SlackwareThis Forum is for the discussion of Slackware Linux.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Originally posted by LNXman HMMM, your config file seems to be fine. I see nothing out of the ordinary.
Things you can try in the mean time:
1.
..........a) Uncomment the #VideoRam 1024 argument line
..........b) Run startx again
..........NOTE: If you want to force a particular depth when running startx
........................then you can run it like this startx -- -bbp <depth>.
..........c) If it fails, continue, else you are done. 8^)
Okay, I uncommented as suggested, and ran startx -bpp 8. I got a blank screen with nothing on it. I tried toggling screen resolutions, then tried backing out of X--nothing happened, so I rebooted. I got the same as before in my fifth post on this thread--no (EE)s, just that (WW).
Quote:
2.
...........a) Instead of using the trident driver as you are doing. Maybe you could
................use the generic VGA driver instead, or the generic VESA driver.
................So, first comment the line Driver "trident" out with the #.
...........b) Below it type Driver "vga"
...........c) Save the file
...........d) run startx again
...........e) If it fails, continue, else you are done. 8^)
Okay. I'll show you my Device section of XF86Config as it stands:
This at least brought up something (with just startx. XFCE loaded up, but the mouse pointer was the size of a Buick. I could see the lower left corner of the desktop. I changed the XF86Config file again to look like this:
With a just a normal startx I got a blank screen, then a portion of XFree86.0.log:
(WW) VGA: Chipset "tgui9680" in Device section "Trident TGUI9660" isn't valid for this driver
(EE) No devices detected.
Fatal Server error:
no screens found
<and then I got the same XIO error>. So I went on to step three.
Quote:
3.
...........a) Follow directions from step 2a from above, but comment the line
.................Driver "vga" out this time.
...........b) Below it type Driver "fbdev"
...........c) Save the file
...........d) run startx again
...........e) if it fails, then "CRAP!!!!!!!, a;sldkjf;alsjdie;j" . . . . . . 8^( , and continue
And everything is still HUGE. I can only see the lower left corner of the desktop again. I chose a default depth of 8 which should have selected the first resolution of 1024x768, but apparently, it went straight to 640x480 (or smaller). I can post my /var/log/XFree86.0.log, but it'll take some time. I have to type it all out (unless I can get my internet connection working through command line--maybe then I can copy and paste somehow?). If you need it, I'll post it up, but give me some time.
try the following using the TRIDENT driver WITHOUT the Chipset declaration.
First, the command should be startx -- -bpp <depth> just like it is written. . . maybe that may help a bit. . .
Second, I would like you to search through your log after you run startx with the trident driver, and search for the following strings:
-The string: "Trident TGUI 96xx" <- this should take you to the info about the PCI slot where the card is found. If you do not have this string anywhere, then give me the line that contains the pci information of your card from running the command lspci -b . It shoud have a format like -> MN:OP:Q VGA compatible controller . . ., etc. (Where M <-> Q, are integers)
What I want to see is if there may be a problem with XFree86 finding the video card. . . What gives me a clue is the output from when you ran the "fbdev" driver where it could not detect any device.
If you feel like you want to go ahead, you can try this out as well:
...........* When you get the line containing the pci information for you card
..............(from the lspci command from above). The begining group of
..............numbers is the Bus I.D. where your card resides in the machine.
..............Take note of those numbers.
...........* Now open up your config file and add the line
..............BusID "PCI:MN:OP:Q" (just as you noted from running
..............lspci) to your Device section where you load the
..............trident driver.
...........* Save and startx again as previously explained.
Hopefully fbdev will behave a bit nicer with you. You can also try this with the other drivers as well.
-The string: "Chipset <chipset_name> found!" (Where <chipset_name> will be the name of the chipset found). This string should give us a better idea of what chipset it thinks it finds. . .
I ran startx -- -bpp 8. I typed it in exactly as that. I got a blank screen. The hard drive would spin for a few seconds, then nothing. I couldn't toggle any settings or back out of X, so I had to reboot.
