FAQ for isapnptools $Author: fox $ $Date: 2001/04/11 21:46:28 $ Answers to Frequently Asked Questions about isapnptools. $Revision: 0.17 $ ______________________________________________________________________ Table of Contents 1. General 1.1 Where can I get the latest FAQ ? 1.2 Support contacts 1.3 What do isapnptools do ? 1.4 How do I use isapnptools to configure my ISA PnP cards ? 1.5 Who wrote/maintains isapnptools ? 1.6 What is the latest version, and where can I get it from ? 1.7 What version do I have ? 1.8 What about security ? 1.9 Can isapnptools be used to configure PCI hardware. 1.10 Is there a version of isapnptools for PCI ? 1.11 NT,WIN95,... works ok now, will running isapnp screw it up ? 1.12 Isn't there a better way to configure PnP cards ? 1.13 How do I find out if my card is supported ? 1.14 Where can I find out more ? 1.15 Is there a mailing list for isapnptools ? 1.16 My card works, but it's not in the Hardware-HOWTO (appendix C). 2. General - Fundamentals 2.1 What is a driver (module) ? 2.2 What is a logical device ? 3. General - Compilation/installation 3.1 What configuration options are there when compiling ? 3.2 I get lots of link errors 3.3 Why does make give a "segmentation violation" ? 3.4 I've got the binary distribution and the binaries won't run. 3.4.1 Linux 3.4.2 DOS 3.5 I get warnings about implicit declaration of snprintf when compiling. 3.6 I get './libisapnp.a: could not read symbols: ... add one' when cross compiling. 3.7 Problems compiling for 2.1.x kernels 3.8 I've upgraded my library/C-compiler and now have problems compiling the programs. 3.9 Problems compiling for 2.2.11 kernels 3.10 Can isapnptools be used with operating systems other than Linux. 3.11 What do I do if I think I've found a bug or got some other problem ? 3.12 How do I make a statically linked binary ? 3.13 How can I make a non-native binary ? 4. General - Running 4.1 Which kernel do I need ? 4.2 I get "Permission denied" when trying to run the programs. 4.3 The program crashes with a "Segmentation fault" 4.4 Where can I get example configuration files ? 4.5 How can I find out how the card is configured ? 4.6 Isapnp fails with error "Error occurred executing request 'ISOLATE' on or around line 13 --- further action aborted" 4.7 What is a good readport address to use ? 4.8 Pnpdump doesn't find all the PnP boards in the system. 4.9 Pnpdump doesn't find a particular board in the system. 4.10 Pnpdump messes up with more than one PnP board. 4.11 Pnpdump cannot see the boards after using the -r option. 4.12 Pnpdump reports "No boards found" 4.13 Pnpdump reports "Ident Byte x, (..) differs from resource data (..)" 4.14 Pnpdump reports "Unknown tag .., Resource data dump aborted" 4.15 What does "Don't know what to do with ... on or around line x" mean ? 4.16 What does "Error occurred executing request 'LD 2' on or around line..." mean ? 4.17 What does "End tag checksum 0xb8 (BAD)" mean (in the pnpdump output) ? 4.18 What does "IO range check attempted while device activated" mean (from isapnp) ? 4.19 What does "IO range check failed for n bytes of IO at addr" mean (from isapnp) ? 4.20 I think I've configured the board correctly, but the driver still can't find it. 4.21 What does "/proc/bus/pci/devices not found, so PCI resource conflict not checked" mean ? 4.22 What causes lots of errors running the script generated by pnpdump -s ? 4.23 Do I have to reboot after running isapnp or pnpdump ? 5. General - Kernel/Driver issues 5.1 Do I need to configure CONFIG_PNP in the kernel ? 5.2 I've configured my ... using isapnptools, but the kernel doesn't recognise it. 5.3 I get the error message "... module support already loaded" when trying to load the ... module. 5.4 I'm having trouble getting interrupts to work. 5.5 I've configured the card, but /proc/interrupts still shows the interrupt as unused. 5.6 The driver for my card cannot be built as a module. 5.7 As soon as I run isapnp, my machine locks up. 5.8 How can I stop isapnp and pnpdump using certain addresses for the READPORT. 6. Upgrading 6.1 I've upgraded isapnptools from my favourite distribution and it has wiped out my isapnp.conf file. 6.2 What should I do to upgrade isapnptools. 6.3 I've upgraded isapnptools, and now isapnp gives errors. 6.4 I've upgraded my motherboard, and now isapnp can't find my boards. 6.5 I've upgraded my kernel to 2.4.x, now isapnp can't initialise my Sondblaster AWE. 7. Resources 7.1 Determining resource usage (linux). 7.2 I've configured the card but the /proc entries haven't changed. 7.3 I've got a ..., which runs fine under Win95, how can I find out the port settings ? 7.4 Examining Win95 resource settings. 7.5 The windows CTPNP.CFG file. 7.6 How can I prevent pnpdump allocating certain resources with the -c flag. 7.7 Common problems 7.8 What does isapnp mean by "Fatal - resource conflict allocating DMA1". 7.9 Do I have to configure all the devices. 7.10 How can I reconfigure a device. 8. Advanced usage. 8.1 Common configurations on multiple machines, Version 1.11 and earlier. 8.2 Common configurations on multiple machines, Version 1.12 and later. 8.3 Sending a RESET CSN command. 8.4 Dumping the boards configuration registers. 8.5 Displaying a list of isapnp boards in the system. 9. Soundcards. 9.1 I've got an AWE32/AWE64, but Linux doesn't recognise it. 9.1.1 The output from pnpdump (before configuration). 9.1.2 The editted configuration file for isapnp. 9.1.3 An extract of the bootup script to configure sound. 9.2 Problems with the Aztech Sound Galaxy 16 III-3D PnP sound card. 9.3 CD sound doesn't work (one channel or no channels). 9.4 What to do about "sb: Interrupt test on IRQ10 failed - device disabled." 9.5 Soundblaster site. 9.6 Sound works if I boot straight into Linux, but not if booting from Win98. 10. Modems. 10.1 See also. 10.2 I've got a WinModem, and I've followed all the instructions, but it still doesn't work. 10.3 I've got a PCI modem - what about that ? 10.4 I've got a modem that works in DOS and Windows, but not Linux. 10.5 I've configured the device, but it is really SLOW. 10.6 I've configured the device, but it doesn't work. 10.7 I've configured the device, it seems to send ok, but not receive. 10.8 What's this about not using /dev/cua1 ? 10.9 How can I check the modem is working ? 10.10 Thunderlink modems (PCT3880). 10.11 USR Robotics Sportster. 10.12 General help on setting up internet dial up link. 11. Network cards. 11.1 The card supports non PnP configuration. 12. IDE ports. 13. PCMCIA cards. 14. Disclaimers etc. ______________________________________________________________________ 11.. GGeenneerraall 11..11.. WWhheerree ccaann II ggeett tthhee llaatteesstt FFAAQQ ?? The latest FAQ is available from the isapnptools home websitein various forms: isapnpfaq as a text file (this is the form it is included as in the distribution), iaspnpfaq as a set of html files , or the iaspnpfaq as sgml source . The text version is designed to be piped through less to give highlighting etc. If you want to strip out the control characters, pipe it through sed thus: cat isapnpfaq.txt | sed 's/^H.//g' > isapnpfaq.clean.txt Where ^H above is control-H. (You may need to prefix it with ^V when entering it on the command line). To see if the one on the website is newer, compare the revision with the revision of this document. You may also find the Plug-and-Play-HOWTO useful. 11..22.. SSuuppppoorrtt ccoonnttaaccttss See ``Disclaimers''. 11..33.. WWhhaatt ddoo iissaappnnppttoooollss ddoo ?? Isapnptools allow you to configure the ISA PnP cards complying with the Plug and Play ISA Specification Version 1.0a, which may be obtained from ftp://ftp.microsoft.com/developr/drg/Plug-and- Play/Pnpspecs/ . These boards use registers instead of jumpers to set the board address, interrupt assignments etc (resource allocations). In addition, the board contains a description of the resources that need to be allocated, with limits on the settings allowable. In other words, isapnptools allow you to examine and change 'virtual jumpers' on the board. Note that you will need to know what other hardware is in your PC and what their 'jumper' settings are, including any devices built into the motherboard. The BIOS setup screens will help with this. The BIOS (or any other PnP configuration program such as Win95 or isapnptools) uses a special protocol described in the specification to discover all the PnP boards, and allocate the resources so that there are no conflicts. Because the BIOS rarely does the job properly, it is better to get it to do as little as possible, and use isapnptools to configure the resource settings. This is not always possible with newer BIOSes, so see also ``Pnpdump reports "No boards found"''. 11..44.. HHooww ddoo II uussee iissaappnnppttoooollss ttoo ccoonnffiigguurree mmyy IISSAA PPnnPP ccaarrddss ?? See the instructions in the INSTALL file for the overall process. See the various README files for examples and specific details. See the FAQ too, particularly for sound cards. The basic idea is you run pnpdump to scan your PnP boards and produce a skeleton configuration file. You then uncomment lines as required to setup the boards. Uncommenting lines has the same function as moving jumpers on a non- PnP card. Generally, for each logical device you will want to uncomment one each of the resources, and maybe change the value for the resource within the range of the comments immediately above, such that when you've finished there are no conflicts. You then run isapnp with the configuration file on each boot, and it will configure the boards like you asked. This is done by editing the boot up scripts to add the appropriate command. You also need to arrange to load the appropriate kernel driver modules, after configuring the boards, to allow programs to access the boards. In some cases the driver may already be in the kernel and you can just run a program to tell it about the new hardware. Again, this is done by adding the appropriate commands to your system boot up scripts. If you don't know how to recompile the kernel to make drivers modules rather than built-in etc, refer to the Kernel-HOWTO . In the event of difficulties, see the FAQ first for answers. 11..55.. WWhhoo wwrroottee//mmaaiinnttaaiinnss iissaappnnppttoooollss ?? Peter Fox, with some help and suggestions from many others. See the README . Contributors with useful suggestions and improvements may have their name and email address added to the README, if you don't want this to happen, please tell me. 11..66.. WWhhaatt iiss tthhee llaatteesstt vveerrssiioonn,, aanndd wwhheerree ccaann II ggeett iitt ffrroomm ?? See the isapnptools home site 11..77.. WWhhaatt vveerrssiioonn ddoo II hhaavvee ?? For isapnptools release 1.19 and later, simply run pnpdump --version, the isapnptools release will be given. For earlier versions, you can run strings `which pnpdump` | grep '\$Id' which should give an output like $Id: isapnpfaq.sgml,v 0.17 2001/04/11 21:46:28 fox Exp $ $Id: isapnpfaq.sgml,v 0.17 2001/04/11 21:46:28 fox Exp $ $Id: isapnpfaq.sgml,v 0.17 2001/04/11 21:46:28 fox Exp $ The isapnptools version will be the same as the pnpdump.c,v version for versions 1.14 and later, and one greater earlier than that. You can compare the version number against the CHANGES file (NEWS file since 1.23). 