l4-hackers November 2006

l4-hackers@os.inf.tu-dresden.de

25 participants
28 discussions

vbemode option in l4dope?
by "Gábor Bérczi (Gabucino)" 05 Dec '06

05 Dec '06

Greetings. I'd like to ask if it's planned to implement l4con's --vbemode command line option in l4dope too? I couldn't get grub's vbeset option to work in QEMU (always ends up in black screen), but l4con-s vbemode option does work. -- Gábor Bérczi

3 9

Re:Timing problem on mem_fpage_unmap() in the fiasco kernel
by WanKeun Jo 24 Nov '06

24 Nov '06

Hi, I prevented the timing problem to occur by simply unset the l4linux-2.6 option "Show timing information on printks" before. When the option is set, then in the function 'vprintk()' of "l4linux-2.6/kernel/printk.c", the system get current jiffies and then converted it to the string and add it to log buffer. The system works well right before the following codes in the function. When I remove the following code, the system works well without adding timing information on log buffer. ------------------------------------------------ tlen = sprintf(tbuf, 590 "<%c>[%5lu.%06lu] ", 591 loglev_char, 592 (unsigned long)t, 593 nanosec_rem/1000); 594 595 for (tp = tbuf; tp < tbuf + tlen; tp++) 596 emit_log_char(*tp); 597 printed_len += tlen; ------------------------------------------------ However, when I add one of the above three statements - (1) tlen = sprintf(), (2) for(), (3) printed_len += tlen - then the timing error occur during the "Freeing INITRD". even one statement of them.. This is so strange to me and also to you. I'm just beginner in this kind of works, so there might something that I overlooked. Is there somebody who can give me an advice for this kind of ERRORs..?? Thank you.