My /var/log/XFree86.0.log shows the following string:
after various resource ranges and modules loaded (like dbe, extmod, type1, freetype, etc.) it loads module "trident":
(II) LoadModule: "trident"
(II) Loading /usr/X11R6/lib/modules/drivers/trident_drv.o
(II) Module trident: vendor="The XFree86 Project"
compiled for 4.2.0, module version = 1.0.0
Module class: XFree86 Video Driver
ABI class: XFree86 Video Driver, version 0.5
I also ran lspci -b just in case. Here's what I got:
(WW) TRIDENT: No matching Device section for instance (BUSID PCI:0:11:0) found
(EE) No devices detected.
Fatal server error:
no screens found
(same XIO error)
Why is the PCI section different? This is odd. I ran it again, but used BusID "PCI:00:0b:0" instead. I got the same thing so I used BusID "PCI:00:11:0 this time. This resulted in a blank screen and a reboot. /var/log/XFree86.0.log showed the following:
(II) Primary Device is: PCI 00:0b:0
(--) Chipset tgui9660 found
This is weird. When I ran lspci -b I got PCI:00:0b.0, and the log shows PCI:00:0b:0. Does this matter? I'm at a loss here. It seems like it can find the card and driver just fine, but it X can't. I don't know what the deal is.
Seems we are getting somewhere since we get no errors.
The 0b and 11 are the same thing (0b being 11 in hex). However, it seems that the XFree86 Server does not like hex number for BUS I.D.'s, so when you actually changed the 0b to an integer, then it was able to find the chipset (as you noted in your last switch configuration switch).
Now that you know it can detect a chipset, try to force the tgui9680 chipset when you run X. Then check the logs to see if in fact it tried to load that chipset as stated in your config file or not.
Also, let's try to give it the board name as well. . . in your case the board name will be "TGUI 96xx"
So, your Device section should look like this:
-------------------------->8-------------------------
Section "Device"
Identifier "Trident TGUI9660"
# Driver "vga"
# Driver "fbdev"
Driver "trident"
BoardName "TGUI 96xx"
Chipset "tgui9680"
VideoRam 1024
BusID "PCI:00:11:0" #NOTE: You had it PCI:00:0b.0" (see '.')
# Insert Clocks lines here if appropriate
EndSection
--------------------->8-----------------------
One last thing to check is whether something else may not be loading right, like for example, your mouse. Sometimes a problem with the mouse will not let XFree86 display, eventhough there were no error. Just look for any lines that may look odd to you from the log . . . BTW, some problems do no start with the usual (EE) in some cases because they sometimes are output within a successfull load.
Another thing to try would be the other two 968x chipsets: "tgui9682" or "tgui9685". If you would like. . .
Thanks. I'll try it later tonight and post the results. Do you think the XFree86 version could be too high for my old box? If this fails, you think I should try another reinstall?
Sorry I didn't respond sooner. The wife was up late last night and she's jealous of my computer. I didn't get to try anything out until now. Here's the what happened:
When I added the BoardName "TGUI 96xx" I got the following:
(EE) TRIDENT(0): No support for "tgui9680"
(EE) Screen(s) found, but none have a usable configuration.
Fatal Server error:
no screens found
(I then got the same XIO error).
When I tried changing the chipset to tgui9682 and tgui 9685 I got the following:
(WW) TRIDENT: Chipset "tgui9682" <or tgui9685 respectively> in Device section "Trident TGUI9660" isn't valid for this driver
(EE) No devices detected
(then I got the no screens found and XIO errors)
/var/log/XFree86.0.log showed no errors other than those above, and everything else looks good. It also lists the avialable support for Trident and both 9660 and 9680 are in that list. This makes no sense. I tried again, but used chipset 9660 and had to reboot. I got this in XFree86.0.log:
#snip the beginning
(WW) Open APM failed (/dev/apm_bios) (No such device)
#snip more stuff that looks good
(**) Chipset override: tgui9660
(**) Chipset tgui9660 found
#snip more stuff that looks good
(WW) System lacks support for changing MTRRs
I tried with tgui9680 as chipset and got the same thing as above (no support for tgui9680 et. al.). I have no idea what's happening now. Any thoughts? Should I go with an older Slackware or version of X? Again, thanks for helping me out on this one.
