Re: compiling a new kernel
Bruce == Bruce Park [EMAIL PROTECTED] writes: Bruce Hello debian users, I'm currently using the 2.4.18-bf2.4 Bruce kernel and I would like to use a new kernel so that I can Bruce have access to my ms sidewinder joystick. Bruce Since I'm new to kernel compiling, I was wondering if Bruce anyone can give me some pointers on this. For example, I Bruce don't even know what new kernel to choose. The current Bruce kernel that I have on my system was chosen for it's ability Bruce with USB devices. I noticed the other flavors didn't have Bruce such a good choice when loading the device modules during Bruce installation. Hi Bruce, Your best bet would be to install the standard 2.4.18 kernel first (I know nothing about the ms sidewinder joystick so I cannot tell if it is supported, but I do see a module called sidewinder.o in the image). The prebuilt Debian kernels are very well put together and you rarely need to build your own unless you have very estoteric needs. The bf2.4 kernel is limited by the desire to keep it on floppies during installation. So, what are your options? Well shyamal@rattler:~$ apt-cache search kernel-image-2.4.18 kernel-headers-2.4.18-bf2.4 - Headers for Linux kernel version 2.4.18 (bf variant) on 386 kernel-image-2.4.18-386 - Linux kernel image for version 2.4.18 on 386. kernel-image-2.4.18-586tsc - Linux kernel image for version 2.4.18 on Pentium-Classic. kernel-image-2.4.18-686 - Linux kernel image 2.4.18 on PPro/Celeron/PII/PIII/PIV. kernel-image-2.4.18-686-smp - Linux kernel image 2.4.18 on PPro/Celeron/PII/PIII/PIV SMP. kernel-image-2.4.18-bf2.4 - Linux kernel image for version 2.4.18 (bf variant) on 386. kernel-image-2.4.18-k6 - Linux kernel image for version 2.4.18 on AMD K6/K6-II/K6-III kernel-image-2.4.18-k7 - Linux kernel image for version 2.4.18 on AMD K7 pcmcia-modules-2.4.18-bf2.4 - PCMCIA Modules for Linux (kernel 2.4.18-bf2.4). shyamal@rattler:~$ Pick the 2.4.18-{arch} package that best matches your needs (use -386 if you are truly lost). Get a root prompt and type 'apt-get install kernel-image-2.4.18-{arch}' and follow the instructions for the initrd line you need to add to your /etc/lilo.conf file and you should be set. As long as you add the initrd stuff as instructed you should just have to reboot to get the new kernel. Likely, you will find your ethernet card or similar hardware might stop working. This is because bf2.4 has stuff built into it that the other images do not, and when you switch to these kernels you need to explicitly load them. If you don't know how to work with /etc/modules to fix this read 'man modules' and look in dmesg to determine what hardware drivers are being used. There is also a tool (modconf?) to do this, but I've never used it. If you don't get anywhere with that, post here for help. Doing a 'modprobe sidewinder' might be enough to get the joystick to work. If so, add it to /etc/modules. But I really do not have a clue about this ;-) Cheers! Shyamal -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: compiling a new kernel
Bruce Park, 2002-Dec-23 21:33 -0500: Hello debian users, I'm currently using the 2.4.18-bf2.4 kernel and I would like to use a new kernel so that I can have access to my ms sidewinder joystick. Since I'm new to kernel compiling, I was wondering if anyone can give me some pointers on this. For example, I don't even know what new kernel to choose. The current kernel that I have on my system was chosen for it's ability with USB devices. I noticed the other flavors didn't have such a good choice when loading the device modules during installation. Anyhow, any help or suggestion is greatly appreciated. I can only layout the way I've been doing it. I use kernel-package, which is the debian way of compiling a kernel and including it in the package system. However, I use the kernel source from www.kernel.org as opposed to the the packaged source. I don't know what the difference is, if any. High-level processes and Notes: 1. get the kernel source, either packaged or from kernel.org, 2.4.20 is the latest stable source 2. install kernel-package along with libncurses5-dev and any dependancies 3. read the kernel-package doc to understand the process in detail: /usr/share/doc/kernel-package/README.gz 4. start with the current kernel's config...copy /boot/config-2.4.18-bf2.4 to /usr/src/linux, then when you make menuconfig/xconfig, load the config and start there. this way you'll be starting with a known good config and can add additional stuff and strip out unnecessary stuff 5. compile, install and reboot and described in the kernel-package doc have fun, jc -- Jeff CoppockSystems Engineer Diggin' Debian Admin and User -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Compiling a new kernel
Hi everybody I hope that somenone can help me. I just followed precisely the instruction for compiling/installing a new kernel (2.4.19). The previeous version was 2.2. tar xvjf "the kernel source" make menuconfig make dep make bzImage make modules make modules_install then I copied the new image in the boot location of the old one /boot/bzImage.(I saved the old one). So I edit the lilo.conf and I specified where to load the new image. After rebooting (everything seemed to go fine) I launched the "kernelversion" command but the answer was 2.2!!! What happened?? I want to precise that I've been using for one monthand thatI'm a very new user! Thanks in advance to everyone with a good solution Emanuele
