- l4-hackers - example.com Site

some questions in the l4 config
by chenggh 29 Aug '07

29 Aug '07

Hi: In the drops home directory we could use `make -C l4 config O=./build` to enter the l4 configuration interface. There are some options I don't understand as following: Target Architecture --- (x86) Target Architecture // This is ok and there are 3 options to be chosen. CPU type: "586" //This option is confused because it depends on our input. For example If my cpu is pentim4 can I input p4 in the CPU type? In this place is there avaiable small cpu type set? (L4.V2) ABI (X) L4.V2 ( ) L4.X2 ( ) L4.Sec.(V2.emu) ( ) Linux There are four options to be chosen. L4.v2, L4.X2 and L4.sec are explained in other documents. But I choose Linux in this place what can we do? Advances: [*] Use system-call entry code in the KIP (absolute) [ ] Use hierarchical task system (EXPERIMENTAL) [ ] Enable RELEASE flag [ ] Do not build anything outside l4 directory Please also explain these 4 options. Thans very much. Cheng Guanghui

2 1

Re: bug report
by chenggh 29 Aug '07

29 Aug '07

> Thank you for your bug report. The installed version of bison is to old. > Please find attached a patch that fixes bison version detection. (You > will have to rerun 'make -f Makefile.drops' in the dice directory.) What is the distribution you used? I use the Ubuntu6.06 with default bison version 2.1. Is the bison 2.3 ok? Thanks. Cheng Guanghui

2 1

bug report
by chenggh 29 Aug '07

29 Aug '07

ranlib libfe.a make[5]: Leaving directory `/home/L4/tudos/dice/src/fe' Making all in parser make[5]: Entering directory `/home/L4/tudos/dice/src/parser' Making all in idl make[6]: Entering directory `/home/L4/tudos/dice/src/parser/idl' bison --report=all -d idl-parser.yy idl-parser.yy:2.1-8: invalid directive: `%require' idl-parser.yy:2.10-15: syntax error, unexpected string make[6]: *** [idl-parser.tab.cc] Error 1 make[6]: Leaving directory `/home/L4/tudos/dice/src/parser/idl' make[5]: *** [all-recursive] Error 1 make[5]: Leaving directory `/home/L4/tudos/dice/src/parser' make[4]: *** [all-recursive] Error 1 make[4]: Leaving directory `/home/L4/tudos/dice/src' make[3]: *** [all-recursive] Error 1 make[3]: Leaving directory `/home/L4/tudos/dice' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/L4/tudos/dice' make[1]: *** [all] Error 2 make[1]: Leaving directory `/home/L4/tudos/dice' make: *** [../dice] Error 2 make: Leaving directory `/home/L4/tudos/l4' OS: Ubuntu6.06 bison (GNU Bison) 2.1 This l4 is downloaded on 2007-08-27. Thanks. Cheng Guanghui

2 1

Re: The ultimate HOW-TO
by Bjoern Doebel 27 Aug '07

27 Aug '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 skitmax81(a)yahoo.it wrote: > Bjoern Doebel wrote: >> -----BEGIN PGP SIGNED MESSAGE----- >> Hash: SHA1 >> >> Hi, >> >>> PS. For a test, could you please post the entry to add at the menu.lst >>> file >>> to try booting L4Linux? >>> Maybe everything is still working.. &-( >> Did you have a look at >> http://os.inf.tu-dresden.de/L4/LinuxOnL4/use-2.6.shtml >> ? >> .... >> > Quoted from: > http://www.nabble.com/The-ultimate-HOW-TO-tf4192072.html#a12351172 > > ------------------------------------------------------------------------------------------------------ > > Yes, I've read the page, but, because I'm newbie, I don't understand how to have a standalone > setup without DOpE.. I would like to have a simple console where compile/build and run <i>lmbench</i> As Adam already pointed out, you shouldn't even use DOpE when doing a benchmark. So let's sum it up: * Build Fiasco and the hello server. Then run it and make sure that the red (performance-critical) warnings at Fiasco bootup are gone. * Build the L4Env packages needed for L4Linux, namely those that appear in the menu.lst which can be found on the "Using L4Linux" page you already know. * Build L4Linux. Disable DOpE support, disable networking and sound stuff, enable VGA text console and hard disk support. * Put all the binaries obtained so far into the boot directory where GRUB gets its modules from, e.g., where your (SUSe?) Linux kernel is. * Create a loader config file for L4Linux, just like the one on the L4Linux page and put it in the same location. You don't have to use a ramdisk, but can just put a "root=/dev/hda1" or something similar there. Basically, you should be able to use your native Linux environment from HD with L4Linux, too. * Add a menu.lst entry such as the one on the L4Linux page to your GRUB menu.lst. Which of the above points is unclear to you? > I'm not intereste to have networking or sound or other stuffs, only a simple console Linux-like.. it's > for a course and I'm a bit frustrated.. two months and no results.. sigh! :-(( Regards, Bjoern -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFG0v6tP5ijxgQLUNkRApeUAJ9jKk/U6xdywJ4Ih1FH4XeWiMT+cgCfXelx LCnf2fk5w8zcNoyxIwOnJm0= =11yE -----END PGP SIGNATURE-----

