l4-hackers August 2014

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

20 participants
35 discussions

Graphical desktop, l4con, mag
by teclis High Elf 02 Jun '17

02 Jun '17

Guten Tag, I've got the Hello app runing with fiasco and l4re in qemu, how do I get the graphical UI and console running that I see on the screen shots page? Do I need to use l4con and mag? I don't see an entry with l4con in modules.list. danke * <http://german.about.com/library/media/Audio/gutentag.wav>*

6 15

shared dataspace for l4re_kernel/ registering additional caps in ned
by Stark, Josef 12 Mar '15

12 Mar '15

Hello all, for a project, I want to be able to load (elf-)binaries not only from the filesystem, but also from the network (receive binary, then start it.) For this, I will write a L4 app (let's call it "manager" for now) which handles the network part and then tells ned to start the received binary via a special "filename". I already modified ned so that its LUA-interface is exposed to other apps via a capability. But here comes the first problem: Normally you specify the capability table of a task and an IPC channel in the lua-config file, which ned reads. Ned then prepares the channel and the caps table and starts the task, which then can register the capability. Otherwise, registering fails (the example would do a printf("Could not register my service, is there a 'calc_server' in the caps table?\n");) So... how do I specify which capabilities ned itself can have/register? Currently, I need to have a lua-config which tells ned to start another instance of ned with the required permission to register the capability. But is there any way to avoid having two ned instances and making the first (and only) instance able to register the cap? Second problem: I found out, that ned (also moe) doesn't start the desired app directly but instead starts the l4re binary (source located in src/l4/pkg/l4re_kernel), passing it the filename of the desired binary. l4re, once started, gets the dataspace of this file and then ultimately starts the app. For my purpose, I would modify l4re_kernel so that if it detects the special "filename" mentioned above, it does not read the file from the filesystem but loads the binary that the "manager" received over the network. To get the data from the manager to the l4re_kernel instance, I thought that a shared dataspace or a shared memory area l4shmc_area_t like in prdocons.c example would be suitable. However, again, both of these things normally require special parameters in the lua config file (as can be seen in the examples), but I have no clue how to do this to l4re_kernel since it is not loaded from such a file, but loaded implicitly each time a new task is started. (so I guess this makes it even harder than the problem with ned since ned is only started once.) Does anyone have ideas on how to solve these? If someone finds that I phrased this too opaque, please let me know. Best regards, Josef Stark

3 16

cannot find -lld-l4
by ba_f 25 Sep '14

25 Sep '14

Hello l4-hackers, I get an error when trying to compile, respectively linking my program. /opt/gcc-linaro-arm-linux-gnueabihf-4.8-2014.03_linux/bin/arm-linux-gnueabihf-ld: cannot find -lld-l4 make[3]: *** [myProg] Fehler 1 My Program does not include this lib. It comes from the L4-make process (prog.mk). I have already compiled smaller programs, without problems. BTW before that missing lib error, i get a lot of "multiple definition of...". Any ideas? I use Debian 7 32bit. Thanks

3 4

l4/sys/syscalls.h: No such file or directory
by Valentin Hauner 18 Sep '14

18 Sep '14

Hi, I'm using the following example provided by Michael Hohmuth in 2000 to create new tasks and threads: http://os.inf.tu-dresden.de/pipermail/l4-hackers/2000/000384.html I've placed the code in a copy of the hello example, but added > REQUIRES_LIBS = l4sys to the Makefile. The directory structure is located at './l4/pkg/taskcreat'. Nevertheless, compilation is terminated due to a fatal error: > ./l4/pkg/taskcreat/server/src/main.c:5:29: fatal error: l4/sys/syscalls.h: No such file or directory Why is syscalls.h missing here? Best regards, Valentin

