules. and nothing from the open sound system (oss) menu because it's deprecated. you might want to try that.

also, kill any sound daemons like esd or arts; those tend to interfere with alsa.

i can tell you how to use the external mouse or dri if you need help.

the intel8x0 module is in the kernel, under

Sound ->
Advanced Linux Sound Architecture ->
PCI Devices ->
<*> Intel i8x0/MX440, SiS 7012; Ali 5455; NForce Audio; AMD768/8111

it's the fourth option from the bottom in the pci device list. i guarantee that it's in both 2.6-test1 and 2.6-test1-mm1 because i've been using straight kernel support for both with no problems.


for DRI:

under character devices, you need

<*> /dev/agpgart (AGP Support)

compiled _directly_ into the kernel. it's important that it's a part of the kernel or else direct rendering won't work. i don't know what specific chipset we have, so i just compiled them all in. no big deal.

below that, you want
[*] Direct Rendering Manager (XFree86 4.1.0 and higher DRI support)

and below that, you want

<M> ATI Radeon

as a _module_. it's important that it's modular, and loaded up at boot, before you start X. also, you need to completely remove the old radeon DRM driver that you used with kernel 2.4 (if you have one) here's why:

i don't know why agpgart needs to be in the kernel, but i know that if it's a module, it will say that agp is not available when starting X.

in 2.5/2.6, the module system is completely different. trying to load an old 2.4 module will not work - in fact, it could mess things up badly. when x starts up and you have the radeon dri driver specified, x will load the module if it's not loaded. if it tries to load the old driver, x will hang with a black screen, requiring a reboot. so if you put support directly into the kernel for the radeon, x will still see that there is no radeon module loaded, try to load the old radeon module, and hang. and naturally, no support at all brings the same results. if you have the radeon driver as a module and loaded on boot, X will see that the radeon module _is_ loaded and it won't try and load the old driver, and will correctly use the new and loaded module.

ati's firegl drivers don't work yet. but since 2.6 is entering testing and a lot more people will be stress testing it, i'm sure ati (or someone else, if not them) will get it working with 2.6.


the synaptics mouse driver is an x driver, and independent of the kernel. the same zaxismapping configuration would work with 2.4, but you just wouldn't be using the event interface device or protocol anymore.