RE: Compiling a new kernel
Title: Message * Maybe you didn't compile the 2.4.19 kernel check if you're compiling this one and not the 2.2 * Maybe you made a mistake in lilo.conf, check the process again -Original Message-From: Emanuele Boieri [mailto:[EMAIL PROTECTED]] Sent: 04 November 2002 13:13To: debianSubject: Compiling a new kernel Hi everybody I hope that somenone can help me. I just followed precisely the instruction for compiling/installing a new kernel (2.4.19). The previeous version was 2.2. tar xvjf "the kernel source" make menuconfig make dep make bzImage make modules make modules_install then I copied the new image in the boot location of the old one /boot/bzImage.(I saved the old one). So I edit the lilo.conf and I specified where to load the new image. After rebooting (everything seemed to go fine) I launched the "kernelversion" command but the answer was 2.2!!! What happened?? I want to precise that I've been using for one monthand thatI'm a very new user! Thanks in advance to everyone with a good solution Emanuele
Re: Compiling a new kernel
On Monday 04 November 2002 14:12, Emanuele Boieri wrote: Hi everybody I hope that somenone can help me. I just followed precisely the instruction for compiling/installing a new kernel (2.4.19). The previeous version was 2.2. 1.. tar xvjf the kernel source 2.. make menuconfig 3.. make dep 4.. make bzImage 5.. make modules 6.. make modules_install then I copied the new image in the boot location of the old one /boot/bzImage.(I saved the old one). So I edit the lilo.conf and I specified where to load the new image. After rebooting (everything seemed to go fine) I launched the kernelversion command but the answer was 2.2!!! What happened?? I want to precise that I've been using for one month and that I'm a very new user! Thanks in advance to everyone with a good solution Emanuele Since you did not mention it: Did you also run lilo ? -- Michael -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Compiling a new kernel
04/11/2002 13:12:57, Emanuele Boieri [EMAIL PROTECTED] wrote: then I copied the new image in the boot location of the old one /boot/ bzImage.(I saved the old one). So I edit the lilo.conf if you copied the new image to the same place as the old one, you shouldnt need to edit lilo.conf and I specified where to load the new image. After rebooting did you actually excecute lilo? hugh -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Compiling a new kernel
Hello, I recently tried to compile a new kernel (2.2.18pre21) on my machine. I used the instructions in Compiling a new kernel from Installing Debian GNU/Linux 2.2 For Intel x86 document: http://www.debian.org/releases/stable/i386/ch-post-install.en.html#s-kernel-baking But something probably went wrong because I compiled my sound driver as a module and tried to insert it and it gives me the error message: bash-2.03# insmod ./es1371.o ./es1371.o: kernel-module version mismatch ./es1371.o was compiled for kernel version 2.2.18pre21 while this kernel is version 2.2.17. The kernel intallation package did exist after cinfiguring it and runnung make-kpkg clean and fakeroot make-kpkg revision=blah.1.0 kernel_image. Upon installing the package said it was going to use my current lilo.conf to make the new kernel bootable. I hope this is enough and not too redundant info. Can someone help me figuering out what is wrong and how to fix it? Thanks, Kaloyan _ Get your FREE download of MSN Explorer at http://explorer.msn.com
Re: Compiling a new kernel
bash-2.03# insmod ./es1371.o ../es1371.o: kernel-module version mismatch ./es1371.o was compiled for kernel version 2.2.18pre21 while this kernel is version 2.2.17. Looks like you're not running the newly compiled kernel 2.2.18pre21, but your old 2.2.17. You can check what kernel you're running with uname -r Have a look at /etc/lilo.conf . There should be at least two image entries. Here is what I have: image=/vmlinuz label=Linux ... and image=/vmlinuz.old label=LinuxOLD ... These entries correspond to symbolic links. Check where they point. In my case, /vmlinuz points to /boot/vmlinuz-2.2.18pre21. It should be the same in your case. If so, you can hit shift at boot time when Lilo appears, then type Linux (or whatever label entry you have), and it runs /boot/vmlinuz-2.2.18pre21. If you make changes to /etc/lilo.conf, you must run lilo on the commandline for the changes to take effect. All this is normally taken care of at the end of the kernel build and install process. It asks you whether you it should run lilo. Maybe you answered no. You can get more information with info lilo and info lilo.conf
Re: Compiling a new kernel
bash-2.03# insmod ./es1371.o ./es1371.o: kernel-module version mismatch ./es1371.o was compiled for kernel version 2.2.18pre21 while this kernel is version 2.2.17. Hi Kaloyan! I think you shold look whether your new kernel is booting. The new vmlinuz should be in /boot or if you made no changes in sourcetree/Makefile in /, your root directory. If vmlinuz is in /boot you have to make shure that your /etc/lilo.conf lokes like mine : default=Linux image=/boot/vmlinuz-2.4.0 label=Linux read-only image=/boot/vmlinuz-2.2.18 label=old read-only other=/dev/hda2 label=dos Then you can be shure that your new kernel will be used at boottime. Note that the default label (default=Linux) will be loaded if you use the Enter key or do nothing while LILO is comming! And donĀ“t forget to type # lilo on console. With this, your new lilo configuration will be written into MBR. Mathias