1 0

Re: The ultimate HOW-TO
by chenggh 25 Aug '07

25 Aug '07

> > API Version: (87) experimental > > Sigma0 config ip:001048a8 sp:002c9760 > > Roottask config ip:00110000 sp:00000000 > > Starting kernel /DROPS/kernel/fiasco/build/fiasco at 0030082c > > > > Comport 0x0001 is not accepted by the uart driver! > > > > Welcome to Fisco (ia32)! > > DD-L4(v2)/ia32 microkernel (C) 1998-2007 TU Dresden > > Rev: rUNKNOWN compiled with gcc 4.1.0 for Intel Pentium [ ] > > Performance-critical config option(s) detected: > > CONFIG_NDEBUG is off > > > > Enabling special fully nested mode for PIC > > > > Assertion failed: '0' > > in /DROPS/kernel/fiasco/src/abi/kip.cpp:126 > > at f00175d1 > > > > Press any key to reboot. > > Founded the option and set it on, I see that the output is something like > #06.. but it reboots and I can't read so fast! > The result is that I can't be more exahustive in the explaination so I > answer: How to have a log file to read the outputs? In the grub/menu.lst you could do like this: module /boot/L4Linux/main -nowait -nokdb -serial_esc -comspeed 115200 -comport 1 Then you could receive the boot message in another machine with connection to the l4 machine by ttyS0 (com1) with comspeed 115200 like this: minicom -m -s -C logfile Good luck to you. Cheng Guanghui

1 0

Re: The ultimate HOW-TO
by chenggh 25 Aug '07

25 Aug '07

> Founded the option and set it on, I see that the output is something like > #06.. but it reboots and I can't read so fast! > The result is that I can't be more exahustive in the explaination so I > answer: How to have a log file to read the outputs? In the grub/menu.lst you could do like this: module /boot/L4Linux/main -nowait -nokdb -serial_esc -comspeed 115200 -comport 1 Then you could receive the boot message by the ttyS0 (com1) with comspeed 115200. Good luck. Cheng Guanghui

2 1

Re: about compile environment
by chenggh 15 Aug '07

15 Aug '07

On Wednesday 15 August 2007 02:19, sunshine007 wrote: > Thanks adam, > > but now i have problem with booting the L4linux by remote serial console. > i follow the pdf kickstart. > but, i still got error like this (after i reboot the machine): > Using IPI shortcut mode > List of all partitions > No file system could not mount root, tried : > kernel panic - not syncing > VFS : unable to mount root fs on unknown block (0,0) These information shows that your Fiasco has no problem and it could drag the booting of L4Linux. But the L4Linux can't find or mount the root file system. If VFS can't mount root fs with unknown blocks. First you should check your l4linux configuration with the filesystem type as ext2, ext3, etc. Basically you should compile the ext2 or ext3 into the linux kernel not compile as modules. > my menu.lst : > #L4linux > title L4Linux-2.6/Fiasco+dope > root (hd0,2) > kernel /boot/L4linux/bootstrap > modaddr 0x02000000 > module /boot/L4linux/fiasco -nowait -nokdb -serial_esc -comspeed 115200 > -comport 1 > module /boot/L4linux/sigma0 > module /boot/L4linux/roottask task modname "bmodfs" attached 4 modules > module /boot/L4linux/events > module /boot/L4linux/names --events > module /boot/L4linux/log --events > module /boot/L4linux/dm_phys --isa=0x00800000 -v --events > module /boot/L4linux/simple_ts -t 300 --events > module /boot/L4linux/rtc --events > module /boot/L4linux/l4io --noirq --events module > module /boot/L4linux/loader --fprov=BMODFS linux26.cfg > module /boot/L4linux/bmodfs > module /boot/L4linux/vmlinuz > module /boot/L4linux/linux26.cfg > module /boot/L4linux/libloader.s.so > module /boot/L4linux/libld-l4.s.so > #end l4linux Here, your root partition is located in /dev/hda3 > #linux26.cfg > modpath "(hd0,2)/boot/L4linux" > task "vmlinuz" "earlyprintk=yes console=ttyS1,115200 init=1 mem=256M > root=/dev/sda3" > all_sects_writable > Here, your l4linux kernel image is located in /dev/hda3 but root partitition is located in /dev/sda3. > # fdisk -l > > Disk /dev/sda: 250.0 GB, 250000000000 bytes > 255 heads, 63 sectors/track, 30394 cylinders Units = cylinders of 16065 * > 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/sda1 1 5 40131 de Dell Utility > /dev/sda2 * 6 2616 20972857+ 7 HPFS/NTFS > /dev/sda3 2617 26931 195310237+ 83 Linux > /dev/sda4 26932 30394 27816547+ 82 Linux swap > > It is i wrongly typed the root partition? > But there is no devive named /dev/hda3. So may you should replace /dev/sda3 with /dev/hda3. You can try. Best regards. Cheng Guanghui