11..88.. WWhhaatt aabboouutt sseeccuurriittyy ?? Isapnptools must be run as root, and should not be made setuid root. The configuration files used, such as /etc/isapnp.conf and /etc/isapnp.gone should be owned by, and writeable only by, root. This should minimize any interest in buffer overflow problems, as you have to be root in the first place. In an attempt to prevent trojans being introduced into the isapnptools programs, I maintain md5sums of all the files on my web site, see md5.txt If you suspect the md5 sums have been tampered with, please tell me, and I can check against the originals. md5sum is part of the GNU textutils package. If you are worried about security, I suggest you maintain isapnptools my compiling from source. After doing an exhaustive check for trojans, you can then simply produce diffs of the new version against the old and check that. 11..99.. CCaann iissaappnnppttoooollss bbee uusseedd ttoo ccoonnffiigguurree PPCCII hhaarrddwwaarree.. No. 11..1100.. IIss tthheerree aa vveerrssiioonn ooff iissaappnnppttoooollss ffoorr PPCCII ?? The nearest equivalent is pciutils. Have a look at freshmeat; see http://freshmeat.net/news/1999/01/28/917541046.html for instance. 11..1111.. NNTT,,WWIINN9955,,...... wwoorrkkss ookk nnooww,, wwiillll rruunnnniinngg iissaappnnpp ssccrreeww iitt uupp ?? No. isapnp only configures the cards for Linux. When you reboot, the BIOS will restore the card's settings. Note that Win95 will carry out the function of the BIOS (in terms of locating and configuring PnP boards) on PCs without a PnP BIOS. Having said that, you must be careful using the REG and POKE keywords that you don't write to some register that corrupts the board resource data. (The resource data is often in EEPROM, and can be written to by suitable software to initialise it). See also ``What does "End tag checksum 0xb8 (BAD)" mean (in the pnpdump output) ?''. 11..1122.. IIssnn''tt tthheerree aa bbeetttteerr wwaayy ttoo ccoonnffiigguurree PPnnPP ccaarrddss ?? Well, I quite like isapnptools as it is predictable, and doesn't make the kernel any bigger (but then, I may be slightly biased). However, you may want to check out a configuration manager for linux which now supports 2.0.30: CM for linux . You also might want to look at Tom Lees' kernel PnP suite at , the files are in . Work is underway to integrate PnP support into the kernel. See http://www.cs.utwente.nl/~vanrein/vha/ . As this involves changing all the drivers, it may take a while. It now looks like linux 2.4.x will support ISA PnP in the kernel if you want to use it. Even now (2.2.14), some drivers do their own PnP support, particularly the ALSA sound drivers. A /proc based interface has been developed for v2.2 kernels, see isapnp-2.2-V0.1.tgz . 11..1133.. HHooww ddoo II ffiinndd oouutt iiff mmyy ccaarrdd iiss ssuuppppoorrtteedd ?? Note that the following are ideas. Linux is a fast moving world, and due to the age of this document, likely to be out of date. New hardware is gaining support all the time. See the Hardware-HOWTO. Check your linux distribution's hardware compatibility list. For example, Redhat 5.2's is at Redhat 5.2 support . Refer to the card vendor and manufacturer for Linux support information. If they won't support linux, tell your friends, and tell them you're telling your friends. Assuming you specified that the card was for use with Linux, you should be able to change it without too much trouble (having been sold a product unfit for purpose etc). See also ``Where can I find out more ?'' See also the device specific sections of this FAQ below. 11..1144.. WWhheerree ccaann II ffiinndd oouutt mmoorree ?? A good place to find out more about PC hardware in general is the pc- hardware-faq, regularly posted to the newsgroups: comp.sys.ibm.pc.hardware.video, comp.sys.ibm.pc.hardware.comm, comp.sys.ibm.pc.hardware.storage, comp.sys.ibm.pc.hardware.cd-rom, comp.sys.ibm.pc.hardware.systems, comp.sys.ibm.pc.hardware.networking, comp.sys.ibm.pc.hardware.chips, comp.sys.ibm.pc.hardware.misc, comp.answers, news.answers. A good place to find out about Linux configuration etc are the HOWTOs and mini-HOWTOS. These can be found on the MetaLab HOWTO website or from the TSX-11 HOWTO ftpsite or try http://www.linux- howto.com/ . These are also normally supplied with standard Linux distributions. A more practical guide to using these tools (and fixing other configuration problems) can be found on the SuSe Linux support database. See /usr/doc/support-db/sdb_e/rb_isapnp.html if you've installed their database, or http://www.suse.de/ . See if you can find any information/statements on the vendor websites. For example the soundblaster linux website . For PnP DOS utilities etc, see the phoenix utilites website . Craig Hart's website also contains useful pnp stuff, including a list of pnp ids. 11..1155.. IIss tthheerree aa mmaaiilliinngg lliisstt ffoorr iissaappnnppttoooollss ?? There used to be a mailing list for all pnp issues, including isapnptools, at pnp-list@redhat.com, but it shut down on 3-Nov-1999. So the answer is not that I am aware of. If you would like to receive release announcements for isapnptools by email (releases are normally announced to comp.os.linux.announce anyway), then send an email to isapnp-announce@roestock.demon.co.uk, wwiitthh tthhee ssuubbjjeecctt ""ssuubbssccrriibbee iissaappnnpp--aannnnoouunnccee"", giving the details of your email address. 11..1166.. MMyy ccaarrdd wwoorrkkss,, bbuutt iitt''ss nnoott iinn tthhee HHaarrddwwaarree--HHOOWWTTOO ((aappppeennddiixx CC)).. Tell the Hardware-HOWTO maintainer ! 22.. GGeenneerraall -- FFuunnddaammeennttaallss 22..11.. WWhhaatt iiss aa ddrriivveerr ((mmoodduullee)) ?? If you don't know what this is, I suggest you start by reading the Kernel-HOWTO available from the MetaLab website Kernel-HOWTO or from the TSX-11 ftpsite Kernel-HOWTO , or from their respective mirrors. See also www.linux-howto.com . 22..22.. WWhhaatt iiss aa llooggiiccaall ddeevviiccee ?? The functionality of a card is split into many different logical devices for convenience. The idea is that each logical device is independent and has it's own driver. This way, the user can decide to use as much or as little of the card functionality as required, and also allows common drivers for many different boards (the drivers being written for each logical device function). So for example a soundcard will typically have at least four different logical devices: 1. The basic soundcard 2. Soundcard extensions such as wave table 3. A game port 4. A CD-rom port. The logical devices are numbered starting from 0, incrementing by 1. The first entry in the resource data applies to logical device 0, the next to logical device 1, and so on. Each logical device has its own set of resources ( IO addresses, INT errupts, DMA channels etc) to program, and isapnp uses the LD command to select the logical device to be configured. 33.. GGeenneerraall -- CCoommppiillaattiioonn//iinnssttaallllaattiioonn 33..11.. WWhhaatt ccoonnffiigguurraattiioonn ooppttiioonnss aarree tthheerree wwhheenn ccoommppiilliinngg ?? Use the --help option to configure to get the latest list. 33..22.. II ggeett lloottss ooff lliinnkk eerrrroorrss Linking gives loads of errors such as pnpdump.o: In function `dumpdata': pnpdump.o(.text+0x11b0): undefined reference to `__outbc' pnpdump.o(.text+0x11c4): undefined reference to `__outb' pnpdump.o(.text+0x11db): undefined reference to `__outbc' pnpdump.o(.text+0x11ef): undefined reference to `__outb' pnpdump.o(.text+0x1374): undefined reference to `__outbc' pnpdump.o(.text+0x1388): undefined reference to `__outb' Chances are the -O2 compiler option is missing. 33..33.. WWhhyy ddooeess mmaakkee ggiivvee aa ""sseeggmmeennttaattiioonn vviioollaattiioonn"" ?? Your make program is not recent enough. I've seen make 3.74 do this. Type make -v to find out what version you have. Make 3.76.1 seems ok. 33..44.. II''vvee ggoott tthhee bbiinnaarryy ddiissttrriibbuuttiioonn aanndd tthhee bbiinnaarriieess wwoonn''tt rruunn.. 33..44..11.. LLiinnuuxx If you get messages such as: bash: ./sbin/pnpdump: No such file or directory or bash: ./isapnp: No such file or directory. when attempting to run the precompiled binaries, then chances are you haven't got the required C libraries. To check this, change to the sbin directory and type ldd pnpdump ldd isapnp You should get something like libc.so.5 => /lib/libc.so.5 for each one. If you get "not found" then you will have to recompile. Note that the precompiled binaries in my binary distributions are compiled for libc5. If you want binaries for glibc systems, Redhat 5.2 on, Suse 6.0 on etc, you'll have to go elsewhere (probably to your distribution's web site). The best thing to do is to delete the binaries in the distribution and try to compile your own anyway. 33..44..22.. DDOOSS If you get a complaint along the lines of cannot find DPMI, then you need to get hold of a DPMI host software. See the FAQ on the DJGPP site for details (follow the path through Running --- Distributing). This is most likely to occur on a pure DOS machine. Win95 for example includes a DPMI host. The free DPMI server (cwsdpmi) is normally part of the djgpp distribution. For example, my ISP mirrors it in the and directories. 33..55.. II ggeett wwaarrnniinnggss aabboouutt iimmpplliicciitt ddeeccllaarraattiioonn ooff ssnnpprriinnttff wwhheenn ccoomm-- ppiilliinngg.. You need to edit the makefile so that -DNEEDSNPRINTF is included in PNPFLAGS. In isapnptools-1.23 on, using configure, this #define has been changed to HAVE_SNPRINTF and has the opposite sense. The configure script should detect the setting automatically when cross compiling. 33..66.. II ggeett ''..//lliibbiissaappnnpp..aa:: ccoouulldd nnoott rreeaadd ssyymmbboollss:: ...... aadddd oonnee'' wwhheenn ccrroossss ccoommppiilliinngg.. You need to set the GNUTARGET environment variable then rebuild: export GNUTARGET=coff-i386 make clean make CC=gcc-dos See also ``How can I make a non-native binary ?''. 33..77.. PPrroobblleemmss ccoommppiilliinngg ffoorr 22..11..xx kkeerrnneellss The latest kernels (2.1.7x ish) have problems with the header files. If you are playing with these kernels, you really ought to know how to fix this yourself. I've been told that adding a -D__KERNEL__ to the Makefile fixes it. 33..88.. II''vvee uuppggrraaddeedd mmyy lliibbrraarryy//CC--ccoommppiilleerr aanndd nnooww hhaavvee pprroobblleemmss ccoomm-- ppiilliinngg tthhee pprrooggrraammss.. If you've change from libc5 to glibc, or want to use EGCS, you could look at http://huizen.dds.nl/~frodol/glibc/problems.html for some help. See also the ``The program crashes with a "Segmentation fault"'' section. 33..99.. PPrroobblleemmss ccoommppiilliinngg ffoorr 22..22..1111 kkeerrnneellss There have been reported some problems getting isapnptools 1.18 to compile with kernel 2.2.11. I quote Juergen Rose: I still played a little bit with your and with the linux sources. If I change in pnpdump.c and isapnp.y the order of # include # include to # include # include isapnptools-1.18 compiles, but with warnings. The warnings are due to the facts, that: - /usr/include/asm/processor.h is included in /usr/include/linux/sched.h at line 74 and needs "struct task_struct" - "struct task_struct" is defined in /usr/include/linux/sched.h at line 218 So I moved the line in /usr/include/linux/sched.h #include /usr/include/asm/processor.h below the definition of "struct task_struct". And could compile isapnptools-1.18 even without warnings. I thought this looked like kernel include file failings. 