Re: Problems rebooting after compiling a new kernel
Hi, Rajesh == Rajesh Saxena [EMAIL PROTECTED] writes: Rajesh (Then I moved /lib/modules/2.2.17 to /lib/modules/2.2.17.old) Please note that this location shall not be used by either your old kernel, nor you new one. May I humbly suggest kernel-package, and specifically point you to the FLavours.gz documentation therein? Rajesh (Then I moved /boot/System.map and /boot/vmlinuz-2.2.17 likewise) cp /usr/src/linux/System.map /boot/map cp /usr/src/linux/arch/i386/boot/bzImage /boot/bzImage Your modules and system.map files are saved, but are not live; your old kernel, regardless of its current name, shall still look for its modules and map file in the original location. Rajesh (I changed the symlinks /vmlinuz and /vmlinuz.old to reflect the changes Rajesh in /boot) Well, you should not be surprised if the latter does not work. Rajesh I edited /etc/lilo.conf to add a new block for /boot/bzImage) /sbin/lilo Rajesh No lilo errors popped up so I rebooted and once it came time for the new Rajesh boot process I got this error.. Rajesh VFS: Cannot open root device 16:01 Rajesh Kernel panic: VFS: Unable to mount root fs on 16:01 Umm. You are compiling the IDE hard disk suport as a module. And you have no SCSI support at all. # CONFIG_BLK_DEV_HD_IDE is not set CONFIG_BLK_DEV_IDEDISK=m CONFIG_BLK_DEV_IDECD=m So your IDE Hard drive is compiled as a module. But the module can't be accessed until we can read the hard drive. Rajesh To try to fix things I tried to reboot with the boot floppy I Rajesh made when I initially installed Debian but I get a *lot* of Rajesh module dependencies problems(I'm assuming because of Rajesh /lib/modules/2.2.17.old )but one which just won't go away and Rajesh blocks the login process is Well, your old kernel, on the floppy, can't find any of its modules, since /lib/modules/2.2.17 now contains a new set of modules. Rajesh Can someone please point out what I did wrong and how I can Rajesh correct it. If it's not reparable then what steps do I need Rajesh to take once I reinstall and before I reboot with the new Rajesh kernel. Thanks in advance. If you have the Debian rescue floppy, just boot into that, switch to virtual console 2 (Alt-F2), and mount your root device under /mnt. Go in, and move the /mnt/lib/modules/2.2.17 to 2.2.17.new, and likewise with System.map (I would not touch the actual images). Exit, remove the rescue disk, and boot back to the old image (which should no longer have module problems). I wish you luck. -- They [preachers] dread the advance of science as witches do the approach of daylight and scowl on the fatal harbinger announcing the subversions of the duperies on which they live. Thomas Jefferson Manoj Srivastava [EMAIL PROTECTED] http://www.debian.org/%7Esrivasta/ 1024R/C7261095 print CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E 1024D/BF24424C print 4966 F272 D093 B493 410B 924B 21BA DABB BF24 424C
Problems rebooting after compiling a new kernel
Hi guys.. I'm having some weird problems when I try to reboot after compiling a kernel to suit my hardware. I've compiled linux kernels in debian several times before but I recently installed Debian Potato on a new intel box. In order to be as detailed as possible I've included some files at the bottom of the message(mainly the .config file I used to compile the kernel)so if the message's length is too long then I apologize. First of all, this is a list of all the actions I took before rebooting #cd /usr/src/linux #make menuconfig #make dep #make clean #make bzImage #make modules (Then I moved /lib/modules/2.2.17 to /lib/modules/2.2.17.old) #make modules_install (Then I moved /boot/System.map and /boot/vmlinuz-2.2.17 likewise) #cp /usr/src/linux/System.map /boot/map #cp /usr/src/linux/arch/i386/boot/bzImage /boot/bzImage (I changed the symlinks /vmlinuz and /vmlinuz.old to reflect the changes in /boot) I edited /etc/lilo.conf to add a new block for /boot/bzImage) #/sbin/lilo No lilo errors popped up so I rebooted and once it came time for the new boot process I got this error.. VFS: Cannot open root device 16:01 Kernel panic: VFS: Unable to mount root fs on 16:01 To try to fix things I tried to reboot with the boot floppy I made when I initially installed Debian but I get a *lot* of module dependencies problems(I'm assuming because of /lib/modules/2.2.17.old )but one which just won't go away and blocks the login process is modprobe: modprobe: Can't locate module net-pf-1 Before I attach the .config file let me say that this computer booted up fine before compiling a new kernel and that I've tried to boot into single user mode but that won't work either because of the module problems. Can someone please point out what I did wrong and how I can correct it. If it's not reparable then what steps do I need to take once I reinstall and before I reboot with the new kernel. Thanks in advance. -- Here's what's in my /usr/src/linux/.config # # Automatically