1 0

about compile environment
by huang 15 Aug '07

15 Aug '07

hi,dear Adam: Could you give me some advices about the operating system on which we can build fiasco and l4env? I have used some different linux such as suse 9 and fedora core 3, fedora core 6 and so on, there always some problem ocurred while building or booting the l4 components. What operating system are you using now? And what operating systems are fully tested for the l4 system? Best regards! huang

6 9

about input in hello
by huang 11 Aug '07

11 Aug '07

hi,dear all: if I want to get some input in the "hello" example, what should I do? I simply use "scanf()", but it does not work! my main.c is as below: #include <l4/sys/types.h> #include <l4/sys/syscalls.h> #include <l4/sys/ipc.h> #include <stdio.h> int main(int argc, char **argv) { l4_threadid_t th; l4_msgdope_t result; l4_umword_t d1, d2; int choice; printf("Hello World\n"); th = l4_myself(); for (;;) { printf("hello: My thread-id is %x.%x\n", th.id.task, th.id.lthread); scanf("%d",&choice); if(choice==0) return 0; } }

2 1

Re: I have a problem, when run Fiasco-UX
by chenggh 10 Aug '07

10 Aug '07

On Friday 10 August 2007 09:27, xiaosan Huang wrote: > Hi, > > I am a new learner of Fiasco-UX. I followed the steps on > http://os.inf.tu-dresden.de/Studium/MkK/uebung/. Everything was ok before > I typed the command : > > ./fiasco -l name -l log -l dm_phys -l hello > > The error information is as follows: > > Fiasco-UX on Linux 2.6.19(i686) > > Mapped 64MB Memory +0 KB Framebuffer + 0 MB Input Area on FD3 > > > > Failed to open file : sigma0 Do you have the sigma0 application in the current directory? cheng

1 0

Re: I have a problem, when run Fiasco-UX
by Udo A. Steinberg 10 Aug '07

10 Aug '07

On Fri, 10 Aug 2007 17:27:34 +0800 xiaosan Huang (XH) wrote: XH> The error information is as follows: XH> XH> Fiasco-UX on Linux 2.6.19(i686) XH> Mapped 64MB Memory +0 KB Framebuffer + 0 MB Input Area on FD3 XH> Failed to open file : sigma0 XH> XH> What is the problem? Please help me. The Problem is that Fiasco-UX cannot find the sigma0 binary file. You can either: * copy sigma0 into the directory where you start Fiasco-UX * create a symlink named "sigma0" in that directory, pointing to the file * specify -S parameter with the full path of sigma0 For some more details, have a look at: http://os.inf.tu-dresden.de/fiasco/ux/using.shtml Cheers, - Udo

1 0

I have a problem, when run Fiasco-UX
by xiaosan Huang 10 Aug '07

10 Aug '07

Hi, I am a new learner of Fiasco-UX. I followed the steps on http://os.inf.tu-dresden.de/Studium/MkK/uebung/. Everything was ok before I typed the command : ./fiasco -l name -l log -l dm_phys -l hello The error information is as follows: Fiasco-UX on Linux 2.6.19(i686) Mapped 64MB Memory +0 KB Framebuffer + 0 MB Input Area on FD3 Failed to open file : sigma0 What is the problem? Please help me. Xiaosan Huang

1 0

L4.sec status ping
by Ernst Rohlicek jun. 07 Aug '07

07 Aug '07

Hello everyone, I would just like to ask about the current status or some roadmap for L4.sec? Is it actively being worked on? Reason is: I need to choose a microkernel to start working on an experimental operating system within this year I need some facts for decision-making - and I would prefer a capability-based one like L4.sec. Thank you, Ernst Rohlicek jun.