33..1100.. CCaann iissaappnnppttoooollss bbee uusseedd wwiitthh ooppeerraattiinngg ssyysstteemmss ootthheerr tthhaann LLiinnuuxx.. No, or at least not easily. Issues that require addressing include 1. Isapnptools needs to be able to get direct IO access privileges. (Not a problem with DOS, hence the DOS port). I suspect that this will require a device driver of some sort under windows NT. 2. Some mechanism is needed to start up the device driver after isapnp has run. 33..1111.. WWhhaatt ddoo II ddoo iiff II tthhiinnkk II''vvee ffoouunndd aa bbuugg oorr ggoott ssoommee ootthheerr pprroobblleemm ?? Check the web site to see if the bug has already been reported. Check the FAQ to see if the problem is operator error. Check all the READMEs and INSTALL file to see if the answer is there. If the program doesn't seem to obey the documentation, please look at the source to identify the cause of the discrepancy, and report the problem. If nothing shows in the above, please report the problem, including the following information as appropriate: +o Output of 'cat /proc/cpuinfo' +o Output of 'cat /proc/pci' +o Output of 'cat /proc/interrupts' +o Output of 'cat /proc/ioports' +o Output of 'cat /proc/devices' +o Output of 'cat /etc/isapnp.gone' +o Your configuration file (/etc/isapnp.conf), before and/or after editting. +o Output of 'cat /usr/include/kernel/autoconf.h' (or the autoconf.h for the kernel you're running). +o Output of 'lsmod' if you use kernel modules +o The command lines used to install the relevant device driver kernel modules, or configure the built in device driver (eg setserial). +o Output of 'dmesg' +o Output of 'gcc -v' +o Output of 'ldd prog' where prog is the program giving trouble. +o Inputs and Outputs of the program giving trouble: prog args.. > prog.stdout 2> prog.stderr +o The patch that cures the problem ! Note that if you provide a useful contribution to isapnptools as a result of you report, you will automatically get added to the announcements mailing list (see ``Is there a mailing list for isapnptools ?''), unless you request otherwise. Also your name may get added to the README list of credits. 33..1122.. HHooww ddoo II mmaakkee aa ssttaattiiccaallllyy lliinnkkeedd bbiinnaarryy ?? A statically linked binary should work on many different linux systems with the same processor type. This is because all the standard library functions required have been extracted and included in the binary executable. The disadvantage is that the binaries are a lot bigger. Add the -static flag to LDFLAGS in the makefile. This could be done for instance by adding the following line to the Makefile: LDFLAGS += -static or my setting it when running make: make LDFLAGS+="-static" 33..1133.. HHooww ccaann II mmaakkee aa nnoonn--nnaattiivvee bbiinnaarryy ?? Isapnptools-1.23 and later use autoconf for configuring the programs for building. This makes it very easy to build programs for machines other than the one you are building on. For example, on my machine I have got a djgpp cross-compiler set up so that I can make DOS executables. In this case, I use the following commands: CC="gcc -b i386-go32-msdos" configure --disable-realtime --disable-proc --disable-pci make GNUTARGET=coff-i386 I also have an old machine which still runs a.out binaries. In that case I use: CC="gcc -b i486-linuxaout" configure --disable-pci --prefix=/ make 44.. GGeenneerraall -- RRuunnnniinngg 44..11.. WWhhiicchh kkeerrnneell ddoo II nneeeedd ?? Any kernel should do. The real time stuff may not exist in older kernels (1.2.13 and before), but just compile without it. If you have new hardware, you may need a later kernel anyway. Note that more and more drivers in newer kernels carry out their own PnP activity, so some ISA PnP cards may not even require isapnptools to configure, though isapnptools can still be useful for diagnosing things. Version 2.4.x of the kernel will basically have the functionality of isapnptools built in. 44..22.. II ggeett ""PPeerrmmiissssiioonn ddeenniieedd"" wwhheenn ttrryyiinngg ttoo rruunn tthhee pprrooggrraammss.. You must be root to run isapnp and pnpdump because the program requires direct access to the IO ports. 44..33.. TThhee pprrooggrraamm ccrraasshheess wwiitthh aa ""SSeeggmmeennttaattiioonn ffaauulltt"" If you compiled the program using gcc2.8.0 this is very likely due to a bug in the compiler's IO port handling. Recompile using gcc2.7.2. If you really must use gcc.2.8.x (like egcs) then you can try recompiling the KERNEL with the ioport.c from one of the later 2.1.xxx kernels, or 2.2.x kernels. Another cause would appear to be using a binary from one linux distribution on another. In this case (and as I recommend everywhere), simply download the latest version and compile it yourself. It really isn't that difficult. See also ``Making a static binary''. 44..44.. WWhheerree ccaann II ggeett eexxaammppllee ccoonnffiigguurraattiioonn ffiilleess ?? The output of pnpdump should give you the examples you need for your hardware. If you run it with the -c (--config) flag, it should even show how to configure your hardware in the output. The output of pnpdump is supposed to be used as the input to isapnp after editing. See also README.modules and INSTALL . 44..55.. HHooww ccaann II ffiinndd oouutt hhooww tthhee ccaarrdd iiss ccoonnffiigguurreedd ?? Use the -di options with pnpdump . The output dumps all the configuration registers for each device, you will have to interpret the results against the PnP spec, but it's mostly obvious. If you do this before running isapnp you can see the BIOS/power up defaults. Note that unused registers generally return all zeros. Unused DMA slots should return 4, though some devices erroneously return 0 (which is a valid used DMA channel). 44..66.. IIssaappnnpp ffaaiillss wwiitthh eerrrroorr ""EErrrroorr ooccccuurrrreedd eexxeeccuuttiinngg rreeqquueesstt ''IISSOO-- LLAATTEE'' oonn oorr aarroouunndd lliinnee 1133 ------ ffuurrtthheerr aaccttiioonn aabboorrtteedd"" This means no boards were found. If you have just added a new card into an ISA bus slot, it may be conflicting with the readport address in your /etc/isapnp.conf file. You will need to check the resource assignments in your configuration file to make sure they don't clash with the new card. Try re-running pnpdump to see if it comes up with a different readport address, then use the new readport address in you configuration file. If the new card is also PnP, you may also need to copy the configuration instructions across too. Similarly, if you have changed your motherboard, it may have something clashing with your readport address. See also ``Pnpdump reports "No boards found"'' 44..77.. WWhhaatt iiss aa ggoooodd rreeaaddppoorrtt aaddddrreessss ttoo uussee ?? A popular READPORT address seems to be 0x273 or 0x277. You shouldn't need to explicitly specify one though as pnpdump can find one automatically. 44..88.. PPnnppdduummpp ddooeessnn''tt ffiinndd aallll tthhee PPnnPP bbooaarrddss iinn tthhee ssyysstteemm.. Are you sure the missing board is configured as PnP ? Many PnP boards come with a DOS utility which allows them to be configured as non PnP, or has jumpers which can be used instead of PnP. See also ``Pnpdump reports "No boards found"'' 44..99.. PPnnppdduummpp ddooeessnn''tt ffiinndd aa ppaarrttiiccuullaarr bbooaarrdd iinn tthhee ssyysstteemm.. Are you sure the board is ISA PnP ? If the board is PCI, it won't be found, neither will it be found if it is not PnP, which could well be the case for older boards. Note that for some cards with similar names, one could be ISA PnP, and one not. See also ``Pnpdump doesn't find all the PnP boards in the system.'' See also ``Pnpdump reports "No boards found"'' 44..1100.. PPnnppdduummpp mmeesssseess uupp wwiitthh mmoorree tthhaann oonnee PPnnPP bbooaarrdd.. There was a bug in pnpdump somewhere between isapnptools version 1.9 and 1.15 which resulted in incorrect handling of multiple boards (the logical device count for each board was not reset, so only the first board found was correct). I suggest you upgrade to the lastest version. 44..1111.. PPnnppdduummpp ccaannnnoott sseeee tthhee bbooaarrddss aafftteerr uussiinngg tthhee --rr ooppttiioonn.. Use the -i (--ignorecsum) flag to pnpdump . This flag was added in version 1.18 of isapnptools. See also ``Pnpdump reports "No boards found"'' 44..1122.. PPnnppdduummpp rreeppoorrttss ""NNoo bbooaarrddss ffoouunndd"" This means that you have no ISA PnP boards in your system. That's easy then, you don't need isapnptools... If, in fact, you think you should have some ISA PnP boards, there are a few possibilities: +o The board is broken --- unlikely. +o There is a conflict in the system resulting in no suitable READPORT address being available. Although this is unlikely, a faulty PnP device can give a fault which causes all READPORT addresses to fail. To see if this is the case, use the -i (--ignorecsum) flag to pnpdump . This flag was added in version 1.18 of isapnptools. +o You specified two arguments to pnpdump , and the READPORT address didn't match the value used by the earlier PnP isolator. If you did this, and you used two arguments to prevent pnpdump from resetting some device's configuration, upgrade to Version 1.13 of isapnptools, which will do it's own isolation without resetting any device's configuration, even with less than two arguments. Alternatively, if you are running the DOS version under windows 95, see the file in the distribution. If you don't care if the boards configuration is reset, just run pnpdump with fewer arguments. +o The board's PnP capability is configurable. This feature is provided by some boards to support legacy systems. It allows a board to be set to fixed address/interrupt settings, or it can be set to PnP mode allowing standard software reconfiguration. If the board has been configured to a fixed setting, pnpdump won't find it. The thing to do is to find the (DOS based probably) configuration software and run it to check the board settings. If the board is this type, there are two courses of action: 1. Use the configuration software to set the board to a suitable set of addresses and interrupts. This will allow the use of built-in kernel drivers as the board will be present at power up. 2. Use the configuration software to set the board to PnP mode, then use isapnptools to configure the board after the kernel is running. +o The board disappears after being configured (by the BIOS). Try the --reset option to pnpdump . +o You have a new computer, like my recent motherboard upgrade, and the BIOS lists the card during bootup, but isapnp can't see it. I've got this too, and it appears that some sort of conflict is occurring during the isolation sequence for the first (of two) boards, causing the checksum to fail for every readport address tried. I've added the -i (--ignorecsum) flag to pnpdump in version 1.18 of isapnptools to allow the isolation process to complete in the presence of this fault. The fault I've seen looks like there is some sort of bus contention going on, with lower priority devices not dropping out like they should during the isolation sequence: this results in a checksum error on every readport address tried, because the first byte read on the isolation sequence looks like the logical OR of each devices first byte. +o Another thing to try is to play with the BIOS PnP settings. I've had a report that setting the 'PnP aware OS' setting to YES made things work (this was for an ASUS TX-97 E motherboard). In my case, I've found that I need to set it to NO to make things work. In fact it's even worse than that: if I set any interrupt to non PnP, except IRQ10 which has my (non PnP) network card on, nothing works even then. And when it is working, if I ever try a full reset, nothing works from then on until I reboot. (My motherboard chipset is ACER LABS ALADDIN IV, reported by cat /proc/pci, with AMI BIOS). +o Try specifying a readport address such as 0x273 manually. It may be that there is a PnP device at the auto selected address which is causing mysterious problems. +o Version 1.19 and later of isapnptools adds the -D --debug-isolate option to pnpdump. This allows you to see the details of the isolation process in action. By studying the results, and the ISA PnP spec, you may be able to figure out what is going wrong. The code for this is in res-access.c. +o It may be worth checking your hardware by seeing if windows 95/98 can see it, if you have the option. +o If the card came with DOS drivers etc, it might be worth checking they can access the card (running under DOS). If so, can you run the DOS programs in DOSEMU and get them to still work (this will be difficult) ? If so, can you trace the IO accesses to see how the DOS program accesses the card (V. difficult) ? If you can do all this, send me the patches to fix isapnptools please ! +o You could try getting a known good ISA PnP card and trying that in the system. This will tell you if there is something wrong with the original board. See also ``I've got a ..., which runs fine under Win95, how can I find out the port settings'' and ``Examining Win95 resource settings''. 