generated make config: don't edit # # # Code maturity level options # CONFIG_EXPERIMENTAL=y # # Processor type and features # # CONFIG_M386 is not set # CONFIG_M486 is not set # CONFIG_M586 is not set # CONFIG_M586TSC is not set CONFIG_M686=y CONFIG_X86_WP_WORKS_OK=y CONFIG_X86_INVLPG=y CONFIG_X86_BSWAP=y CONFIG_X86_POPAD_OK=y CONFIG_X86_TSC=y CONFIG_X86_GOOD_APIC=y CONFIG_1GB=y # CONFIG_2GB is not set # CONFIG_MATH_EMULATION is not set CONFIG_MTRR=y # CONFIG_SMP is not set # # Loadable module support # CONFIG_MODULES=y # CONFIG_MODVERSIONS is not set # CONFIG_KMOD is not set # # General setup # CONFIG_NET=y CONFIG_PCI=y # CONFIG_PCI_GOBIOS is not set # CONFIG_PCI_GODIRECT is not set CONFIG_PCI_GOANY=y CONFIG_PCI_BIOS=y CONFIG_PCI_DIRECT=y CONFIG_PCI_QUIRKS=y # CONFIG_PCI_OPTIMIZE is not set CONFIG_PCI_OLD_PROC=y # CONFIG_MCA is not set # CONFIG_VISWS is not set CONFIG_SYSVIPC=y CONFIG_BSD_PROCESS_ACCT=y CONFIG_SYSCTL=y CONFIG_BINFMT_AOUT=y CONFIG_BINFMT_ELF=y CONFIG_BINFMT_MISC=y # CONFIG_BINFMT_JAVA is not set CONFIG_PARPORT=m CONFIG_PARPORT_PC=m # CONFIG_PARPORT_OTHER is not set # CONFIG_APM is not set # # Plug and Play support # # CONFIG_PNP is not set # # Block devices # CONFIG_BLK_DEV_FD=y CONFIG_BLK_DEV_IDE=m # # Please see Documentation/ide.txt for help/info on IDE drives # # CONFIG_BLK_DEV_HD_IDE is not set CONFIG_BLK_DEV_IDEDISK=m CONFIG_BLK_DEV_IDECD=m # CONFIG_BLK_DEV_IDETAPE is not set CONFIG_BLK_DEV_IDEFLOPPY=m # CONFIG_BLK_DEV_IDESCSI is not set CONFIG_BLK_DEV_CMD640=y # CONFIG_BLK_DEV_CMD640_ENHANCED is not set CONFIG_BLK_DEV_RZ1000=y CONFIG_BLK_DEV_IDEPCI=y CONFIG_BLK_DEV_IDEDMA=y # CONFIG_BLK_DEV_OFFBOARD is not set CONFIG_IDEDMA_AUTO=y # CONFIG_BLK_DEV_OPTI621 is not set # CONFIG_BLK_DEV_TRM290 is not set # CONFIG_BLK_DEV_NS87415 is not set # CONFIG_BLK_DEV_VIA82C586 is not set # CONFIG_BLK_DEV_CMD646 is not set # CONFIG_BLK_DEV_CS5530 is not set # CONFIG_IDE_CHIPSETS is not set # # Additional Block Devices # # CONFIG_BLK_DEV_LOOP is not set # CONFIG_BLK_DEV_NBD is not set # CONFIG_BLK_DEV_MD is not set # CONFIG_BLK_DEV_RAM is not set # CONFIG_BLK_DEV_XD is not set # CONFIG_BLK_DEV_DAC960 is not set CONFIG_PARIDE_PARPORT=m # CONFIG_PARIDE is not set # CONFIG_BLK_CPQ_DA is not set # CONFIG_BLK_DEV_HD is not set # # Networking options # CONFIG_PACKET=y CONFIG_NETLINK=y # CONFIG_RTNETLINK is not set CONFIG_NETLINK_DEV=y # CONFIG_FIREWALL is not set # CONFIG_FILTER is not set CONFIG_UNIX=y CONFIG_INET=y # CONFIG_IP_MULTICAST is not set # CONFIG_IP_ADVANCED_ROUTER is not set # CONFIG_IP_PNP is not set # CONFIG_IP_ROUTER is not set # CONFIG_NET_IPIP is not set # CONFIG_NET_IPGRE is not set # CONFIG_IP_ALIAS is not set CONFIG_SYN_COOKIES=y # # (it is safe to leave these untouched) # # CONFIG_INET_RARP is not set CONFIG_SKB_LARGE=y # CONFIG_IPV6 is not set # # # # CONFIG_IPX is not set # CONFIG_ATALK is not set # CONFIG_X25 is not set # CONFIG_LAPB is not set # CONFIG_BRIDGE
Compiling a new kernel and dselect problems
Hi, I'm a new entry in the worderful Debian world. Previously i'd used Red Hat 6.x ad Mandrake but i thought i' was time to make a jump in the real Linux world and started using a potato release Well, my problem arise when i want to build a custom kernel the Debian-way. When i built it i found a .deb package ready to install. I dpkg -i the kernel image. My system warns me that there is a kernel image with the same name installed yet (my version has a revision of my own) but i proceed. All goes well but whe i try to install a new package using dselect, in the Install session, dselect tells me he wants to upgrade my custom kernel with the standard kernel having the same name kernel so i've to interrupt the installation of my packages. Can somebody tell me ho to solve this problem ? Can i build the custom kernel with a different name (not only the revision name) ? Thanks in advance Luca De Giorgi
Re: Compiling a new kernel and dselect problems
On Thu, Jun 29, 2000 at 10:28:12AM +0200, Luca De Giorgi wrote: Hi, I'm a new entry in the worderful Debian world. Previously i'd used Red Hat 6.x ad Mandrake but i thought i' was time to make a jump in the real Linux world and started using a potato release Well, my problem arise when i want to build a custom kernel the Debian-way. When i built it i found a .deb package ready to install. I dpkg -i the kernel image. My system warns me that there is a kernel image with the same name installed yet (my version has a revision of my own) but i proceed. All goes well but whe i try to install a new package using dselect, in the Install session, dselect tells me he wants to upgrade my custom kernel with the standard kernel having the same name kernel so i've to interrupt the installation of my packages. Can somebody tell me ho to solve this problem ? Can i build the custom kernel with a different name (not only the revision name) ? What is happening is that Debian has a version of the same kernel with a revision number higher than what you are using for your custom kernel. To prevent an upgrade, you need to specify an epoch on the make-kpkg command line, such as: make-kpkg --revision=3:custom.1.0 kernel_image This is discussed in /usr/share/doc/kernel-package/README.gz. -- Bob Nielsen, N7XY [EMAIL PROTECTED] Bainbridge Island, WA http://www.oz.net/~nielsen
compiling a new kernel ....