7 16

Status of idl4?
by Jonathan S. Shapiro 06 Aug '07

06 Aug '07

Just as a sanity check, I ran a trivial IDL file through idl4. The generated server header does not provide any obvious means to get the reply send, because the message block is passed **by value**. Is this intentional? Is DICE now preferred? -- Jonathan S. Shapiro, Ph.D. Managing Director The EROS Group, LLC

2 1

NMI during irq_interrupt() issue
by Stefan Scheler 05 Aug '07

05 Aug '07

Hello, I implemented a statistical profiler for fiasco using performance monitoring counters (nmi handler is attached to handle_slow_trap()). It works fairly well for "mathematical" workloads but as soon as interrupt activity gets involved to a greater extent, the system freezes. For example, keeping a button pressed while sampling will freeze the machine within a few seconds. The NMI watchdog does not trigger during that freeze. Higher sampling rates also cause these freezes. So imho there seems to be a problem when irq_interrupt() or Irq::hit() is interrupted by an NMI. What do you think? Any hints on how to fix this or on how to debug this will be greatly appreciated. Thanks in advance. I will happily provide further information if needed. -- Stefan Scheler

3 4

[DICE] Compilation error with two flexpages
by Clermont, Nicolas 01 Aug '07

01 Aug '07

Hi everybody, I am currently playing with flexpages. I have written simple client/server application which sends data to each other using flexpages. I have written a simple IDL file for the server interface. The first version was the following : library test_server { interface worker { int send_data([in] flexpage page, [out, max_is(4096)] char reply[]); }; }; The client sends data through the flexpage, and the server replies using string ipc. This way the DICE generated code compiles without error, and my application works. Then I wanted the server to reply also through a flexpage, but another one than the page provided by the client application. So I modify my IDL this way : library test_server { interface worker { int send_data([in] flexpage page, [out, in] flexpage reply); }; }; I use [out, in] because this way the client can first assign the snd_base before sending the request. After modifying my client lib and the component function accordingly, I get a compile error in a DICE generated file : /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c: In function `test_server_worker_wait_any': /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c:91: error: `_dice_tmp_offset' undeclared (first use in this function) /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c:91: error: (Each undeclared identifier is reported only once /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c:91: error: for each function it appears in.) /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c: In function `test_server_worker_reply_and_wait': /home/nico/build-tudos/pkg/test_server/idl/OBJ-x86-l4v2/test_server_worker-s erver.c:172: error: `_dice_tmp_offset' undeclared (first use in this function) What I see in the generated file : filter_server_worker_wait_any (...) { ... if (l4_ipc_fpage_received(_dice_result)) { _dice_tmp_offset = 0; while ((_dice_msg_buffer->_word._word[0][_dice_tmp_offset++] != 0) && (_dice_msg_buffer->_word._word[0][_dice_tmp_offset++] != 0)) /* empty */; /* skip zero fpage */ _dice_tmp_offset += 2; _dice_opcode = _dice_msg_buffer->_word._word[0]; } else _dice_opcode = _dice_msg_buffer->_word._word[0]; return _dice_opcode; } The _dice_tmp_offset is not declared, the second error (line 172) is about the same thing (it is exactly the same code in reply_and_wait()). So I have tried to declare _dice_tmp_offset as an unsigned int, then I have : /home/clermont/build-tudos/pkg/filter_server/idl/OBJ-x86-l4v2/filter_server_ worker-server.c: In function `filter_server_worker_wait_any': /home/clermont/build-tudos/pkg/filter_server/idl/OBJ-x86-l4v2/filter_server_ worker-server.c:92: error: subscripted value is neither array nor pointer /home/clermont/build-tudos/pkg/filter_server/idl/OBJ-x86-l4v2/filter_server_ worker-server.c:92: error: subscripted value is neither array nor pointer (The line 92 is the one with while(...)). Indeed after looking the code I find it really strange, as _word[0] is a register.... More _dice_tmp_offset is incremented by two after the while loop and never used after... I really don't understand this code... If I only set the reply flexpage as [in], I have the same behavior... Am I doing something wrong ? Could it be a bug in DICE ? Is it permitted to use one in-flexpage and one out-flexpage for the same function ? Is there something special to do ? Thanks, -- nico