Sorry I have not gotten to you earlier, but I have been busy the last couple of days, and right now I am catching up with email . . .
Anyhow, I did a bit of research and it seems that some old cards support have been broken with the 4.2.X versions. Mainly the the video cards that are considered very, very, very, very. . . , very old by today's standards.
So now, you only have a choice. As you have previously questioned about reverting back a 3.3.X release of XFree86, now, I believe, reverting back is your only choice.
I checked the 3.3.X release version of XFree86 in Mandrake(since you mentioned it worked), and found out they have been using the last known 3.3.X version of XFree86 release for old video cards detected during installation. It being version 3.3.6.
If you would like to install the 3.3.6 version of XFree for Slackware, you are going to have to obtain the packages from the 7.1 version release of Slackware under the x1 directory category. A mirror link is here .
First uninstall ALL the packages that are relative to your current XFree86 installation. If you are not sure which are they, then check the contents of your install cd under slackware/x.
Then, install the 3.3.6 version (and its related packages) that come in the 7.1 release version of Slackware.
Once you install it, you can run the xf86config to setup your XFree86 server, and hopefully you should be on your way. . .
Make sure you make a copy of your current XF86Config file from your current version of XFree86 (I recommend to copy it to the home directory) so that you can have a reference if you need to when configuring 3.3.6 . . .
Thanks for helping me through this whole ordeal. I really appreciate you taking the time to follow along. I'll try do as you suggested tonight when I get home.
I just checked out that link you provided. I don't know if I'll be able to get the files since my box isn't hooked up to the net yet (I can't get my pppsetup to actually work, although, when I was able to open the GUI in HUGE resolution, I could setup Kppp with all my info and it worked fine).
Is there a way to grab all those files at once or do I have to select them individually?
Another thing, I'll have to do some reading up on installing/uninstalling and copying files as I've not done anything with files yet (only uninstall and install the OS and play around with settings in GUI, etc.). Hopefully it's not too painful.
Since I can't get the GUI to work right, any hints on pppsetup? I've gone through it a few times setting up the "expect ***" and "say ***" but for some reason it just doesn't fly. I know my user name, password, DNS, gateway, etc. are all correct, but it doesn't seem to take. I'll try it again when I get home. Is there anyway to run Kppp from command line? Maybe it will work if I do it that way? I've never connected to the internet from command line so this will be a new experience for me. If I like it, I might just keep it that way.
You can setup your ppp via cli by using the pppsetup application provided for slackware. Just make sure you read what each step tells/asks you, and how to run it as well. If you want to get a heads up, you can check out the /usr/doc/Linux-HOWTOs/PPP-HOWTO that comes with your distribution. It should not be that difficult to achieve.
WRT an ftp client for cli, I tent to use ncftp. It is very nice, and easy to use. Just make sure you read about it. Either search the net once you connect for tutorials, or just check its manual.
If you want to learn more about how to install/remove packages. Check the manuals for installpkg, removepkg, and pkgtool (i.e. man <application>).
Thanks for all the suggestions. I read the HOWTOs for ppp and lots of stuff on www.tldp.org. I also read the man pages for the package tools you suggested. I the ran the pppsetup again. It said I need to add ppp to my kernel. I tried to run a make config and a make menuconfig but "make" is not found. The pppsetup went very well though. It wrote all the necessary files. I checked through some files and one of them (I can't remember which one unfortunately) said that make should be found in /bin/bash or /bin/sh. I cd'd to /bin and ls'd to find both bash and sh. Neither were there. I'm thinking I'm going to need to reinstall again. I had tried pppsetup in the past a few times with Slack 8.1 and was able to successfully make config but not "make menuconfig" at least one of those tools was available. I don't know why it can't find it this time. You think I should reinstall again or is there another way to get the "makes" setup correctly?
I just thought I would let you know that before I posted this problem I had done everything I could think of to find a solution. I had searched various messageboards, websites, books (Running Linux, Linux in a Nutshell) and couldn't find anything that was the same as or similar to my problem. I even reinstalled Slackware. I had posted this exact same problem on another Linux help website on 10-16, and still have had no responses--but almost 20 reads. You are the only person who's taken the time to help me out with all this and I wanted to thank you for your time and patience. I hope I can return the favor in the future.