hello, i did just upgrade my laptop to potato, noticed that an old 2.2.1 kernel was running and launched compile... but it aborted with the message: erm7:/usr/src/linux# make-kpkg binary ... (cd debian/tmp-source/usr/src/; tar zcf kernel-source-2.2.3.tar.gz kernel-source-2.2.3;\ rm -rf kernel-source-2.2.3;) dpkg-gencontrol -isp -pkernel-source-2.2.3 -Pdebian/tmp-source/ dpkg-gencontrol: error: package kernel-source-2.2.3 not in control info make: *** [stamp-source] Error 29 strange enough on my workstation beneath (which has exactly the same sources.list, and was updated at the very same time) the 'make-kpkg binary' still runs... Workstation is a PPro200, Laptop is a PII-233... any hint on what's going wronf and how to resolve? (meaning using the debian kernel package...) ciao bboett == acount at earthling net http://erm6.u-strasbg.fr/~bboett === Unsolicited commercial email is NOT welcome at this email address To contact me replace acount by bboett in above addresses
Re: Compiling a new kernel.
JRL == James R Lunsford [EMAIL PROTECTED] writes: JRL /usr/src/linux is linked to the 2.0.34 directory. Am I correct JRL in assuming that I have to relink /usr/src/linux to the 2.0.35 JRL directory? Yes. JRL Anyway, can someone give me a down and dirty way to do this? Also, what JRL are the steps for compiling a kernel? I've done it before, but I'm a Install the kernel-package package and check /usr/doc/kernel-package/README.gz Ciao, Martin
Compiling a new kernel.
I'd like to go from 2.0.34 to 2.0.35. I'm moving kind of slow because I'm using the SuSE 3DLabs xserver for my Creative Labs video card, and I've heard that it won't work with kernels above 2.0.34 and I recently heard that it WILL work with 2.0.35. Anyway, I've had to recompile my existing kernel a few times, once for sound and once for the hell of it, so I've done that. But whenever I did it I just changed to '/usr/src/linux' and did the 'make menuconfig' config thing and configured the kernel. I've got the 2.0.35 source and I've got the updated kernel package for Debian but I noticed one thing: /usr/src/linux is linked to the 2.0.34 directory. Am I correct in assuming that I have to relink /usr/src/linux to the 2.0.35 directory? Anyway, can someone give me a down and dirty way to do this? Also, what are the steps for compiling a kernel? I've done it before, but I'm a scrap of paper kind of guy. I jot down notes as I go along from readmes, how-to's, whatever and at one time I had the process pretty well described on 3 or 4 scraps of paper, but I think I'm missing a scrap or two. I promise if someone gives me the steps, I'll organize them and keep 'em. Or better yet, is the process described pretty well in the readme's for the source (2.0.35)? I'm pretty sure that that's pretty close to the way that I did it the other 2 times, but the procedure is described differently in the kernel package readmes. TIA -- James R. Lunsford Email - mailto:[EMAIL PROTECTED] Home Page - http://www.comports.com/jrl007 ICQ - 2114258
Re: Thanks for the help with compiling a new kernel... BUT!
Thalia L. Hooker wrote: Hi, ... I tried the kernel package and compilation went fine except it didn't seem to compile any of the new modules even though I requested SCSI support, SCSI disk support, and the driver AHA152x. I say this because when I noticed it had not detected any scsi hosts, I searched for a file: aha152x.o and didn't see anything like it on the whole hard drive. ... CONFIG_SCSI_AHA152X=Y Looks like you've compiled it in, shouldn't it be M for module? It still doesn't explain why SCSI is not being detected, perhaps you need to add some command-line options? ... - If I add new RAM to my system, do I need to recompile the kernel for it to be able to use all of it? You may need to tell kernel mem=24M (see LILO docs) if your bios doesn't report 16 Mb. HTH -- Dimitri emaziuk at curtin dot edu dot au Entia non sunt multiplicanda praeter necessitatem -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Thanks for the help with compiling a new kernel... BUT!