2 1

Re: about l4linux on fiasco-UX
by Adam Lackorzynski 29 Jul '07

29 Jul '07

On Wed Jul 25, 2007 at 16:49:51 +0800, kashin Lin wrote: > 2007/7/25, Adam Lackorzynski <adam(a)os.inf.tu-dresden.de>: > > > > > >On Tue Jul 24, 2007 at 10:10:23 +0800, kashin Lin wrote: > >> 2007/7/24, Adam Lackorzynski <adam(a)os.inf.tu-dresden.de>: > >> > > >> > > >> >On Sun Jul 22, 2007 at 23:09:10 +0800, kashin Lin wrote: > >> >> 2007/7/22, Adam Lackorzynski <adam(a)os.inf.tu-dresden.de>: > >> >> > > >> >> > > >> >> >On Sun Jul 22, 2007 at 01:00:44 +0800, kashin Lin wrote: > >> >> >> 2007/7/21, kashin Lin <kashin08(a)gmail.com>: > >> >> >> > > >> >> >> > > >> >> >> > > >> >> >> >2007/7/21, Adam Lackorzynski <adam(a)os.inf.tu-dresden.de>: > >> >> >> >> > >> >> >> >> > >> >> >> >> On Fri Jul 20, 2007 at 15:54:06 +0800, kashin Lin wrote: > >> >> >> >> > when i runed l4linux on fiasco-UX by "./lx con", it stoped > >> >> >> >> > after following messages: > >> >> >> >> > >> >> >> >> We need to find out by adding earlyprintk=keep to the L4Linux > >> >kernel > >> >> >> >> command line. Then there should be more text. > >> >> >> > > >> >> >> > > >> >> >> > >> >> >> > >> >> >> Here are the results after adding "earlyprintk=keep" : > >> >> > > >> >> >> >con | vc[1] 640x480@16, bpl:1280, gmode:0x15, evprt:E.03save:0 > >> >> >> >run | contxt_init(): 640x462, cols:80, lines:33, > >sb_lines:1033 > >> >> >> >l4lx | INITRD: Size of RAMdisk is 16384KiB > >> >> >> >l4lx | RAMdisk from 03400000 to 04400000 [16384KiB] > >> >> >> >[ 0.000000] Linux version 2.6.21-l4 (kashin08@kashin08-desktop > >) > >> >(gcc > >> >> >> >version 4.1.2 (Ubuntu 4.1.2-0ubuntu4)) #1 Fri Jul 20 12:14:47 CST > >> >2007 > >> >> >> >[ 0.000000] CPU: Vendor unknown, using generic init. > >> >> >> >[ 0.000000] CPU: Your system may be unstable. > >> >> >> >[ 0.000000] dm_phys: 0000000000800000 - 0000000003400000 > >> >(usable) > >> >> >> >[ 0.000000] 44MB memory available. > >> >> >> >[ 0.000000] INITRD: 03400000 - 04400000 > >> >> >> >[ 0.000000] Zone PFN ranges: > >> >> >> >[ 0.000000] DMA 0 -> 512 > >> >> >> >[ 0.000000] Normal 512 -> 13312 > >> >> >> >[ 0.000000] early_node_map[1] active PFN ranges > >> >> >> >[ 0.000000] 0: 0 -> 13312 > >> >> >> >[ 0.000000] Allocating PCI resources starting at 10000000 > >(gap: > >> >> >> >03400000:fcc00000) > >> >> >> >[ 0.000000] Built 1 zonelists. Total pages: 13208 > >> >> >> >[ 0.000000] Kernel command line: mem=44M earlyprintk=keep > >> >root=1:0 > >> >> >> >load_ramdisk=1 ramdisk_size=16384 l4env_rd=../../../test_lx/rd1 > >> >> >> >panicblink=0 > >> >> >> >lang=us > >> >> >> >[ 0.000000] Initializing CPU#0 > >> >> >> >[ 0.000000] PID hash table entries: 256 (order: 8, 1024 bytes) > >> >> >> >[ 0.000000] Detected 2010.311 MHz processor. > >> >> >> >[ 1236.204905] keeping early console > >> >> >> >l4lx | panic: going to sleep forever, bye > >> >> > > >> >> >Can you please post your .config? > >> >> > >> >> > >> >> following is the content of my ".config". (also in attached file) > >> > > >> >So many options enabled... Can you please use the ux default > >> >configuration for a start? (arch/l4/configs/x86-ux_defconfig, > >> >make x86-ux_defconfig will also do). > >> > >> > >> after using "make x86-ux_defconfig", it still doesn't work. > >> but the error messages are different this time : > > > >Ok, different. Can you please add showpfexc=1 to the L4Linux kernel > >command line and post the last part of log again?! There should be more > >output then. Then please look up the code at the PC that is printed, or > >send me (only me) your vmlinuz file. > > > following are the messages after i added "showpfexc=1". > NET: Registered protocol family 16 > l4lx | EX: F.04: pc = 0040acf5 trapno = 0xd err = 0x402 > Setting up standard PCI resources > l4lx | EX: F.04: pc = 0040acf5 trapno = 0xd err = 0x402 > l4lx | EX: F.04: pc = 00408d36 trapno = 0xd err = 0x0 > Trap: 13: 0000 [#1] It's trapping on segment access, I wonder how this can happen. What's the version of your host Linux kernel? > another question i want to ask : > according to "l4-kickstart.pdf", there are some configuration options should > been enable. > for instance: > > Graphics support > <*> Support for frame buffer devices > Console display driver support ---> > [ ] VGA text console > < > MDA text console (dual-headed) (EXPERIMENTAL) > <*> Framebuffer Console support > [*] Framebuffer Console Rotation > [ ] Select compiled-in fonts > > but the default configuration in "arch/l4/configs/x86-ux_defconfig" doesn't > obey that. x86-ux_defconfig enables all of the three options, or do you mean that the pdf says other options should be enabled? > should i follow the "l4-kickstart.pdf" or arch/l4/configs/x86-ux_defconfig > is fine? x86-ux_defconfig, this should always be the most recent version. Adam -- Adam adam(a)os.inf.tu-dresden.de Lackorzynski http://os.inf.tu-dresden.de/~adam/