4 52

Question about dec_lock_cnt method in Context class
by Yuxin Ren 16 Sep '14

16 Sep '14

Hi All, I have a question about the dec_lock_cnt method in Context class under multiple processor. In its implementation, it checks if thread's home cpu is equal to current cpu. If not, it does not unset "_running_under_lock" variable, even if the _lock_cnt is 0. Why does it check if home cpu is equal to current cpu? Consider the following case: Thread A is running on core 1 holding a helping lock, but is preempted by another high priority thread. Thread B is running on core 2, and tries to take the same helping lock. It finds the lock is held by A, so it helps A to run by migrating A to core 2. When A releases its lock, finding the current cpu is not its home cpu, so it does not set "_running_under_lock" to false. Then A continues to run on core 1 and tries to take that lock again. But now it can never take the lock, as the implementation of helping lock. PRIVATE inline bool NO_INSTRUMENT Switch_lock::set_lock_owner(Context *o) { bool have_no_locks = o->_lock_cnt < 1; if (have_no_locks) { assert_kdb (current_cpu() == o->home_cpu()); for (;;) { if (EXPECT_FALSE(access_once(&o->_running_under_lock))) continue; if (EXPECT_TRUE(mp_cas(&o->_running_under_lock, Mword(false), Mword(true)))) break; } } ... } Do I misunderstand anything here? Thank you very much. Best, Yuxin

2 8

Checking executables before running them in L4re/fiasco
by Masti Ramya Jayaram 09 Sep '14

09 Sep '14

Dear all, I am using fiasco/l4re to run applications (say hello world) and L4linux. The module list in the two cases looks like this: entry hello-cfg kernel fiasco -serial_esc -freq=533000 roottask moe rom/hello.cfg module l4re module ned module hello.cfg module hello entry l4linux kernel fiasco -serial_esc -freq=533000 roottask moe rom/l4linux.cfg module l4re module ned module l4linux.cfg module io module vmlinuz module ramdisk-x86.rd I would like to add some functionality that performs a few checks on hello and vmlinuz before they are executed. In this context, I have the following questions: a. Where is it best to implement such functionality - moe or ned? Or should I implement a separate server and do this? b. I need access to an special range of IO memory when I perform the checks. i guess I could always use the io module and do "l4io_request_iomem()" . Would that be the right way to do it from moe/ned? c. Is there a way to allocate/reserve a dedicated range of physical memory to these applications (hello/vmlinuz) so that they get allocated the same chunk everytime (of physical memory)? Thanks in advance, ramya

2 12

Minimal requirements to run L4linux using only a serial console
by Masti Ramya Jayaram 30 Aug '14

30 Aug '14

Dear all, I am trying to run L4linux on fiasco/l4re but with the keyboard and VGA disabled. I can get all the way to the command prompt (of the shell) in the initramfs but then, on pressing a key, I get a page fault and I cannot debug the system anymore using the serial_esc. I suspect it is my configuration which I have provided below. I was wondering what the minimal packages are to achieve this. modules.list: entry l4linux kernel fiasco -serial_esc -freq=533000 roottask moe rom/l4linux.cfg module l4re module ned module l4linux.cfg module io module vmlinuz module drops-rd.rd I have disabled the framebuffer and input server in the L4linux stub drivers configuration and enabled the serial console. l4linux.cfg: --vim:set ft=lua: L4.default_loader:start({ caps = { log = L4.Env.log:m("rws"), }, l4re_dbg = L4.Dbg.Warn, log = { "l4linux", "yellow" }, }, "rom/io -vvvv rom/vmlinuz mem=64M console=ttyLv0 l4x_rd=rom/drops-rd.rd root=1:0 ramdisk$ I also specify a custom frequency for my board to fiasco.oc, do I have to make similar modifications to l4linux (e.g., Timer frequency)? Thanks, Ramya

3 14

RE: Minimal requirements to run L4linux using only a serial console
by Masti Ramya Jayaram 29 Aug '14