Distribution: Slackware, (Non-Linux: Solaris 7,8,9; OSX; BeOS)
Posts: 1,152
Rep:
In order to have make, you need to install the development tools. You can
do this using pkgtool (man pkgtool). Mount your 8.1 install disk and go to
the appropriate directory that contains the development tools (you can
read the file list to find where the dev tools are).
Are you now working on a different problem from what you began with? It
looks like you still have the X problem, but are having a problem with
getting connected via PPP?
I believe the XFree86.org FTP site allows one to download a directory
with a command like:
ftp> get 3.3.6-directory.tar
where 3.3.6-directory is the directory that contains all the files you need.
If not, there may be mirrors that allow this (it's server dependent).
Maybe the packages that come with the tools were not installed during installation. Did you choose everything when you installed? Or did you let Slackware select?
To be sure of what you have installed in your system, you can check the contents of /var/log/packages. In that location you will find files with the package names you have installed in your system. If a package name you need is not present, they you are going to have to install it with installpkg. Like this:
installpkg <package_name>
Another way to find out whether a specific binary has been installed in your system (whether an application, or library) is to grep for its name. Like this (assuming you are currently in /var/log/packages):
grep <string_to_find> * | more
where <string_to_find> is the name of the application/library you are trying to see if is installed.
If the above command finds it, then you will see the filename that contains your <string_to_find> is the package name that you installed from the cd.
From that point on you make your own decisions/approaches. You can either removepkg and then installpkg the same package to try to fix the problem if it you feel there may been a bad install.
Or, if it does continues not to work after you re-install, you might check whether all the necessary libraries have been installed for that particular binary you want to run. Example:
Let's say you are trying to run foo, but foo does not run. So you search in /var/log/packages to see if it is installed using the grep command as explained above. Then you find it is installed under /usr/local/bin from your query. So, you say to yourself: "WTF?!?!"
So now you wonder if maybe all the depencies for that binary have been met, because maybe one of them is missing and causes foo not to run.
To check, you would run the following:
ldd /usr/local/bin/foo
Where, "ldd" is a command which will show you whether all library depencies are present for foo to be able to run. When you run it, you get the following output (remember this is a made up example):
librt.so.1 => /lib/librt.so.1 (0x4002d000)
libc.so.6 => /lib/libc.so.6 (0x4003f000)
libpthread.so.0 => Not Found
/lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x40000000)
Then you notice that libpthread.so.0 is not found. Which means the package containing libpthread.so.0 has not been installed. Now you are left with the option to find which package contains libpthread.so.0 in your installation cdrom. To make it easy, you can accomplish this by running the following little script I just wrote:
if [ ! "$1" -o ! "$2" ]; then
echo "USAGE: $0 <slack_package_repository> <pattern>";
echo "DESCRIPTION: simple script that locates files withing "
echo "..............a slackware packages, or any archive with the"
echo "...............tgz extension"
exit 1
fi
for file in `find $1 -type f -iregex .*tgz -print`;
do
result="`tar ztf $file | grep $2`";
if [ "$result" ]; then
echo "$result";
echo "|-> $file";
fi
done
- Transfer it to a file and name it whatever you want.
- Make sure the file has executable permission (i.e. chmod 755 <file>)
- Run it like this:
...../<file> in the command line, so that you learn what arguments it
....takes.
....(NOTE: <file> is the filename you gave the script)
Once you are familiar with this, you can mount your cd. For example: mount -t iso9660 /dev/<device> /mnt/cdrom
And then run the script: ./<file> /mnt/cdrom/slackware libpthread.so | more
Finally if it finds it, you will get an output sort of like the following:
Where each line begining with |-> will contain the path of the package that may contain what you are looking for.
Then, for the example search, we see that libpthread.so is found in the libpthread-1.0.3-i386-i.tgz package under: /mnt/cdrom/slackware/yy/
So now, all you would have to do is install it. . .
Hope this helps you a bit with your XFree86 version 3.3.6 install . . .
ALSO, make sure your have the ppp package installed from your cdrom (usually under slackware/n). PPP support should be already active in your kernel if you have not changed it and recompiled it . . .