2 3

Writing about L4 and the future Information Society
by John van V. 29 Jul '07

29 Jul '07

Hello all, I am presently starting to write about L4; I was an admin in financial technology from 1999 to 2002 and I specifically worked to integrate OSs to enable the workforces of the world, and I also attempted to help talented youth learn to operate systems, and also hoped to help impoverished nations join the Information Society. Here is my personal website where my writing will be presented. You can see my motivations: http://thinman.com (it is presently in its "capstone degree project" format, which earned me an A. I will be attempting to evolve it in the near future into a mix of operating systems, hardware platforms, education, and fine art to create the kind of mental diversity we find in a healthy society. In 2000, I strongly felt that Linux needed to evolve, and that new OSs needed to be introduced to the public domain. I looked at ExOS, and I liked it a lot, but it had the problem of being tied to only a single CPU architecture. In 2002 I got fired along w/ all my technology collaborators and fellow club members because of the tech crash, and then I also started to suffer the traumatic after affects from being a volunteer rescuer during the terror attack on the World Trade Center on 9/11/2001. Probably a hundred or more people I knew were killed, including 1/2 of my local fire department whom I knew personally. Also former coworkers of mine as well as many of the Sun techs and sales staff I knew had perished. The Sybase meeting rooms that had been frequent meeting places for my various clubs has also been destroyed. What I did afterwards to get past my grief issues was become a long haul truck driver, and I drove the roads of the US for a few years, which actually was an incredibly enlightening experience. In 2005, when I was about to start investing in my own truck fleet, I had an epiphany -- to teach middle school kids science, and especially technology. I believe that middle school and high school students, from my experiences, are the best potential technology core from the perspective of the future-- and they work for free. Also, I should mention that many people are saying that OS development in general has slowed to a near halt. This appears to me to be so, though all of you involved in L4 are not to blame! I feel that L4, as the leading new OS, can be a focus for, or perhaps locus of, all the things we need to do, and all the reflections we need to make based on the nearly two decades of successful, yet ultimately heartbreaking experiences, in the new Information Society. Any comments would be greatly appreciated. Thanks in advance, John -- Empathy http://thinman.com/empathy Photography http://thinman.com/photography Technology http://thinman.com