44..1133.. PPnnppdduummpp rreeppoorrttss ""IIddeenntt BByyttee xx,, ((....)) ddiiffffeerrss ffrroomm rreessoouurrccee ddaattaa ((....))"" and goes on .. "Assuming the card is broken". What is supposed to happen, is that when reading the resource data, the first few bytes should be the same as the serial identifier (see section 4.5 of the PnP spec). (In the hardware they come from the same eeprom). Some cards are broken, and don't reset the eeprom read pointer when reading after the sleep state, in this case (detected by the ident byte 0 differs from resource data), we assume the ident stuff has been skipped and we are into resource data proper. 44..1144.. PPnnppdduummpp rreeppoorrttss ""UUnnkknnoowwnn ttaagg ....,, RReessoouurrccee ddaattaa dduummpp aabboorrtteedd"" This often comes after ``Pnpdump reports "Ident Byte x, (..) differs from resource data (..)"''. This means that an invalid tag was encountered in the board's resource data. There is no mechanism for re-synchronising, so the resource data dump for that card has been aborted. There's not much you can do about this. See also ``What does "End tag checksum 0xb8 (BAD)" mean (in the pnpdump output) ?''. 44..1155.. WWhhaatt ddooeess ""DDoonn''tt kknnooww wwhhaatt ttoo ddoo wwiitthh ...... oonn oorr aarroouunndd lliinnee xx"" mmeeaann ?? If the word is a valid one, then it could just be that the braces are mismatched in your configuration file. The best solution is to strip out all the comments and examine what is left (the problem may then be obvious). You can do this easily: cat isapnp.conf | grep -v '^#' | sed '/^$/d' | less 44..1166.. WWhhaatt ddooeess ""EErrrroorr ooccccuurrrreedd eexxeeccuuttiinngg rreeqquueesstt ''LLDD 22'' oonn oorr aarroouunndd lliinnee......"" mmeeaann ?? This means that attempting to read back the logical device number failed. The specification is rather unclear on whether this is guaranteed to work, and in any event, it doesn't appear to work with some devices. There are two solutions: 1. Get isapnp version 1.10 or later which supports VERIFYLD . 2. Use direct register access to select the logical device. So instead of configuring each logical device as normal: (CONFIGURE CTL009c/481377 (LD 0 ... LD 0 settings (ACT Y)) (LD 1 ... LD 1 settings (ACT Y)) (LD 2 ... LD 2 settings (ACT Y)) (LD 3 ... LD 3 settings (ACT Y)) ) pretend they are all logical device 0, and then change the logical device by directly POKEing the logical device register (REGister 7). (CONFIGURE CTL009c/481377 (LD 0 ... LD 0 settings (ACT Y) (REG 7 (POKE 1) (PEEK)) # Set logical device 1, but no check ... LD 1 settings (ACT Y) (REG 7 (POKE 2) (PEEK)) # Set logical device 2, but no check ... LD 2 settings (ACT Y) (REG 7 (POKE 3) (PEEK)) # Set logical device 3, but no check ... LD 3 settings (ACT Y) )) 44..1177.. WWhhaatt ddooeess ""EEnndd ttaagg cchheecckkssuumm 00xxbb88 ((BBAADD))"" mmeeaann ((iinn tthhee ppnnppdduummpp oouuttppuutt)) ?? It means that the resource data read from the board was corrupt, which in turn means that the resource data dump may be wrong. If you know what the card resources really are, you can still configure it with isapnp. The resource data is often stored in EEPROM, and can be written to via the appropriate accesses. It can be corrupted by a faulty windows 95 driver, or driver crash for example. See also ``NT,WIN95,... works ok now, will running isapnp screw it up ?''. 44..1188.. WWhhaatt ddooeess ""IIOO rraannggee cchheecckk aatttteemmpptteedd wwhhiillee ddeevviiccee aaccttiivvaatteedd"" mmeeaann ((ffrroomm iissaappnnpp)) ?? This means you have attempted to CHECK the device resource allocation for conflicts using the IO range feature, but the device has already been activated. Either remove the check, or put (ACT N) before the IO allocation. Note that activated here means the board has been configured and is available for use, it does not mean it is in use. It may or may not have a driver installed in the kernel to use it, and there is no easy way to find out. If you rerun isapnp, this is very likely to occur. You must first deactivate any drivers using the board. Then, when you rerun isapnp, make sure there is an (ACT N) at the beginning of the device configuration, configure and check the settings, then (ACT Y) and the end to reactivate it. You can then restart the drivers for the board. 44..1199.. WWhhaatt ddooeess ""IIOO rraannggee cchheecckk ffaaiilleedd ffoorr nn bbyytteess ooff IIOO aatt aaddddrr"" mmeeaann ((ffrroomm iissaappnnpp)) ?? This means you have attempted to CHECK the device resource allocation for conflicts using the IO range feature, but the check failed. Either: the device doesn't support IO range checks (which you can find out from the pnpdump output), in which case remove the check; or the device SIZE is too large, in which case correct it; or the test has failed due to a resource conflict, in which case change the IO BASE address. In versions of isapnp in isapnptools-1.18 and earlier, there was a bug which caused this to occur anyway. A patch to fix it was produced in June 1999. Later versions should work properly. 44..2200.. II tthhiinnkk II''vvee ccoonnffiigguurreedd tthhee bbooaarrdd ccoorrrreeccttllyy,, bbuutt tthhee ddrriivveerr ssttiillll ccaann''tt ffiinndd iitt.. The configuration file must end in WAITFORKEY . pnpdump from isapnptools 1.9 neglected to put this at the end of the generated configuration script. 44..2211.. WWhhaatt ddooeess ""//pprroocc//bbuuss//ppccii//ddeevviicceess nnoott ffoouunndd,, ssoo PPCCII rreessoouurrccee ccoonnfflliicctt nnoott cchheecckkeedd"" mmeeaann ?? This message occurs if running kernels before the later 2.1.xxx series. /proc/bus/pci/devices appears in later kernels to allow the PCI device information to be read. If you have disabled the proc filesystem, or are running an older kernel, then the isapnptools programs cannot check resource allocations against those assigned to PCI devices. This is unlikely to be a problem, because the resources are likely to be shown used in /proc/interrupts and /proc/ioports etc if a driver is using them. If this isn't good enough, and you still get a problem with resource clashes, just add the PCI resources used to the /etc/isapnp.gone file. 44..2222.. WWhhaatt ccaauusseess lloottss ooff eerrrroorrss rruunnnniinngg tthhee ssccrriipptt ggeenneerraatteedd bbyy ppnnppdduummpp --ss ?? Running a shell script generated by pnpdump -s gives lots of errors like: ./pnp.sh: dev_ids[0]=CTL0043: command not found ./pnp.sh: irq[0]=9: command not found ./pnp.sh: irq_flags[0]=0x0a: command not found ./pnp.sh: dma[0]=0: command not found ./pnp.sh: dma_flags[0]=0x08: command not found ./pnp.sh: io_start[0]=0x0240: command not found ./pnp.sh: io_len[0]=16: command not found ./pnp.sh: io_flags[0]=0x01: command not found ./pnp.sh: ${dev_ids[$i]}: bad substitution This is due to the shell not understanding array variables. You need a newer shell, such as bash 2.0.x. 44..2233.. DDoo II hhaavvee ttoo rreebboooott aafftteerr rruunnnniinngg iissaappnnpp oorr ppnnppdduummpp ?? No. You won't normally need to reboot. Only if you do something silly like reset the PnP hardware while a driver is trying to use it. The thing to do is to do all your experimenting with the programs and configuration files first. Only when you have the right settings, load the kernel drivers etc. If you want to experiment some more, unload the drivers first. 55.. GGeenneerraall -- KKeerrnneell//DDrriivveerr iissssuueess 55..11.. DDoo II nneeeedd ttoo ccoonnffiigguurree CCOONNFFIIGG__PPNNPP iinn tthhee kkeerrnneell ?? This applies to 2.1.xxx and later kernels. They now include a configuration option CONFIG_PNP. This is intended to enable the kernel to configure some PnP devices, but there is little or no support in the kernel (2.1.131). If you're using isapnptools, it would probably be better to set this value to N, to prevent bad interactions when it does do something. 55..22.. II''vvee ccoonnffiigguurreedd mmyy ...... uussiinngg iissaappnnppttoooollss,, bbuutt tthhee kkeerrnneell ddooeessnn''tt rreeccooggnniissee iitt.. There are lots of possibilities for this, just remember: 1. isapnp must be run before the driver initialises, unless a separate program is available to initialise drivers later (eg. _s_e_t_s_e_r_i_a_l). (See the file provided with the distribution). 2. The settings for the board specified to isapnp must match those of the driver for the board. 3. You must load a driver for the card. It is not sufficient to have an entry in dev/. 55..33.. II ggeett tthhee eerrrroorr mmeessssaaggee ""...... mmoodduullee ssuuppppoorrtt aallrreeaaddyy llooaaddeedd"" wwhheenn ttrryyiinngg ttoo llooaadd tthhee ...... mmoodduullee.. This means that the driver module has already been loaded or is built into the kernel. Try typing "cat proc/devices/" to see what devices are already loaded before insmodding the driver. This could also be due to not rebooting after rebuilding the kernel with module support for the required driver. 55..44.. II''mm hhaavviinngg ttrroouubbllee ggeettttiinngg iinntteerrrruuppttss ttoo wwoorrkk.. A bug has been discovered in isapnp for Version 1.11 of isapnptools, relating to the setting of interrupt edges. This is unlikely to be a problem for most hardware because the hardware doesn't allow anything other that +ve edge interrupts, however, as Kevin Bowman has pointed out, some hardware, such as that using the Fujitsu MB86701 does allow the interrupt line to be programmed. In this case, try upgrading to Version 1.13 or later. 55..55.. II''vvee ccoonnffiigguurreedd tthhee ccaarrdd,, bbuutt //pprroocc//iinntteerrrruuppttss ssttiillll sshhoowwss tthhee iinntteerrrruupptt aass uunnuusseedd.. /proc/interupts still lists the IRQ I am choosing as unused and the same for the IO ports. Does this mean the installation has failed ? Not necessarily. Until you load a device driver that uses them, the resources remain unused. You can use pnpdump -d to check the programming if you have a version that supports the --dumpregs option. 55..66.. TThhee ddrriivveerr ffoorr mmyy ccaarrdd ccaannnnoott bbee bbuuiilltt aass aa mmoodduullee.. If the driver you want to use can only be built into the kernel, then you may want to consider using the DOS version of isapnptools, then booting Linux via loadlin. See in the distribution for details. 