I am still having a lot of trouble! I have no trouble running: 'make xconfig', followed by 'make kpkg-clean', 'make-kpkg -r custom.1.0 kernel_image' 'dpkg -i /usr/src/kernel-image-2.0.30_custom.1.0_i386.deb' 'make modules' 'make modules_install' to create a custom kernel boot disk. But I tried specifying mouse/serial/parallel support as modules, and during boot the system bitches that there are no modules/drivers available. Then the systems init xdm and the screen goes black (though the computer still responds to keyboard input.) If anyone has any additional help, I'd greatly appreciate it! Stan -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Thanks for the help with compiling a new kernel... BUT!
On Tue, 28 Oct 1997, Ignus Fast wrote: I am still having a lot of trouble! I have no trouble running: 'make xconfig', followed by 'make kpkg-clean', 'make-kpkg -r custom.1.0 kernel_image' 'dpkg -i /usr/src/kernel-image-2.0.30_custom.1.0_i386.deb' Ready?!? 'make modules' 'make modules_install' AFAIK, this is already done by the first 4 commands Just run 'dpkg -i /usr/src/kernel-image-2.0.30_custom.1.0_i386.deb' again to fix the modules. to create a custom kernel boot disk. But I tried specifying mouse/serial/parallel support as modules, and during boot the system bitches that there are no modules/drivers available. Then the systems init xdm and the screen goes black (though the computer still responds to keyboard input.) Do you mean that X doesn't start properly anymore since you built this new kernel? You can read back the startup messages with the `dmesg` command. Cheers, Joost -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Thanks for the help with compiling a new kernel... BUT!
You might have better success it you make your modules using make-kpkg rather than make-ing in the source directory. Try executing `make-kpkg --targets' to get a list of targets (which includes modules). Caveat - I haven't made modules with make-kpkg myself, so I don't know what pitfalls might await you along the way... I'm thinking of starting a Debian custom kernel faq that would address issues such as Ignus' questions and my own experience with pcmcia and the kernel. Any one interested in contributing? Regards, Mike -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Thanks for the help with compiling a new kernel... BUT!
Hi, Ignus == Ignus Fast [EMAIL PROTECTED] writes: Ignus I am still having a lot of trouble! I have no trouble running: Ignus 'make xconfig', followed by Ignus 'make kpkg-clean', This is the wrong order. Try make-kpkg clean, followed by make xconfig Ignus 'make-kpkg -r custom.1.0 kernel_image' Ignus 'dpkg -i /usr/src/kernel-image-2.0.30_custom.1.0_i386.deb' This makes and installs the modules for you. You are done. make-kpkg does more than you thought ;-) Ignus 'make modules' 'make modules_install' This may be your problem. You compiled and installed a kernel and modules, and you then clobbered modules with another set. *Don't* do this. The .deb file is all you need. Ignus to create a custom kernel boot disk. But I tried specifying Ignus mouse/serial/parallel support as modules, and during boot the Ignus system bitches that there are no modules/drivers available. Ignus Then the systems init xdm and the screen goes black (though the Ignus computer still responds to keyboard input.) Ignus If anyone has any additional help, I'd greatly appreciate it! Hope this helps manoj -- If anything can go wrong, it will. Edsel Murphy Manoj Srivastava [EMAIL PROTECTED] http://www.datasync.com/%7Esrivasta/ Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Thanks for the help with compiling a new kernel... BUT!
Hi, I finally got over my signal 11 problems and can now recompile a kernel. The problems seem to have been due to bad RAM. I reinitialized my partitions because I was afraid I would run into filesystem problems due to the numerous times my computer froze when I was trying to recompile a kernel. I tried the kernel package and compilation went fine except it didn't seem to compile any of the new modules even though I requested SCSI support, SCSI disk support, and the driver AHA152x. I say this because when I noticed it had not detected any scsi hosts, I searched for a file: aha152x.o and didn't see anything like it on the whole hard drive. Lines from my .config file: CONFIG_MODULES=Y #CONFIG_MOVERSIONS is not set CONFIG_KERNELD=y ... #CONFIG_MAX_16M=Y ... CONFIG_SCSI=Y CONFIG_BLK_DEV_SD=Y CONFIG_CHAR_DEV_ST=Y CONFIG_SCSI_AHA152X=Y The only thing I did different from below was that I used 'make config' Could this have anything to do w/answering NO to the question about modules version numbers? I am still having a lot of trouble! I have no trouble running: 'make xconfig', followed by 'make kpkg-clean', 'make-kpkg -r custom.1.0 kernel_image' 'dpkg -i /usr/src/kernel-image-2.0.30_custom.1.0_i386.deb Couple other newbie questions for you: - Since my kernel isn't working right yet, can I recompile a new kernel from the same directory I compiled the previous one? Do I need to change the numbering in custom.1.? - Since I didn't limit the kernel to 16 MB kernel-kpkg defaults to bZimage, should my new kernel detect 24 MB? My boot messages say 16MB after recompiling. - If I add new RAM to my system, do I need to recompile the kernel for it to be able to use all of it? LThanks, Thalia -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Thanks for the help with compiling a new kernel... BUT!