2 1

fiasco compilation problems
by Hendrik Tews 24 Nov '06

24 Nov '06

I just failed in trying to build fiasco-ux: First I followed the fiasco building and using page (the hello module variant) using the public cvs: - The following dependencies are not mentioned: libncurses5-dev, autoconf, automake1.9, doxygen I don't understand autoconf: the configure scripts should be generated at the distributors site? When autoconf was missing, I got a message saying I need autoconf 2.50, it didn't say I need automake1.9. Here autoconf comes by default with 1.8. Further I don't understand doxygen: Wouldn't it be better to have a separate doc goal and the main make goal independent from doxygen? I appreciate the different config tools in l4 and fiasco as sign of big freedom in this world (and in Dresden in particular). ;-) However, it would make live easier for the casual l4 configurator (like me) if keybindings would agree and fiasco config had an exit button. - the fiasco-ux site says Enable at least the following configuration options: * Target platform (choose UX) * Assembler IPC shortcut * Generate inline code I could neither find IPC shortcut nor inline code. - Then I got the following message: Flex and Byacc. (If you don't have Byacc, but Bison, try setting "bison -y" in the L4 configuration menu under "Compilers and Tools". Other Yacc implementations might do it as well.) but neither the message nor the website says where to find the L4 config menu. Wouldn't it be possible to let the config automatically choose bison if byacc is not found? - the "-C l4" in in 2.a seems to prohibit relativ path for the builddir. Why not cd l4 before? - Point 3.a says "cd l4", but there is no l4 inside l4 - Somebody told me that the preferred way to download fiasco nowadays is to use one of the opentc snapshots. If thats true if would be nice to have a link to these snapshots. - Finally fiasco-ux and all the modules were compiled: tandem l4v2 38> ./fiasco Fiasco-UX on Linux 2.6.17-2-k7 (i686) Mapped 64 MB Memory + 0 KB Framebuffer + 0 MB Input Area on FD 3 Loading Module 0x00090000-0x00096d44 [sigma0-ux] Loading Module 0x00120000-0x002bd9f4 [roottask] Bootstrapping... Welcome to Fiasco(ux)! DD-L4(v2)/x86 microkernel (C) 1998-2006 TU Dresden Rev: Thu Nov 23 10:27:00 2006 compiled with gcc 4.1.2 for Intel Pentium CPU: AuthenticAMD (F:23:2:0) Model: Athlon 64 X2 (Toledo) at 2418 MHz 32/512 Entry I TLB (4K pages) 8 Entry I TLB (4M pages) 32/512 Entry D TLB (4K pages) 8 Entry D TLB (4M pages) 64 KB L1 I Cache (2-way associative, 64 bytes per line) 64 KB L1 D Cache (2-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) ASSERTION_FAILED (((((*(int *) &(status))) & 0xff) == 0x7f) && ((((*(int *) &(status))) & 0xff00) >> 8) == 10) in function static unsigned int Hostproc::create(unsigned int) in file /home/tews/Fiasco/current/l4/kernel/fiasco/src/kern/ux/hostproc.cpp:120 Now I tried the latest opentc snapshot. I ingored the opentc build instructions and applied the fiasco build instructions to this tree. At some stage it was complaining about gcc-4.1.2 being the wrong compiler version. I changed the compilers to gcc-3.4/g++-3.4 in the l4 config. After that make skipped a lot of dirs, because the goals were done before already. Shouldn't it recompile everything after changing the compiler? So I did make clean then and rebuild. But still I saw a lot of messages with ``g++'' (which would be 4.1.2). That's unproblematic? The compilation stoped eventually with make[3]: Entering directory `/home/tews/Fiasco/opentc-snap/l4/build/doc/building-howto' cp building.tex drops-building.tex cp: cannot stat `building.tex': No such file or directory make[3]: *** [drops-building.tex] Error 1 make[3]: Leaving directory `/home/tews/Fiasco/opentc-snap/l4/build/doc/building-howto' make[2]: *** [all] Error 2 make[2]: Leaving directory `/home/tews/Fiasco/opentc-snap/l4/doc/building-howto' make[1]: *** [building-howto] Error 2 make[1]: Leaving directory `/home/tews/Fiasco/opentc-snap/l4/doc' make: *** [doc] Error 2 There seems a problem with the builddir, because tandem l4 29> find . -name building.tex ./doc/building-howto/building.tex However, the build/bin/x86_586/l4v2 dir looked already very full, so I decided to ignore the build instructions and tried: tandem l4v2 7> ./fiasco Fiasco-UX on Linux 2.6.17-2-k7 (i686) Mapped 64 MB Memory + 0 KB Framebuffer + 0 MB Input Area on FD 3 Loading Module 0x00090000-0x00096ce4 [sigma0-ux] Loading Module 0x00120000-0x002bda14 [roottask] Bootstrapping... Welcome to Fiasco(ux)! DD-L4(v2)/x86 microkernel (C) 1998-2006 TU Dresden Rev: Unknown compiled with gcc 4.1.2 for Intel Pentium CPU: AuthenticAMD (F:23:2:0) Model: Athlon 64 X2 (Toledo) at 2418 MHz 32/512 Entry I TLB (4K pages) 8 Entry I TLB (4M pages) 32/512 Entry D TLB (4K pages) 8 Entry D TLB (4M pages) 64 KB L1 I Cache (2-way associative, 64 bytes per line) 64 KB L1 D Cache (2-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) ASSERTION_FAILED (((((*(int *) &(status))) & 0xff) == 0x7f) && ((((*(int *) &(status))) & 0xff00) >> 8) == 10) in function static unsigned int Hostproc::create(unsigned int) in file /home/tews/Fiasco/opentc-snap/l4/kernel/fiasco/src/kern/ux/hostproc.cpp:120 which is the same problem as before. Bye, Hendrik

2 1

TC and L4
by Burak OÐUZ 23 Nov '06

23 Nov '06

Hi, I have a few questions about L4 and trusted platforms. Are there any implementation of L4 which supports TSS 1.2 stack? Are there any interfaces supplied by L4 to access TPM? Are there any trusted platform application on L4? Thanks in advance.. Best regards.. -- burak() (ps: bf) METU CENG '06

4 3

Performance counter and NMI handling
by Jugwan Eom 22 Nov '06

22 Nov '06

Hi, I'm working to implement the Linux Oprofile functionality on L4/Fiasco environment. OProfile configures the performance counter to raise a NMI whenever the counter overflows. I modified the NMI handling codes in handle_slow_trap() to collect samples but I encountered an strange system behavior, REBOOTING without any information. If there is no user application activity, the system looks fine. Otherwise, the system reboots even if a very simple application like hello runs and the required time before reboot varies. I don't know what causes this behavior. IMHO, it seems a kind of triple fault. Is there any one give me an advice? Thanks in advance. -- Jugwan Eom