2 1

problem about using Fiasco-UX
by 王涛 27 Jul '07

27 Jul '07

Hi all. after i built the kernel ,first i prepare as follow: Mkdir fiasco_test Cp kernel.image fiasco_test/fiasco Cp irq0 fiasco_test/ Cp roottask fiasco_test/ Cp sigma0-ux fiasco_test Cp bootstrap fiasco_test/ then ./fiasco i got the error message: Loading Module 0x00090000-0x00096cc4 [sigma0-ux] Loading Module 0x00120000-0x002bd9f4 [roottask] Bootstrapping... [1;32mWelcome to Fiasco(ux)! [0;32mDD-L4(v2)/x86 microkernel (C) 1998-2006 TU Dresden Rev: Tue Oct 17 17:00:30 2006 compiled with gcc 4.1.1 for Intel Pentium[0m CPU: GenuineIntel (F:4:7:0) Model: Pentium D (Smithfield) at 2792 MHz 128 Entry I TLB (4K or 4M pages) 64 Entry D TLB (4k or 4M pages) 12K ?ops T Cache (8-way associative) 16 KB L1 D Cache (8-way associative, 64 bytes per line) 1024 KB L2 U Cache (8-way associative, 64 bytes per line) Freeing init code/data: 24576 bytes (6 pages) segmentation fault Fiasco-UX on Linux 2.6.18-1.2798.fc6 (i686) gcc/g++ version:4.1.1 i will appreciate that you can help me solve the problem

2 1

About build fiasco for xScale
by kashin Lin 25 Jul '07

25 Jul '07

Hi~ When i cross-compile fiasco for xScale, it says 'kernel.image': No such file. i checked the files in kernel/fiasco/build. no file named "kernel.image" but there is a file named "fiasco.image". i am not sure what's the relationship between them. how to fix it? following is part of the output message: make[3]: Entering directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build' Checking compiler config... unchanged. make[3]: Leaving directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build' make srcdir=/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/src objbase=/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build -f /home/kashin08/L4Linux/L4_ARM/kernel/fiasco/src/Makefile.sub2 all make[3]: Entering directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build' -e ... Generating version information Checking initcalls arm-linux-nm: 'kernel.image': No such file start (_initcall_start) or end (_initcall_end) symbol not found! at /home/kashin08/L4Linux/L4_ARM/kernel/fiasco/src/../tool/checkinitcalls line 67. make[3]: *** [Checkinitcalls] Error 1 make[3]: Leaving directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco/build' make[1]: *** [build] Error 2 make[1]: Leaving directory `/home/kashin08/L4Linux/L4_ARM/kernel/fiasco' make: *** [../kernel/fiasco] Error 2

2 9

about l4linux on fiasco-UX
by kashin Lin 24 Jul '07

24 Jul '07

Hi~ according to http://os.inf.tu-dresden.de/fiasco/ux/status.shtml, it says one of the differences between native fiasco and fiasco-UX is : Port I/O and I/O flexpages. so is it possible to run a multimedia application in l4linux on fiasco-UX? (ex: jpeg, mp3 ... etc) this kind of application seems that it could access hardware directly( ex: speaker or monitor )? i am not sure is this guessing right.

3 13

About download GRUB
by kashin Lin 24 Jul '07

24 Jul '07

Hi~ according to this website " http://os.inf.tu-dresden.de/drops/download.html", the GRUB to boot l4linux/fiasco is not the normal version. but it can't be downloaded from SVN now. how can i get that?

