- l4-hackers - example.com Site

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 23 Jul '07

23 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