29 Aug '14

Hi Martin, That was a very useful tip. I simply changed the default value in kernel/fiasco/src/kern/config.cpp from SERIAL_ESC_IRQ to SERIAL_ESC_NOIRQ which leads to polling (I am guessing) and voila! It works! I even tried on the actual SCC and it works! I can break into the debugger as well as input to l4linux. Thanks a lot for your help for your help. Best, Ramya ________________________________________ From: Masti Ramya Jayaram Sent: 29 August 2014 17:48 To: Martin Schröder; l4-hackers(a)os.inf.tu-dresden.de Subject: RE: Minimal requirements to run L4linux using only a serial console Also with acpi on, it does enter the debugger in qemu. Best, Ramya ________________________________________ From: Masti Ramya Jayaram Sent: 29 August 2014 17:45 To: Martin Schröder; l4-hackers(a)os.inf.tu-dresden.de Subject: RE: Minimal requirements to run L4linux using only a serial console Hey Martin, No I just realized that ESC does not break into the debugger. But I can clearly see that the process instead of printing "Hello world" prints "ello World" :) What could the reason be? Thanks, Ramya ________________________________________ From: l4-hackers [l4-hackers-bounces(a)os.inf.tu-dresden.de] on behalf of Martin Schröder [martin.schroeder(a)openlimit.com] Sent: 29 August 2014 17:35 To: l4-hackers(a)os.inf.tu-dresden.de Subject: Re: Minimal requirements to run L4linux using only a serial console Hi Ramya, a: No, its not a must. The UART could also be polled (by a timer) to see if something arrived. b: At least regarding l4linux there is nothing more to it - afaik. But let's go back to the hello word example on the SCC. What happens if you press the <ESC> key? Does it break into the JDB debugger? Best regards, Martin Am 29.08.2014 um 16:56 schrieb Masti Ramya Jayaram: > Hey Martin, > > I am trying to run l4linux/fiasco on the Intel SCC which does have a pic-8259. The Intel SCC does not have a BIOS and does not have any ACPI tables which I think qemu generates on the fly. It also does not use a Screen or keyboard. > > I obtained a patch from Markus Partheymueller which they used to boot fiasco on the SCC. With it I am able to run hello world both in qemu with no-acpi as well as on the SCC successfully. > > The patch to the best of my knowledge does the following: > > a. Makes the cmdline also parse an argument so that the frequency is an input (533Mhz) > b. Uses the apic timer for scheduling. > c. Disables the pic-8259 because it does not exist. Instead the core in the SCC has two interrupt lines and this patch initializes these lines as interrupts to the core. > d. Disables keyboard and screen and enables a serial line for communication instead > e. Disables super pages because the SCC does not support it. > > So, when I try the patched fiasco/L4linux on qemu, it works when I do not disable acpi. But since the SCC does not have any ACPI/IO-APIC, when I try to run the same thing on qemu with ACPI off, l4linux boots but is not responsive to input. > > In short: > > when I run with I (below) , my l4linux responds to input and with II (below) not. > > qemu-system-i386 -m 256 -serial stdio -kernel bootstrap_l4linux.elf -append "-arg=fiasco,,-serial_esc -freq=533000" > > qemu-system-i386 -no-acpi -m 256 -serial stdio -kernel bootstrap_l4linux.elf -append "-arg=fiasco,,-serial_esc -freq=533000" > > Now the question is: > > a. Is the IO-APIC a must for the serial line? > b. Are ACPI tables just a matter of writing data at the right memory locations (e..g, qemu lets you give the acpi-table as input I noticed)? Or is there more to it? > > Thanks for your patience, > ramya > > > ________________________________________ > From: l4-hackers [l4-hackers-bounces(a)os.inf.tu-dresden.de] on behalf of Martin Schröder [martin.schroeder(a)openlimit.com] > Sent: 29 August 2014 16:29 > To: l4-hackers(a)os.inf.tu-dresden.de > Subject: Re: Minimal requirements to run L4linux using only a serial console > > Hi Ramya, > > Why did you disable the PIC? > What is your aim with ACPI and why should this lead to a change in the UART driver? > And what has the serial console to do with ACPI tables? > > For me the topics have at the moment very little in common or I am missing something. > > Best regards, > Martin > > > Am 29.08.2014 um 15:58 schrieb Masti Ramya Jayaram: >> Also, given that the thing works with ACPI on, wouldn't it be better to include ACPI information in the boot instead of changing the driver? Or is this impossible? >> >> Best, >> Ramya >> >> >> ________________________________________ >> From: Masti Ramya Jayaram >> Sent: 29 August 2014 15:55 >> To: Martin Schröder; l4-hackers(a)os.inf.tu-dresden.de >> Subject: RE: Minimal requirements to run L4linux using only a serial console >> >> Hey Martin, >> >> Back on the