2 1

About build of L4Linux
by 林和宏 22 Jul '07

22 Jul '07

Hello. I am a university student in japan. I have a question about an error in building L4Linux 2.6. I want to run L4Linux with L4Ka::Pistachio on an IA32 architecture machine. I got a following error when I tried to make L4Linux. ------------------------------------------------------------- ..... LD .tmp_vmlinux1 KSYM .tmp_kallsyms1.S AS .tmp_kallsyms1.o LD .tmp_vmlinux2 KSYM .tmp_kallsyms2.S AS .tmp_kallsyms2.o LD .tmp_vmlinux3 KSYM .tmp_kallsyms3.S AS .tmp_kallsyms3.o CHK include/linux/version.h *** Error: Headers not exportable for this architecture (l4) <--- this error make[1]: *** [headers_install] error 1 make: *** [vmlinux] error 2 ------------------------------------------------------------- My build way is based on the official website (http://os.inf.tu-dresden.de/L4/LinuxOnL4/build-2.6.shtml) First, I download the source of L4Linux 2.6 and the l4linux-2.6-env module using SVN repository. [root@lin root]# svn co http://svn.tudos.org/repos/l4linux/trunk l4linux-2.6 [root@lin root]# svn co -N http://svn.tudos.org/repos/tudos/trunk tudos [root@lin tudos]# ./repomgr checkout l4linux-2.6-env The directory structure is following. tudos/ - build/ - dice/ - l4/ - l4linux-2.6/ - repomgr - tools Next, I configure and make L4 enviroment. [root@lin tudos]# make -C l4 O=/root/tudos/build/ config [root@lin tudos]# make -C l4 O=/root/tudos/build/ In DROPS Configuration Tool, I set the path to the DROPS tree. Other parameters are default. Paths and Directories ---> DROPS_STDDIR: "/root/tudos/" Finally, I go to the l4linux-2.6 directory and configure and make. [root@lin l4linux-2.6]# make menuconfig [root@lin l4linux-2.6]# make I change a lot of parameters in L4Linux kernel configuration to avoid many other errors in start and end of make sequence. L4Linux configuration ---> - (/root/tudos/build/) L4 tree build directory Proccessor type and features ---> - [ ] Symmetric multi-proccessing support - Subarchitecture Type (PC-compatible) ---> - [ ] HPET Timer Support - Preemption Model (No Forced Preemption (Server)) ---> - [ ] Local APIC support on uniproccessors - [ ] Machine Check Exception - High Memory Support (off) ---> - [ ] MTRR (Memory Type Range Register) support - [ ] kexec system call Power Management option (APIC, APM) ---> - [ ] Power Management support Bus option (PCI, PCMCIA, EISA, MCA, ISA) ---> - [*] PCI access support PCI access mode (Direct) ---> File systems ---> - Pseudo filesystems ---> - [ ] HugeTLB file system support The error massage may be printed by l4linux-2.6/Makefile (line 928). If there is not file "include/asm-$(ARCH)/Kbuild", it print the error in make. But I seem not to exist Kbuild in include/asm-l4 direcory ... What is the cause of the error ? I think that some settings about L4 environment are lacked. But I cannot understand the sure way to build L4Linux and its environment yet... Please help me ! Thanks and best regards, Lin

5 31

about build l4linux
by kashin Lin 19 Jul '07

19 Jul '07

Hi~ when i issue make to build l4linux in directory l4linux-2.6/, i got some error message: kashin08@kashin08-desktop:~/L4Linux/L4_x86/l4linux-2.6$ make CHK include/linux/version.h CHK include/linux/utsrelease.h CC arch/l4/kernel/asm-offsets.s In file included from include/linux/irqflags.h:46, from include/asm-l4/l4-arch/asm/system.h:507, from include/asm-l4/l4-arch/asm/processor.h:18, from include/asm-l4/i386-arch/asm/atomic.h:5, from include/linux/crypto.h:20, from arch/l4/kernel/asm-offsets.c:8: include/asm-l4/l4-arch/asm/irqflags.h: in function 「raw_irqs_disabled_flags」： include/asm-l4/l4-arch/asm/irqflags.h:168: Error：「L4_IRQ_DISABLED」 not declare In file included from include/asm-l4/l4-arch/asm/suspend.h:7, from include/linux/suspend.h:5, from arch/l4/kernel/asm-offsets.c:12: include/asm-l4/i386-arch/asm/i387.h: in function 「save_init_fpu」： include/asm-l4/i386-arch/asm/i387.h:101: Error：「l4lx_fpu_enabled」 not declare make[1]: *** [arch/l4/kernel/asm-offsets.s] Error 1 make: *** [prepare0] Error 2 my gcc and g++ and make version are: gcc 4.1.2 g++ 4.1.2 make 3.81 how to fix it? i have tried to trace the code and think this error should never appear.

2 9

building l4linux
by kashin Lin 15 Jul '07

15 Jul '07

Hi~ i have some questions again. my building sequence is : ====================================== in l4_all/kernel/fiasco 1. make BUILDDIR=../../build/fiasco in l4_all/build/fiasco 2. make menuconfig in l4_all/l4 3. make config O=../build/l4env 4. make O=../build/l4env ====================================== my questions : 1. when i issue make in l4/ to build the L4 userland, does it automatic build fiasco subsequently？ 2. if yes, is its default output path for fiasco l4_all/kernel/fiasco/build? ( after the operations mentioned above, i find there are some new files in l4_all/kernel/fiasco/build ) but my configuration for fiasco is in l4_all/build/fiasco, is this auto build operation based on my config? or i should set BUILDDIR to l4_all/kernel/fiasco/build? when config l4linux: according to http://os.inf.tu-dresden.de/L4/LinuxOnL4/build-2.6.shtml, there are some options we shouldn't enable ex: like ACPI, SMP, preemption, apic/ioapic, HPET, highmem, MTRR, MCE, power management and similar. 1. does those options are default unable? or we should unable manually? 2. what does "similar" mean? thanks for answering.

2 5