Hi, Mike == Mike Miller [EMAIL PROTECTED] writes: Mike I'm thinking of starting a Debian custom kernel faq that would Mike address issues such as Ignus' questions and my own experience Mike with pcmcia and the kernel. Any one interested in contributing? Please include the Problems file from kernel-package (I'm including a copy below, in case you don't have access to the latest version. You may be able to reuse bits of the kernel-package README (which I posted here recently) for the FAQ. And, if you feed the FAQ back to me, I'll include it in the next kernel-package package so the information is right there for people to use. Also, feel free to ask me any time you have a question about the compilation itself (others can help more than I can about indvidual drivers and patches). manoj -- Knowing that one is dear to oneself, one should guard oneself well. For one out of the three watches of the night a wise man should keep watch. 157 Manoj Srivastava [EMAIL PROTECTED] http://www.datasync.com/%7Esrivasta/ Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E == $Id: Problems,v 1.4 1997/10/22 07:47:57 srivasta Exp $ This is an (incomplete) list of problems that people have encountered in the past while using kernel-package. Please remember to configure the kernel for your machine using make menuconfig, and to clean the source tree before compiling a new image using make-kpkg clean. a) Failure to format a floppy disk while installing kernel image. fdformat from the obsolete package miscutils sometimes has problems formatting floppies. Install the package fdutils instead. Also, sometimes the new fdformat fails for /dev/fd0 unless the parameters are set using setfdprm -- in short, make sure that you can format floppies manually before asking the kernel image postinst to do so. b) warning, 'debian/tmp-image/DEBIAN/control' contains user-defined field 'Installed-Size'warning, dpkg-deb: unable to create '..': is a directory The problem is actually that the version of dpkg being used is too old for kernel-package; kernel-package version 3.X needs dpkg 1.4.0.0 at least. c) dpkg-gencontrol fails with the error message failure: chown new files list file: Illegal seek This is an error in older versions of dpkg-dev. Upgrading to version 1.4.0.9 should help. d) install: debian/changelog: No such file or directory Remember to do make-kpkg clean with a patched/old kernel source tree (if this is not a clean tree, clean it first). e) make-kpkg goes into an infinite loop when trying to make oldconfig. This is _not_ a bug with kernel-package, it is a well known incompatibility between the new version of expr (which has suddenly become POSIX compliant) and the kernel sources (which did not expect expr to behave this way). The fix is to apply the following patch to the kernel sources. --- scripts/Configure.dist Mon Jan 20 14:43:24 1997 +++ scripts/Configure Tue Jan 21 05:41:30 1997 @@ -288,7 +288,7 @@ def=${old:-$3} while :; do readln $1 ($2) [$def] $def $old - if expr $ans : '0$\|-?[1-9][0-9]*$' /dev/null; then + if expr $ans : '0$\|-\?[1-9][0-9]*$' /dev/null; then define_int $2 $ans break else @@ -319,7 +319,7 @@ while :; do readln $1 ($2) [$def] $def $old ans=${ans#*[x,X]} -if expr $ans : '[0-9a-fA-F]+$' /dev/null; then +if expr $ans : '[0-9a-fA-F]\+$' /dev/null; then define_hex $2 $ans break else f) % depmod -a modprobe: error reading ELF header: No such file or directory This is a problem with the newly changed behavior of depmod and friends, who suddenly stopped liking non .o files in /lib/modules/version. Newer versions of kernel package, like this one, handle that right. For older image packages, the test is: % find /lib/modules/2.0.30/ -type f -exec file {} \; | grep -v 'ELF 32-bit' /lib/modules/2.0.30/modules.dep: ASCII text Anything other than modules.dep showing up is something that depmod can no longer tolerate. Remove those files, and things should be fine. g) dpkg-gencontrol fails if LC_ALL is set to de_DE. Actually, I can't really confirm this works for anything except en_US. This is true as of dpkg/dpkg-dev 1.4.0.19. The error actually reported seems to be either that debian/substvars file can not be found, or that the changelog file is empty (eben when the file is not empty). The work around seems to be to set the LC_ALL to en_US while compiling the kernel. This should be fixed soon. h) dpkg upgrades the custom kernel to a new standard kernel. This means that epochs were used;
Re: Thanks for the help with compiling a new kernel... BUT!