55..77.. AAss ssoooonn aass II rruunn iissaappnnpp,, mmyy mmaacchhiinnee lloocckkss uupp.. You have SCSI disks on a PnP adapter card - right ? In this case you must be doing a full reset of the ISA PnP hardware, which is making the disks disappear. Alternatively, you may have a PnP network card and a built-in kernel driver. Make sure your configuration file doesn't contain (ISOLATE PRESERVE). For the network card case, you can still do a full ISA PnP reset if you remove the network card driver from the kernel and use a module instead (and load it after running isapnp). 55..88.. HHooww ccaann II ssttoopp iissaappnnpp aanndd ppnnppdduummpp uussiinngg cceerrttaaiinn aaddddrreesssseess ffoorr tthhee RREEAADDPPOORRTT .. Get isapnptools 1.16 or later, and put the port addresses you want to avoid in the /etc/isapnp.gone file. This assumes you want to try lots of addresses, but skip certain ones. If you just want to use a specific address, specify the address as the last argument. 66.. UUppggrraaddiinngg 66..11.. II''vvee uuppggrraaddeedd iissaappnnppttoooollss ffrroomm mmyy ffaavvoouurriittee ddiissttrriibbuuttiioonn aanndd iitt hhaass wwiippeedd oouutt mmyy iissaappnnpp..ccoonnff ffiillee.. This has been reported as a side effect of upgrading Slackware 3.4 to 3.5. I'm sorry, but it must be a bug in the distribution, as upgrading from the isapnptools source will not do this. Please notify your distribution supplier as they should not stomp on user generated files during a package update. If you get the latest isapnptools, it should be possible to generate a good isapnp.conf file without editting, though you may need cleverer boot up scripts to discover the resource allocations. 66..22.. WWhhaatt sshhoouulldd II ddoo ttoo uuppggrraaddee iissaappnnppttoooollss.. If you are updating using a non source version of isapnptools, I suggest you make a copy of the /etc/isapnp.conf and /etc/isapnp.gone files if present, just in case they get overwritten. You may want to regenerate the /etc/isapnp.conf file to take advantage of new keywords which may have been introduced. In this case run pnpdump -c and capture the output, then edit it in comparison to the previous /etc/isapnp.conf file to make the hardware configuration the same. 66..33.. II''vvee uuppggrraaddeedd iissaappnnppttoooollss,, aanndd nnooww iissaappnnpp ggiivveess eerrrroorrss.. Since version 1.16 of isapnptools, resource checking has been added. If your isapnp.conf file contained resource errors, these would have been invisible until you upgraded isapnptools. Note that due to bugs in the resource checking in versions prior to 1.19, resource conflicts may be indicated erroneously. Adding the line (CONFLICT (IO WARNING)(IRQ WARNING)(DMA WARNING)(MEM WARNING)) near the beginning of the isapnp.conf file would allow you to restore the previous behaviour. 66..44.. II''vvee uuppggrraaddeedd mmyy mmootthheerrbbooaarrdd,, aanndd nnooww iissaappnnpp ccaann''tt ffiinndd mmyy bbooaarrddss.. This could be due to you specifying a READPORT address which is no longer free. See also ``Pnpdump reports "No boards found"'' 66..55.. II''vvee uuppggrraaddeedd mmyy kkeerrnneell ttoo 22..44..xx,, nnooww iissaappnnpp ccaann''tt iinniittiiaalliissee mmyy SSoonnddbbllaasstteerr AAWWEE.. This has been reported with isapnp v1.18. It appears the default readport of 0x203 clashes somehow with the 2.4.1 kernel, even though it has been compiled without kernel PnP support. The solution was to change the readport address to 0x273 (which has been the default since version 1.19). 77.. RReessoouurrcceess It is very important that the resource settings you specify in the isapnp.conf file avoid conflict with any other devices, Plug and Play or otherwise. If you fail to do this, the system may fail to work properly, crash, or lock up. 77..11.. DDeetteerrmmiinniinngg rreessoouurrccee uussaaggee ((lliinnuuxx)).. The easiest way to avoid conflict is to make a list of all the devices in your system and the resources they use. To help in this task, try booting the system with no PnP devices configured, then examine the resources the kernel is using via the proc filesystem (which must be mounted, obviously). You can also follow these instructions after configuring the PnP hardware and loading the appropriate driver modules to check that the drivers have been configured properly. To check the proc file system is mounted, type mount at the shell prompt, and somewhere in the response should be a line similar to the one in the example below: $ mount ... /proc on /proc type proc (rw) ... $ From now on, I shall assume you have the proc filesystem is mounted on proc/. In addition, if you use driver modules you should ensure that they are all loaded so that they grab the resources they require. You may want to print the results obtained below: To see what IIOO aaddddrreesssseess are used, type "catproc/ioports/". This will provide a list of IO ports and the drivers that use them. To see what iinntteerrrruuppttss are used, type "catproc/interrupts/". This must be done with all your devices in use, (for example, mount a floppy disk so that the floppy disk interrupt is in use, and open all the serial ports, for example using the command "(cat < dev/ttyS1) &/"). To see what DDMMAA cchhaannnneellss are used, type "catproc/dma/". Note also that some device drivers do not allocate resources unless a process has ooppeenneedd tthhee ddeevviiccee. This is normally the case for the floppy driver and serial drivers for example. This means that the above commands will probably not locate all the resources used by all the hardware, so you may still have to change some resource allocations to get things to work. Another useful command to try is dmesg | grep -i irq which will trawl through the kernel bootup messages for lines containing irq. This finds resources used by kernel drivers. These commands will give a good idea of which resources are in use, but they may miss something if the device doesn't have a driver for example, or the BIOS has allocated the resources for some other purpose. So it is a good idea to check the BIOS settings when rebooting - particularly look for interrupt lines allocated to PCI devices, you may want to deallocate some of these if your PnP card needs a interrupt. Another possibility is to use the --dumpregs command to pnpdump to see what the BIOS has allocated. For further information on standard motherboard resource settings and interrupt assignments, see the pc-hardware-faq (posted regularly to the comp.sys.ibm.pc.hardware.* news groups). 77..22.. II''vvee ccoonnffiigguurreedd tthhee ccaarrdd bbuutt tthhee //pprroocc eennttrriieess hhaavveenn''tt cchhaannggeedd.. The /proc filesystem only reports resources actually in use by a driver. So for example, if you configure a modem, nothing will change. If you then run setserial, to tell the kernel driver about the new port, you will see the IO ports in use in /proc/ioports, but there will still be no change in /proc/interrupts. It's only after you OPEN the serial port for the modem that the modem IRQ will be registered as in use. Similarly, if you configure a device which uses a kernel module driver, nothing will change until you load the driver module, and even then, not all the resources the device uses may be shown as used. Some resources (particularly IRQ and DMA channels) will only be shown in use after a program opens the device and starts using it. 77..33.. II''vvee ggoott aa ......,, wwhhiicchh rruunnss ffiinnee uunnddeerr WWiinn9955,, hhooww ccaann II ffiinndd oouutt tthhee ppoorrtt sseettttiinnggss ?? Using the PEEK command in an isapnp script allows the existing configuration register settings to be examined. Assuming that the settings don't change between booting Win95 and booting Linux, this script can be used to find the settings. IIff yyoouu bboooott ssttrraaiigghhtt iinnttoo LLiinnuuxx,, ssoommee ddeevviicceess mmaayy nnoott bbee ccoonnffiigguurreedd. To find out the card settings, and then to generate a script to reproduce them, follow these instructions: 1. To find out what configuration registers the card has, you must first run pnpdump with the two optional arguments. With pnpdump from release 1.12 or later you don't need the additional two arguments, though a single argument may be useful to set the READPORT address. See for instructions on how to obtain the READPORT address that Win95 uses. 2. Then examine the output to identify the number of logical devices, and the resources used by each one. For this it is best to pipe the pnpdump output through grep to remove most of the comments thus: pnpdump 2 0x3bb | grep -v '# [ A-Z]' | grep -v '^#$' | grep -v '^$' > pnpdump.res On my system I have two cards, and pnpdump.res looks like this: # # (DEBUG) (READPORT 0x03bb) (CSN 2) (IDENTIFY *) (CONFIGURE EDI0119/236861364 (LD 0 # (IO 0 (BASE 0x0240)) # (INT 0 (IRQ 3 (MODE +E))) # (MEM 0 (BASE 0x0c0000) (MODE bu) (UPPER 0x0c4000)) # (MEM 0 (BASE 0x0c0000) (MODE br) (UPPER 0x004000)) # (ACT Y) )) (CONFIGURE DFX0000/1493 (LD 0 # (IO 0 (BASE 0x02f8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x02f8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x02f8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x03e8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x03e8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x02e8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x02e8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x03f8)) # (INT 0 (IRQ 4 (MODE +E))) # (IO 0 (BASE 0x03f8)) # (INT 0 (IRQ 3 (MODE +E))) # (IO 0 (BASE 0x03f8)) # (INT 0 (IRQ 3 (MODE +E))) # (ACT Y) )) (CONFIGURE DFX0000/1493 (LD 1 # (DMA 0 (CHANNEL 3)) # (DMA 0 (CHANNEL 1)) # (DMA 0 (CHANNEL 4)) # (ACT Y) )) 3. Edit this to remove the duplicates, remove the (ACT ) commands, uncomment each resource and change the resource assignment commands to (PEEK ), thus: (READPORT 0x03bb) (CSN 2) (IDENTIFY *) (CONFIGURE EDI0119/236861364 (LD 0 (IO 0 (PEEK)) (INT 0 (PEEK)) (MEM 0 (PEEK)) )) (CONFIGURE DFX0000/1493 (LD 0 (IO 0 (PEEK)) (INT 0 (PEEK)) )) (CONFIGURE DFX0000/1493 (LD 1 (DMA 0 (PEEK)) )) This shows the ethernet card (EDI0119/236861364) has a single logical device which uses one each IO , INT , and MEM resource, while the PnP modem (DFX0000/1493) has two logical devices: a. (LD 0), which uses an IO resource, and an INT resource, b. (LD 1), which uses a single 8 bit DMA channel. 4. Run isapnp on this script, and the following output is obtained: Board 1 has Identity 13 0e 1e 37 b4 19 01 89 14: EDI0119 Serial No 236861364 [checksum 13] Board 2 has Identity 6e 00 00 05 d5 00 00 d8 10: DFX0000 Serial No 1493 [checksum 6e] Peek(EDI0119/236861364)[0][0x60](IO 0) is 0x0340, (3,64) Peek(EDI0119/236861364)[0][0x70](INT 0) is 0x0A02, (10,2) Peek(EDI0119/236861364)[0][0x40](MEM 0) is 0x0000000000000000, (0,0,0,0,0,0,0,0) Peek(DFX0000/1493)[0][0x60](IO 0) is 0x03E8, (3,232) Peek(DFX0000/1493)[0][0x70](INT 0) is 0x0C02, (12,2) Peek(DFX0000/1493)[1][0x74](DMA 0) is 0x04, (4) From this you can see that the modem card is configured at IO address 0x3E8, and uses interrupt 12. The 2 in the bottom half of the INT register means +ve edge trigger and is unlikely to ever be different. The DMA channel is set to 4, which means DMA not used. Also you can see that the ethernet card is configured at IO address 0x340, and uses interrupt 10. The MEM resource is all 0, which means it is not used. 5. So to program these cards like this, use the script: (READPORT 0x3bb) (CSN 2) (IDENTIFY *) (CONFIGURE EDI0119/236861364 (LD 0 (IO 0 (BASE 0x0340)) (INT 0 (IRQ 10 (MODE +E))) (ACT Y))) (CONFIGURE DFX0000/1493 (LD 0 (IO 0 (BASE 0x3e8)) (INT 0 (IRQ 12 (MODE +E))) (ACT Y))) (WAITFORKEY) 77..44.. EExxaammiinniinngg WWiinn9955 rreessoouurrccee sseettttiinnggss.. Alternatively, for the device of interest: 1. Right click on My Computer 2. Click on Properties->Device Manager 3. Select the device of interest, then click on Properties->Resources. Make a note of the resources used. 77..55.. TThhee wwiinnddoowwss CCTTPPNNPP..CCFFGG ffiillee.. Some machines may have a CTPNP.CFG file containing PnP resource information. Have a look. If you've got one, it may be best to use the READPORT address in there. 