list ;-) >> >> Ok, I can confirm that when I busy loop during fiasco init, I can actually see the character that I typed. So would I have to change the l4ser driver in Linux or the main driver in under pkg/l4/bootstrap or under fiasco in vkey.cpp? :) >> >> Thanks, >> ramya >> ________________________________________ >> From: Masti Ramya Jayaram >> Sent: 29 August 2014 12:15 >> To: Martin Schröder; l4-hackers(a)os.inf.tu-dresden.de >> Subject: RE: Minimal requirements to run L4linux using only a serial console >> >> Hey Martin, >> >> I cleaned up my build and it worked - the mistake was mine. I had a question regarding ACPI. The target machine I am using has no BIOS/ACPi support. So, when I boot linux on it my console input does not work.This is similar to booting in qemu with the -no-acpi option. Serial console output works (I can see the prompt of L4Linux) but input does not. >> >> Is there something special to do when booting with a serial console but no ACPI tables? >> >> Thanks, >> ramya >> ________________________________________ >> From: Martin Schröder [martin.schroeder(a)openlimit.com] >> Sent: 27 August 2014 18:32 >> To: Masti Ramya Jayaram; l4-hackers(a)os.inf.tu-dresden.de >> Subject: Re: Minimal requirements to run L4linux using only a serial console >> >> Hi Masti, >> >> you have all of them already! I did not alter any other file beside the files you already sent. I inserted them into a fresh >> snapshot (http://l4re.org/download/snapshots-oc/l4re-snapshot-2014053111.tar.xz) and started building. Since "drops-rd.rd" was not >> found in the snapshot, I just renamed "ramdisk-x86.rd" to "drops-rd.rd". Nothing more. >> >> Martin >> >> >> >> Am 27.08.2014 um 18:02 schrieb Masti Ramya Jayaram: >>> Hey Martin, >>> >>> thanks for the quick response. Could you send me your configs (l4linux, fiasco, l4re) so that I can try it with that? >>> >>> Thanks, >>> Ramya >>> ________________________________________ >>> From: l4-hackers [l4-hackers-bounces(a)os.inf.tu-dresden.de] on behalf of Martin Schröder [martin.schroeder(a)openlimit.com] >>> Sent: 27 August 2014 16:28 >>> To: l4-hackers(a)os.inf.tu-dresden.de >>> Subject: Re: Minimal requirements to run L4linux using only a serial console >>> >>> Hi Masti, >>> >>> I just applied your config, replaced "drops-rd.rd" by "ramdisk-x86.rd" and everything (command prompt, debugger) runs fine as >>> expected in Qemu. Maybe you ran into a similar problem as described in this thread: >>> >>> http://os.inf.tu-dresden.de/pipermail/l4-hackers/2012/thread.html#5165 >>> >>> >>> Best regards, >>> Martin >>> >>> >>> Am 27.08.2014 um 15:18 schrieb Masti Ramya Jayaram: >>>> Sorry, a couple of typos in the l4linux.cfg >>>> --vim:set ft=lua: >>>> >>>> >>>> L4.default_loader:start({ caps = { >>>> log = L4.Env.log:m("rws"), >>>> }, >>>> l4re_dbg = L4.Dbg.Warn, >>>> log = { "l4linux", "yellow" }, >>>> }, "rom/vmlinuz mem=64M console=ttyLv0 l4x_rd=rom/drops-rd.rd root=1:0 ramdisk_size=4000 l4ser.vkey_enable=1"); >>>> >>>> Thanks in advance, >>>> ramya >>>> >>>> ________________________________________ >>>> From: l4-hackers [l4-hackers-bounces(a)os.inf.tu-dresden.de] on behalf of Masti Ramya Jayaram [rmasti(a)inf.ethz.ch] >>>> Sent: 27 August 2014 14:07 >>>> To: l4-hackers(a)os.inf.tu-dresden.de >>>> Subject: Minimal requirements to run L4linux using only a serial console >>>> >>>> Dear all, >>>> >>>> I am trying to run L4linux on fiasco/l4re but with the keyboard and VGA disabled. I can get all the way to the command prompt (of the shell) in the initramfs but then, on pressing a key, I get a page fault and I cannot debug the system anymore using the serial_esc. I suspect it is my configuration which I have provided below. I was wondering what the minimal packages are to achieve this. >>>> >>>> modules.list: >>>> entry l4linux >>>> kernel fiasco -serial_esc -freq=533000 >>>> roottask moe rom/l4linux.cfg >>>> module l4re >>>> module ned >>>> module l4linux.cfg >>>> module io >>>> module vmlinuz >>>> module drops-rd.rd >>>> >>>> I have disabled the framebuffer and input server in the L4linux stub drivers configuration and enabled the serial console. >>>> >>>> l4linux.cfg: >>>> >>>> --vim:set ft=lua: >>>> >>>> L4.default_loader:start({ caps = { >>>> log = L4.Env.log:m("rws"), >>>> }, >>>> l4re_dbg = L4.Dbg.Warn, >>>> log = { "l4linux", "yellow" }, >>>> }, "rom/io -vvvv rom/vmlinuz mem=64M console=ttyLv0 l4x_rd=rom/drops-rd.rd root=1:0 ramdisk$ >>>> >>>> >>>> I also specify a custom frequency for my board to fiasco.oc, do I have to make similar modifications to l4linux (e.g., Timer frequency)? >>>> >>>> Thanks, >>>> Ramya _______________________________________________ l4-hackers mailing list l4-hackers(a)os.inf.tu-dresden.de http://os.inf.tu-dresden.de/mailman/listinfo/l4-hackers