2 7

(no subject)
by Christian Roepke 21 Nov '06

21 Nov '06

Hello Ryan, try this one: http://www.ertos.nicta.com.au/downloads/tools/arm-linux-3.4.4.tar.gz More information here: http://www.ertos.nicta.com.au/software/prebuilt/binaries.pml Kind regards, Hans > On Mon Nov 20, 2006 at 10:30:59 -0700, Ryan C. Spring wrote: >> I am cross-compiling the DROPS system for ARM and get compiler errors >> about "too many initializers" when it compiles the following statement >> from syscalls_impl.h (line 249) >> >> >> l4_threadid_t preempter = L4_INVALID_ID; >> l4_threadid_t pager = L4_INVALID_ID; >> >> >> When I initialize the unions without the defines, by hand as, it compiles. >> >> l4_threadid_t preempter; >> preempter.raw = (l4_umword_t)-1; >> l4_threadid_t pager; >> pager.raw = (l4_umword_t)-1; >> >> I'm using a TI-provided GCC 3.3.2 ARM crosscompiler. Is my >> cross-compiler to blame here? If so, could someone recommend a >> cross-compiler toolchain for ARM that doesn't error on this syntax? > > I'm using a self-made one, 3.4 based, which works. 4.1 is also known to > work. I can't remember we have tried 3.3 in the past. > > > > Adam -- Hans Christian Röpke Network Administration Applied Data Security Group Ruhr-Universität Bochum, IC 4/42 Universitätsstrasse 150 44780 Bochum, Germany http://www.prosec.rub.de

1 0

Re: Re: Timing problem on mem_fpage_unmap() in the fiasco kernel
by 조완근 21 Nov '06

21 Nov '06

Hi, I solved the timing problem. I just unset the "Show timing information on printks" in "kernel hacking". I found that when the system tries to free INITRD and vprintk() function tries to access any local buffer such as "char tbuf[50]", the system hangs on. I also found all the config files in "l4linux-2.6/arch/arm/configs" unset the option. The option can't work for arm architecture??? Do you know anything about that?? Thank you.

2 1

bug in cross-compiler?
by Ryan C. Spring 20 Nov '06

20 Nov '06

I am cross-compiling the DROPS system for ARM and get compiler errors about "too many initializers" when it compiles the following statement from syscalls_impl.h (line 249) l4_threadid_t preempter = L4_INVALID_ID; l4_threadid_t pager = L4_INVALID_ID; When I initialize the unions without the defines, by hand as, it compiles. l4_threadid_t preempter; preempter.raw = (l4_umword_t)-1; l4_threadid_t pager; pager.raw = (l4_umword_t)-1; I'm using a TI-provided GCC 3.3.2 ARM crosscompiler. Is my cross-compiler to blame here? If so, could someone recommend a cross-compiler toolchain for ARM that doesn't error on this syntax? Regards, Ryan -- Ryan C. Spring Doctoral Student University of Texas at El Paso +1 915-539-7311 www.spring-west.net/ryan