77..66.. HHooww ccaann II pprreevveenntt ppnnppdduummpp aallllooccaattiinngg cceerrttaaiinn rreessoouurrcceess wwiitthh tthhee --cc ffllaagg.. Get isapnptools 1.16 or later, and put the resources you want to avoid in the /etc/isapnp.gone file. 77..77.. CCoommmmoonn pprroobblleemmss The usual source of conflicts in the sound card with something else, tyically the network card. This is because network cards like to be at addresses 0x300,0x320,0x340,etc, and sound cards use IO addresses all over the place, including (in my system) 0x0220-022f, 0x0330-0333, 0x0388-038b. If you have a system with a network card, and then add a sound card later, it is quite likely that the network card is using addresses 0x0320-033f which will be clobbered by the sound card. The solution is to move the network card, and in my system I have it at 0x0340-035f. Another common problem is to forget to configure one or more of the cards resources. This is particularly likely with sound cards which require so many. See ``How can I find out how the card is configured ?'' to see how you can check these. 77..88.. WWhhaatt ddooeess iissaappnnpp mmeeaann bbyy ""FFaattaall -- rreessoouurrccee ccoonnfflliicctt aallllooccaattiinngg DDMMAA11"".. This means you are trying to allocate DMA1 to a PnP device, but the DMA resource has already been allocated to another device. Similar messages can occur for IO ports or interrupts. You need to find out what other device is using the resource, in case it turns out that the device you are trying to configure has already been configured by the BIOS, and a kernel driver has found it and is using it. To find out what is using it, look in /proc/dma, /proc/ioports, /proc/interrupts as appropriate, or /etc/isapnp.gone. If it is in use by another (different) device, then you must change your resource allocation. If is in use by the device you are trying to configure, then you probably don't need to configure the device using isapnptools. Simply make sure the kernel driver settings match those allocated by the BIOS. 77..99.. DDoo II hhaavvee ttoo ccoonnffiigguurree aallll tthhee ddeevviicceess.. No. More modern BIOSes seem to configure almost all devices, so you don't need to run isapnptools at all. If a device is not configured how you want, you can just reconfigure that one. Note that if you do reconfigure a device, it must not have a driver running on it. This means the driver must not be built into the kernel, and any driver modules for the device must be unloaded. 77..1100.. HHooww ccaann II rreeccoonnffiigguurree aa ddeevviiccee.. See also ``Do I have to configure all the devices.'' Reconfiguring a device is almost the same as configuring one. Points to watch out for are: 1. A device driver must not be using the device. This means you must unload the driver module. If the device driver is built into the kernel, and it has detected the hardware, you must not reconfigure the device. 2. If you want to IO range CHECK the device, you must remember to deactivate it first. See ``What does "IO range check attempted while device activated" mean (from isapnp) ?''. 88.. AAddvvaanncceedd uussaaggee.. 88..11.. CCoommmmoonn ccoonnffiigguurraattiioonnss oonn mmuullttiippllee mmaacchhiinneess,, VVeerrssiioonn 11..1111 aanndd eeaarrlliieerr.. It is not possible to use the same configuration file on multiple machines with these earlier versions of isapnptools. However, it is possible to concoct something with a few scripts etc. The key is to split off each board's configuration into it's own file, then use an empty run of isapnp with just an IDENTIFY to discover which board configurations need to be used. Here is what I said to someone with this problem: I suggest you run isapnp twice (or more!): 1. The first time, do a probe to find out which cards are present, ie, remove all the (CONFIGURE ...) sections. Use the output of the (IDENTIFY ) command in the isapnp script to choose the next script. 2. Run the script selected by the above. Note that the second isapnp scripts don't need to run (ISOLATE ), just do a (CSN x)(IDENTIFY ) instead. If you do this, it may be advisable to fix the READPORT in all four scripts. Using this method, you could have an individual script for each board, and if you named them by the compacted ident, you could get a list of files to configure thus: BOARDS=`isapnp probe.conf | grep '^Board' | cut -c22,23,25,26,28,29,31,32,34,35,37,38,40,41,43,44,46,47` for board in $BOARDS do isapnp $board.conf done where probe.conf is little more than (READPORT 0x3bb) # or whatever (ISOLATE) (IDENTIFY *) (WAITFORKEY) Note that if you were to use this method with isapnptools version 1.12 or later, you could skip the grep and cut by using an appropriate IDENTIFYFORMAT command. And each of the board scripts fit the following skeleton: (READPORT 0x3bb) # or whatever (CSN 10) # or whatever (IDENTIFY *) (CONFIGURE ....) (WAITFORKEY) 88..22.. CCoommmmoonn ccoonnffiigguurraattiioonnss oonn mmuullttiippllee mmaacchhiinneess,, VVeerrssiioonn 11..1122 aanndd llaatteerr.. With version 1.12 and later, CONFIGURE blocks for boards not present will be skipped, so you can just have a single file containing the configurations for all boards in all systems. If you have machines with the same or similar setup of PnP cards, then this can be further optimised by having a single configuration for each type of card, using #1,#2,etc instead of the real serial numbers. Any systems with a non-standard setup could be overridden by a later CONFIGURE specifying the card's full serial number. 88..33.. SSeennddiinngg aa RREESSEETT CCSSNN ccoommmmaanndd.. Is there a way to issue a "Reset CSN Command", as described in Section 4.3 (p15) of the "Plug and Play ISA Specification 1.0a" using isapnptools? Sure, just poke the register using the REG command: after selecting a device and logical device (it doesn't matter which !) do (REG 2 (POKE 4)) See section A.1 on page 51 also. This might be a good use for CSN , as in (CSN 1 (LD 0 (REG 2 (POKE 4)))) You may have to reconfigure the cards again afterwards if you want to fiddle with the PnP registers. 88..44.. DDuummppiinngg tthhee bbooaarrddss ccoonnffiigguurraattiioonn rreeggiisstteerrss.. For isapnptools up to 1.15 you can modify the Makefile to uncomment the line PNPDUMPFLAGS += -DDUMPREGS this will cause pnpdump will dump all the standard configuration registers for each board. For isapnptools 1.16 and later, just give the option --dumpregs to pnpdump to get this dump. Note that this dump is dumping the physical registers, and will thus show the settings that have been put in there by the BIOS, or some cards will put a default setting in. Unused registers read back as 0. But for example, the missing two IO resources on an AWE32 have been reported to show up this way. It is also handy to check that isapnp has done the right thing. 88..55.. DDiissppllaayyiinngg aa lliisstt ooff iissaappnnpp bbooaarrddss iinn tthhee ssyysstteemm.. The following shell script will display a list of ISA PnP boards in your system: #!/bin/bash cat << _EOT | isapnp - #(DEBUG) (IGNORECRC) (IDENTIFY-FORMAT "\"%v/%s\"\n" ) (ISOLATE PRESERVE) (IDENTIFY *) (WAITFORKEY) _EOT on my system this gives: "DFX0000/1493" "CMI0001/16777472" 99.. SSoouunnddccaarrddss.. These seem to be the cause of most problems. Just remember that for a card to work, the card must be accessible to the kernel (that's what isapnp is for) aanndd it must have the correct driver aanndd it mustn't conflict with any other devices. If you suspect driver problems etc, you may wish to join the sound mailing list at linux-sound@vger.rutgers.edu. The later 2.1 kernels have a new sound driver, see the documentation in the /usr/src/linux/Documentation/sound directory. Sound drivers have greatly improved with the v2.2 kernels, so you may find it worthwhile investigating these. Look in the kernel sources at Documentation/sound/Introduction, and any other documentation in there relevant to your card. See also the http://www.linux.org.uk/OSS/ for sound drivers etc. For the latest sound drivers, look at http://www.alsa-project.org/ . These drivers can be built to do their own PnP, so if all you have is a sound card, you may not even need to use isapnptools. There is an article on configuring ISA PnP sound cards at http://www2.linuxjournal.com/lj-issues/issue64/3269.html , in the August 1999 on-line issue of the Linux Journal (issue #64). See also ``Common problems''. 99..11.. II''vvee ggoott aann AAWWEE3322//AAWWEE6644,, bbuutt LLiinnuuxx ddooeessnn''tt rreeccooggnniissee iitt.. [According to Stefan Markgraf, the AWE64 suffers the same problem described below, where the two ports at 0xA20 and 0xE20 are not mentioned in the resource dump.] It turns out tthhiiss iiss tthhee ccaassee ffoorr aallll CCTTLL00002211 ddeevviicceess whether on an AWE32, AWE64 or other. You will have to manually add the missing ports. I've had some great feedback from agriffin@cpcug.org on this: (and a couple of corrections from Takashi Iwai ) I have been able to get my SoundBlaster AWE 32 pnp configured and with drivers from Takashi Iwai, I now get some of the best sounds ever. (At least on any of the machines here.) I have included both the original output from pnpdump and the conf that I use to get things working. NNoottee that the dump does not detect the I/O ports at 0xA20 and 0xE20. The manual that came with the card lists these ports and a note in one of the newsgroups show that all three of the AWE I/O ports need to be configured. Before adding the two I/O ports, the isapnp /etc/ does not report anything unusual, but when I modprobe -a sound, syslog contains two lines: AWE32 not found AWE32: not detected and the output from cat /dev/sndstst in this case is: Sound Driver:3.5.4-960630 (Tue Jan 28 21:37:47 EST 1997 root, Linux palis 2.1.23 #1 Tue Jan 28 19:29:03 EST 1997 i586) Kernel: Linux palis 2.0.28 #3 Tue Jan 28 21:38:48 EST 1997 i586 Config options: a80002 Installed drivers: Type 1: OPL-2/OPL-3 FM Type 2: Sound Blaster Type 7: SB MPU-401 Card config: Sound Blaster at 0x220 irq 5 drq 1,5 SB MPU-401 at 0x330 irq 5 drq 0 OPL-2/OPL-3 FM at 0x388 drq 0 Audio devices: 0: Sound Blaster 16 (4.13) Synth devices: 0: Yamaha OPL-3 Midi devices: 0: Sound Blaster 16 Timers: 0: System clock Mixers: 0: Sound Blaster Now, when the two I/O ports (0xA20 and 0XE20) are included in the isapnp.conf, syslog has: AWE32 Sound Driver v0.3.2 (DRAM 2048K) and the output from /dev/sndstat shows: Sound Driver:3.5.4-960630 (Tue Jan 28 21:37:47 EST 1997 root, Linux palis 2.1.23 #1 Tue Jan 28 19:29:03 EST 1997 i586) Kernel: Linux palis 2.0.28 #3 Tue Jan 28 21:38:48 EST 1997 i586 Config options: a80002 Installed drivers: Type 1: OPL-2/OPL-3 FM Type 2: Sound Blaster Type 7: SB MPU-401 Card config: Sound Blaster at 0x220 irq 5 drq 1,5 SB MPU-401 at 0x330 irq 5 drq 0 OPL-2/OPL-3 FM at 0x388 drq 0 Audio devices: 0: Sound Blaster 16 (4.13) Synth devices: 0: Yamaha OPL-3 1: AWE32 Driver v0.3.2 (DRAM 2048k) Midi devices: 0: Sound Blaster 16 Timers: 0: System clock Mixers: 0: Sound Blaster 99..11..11.. TThhee oouuttppuutt ffrroomm ppnnppdduummpp ((bbeeffoorree ccoonnffiigguurraattiioonn)).. # This is free software, see the sources for details. # This software has NO WARRANTY, use at your OWN RISK # # For details of this file format, see isapnp.conf(5) # # Compiler flags: -DREALTIME -DNEEDSETSCHEDULER # # Trying port address 0203 # Board 1 has serial identifier 0d 00 00 d5 88 43 00 8c 0e # (DEBUG) (READPORT 0x0203) (ISOLATE) (IDENTIFY *) # Card 1: (serial identifier 0d 00 00 d5 88 43 00 8c 0e) # CTL0043 Serial No 54664 [checksum 0d] # Version 1.0, Vendor version 1.0 # ANSI string -->Creative SB AWE32 PnP<-- # # Logical device id CTL0031 # # Edit the entries below to uncomment out the configuration required. # Note that only the first value of any range is given, this may be changed if required # Don't forget to uncomment the activate (ACT Y) when happy (CONFIGURE CTL0043/54664 (LD 0 # ANSI string -->Audio<-- # Multiple choice time, choose one only ! # Start dependent functions: priority preferred # IRQ 5. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 1. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 1)) # Next DMA channel 5. # 16 bit DMA only # Logical device is not a bus master # DMA may not execute in count by byte mode # DMA may execute in count by word mode # DMA channel speed in compatible mode # (DMA 1 (CHANNEL 5)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0220 # IO base alignment 1 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0330 # Maximum IO base address 0x0330 # IO base alignment 1 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0330)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0388 # Maximum IO base address 0x0388 # IO base alignment 1 bytes # Number of IO addresses required: 4 # (IO 2 (BASE 0x0388)) # Start dependent functions: priority acceptable # IRQ 5, 7 or 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Next DMA channel 5, 6 or 7. # 16 bit DMA only # Logical device is not a bus master # DMA may not execute in count by byte mode # DMA may execute in count by word mode # DMA channel speed in compatible mode # (DMA 1 (CHANNEL 5)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0300 # Maximum IO base address 0x0330 # IO base alignment 48 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0300)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0388 # Maximum IO base address 0x0388 # IO base alignment 1 bytes # Number of IO addresses required: 4 # (IO 2 (BASE 0x0388)) # Start dependent functions: priority acceptable # IRQ 5, 7 or 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Next DMA channel 5, 6 or 7. # 16 bit DMA only # Logical device is not a bus master # DMA may not execute in count by byte mode # DMA may execute in count by word mode # DMA channel speed in compatible mode # (DMA 1 (CHANNEL 5)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0300 # Maximum IO base address 0x0330 # IO base alignment 48 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0300)) # Start dependent functions: priority functional # IRQ 5, 7 or 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Next DMA channel 5, 6 or 7. # 16 bit DMA only # Logical device is not a bus master # DMA may not execute in count by byte mode # DMA may execute in count by word mode # DMA channel speed in compatible mode # (DMA 1 (CHANNEL 5)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Start dependent functions: priority functional # IRQ 5, 7 or 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0300 # Maximum IO base address 0x0330 # IO base alignment 48 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0300)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0388 # Maximum IO base address 0x0388 # IO base alignment 1 bytes # Number of IO addresses required: 4 # (IO 2 (BASE 0x0388)) # Start dependent functions: priority functional # IRQ 5, 7 or 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0300 # Maximum IO base address 0x0330 # IO base alignment 48 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0300)) # Start dependent functions: priority functional # IRQ 5, 7, 10 or 11. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 5 (MODE +E))) # First DMA channel 0, 1 or 3. # 8 bit DMA only # Logical device is not a bus master # DMA may execute in count by byte mode # DMA may not execute in count by word mode # DMA channel speed in compatible mode # (DMA 0 (CHANNEL 0)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0220 # Maximum IO base address 0x0280 # IO base alignment 32 bytes # Number of IO addresses required: 16 # (IO 0 (BASE 0x0220)) # End dependent functions # (ACT Y) )) # # Logical device id CTL2011 # # Edit the entries below to uncomment out the configuration required. # Note that only the first value of any range is given, this may be changed if required # Don't forget to uncomment the activate (ACT Y) when happy (CONFIGURE CTL0043/54664 (LD 1 # Compatible device id PNP0600 # ANSI string -->IDE<-- # Multiple choice time, choose one only ! # Start dependent functions: priority preferred # IRQ 10. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 10 (MODE +E))) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0168 # Maximum IO base address 0x0168 # IO base alignment 1 bytes # Number of IO addresses required: 8 # (IO 0 (BASE 0x0168)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x036e # Maximum IO base address 0x036e # IO base alignment 1 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x036e)) # Start dependent functions: priority acceptable # IRQ 11. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 11 (MODE +E))) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x01e8 # Maximum IO base address 0x01e8 # IO base alignment 1 bytes # Number of IO addresses required: 8 # (IO 0 (BASE 0x01e8)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x03ee # Maximum IO base address 0x03ee # IO base alignment 1 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x03ee)) # Start dependent functions: priority acceptable # IRQ 10, 11, 12 or 15. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 10 (MODE +E))) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0100 # Maximum IO base address 0x01f8 # IO base alignment 8 bytes # Number of IO addresses required: 8 # (IO 0 (BASE 0x0100)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0300 # Maximum IO base address 0x03fe # IO base alignment 2 bytes # Number of IO addresses required: 2 # (IO 1 (BASE 0x0300)) # Start dependent functions: priority functional # IRQ 15. # High true, edge sensitive interrupt (by default) # (INT 0 (IRQ 15 (MODE +E))) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0170 # Maximum IO base address 0x0170 # IO base alignment 1 bytes # Number of IO addresses required: 8 # (IO 0 (BASE 0x0170)) # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0376 # Maximum IO base address 0x0376 # IO base alignment 1 bytes # Number of IO addresses required: 1 # (IO 1 (BASE 0x0376)) # End dependent functions # (ACT Y) )) # # Logical device id CTL0021 # # Edit the entries below to uncomment out the configuration required. # Note that only the first value of any range is given, this may be changed if required # Don't forget to uncomment the activate (ACT Y) when happy (CONFIGURE CTL0043/54664 (LD 2 # ANSI string -->WaveTable<-- # Multiple choice time, choose one only ! # Start dependent functions: priority preferred # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0620 # Maximum IO base address 0x0620 # IO base alignment 1 bytes # Number of IO addresses required: 4 # (IO 0 (BASE 0x0620)) # Start dependent functions: priority acceptable # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0620 # Maximum IO base address 0x0680 # IO base alignment 32 bytes # Number of IO addresses required: 4 # (IO 0 (BASE 0x0620)) # End dependent functions # (ACT Y) )) # # Logical device id CTL7001 # # Edit the entries below to uncomment out the configuration required. # Note that only the first value of any range is given, this may be changed if required # Don't forget to uncomment the activate (ACT Y) when happy (CONFIGURE CTL0043/54664 (LD 3 # Compatible device id PNPb02f # ANSI string -->Game<-- # Logical device decodes 16 bit IO address lines # Minimum IO base address 0x0200 # Maximum IO base address 0x0200 # IO base alignment 1 bytes # Number of IO addresses required: 8 # (IO 0 (BASE 0x0200)) # (ACT Y) )) # End tag... Checksum 0x00 (OK) 99..11..22.. TThhee eeddiitttteedd ccoonnffiigguurraattiioonn ffiillee ffoorr iissaappnnpp.. # This is free software, see the sources for details. # This software has NO WARRANTY, use at your OWN RISK # # For details of this file format, see isapnp.conf(5) # # Compiler flags: -DREALTIME -DNEEDSETSCHEDULER # # Edited by jag 27 Jan 1997 to add IO Ports A20 and E20 # # Trying port address 0203 # Board 1 has serial identifier 0d 00 00 d5 88 43 00 8c 0e (DEBUG) (READPORT 0x0203) (ISOLATE) (IDENTIFY *) # Card 1: (serial identifier 0d 00 00 d5 88 43 00 8c 0e) # CTL0043 Serial No 54664 [checksum 0d] # Version 1.0, Vendor version 1.0 # ANSI string -->Creative SB AWE32 PnP<-- # # Logical device id CTL0031 # (CONFIGURE CTL0043/54664 (LD 0 # ANSI string -->Audio<--

(INT 0 (IRQ 5 (MODE +E))) # First DMA channel 1. (DMA 0 (CHANNEL 1)) # Next DMA channel 5. (DMA 1 (CHANNEL 5)) (IO 0 (BASE 0x0220)) (IO 1 (BASE 0x0330)) (IO 2 (BASE 0x0388)) # End dependent functions (ACT Y) )) # # Logical device id CTL0021 # (CONFIGURE CTL0043/54664 (LD 2 # ANSI string -->WaveTable<-- (IO 0 (BASE 0x0620)) (IO 1 (BASE 0x0A20)) (IO 2 (BASE 0x0E20)) # End dependent functions (ACT Y) )) # # Logical device id CTL7001 # (CONFIGURE CTL0043/54664 (LD 3 # Compatible device id PNPb02f # ANSI string -->Game<-- (IO 0 (BASE 0x0200)) (ACT Y) )) # End tag... Checksum 0x00 (OK) 99..11..33.. AAnn eexxttrraacctt ooff tthhee bboooottuupp ssccrriipptt ttoo ccoonnffiigguurree ssoouunndd.. ####################################################################### # Add local action for palis.athena.inc 27 Jan 1997 jag # SoundBlaster AWE 32 pnp initlization # Load sound "fonts". NOTE: the synthgs.sbk from the SB install CDROM # Load sound.o module. This should be loaded before the "fonts". ####################################################################### if [ -f /sbin/isapnp ]; then echo "Initializing the SouncBlaster AWE 32 pnp:" /sbin/isapnp /etc/isapnp.conf > /dev/null else echo "No pnp initialization code available.: exit 1 fi # # According to Takashi IWAI, modprobe must come before sfxload # /sbin/modprobe -a sound if [ -f /usr/local/bin/sfxload ]; then echo "Loading Sound Fonts for General Midi:" /usr/local/bin/sfxload -Ii /usr/local/lib/awe/synthgs.sbk else echo "No Sound Font loading available.: fi 99..22.. PPrroobblleemmss wwiitthh tthhee AAzztteecchh SSoouunndd GGaallaaxxyy 1166 IIIIII--33DD PPnnPP ssoouunndd ccaarrdd.. Stefan Markgraf has reported being unable to get the 16 bit MSS compatibility working with this card (Jul 97), and ended up changing it for a AWE64. The 8-bit Sound Blaster pro compatibility, MPU-401 and OPL-3 functions work fine. 99..33.. CCDD ssoouunndd ddooeessnn''tt wwoorrkk ((oonnee cchhaannnneell oorr nnoo cchhaannnneellss)).. There are two orientations for the CD player sound-card signals, try reversing the connector on the sound card if the CD player sound doesn't work. (From an email from kjcox@students.wisc.edu). After much fiddling, I finally got it to work but in the strangest manner. I reversed the cable from the CD-ROM to the sound card 180 degrees while the computer was up and running. I had to first remove the white box on the sound card into which the cable fits. I tried it both ways and then when I put it back while the computer was running (hey, I got desperate) it suddenly decided to work. It took awhile though to get there. Anyway, something to add to your FAQ. Try removing the white box and reversing the cord. You can either turn the box around or switch it altogether. 