To check if ppp is available for your kernel, run the command: modprobe ppp
If you do not get any errors, run the following: lsmod
And see whether ppp is loaded.
If it loads, then activate it for reboot by uncommenting the ppp module activation line(s) in the file "/etc/rc.d/rc.modules".
Moses, thanks for joining in. I'll definitely check out what XFree86.org has got. I tried to mount the cdrom and it couldn't find any media. I checked /etc/fstab and /etc/mtab and it was listed there perfectly. I don't know what happened. It just couldn't see anything on my 8.1 cd.
LNXman, again I thank you for your support. I checked /etc/rc.d/rc.modules and uncommented /sbin/modeprobe ppp and rebooted. It worked, but it couldn't find ppp module. To answer your question--I did an install that allowed me to pick and choose programs (probably a mistake). As for all the other stuff you mentioned, I'll have to do some more reading on it as I'm extremely new to Linux. I've had it installed for a few months now, but am VERY new to it still. Although, all the troubles I've had with Slackware has helped me learn more than I think I would have without them. I've learned mostly about the file structures, and how to edit files a bit. However, I've also learned alot about finding solutions to problems. The steps you mentioned are still a bit over my head, but I've got this page bookmarked for future reference.
I'm going to reinstall again as I'm not having much luck. This time, I'm going to let Slackware do a full install and I'll just remove stuff I don't need later. This way, I'll be starting out with everything I need and some stuff I don't. Hopefully I'll be able to fix my problems faster this time around with a new install of everything.
I'll be using both of your suggestions after the install to obtain an older version of X and for package handling.
Distribution: Slackware, (Non-Linux: Solaris 7,8,9; OSX; BeOS)
Posts: 1,152
Rep:
If the CDROM drive is listed in /etc/mtab, the device is mounted, or the
OS thinks it's still mounted for some reason. Did you get an error that
said something like "media not found"?
Post the result of:
(as root)
mount /dev/cdrom /mnt/cdrom
and
mount
DON'T reinstall! If you do that, you'll just start over with many of the
problems you may have already solved. . .
To do a "full" install after you've already installed Slack, you can go to
your /mnt/cdrom (or wherever your CD is mounted, once you get it
mounted), and install packages as you need them. This is, in my
opinion, a better solution than reinstalling, as it allows you to become
familiar with the pkgtools of Slackware (and more generally, whatever
distro you are using). It also allows you to become familiar with the
various packages that make up the full distribution that is Slackware.
I had already started reinstalling before your post. However, I've taken a break from Linux for the past few days as my reinstall had some errors. I figured out that one should never attempt to do anything with Linux after having slept for 3 hours and been at work for 10 hours. I will be doing another reinstall soon as the last one was botched. I know it was my mistake because I've installed Mandrake 8.1 3 times, Slackware 8.1 about 15 times and never have gotten this error. I don't remember the error at this time, but it won't matter after I format, repartition and reinstall.
Now, some news for both you and LNXman:
I might be getting another PC soon (I don't think I can pass it up). It's a pentium 266MHz, 192MB RAM, 8 or 16 MB vid card, 10GB hard drive, A 24 or 56x CD-ROM, a new sound card, and the only thing on board is the NIC. I don't have all the specs on the makes and models of the hardware, but if it's Linux compatible, I'll be scooping up this PC for $75 Canadian and installing Slackware 8.1 on it. Hopefully, since it will be newer than my current PC (Pentium 120MHz, 80MB RAM, 1MB vid card, 4.3GB hard drive, 8x CD-ROM, ancient sound card, old 33.6 Kbps modem), it will be less of a hassle to work with the hardware. I've read countless threads and websites where people have had problems with the Trident vid cards and hopefully this "new" PC will have a different one. If all goes well, I'll be getting this PC next week and be able to install Slackware, get cable internet access, setup my PC as a firewall and server, and network the 3 PCs. This will take some time for me to understand Linux enough to start doing this, but it will be worth it.
I really appreciate your help in all of this, and I will definitely be returning to this thread in the future for reference when problems arise. You guys have been great, and hopefully I'll be able to return the favor someday. Thanks again!
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.