Hi, Thalia == Thalia L Hooker [EMAIL PROTECTED] writes: Thalia Hi, I finally got over my signal 11 problems and can now Thalia recompile a kernel. The problems seem to have been due to bad Thalia RAM. I reinitialized my partitions because I was afraid I Thalia would run into filesystem problems due to the numerous times Thalia my computer froze when I was trying to recompile a kernel. Thalia I tried the kernel package and compilation went fine except it Thalia didn't seem to compile any of the new modules even though I Thalia requested SCSI support, SCSI disk support, and the driver Thalia AHA152x. I say this because when I noticed it had not detected Thalia any scsi hosts, I searched for a file: aha152x.o and didn't Thalia see anything like it on the whole hard drive. The modules should have been installed under /lib/modules/2.0.30. (you did install the new kernel using dpkg -i, right?) If you wish to test whether the modules are in the new kernel without installing it, you can try: % dpkg -c kernel-image-2.0.30_custom.1.0_i286.deb and look if there is anything in /lib/modules. Another thing to look for, after installation of the new kernel, is if depmod -a (run as root) succeeds. Thalia The only thing I did different from below was that I used Thalia 'make config' That should be fine. Thalia Could this have anything to do w/answering NO to the question Thalia about modules version numbers? Hmm. This _should not_ make a difference, but I have not answered no for about a year now, and my meory is going. Thalia Couple other newbie questions for you: Since my kernel isn't Thalia working right yet, can I recompile a new kernel from the same Thalia directory I compiled the previous one? Do I need to change the Thalia numbering in custom.1.? Remember to do a make-kpkg clean (not a make distclean, that wipes your config file) before you do anything. And yes, do bump up the number every time you put in a new kernel. However, people have reported a problem upgrading the same kernel version as they are running. See, suppose you are running 2.0.30, and you have module xyz.o in there. You reconfigure the kernel (using make config (or xconfig or menuconfig), and instead of module xyz.o, you choose abc.o. Possibly because you have choosen to have xyz.o as a builtin, rather than a module. When you install the new version of the kernel-image, dpkg shall remove xyz.o and add abc.o. Now, if you need xyz.o (which your running kernel thinks is there as a module), there shall be a problem, since your new kernel knows better than to try to load xyz.o, your old kernel was used to doing so, and it does not like you siping the module from under it's fingers. It gets worse. The new modules may not load into the old kernel (I think they should, they just don't seem to want to). Your old kernel may no longer be very stable once you install the new image. So it is a good idea to always reboot imeediately if you have installed a kernel with the same version as the one you were running. Me, I keep a few different versions around, and I boot to 2.1.X, purge kernel-image-2.0.30, install the new 2.0.30, and reboot into 2.0.30. Cumbersome, but I've never had a crash. (this may also be overkill). Thalia - Since I didn't limit the kernel to 16 MB kernel-kpkg Thalia defaults to bZimage, should my new kernel detect 24 MB? My Thalia boot messages say 16MB after recompiling. I think so. My kernel correctly reports my 96MB. Thalia - If I add new RAM to my system, do I need to recompile the Thalia kernel for it to be able to use all of it? You do not need to recompile the kernel. Your BIOS, on the other hand, may need to be changed, and make sure your BIOS does report the new memory (or else reseat the RAM and try again). I hope this helps. manoj -- Where shall I begin, please your Majesty? he asked. Begin at the beginning, the King said, gravely, and go on till you come to the end: then stop. _Alice's Adventures in Wonderland_, Lewis Carroll Manoj Srivastava [EMAIL PROTECTED] http://www.datasync.com/%7Esrivasta/ Key C7261095 fingerprint = CB D9 F4 12 68 07 E4 05 CC 2D 27 12 1D F5 E8 6E -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Compiling a new kernel
In the instructions that came with the kernel source package, Linus recommends making symlinks in the include directory to directories in the usr/src/linux/arch directory. I also remember reading that libc5-dev contains a set of headers. Someone on the list recommended that the directories in the include directory shoudln't be deleted and made into symlinks. Is this correct? Also is glibc the same as libc6? if so why? Bg -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .
Re: Compiling a new kernel
In the instructions that came with the kernel source package, Linus recommends making symlinks in the include directory to directories in the usr/src/linux/arch directory. I also remember reading that libc5-dev contains a set of headers. Someone on the list recommended that the directories in the include directory shoudln't be deleted and made into symlinks. Is this correct? Yes, it's correct. But it's got very little (actually nothing) to do with compiling the kernel (the symlinks you are talking about is so that user-programmes can see parts of the kernel, something that nowadays isn't very usefull any more. That's why Debian libc5, and anyone's glibc(=libc6) do it differently). Also is glibc the same as libc6? Yes. if so why? Basically, because glibc comes after libc5. The soname of libc5 is, you guessed it, 5. Because gnu libc is a major step forward, the soname had to be changed (increased). This leaves us with a gnu libc that has a soname 6, and thus we (and the rest of the linux world) call it libc6. -- joost witteveen, [EMAIL PROTECTED] #!/usr/bin/perl -sp0777iX+d*lMLa^*lN%0]dsXx++lMlN/dsM0j]dsj $/=unpack('H*',$_);$_=`echo 16dio\U$kSK$/SM$n\EsN0p[lN*1 lK[d2%Sa2/d0$^Ixp|dc`;s/\W//g;$_=pack('H*',/((..)*)$/) #what's this? see http://www.dcs.ex.ac.uk/~aba/rsa/ -- TO UNSUBSCRIBE FROM THIS MAILING LIST: e-mail the word unsubscribe to [EMAIL PROTECTED] . Trouble? e-mail to [EMAIL PROTECTED] .