99..44.. WWhhaatt ttoo ddoo aabboouutt ""ssbb:: IInntteerrrruupptt tteesstt oonn IIRRQQ1100 ffaaiilleedd -- ddeevviiccee ddiissaabblleedd.."" This kernel message means that the sound driver couldn't get an interrupt from the sound card on the channel you told it to use. This could be due to the soundcard using a different interrupt, or more than one device trying to use the same interrupt. I get this one too. I managed to get rid of it by changing the preallocated ISA interrupts in the BIOS but, unfortunately, when I got the sound working, all the other PnP stuff stopped working. In my case, the sound is built into the motherboard, and there seems to be no way to configure it directly in the BIOS, so it's guesswork to try to find the interrupt allocation. If I had Win95, it might be possible to find it by looking at the Win95 allocation. This is where an ESCD BIOS interface would be useful - any takers ? 99..55.. SSoouunnddbbllaasstteerr ssiittee.. See also the soundblaster linux website . 99..66.. SSoouunndd wwoorrkkss iiff II bboooott ssttrraaiigghhtt iinnttoo LLiinnuuxx,, bbuutt nnoott iiff bboooottiinngg ffrroomm WWiinn9988.. If you use (ISOLATE CLEAR) then all ISA PnP hardware gets reset, and only those devices you configure will get set up. This is DANGEROUS if you have ISA PnP device drivers configured into the kernel (which work because the BIOS configured them - such as disk interfaces or network cards). If you use (ISOLATE PRESERVE), which is the default, then any devices you don't configure will probably be unconfigured after a direct boot from the BIOS, but they will remain configured by Win98 if you boot via Win98. In this case they may then conflict with the settings you have chosen for the devices you configure for linux. I suggest you look at the PCI interrupt allocations too, they are PnP and can be changed under software control (though normally only the BIOS does this). What you should perhaps do is run pnpdump -di > /tmp/pnpdump.txt before you run isapnp during linux boot up. This will allow you to see what the hardware configuration is - and by doing this booting directly, saving the file, and rebooting via Win98, you can compare the results. The other alternative is to browse the hardware configuration settings in Win98, then change your isapnp.conf and driver settings to match. 1100.. MMooddeemmss.. These are the easiest PnP devices to get installed, due to the ability to configure devices (using _s_e_t_s_e_r_i_a_l) after the kernel has booted. This allows experiments to be made with the settings without having to reboot. There are a few gotchas though: Beware of products using phrases such as "Winmodem" or "Specially designed to work with Windows or Windows 95". These may well not work as standalone products, and require the driver provided with Windows to work. See also the Modem-HOWTO , some other Winmodem information , more on Winmodems and the up and coming Linux winmodem support , also the Linmodem-HOWTO . For a good rant, look at . In a similar way some newer modems may not have the DSP code builtin, but rely on a program during boot to download the code into them. These will work in Linux if it is booted after running DOS/Windows, but they won't work directly after a power up into Linux. Watch out for PCI modems. These often don't have UARTs, and so will require a special driver to work with Linux. (They cannot be configured using isapnptools anyway, the BIOS will have done it). If you have a Lucent/Agere PCI chipset winmodem, look at Lucent winmodem drivers , it also has links for PC-Tel winmodems. 1100..11.. SSeeee aallssoo.. See the file in the distribution for an example of how to get a PnP modem card running. See also the Serial-HOWTO available from MetaLab website Serial-HOWTO or from TSX-11 ftpsite Serial-HOWTO , or from their respective mirrors. 1100..22.. II''vvee ggoott aa WWiinnMMooddeemm,, aanndd II''vvee ffoolllloowweedd aallll tthhee iinnssttrruuccttiioonnss,, bbuutt iitt ssttiillll ddooeessnn''tt wwoorrkk.. Unfortunately, a "WinModem" is only half a modem. A significant proportion of the modem functionality is in the Windows driver, which is not present when trying to use it with Linux. The easiest solution at present is to take it back and get a traditional type of modem which can be used on Unix systems. (Of course, if you have the ability, inclination, and required technical information, I'm sure there are many who would like to see a Linux driver for these beasts). 1100..33.. II''vvee ggoott aa PPCCII mmooddeemm -- wwhhaatt aabboouutt tthhaatt ?? PCI modems are generally winmodems. So see above. In any event isapnptools won't do anything for you. 1100..44.. II''vvee ggoott aa mmooddeemm tthhaatt wwoorrkkss iinn DDOOSS aanndd WWiinnddoowwss,, bbuutt nnoott LLiinnuuxx.. It may be that the card needs DSP code downloading onto it before it'll work. This can sometimes be done by booting via DOS and loadlin. 1100..55.. II''vvee ccoonnffiigguurreedd tthhee ddeevviiccee,, bbuutt iitt iiss rreeaallllyy SSLLOOWW.. Very slow activity can be due to interrupt clashes. When the program that uses the port is running, what does proc/interrupts/ say ? If the interrupt count is always 0 for the serial port, you probably have a hardware clash with two devices using the same interrupt line. Check your BIOS PCI settings for preallocated interrupt lines too. If this is the case, reassign the interrupts so that they no longer clash. 1100..66.. II''vvee ccoonnffiigguurreedd tthhee ddeevviiccee,, bbuutt iitt ddooeessnn''tt wwoorrkk.. Try running your modem communication program under strace to see what it is doing. 1100..77.. II''vvee ccoonnffiigguurreedd tthhee ddeevviiccee,, iitt sseeeemmss ttoo sseenndd ookk,, bbuutt nnoott rreecceeiivvee.. Check that the modem has been configured to the correct IRQ using setserial. Even if you think it has. You must check that the IRQ reported by setserial is the same as the one you configured the modem to use. Use setserial to fix the IRQ if it is wrong. See also ``I've configured the device, but it is really SLOW''. 1100..88.. WWhhaatt''ss tthhiiss aabboouutt nnoott uussiinngg //ddeevv//ccuuaa11 ?? It's just something I've read in the mgetty+sendfax documentation. The two devices behave differently on file open, and were intended for dealing with modems set to auto answer. If you want to send faxes and/or allow dialing in, I suggest you get and install mgetty+sendfax, it includes an email from the writer of the serial device driver on why only ttySx should be used. I've seen hints that the cuax devices may be going in future kernels. The mgetty program listens to the modem and answers it manually, thus preventing a wasted call which would have occurred with an auto-answer modem and crashed computer. 1100..99.. HHooww ccaann II cchheecckk tthhee mmooddeemm iiss wwoorrkkiinngg ?? I use cu to talk to the modem. It replies with OK when I type . After a few transfers like this, check /proc/interrupts has increased by the number of characters transferred. If it hasn't, there is a resource clash, so try another interrupt. Note that IRQ3 and IRQ4 are almost guaranteed to be used by the motherboard, so don't use these for PnP modems. I've also written a program called serialmon which allows you to observe the modem control lines etc. 1100..1100.. TThhuunnddeerrlliinnkk mmooddeemmss ((PPCCTT33888800)).. It has been reported that thunderlink modems are not supported for linux or DOS. It gives the following identification: PCT3880/0[0]{PC-TEL HSP K56FLEX PNP D/F/V MODEM} 1100..1111.. UUSSRR RRoobboottiiccss SSppoorrttsstteerr.. If it has jumpers, use them to set the device settings if you can't get PnP configuration to work. 1100..1122.. GGeenneerraall hheellpp oonn sseettttiinngg uupp iinntteerrnneett ddiiaall uupp lliinnkk.. See for help on getting internet links up and running. 1111.. NNeettwwoorrkk ccaarrddss.. See the file in the distribution for an example of how to get a PnP ethernet card running, using a loadable kernel module driver. See also the Ethernet-HOWTO available from MetaLab website Ethernet- HOWTO or from TSX-11 ftpsite Ethernet-HOWTO , or from their respective mirrors. 1111..11.. TThhee ccaarrdd ssuuppppoorrttss nnoonn PPnnPP ccoonnffiigguurraattiioonn.. If the network card supports jumperless mode (fix the card's configuration using a DOS program that comes with it), or has jumpers to disable PnP mode and configure the device; if PnP settings don't work, the simplest thing to do is to configure the card in jumperless (non PnP) mode, or using the jumpers, to a fixed address and interrupt (I use 0x340 and IRQ10). In the BIOS, assign IRQ10 to legacy ISA. If you have other PnP devices which require configuring, remember to add the network card resources used to the /etc/isapnp.gone file. Build the network driver as a module, then load it during bootup with commands such as (for my NE2000 clone for example): # # And the network card # MODULESDIR="/usr/src/linux.`uname -r`.pf/modules" rm /modules ln -s $MODULESDIR /modules insmod $MODULESDIR/8390.o insmod $MODULESDIR/ne.o io=0x340 This method won't require isapnp to configure the network card (in fact, isapnp won't be able to see it). 1122.. IIDDEE ppoorrttss.. See the file in the distribution for some ideas on how to get a PnP IDE interface running. These are often found on sound cards. If you are running a 2.1.x kernel, it is apparently possible to rescan the IDE interface after boot up, but only if no devices are in use on that controller. If you are using a PCMCIA IDE interface, see also ``PCMCIA cards''. 1133.. PPCCMMCCIIAA ccaarrddss.. If you want to know more about PCMCIA, then you might like to look at the www.linuxorbit.com pcmcia feature . You should almost certainly look at the PCMCIA-HOWTO . 1144.. DDiissccllaaiimmeerrss eettcc.. Information is provided AS IS. No guarantee of accuracy or usefulness is provided. There is no intention to provide legal advice in this document either. If you think there is some, I assure you there isn't, I suggest you seek local advice. (R) Linux is a registered trademark of Linux Torvalds. Other trademarks are the property of their respective owners. Spam email will be forwarded to . Questions and queries about isapnptools may be sent to , and the subject line must contain the word 'isapnp' to get past the spam filter. Note that contacting the above email address is not guaranteed to generate a reply; any reply obtained has no guarantee of accuracy or usefulness. To improve your chances of a useful reply, you should include the following information, as appropriate: 1. Output of 'pnpdump -v' (should print out header stuff and usage) 2. Output of 'cat /proc/cpuinfo' 3. Output of 'cat /proc/pci' 4. Output of 'cat /proc/interrupts' 5. Output of 'cat /proc/ioports' 6. Output of 'cat /proc/devices' 7. Output of 'cat /usr/include/kernel/autoconf.h' (or the autoconf.h for the kernel you're running). 8. Output of 'lsmod' if you use kernel modules 9. The command lines used to install the relevant device driver kernel modules, or configure the built in device driver (eg setserial). 10. Output of 'dmesg' 11. Output of 'gcc -v' 12. Output of 'ldd prog' where prog is the program giving trouble. 13. Inputs and Outputs of the program giving trouble: 'prog args.. > prog.stdout 2> prog.stderr' 14. The patch that cures the problem ! If you supply a useful patch, or useful idea for improvement, you will automatically be added to AUTHORS and the announcement mailing list, unless you say otherwise.