1 0

Re: Minimal requirements to run L4linux using only a serial console
by Martin Schröder 29 Aug '14

29 Aug '14

Hi Ramya, Am 29.08.2014 um 14:29 schrieb Masti Ramya Jayaram: > Hi Martin, > > Thanks for checking that. I also see that my build works provided I don't apply a special patch meant for my board (Intel SCC). The patch essentially only alters the frequency and disables the PIC. If you disable the PIC you won't get any character into your system since UART reception is normally interrupt based. If you want to change this you have to poll the UART FIFO in the driver every second or so if there is any character available. Best regards, Martin.

1 0

Re: advice on using ddelinux
by Björn Döbel 28 Aug '14

28 Aug '14

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 28.08.2014 12:14, Korbinian Ederer wrote: > Yes the symlinks are build properly, but you're right with the > CFlags. The include dir is not listed. How do I tell the make > process to do so? My bad, the variable is called PRIVATE_INCDIR as you can see in all the example Makefiles around the l4/pkg/dde subdirectory. Cheers, Bjoern - -- Dipl.-Inf. Bjoern Doebel Mail: doebel(a)tudos.org TU Dresden, OS Chair Phone: +49 351 463 38 799 Noethnitzer Str. 46 Fax: +49 351 463 38 284 01187 Dresden, Germany WWW: http://www.tudos.org/~doebel - -- "When the seagulls follow the trawler, it's because they think sardines will be thrown into the sea." (Eric Cantona) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1 iEYEARECAAYFAlP/E0sACgkQP5ijxgQLUNmhTgCfbaw/Y+IIysCiwasxjvuVSxEP BdkAoJWEcSHWRfTedvXj0VotJv+1zVvn =o8Pq -----END PGP SIGNATURE-----

1 0

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 August 2014