2 1

Re: Does the L4/Fiasco boot from CF card?
by chenggh 18 Nov '06

18 Nov '06

On Saturday 18 November 2006 18:42, Martin Pohlack wrote: > Hi Cheng, > > chenggh wrote: > > Hi, > > Today I want to boot the L4/Fiasco/L4Linux in the CF card. I copy all > > the necessary files into the CF card and install the Grub-0.97-os.1 in > > this CF card. My test machine is a VIA development board and in this > > board the CF card is connected with this machine by the IDE plugin. > > My grub menu.lst is as following: > > > > title L4Linux26/Fiasco+dope > > kernel (hd0,0)/boot/L4Linux/bootstrap > > modaddr 0x06000000 > > module (hd0,0)/boot/L4Linux/main -nowait -nokdb -serial_esc > > -comspeed 115200 -comport 1 > > module (hd0,0)/boot/L4Linux/sigma0 > > module (hd0,0)/boot/L4Linux/roottask task modname "bmodfs" > > attached 4 modules > > module (hd0,0)/boot/L4Linux/events > > module (hd0,0)/boot/L4Linux/names --events > > module (hd0,0)/boot/L4Linux/log --events > > module (hd0,0)/boot/L4Linux/dm_phys --isa=0x00800000 -v --events > > module (hd0,0)/boot/L4Linux/simple_ts -t 300 --events > > module (hd0,0)/boot/L4Linux/rtc --events > > module (hd0,0)/boot/L4Linux/l4io --noirq --events > > module (hd0,0)/boot/L4Linux/l4exec --events > > module (hd0,0)/boot/L4Linux/l4dope --l4io --menubar > > --transparency module (hd0,0)/boot/L4Linux/loader --fprov=BMODFS > > linux26.cfg module (hd0,0)/boot/L4Linux/bmodfs > > module (hd0,0)/boot/L4Linux/vmlinuz26 > > module (hd0,0)/boot/L4Linux/linux26.cfg > > module (hd0,0)/boot/L4Linux/libloader.s.so > > module (hd0,0)/boot/L4Linux/libld-l4.s.so > > vbeset 0x117 506070 > > > > In the PC I have tested this configuration file and it is ok. Wh en I > > boot the VIA machine I choose the only one option to have this results. > > Booting L4Linux26/Fiasco+dope > > kernel (hd0,0)/boot/L4Linux/bootstrap > > > > Error 13 Invalid or unsupported executable format > > > > pass anykey to continue... > > Maybe the CF card reader is not fully ATA compatible? Maybe it emulates > SCSI? Even if it emulates an ATA drive, are you sure it is visible as > (hd0) to grub? Maybe you have to change the paths. I think the hd0 is visible to the grub. when I modify the grub option when booting I could use Tab-Completion in the grub command interface. But I change the path again. Thank you. cheng

1 0

Does the L4/Fiasco boot from CF card?
by chenggh 18 Nov '06

18 Nov '06

Hi, Today I want to boot the L4/Fiasco/L4Linux in the CF card. I copy all the necessary files into the CF card and install the Grub-0.97-os.1 in this CF card. My test machine is a VIA development board and in this board the CF card is connected with this machine by the IDE plugin. My grub menu.lst is as following: title L4Linux26/Fiasco+dope kernel (hd0,0)/boot/L4Linux/bootstrap modaddr 0x06000000 module (hd0,0)/boot/L4Linux/main -nowait -nokdb -serial_esc -comspeed 115200 -comport 1 module (hd0,0)/boot/L4Linux/sigma0 module (hd0,0)/boot/L4Linux/roottask task modname "bmodfs" attached 4 modules module (hd0,0)/boot/L4Linux/events module (hd0,0)/boot/L4Linux/names --events module (hd0,0)/boot/L4Linux/log --events module (hd0,0)/boot/L4Linux/dm_phys --isa=0x00800000 -v --events module (hd0,0)/boot/L4Linux/simple_ts -t 300 --events module (hd0,0)/boot/L4Linux/rtc --events module (hd0,0)/boot/L4Linux/l4io --noirq --events module (hd0,0)/boot/L4Linux/l4exec --events module (hd0,0)/boot/L4Linux/l4dope --l4io --menubar --transparency module (hd0,0)/boot/L4Linux/loader --fprov=BMODFS linux26.cfg module (hd0,0)/boot/L4Linux/bmodfs module (hd0,0)/boot/L4Linux/vmlinuz26 module (hd0,0)/boot/L4Linux/linux26.cfg module (hd0,0)/boot/L4Linux/libloader.s.so module (hd0,0)/boot/L4Linux/libld-l4.s.so vbeset 0x117 506070 In the PC I have tested this configuration file and it is ok. Wh en I boot the VIA machine I choose the only one option to have this results. Booting L4Linux26/Fiasco+dope kernel (hd0,0)/boot/L4Linux/bootstrap Error 13 Invalid or unsupported executable format pass anykey to continue... I tried it many times and I think it could recognize the file and file system (now it is ext3 and I have ever tried ext2). But I don't know why is this prolem. Best wishes. Thank you. cheng

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

l4-hackers November 2006