August 2009 - l4-hackers - example.com Site

Back to UTCB IPC
by Marc CHALAND 11 Oct '09

11 Oct '09

Hi, I am back with my problem with UTCB IPC. To sum up the context : I use pure l4env applications. Here is the scenario between my server and my client : 1. client makes an UTCB IPC to manager. Manager does an l4rm_area_reserve, creates a rcv_fpage, creates worker and sends back worker thread id to client. 2. client allocates memory and associated flexpage and sends it to server. 3. UTCB IPC is sent to worker with 3 parameters : - on client side : utcb->values=0xbff00200: arg1=0x2 arg2=0x1 arg3=0x6 - on server side : utcb->values=0xbff00200: arg1=0x241401 arg2=0x0 arg3=0x7069776c It seems that 0xbff00200 is not mapped on the same memory on client and server side. As there is no problem for first IPC (utcb->values is also 0xbff00200), is there any potential problem with flexpage ? As utcb->values is still the same, what happens if several threads try to make UTCB IPC ? Another question : when I link my application with liblogserver.a, I have strange logs : .rm.sem.mainMain function returned. It seems that .rm .sem and .main are added by l4env, but I didn't find where. Regards Marc

4 9

Re: can server side get the thread id of client side?
by fang zhengshu 30 Aug '09

30 Aug '09

hi all, 3. Re: can server side get the thread id of client side? > (Bj?rn D?bel) > Message: 3 > Date: Fri, 28 Aug 2009 09:40:17 +0200 > From: Bj?rn D?bel <doebel(a)os.inf.tu-dresden.de> > To: L4 Hackers <l4-hackers(a)os.inf.tu-dresden.de> > Subject: Re: can server side get the thread id of client side? > Message-ID: <4A9789E1.7080503(a)os.inf.tu-dresden.de> > Content-Type: text/plain; charset=ISO-8859-1 > > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi, > > > I know that the client side can get the server's thread_id throw name > > service. I want to ask can the server side get the client's task_id or > > thread_id? > > or must the client side send the thread_id to server? > > your server will call either l4_ipc_wait() or l4_ipc_reply_and_wait() in > order to receive the message from the client. In both cases there is a > src parameter which will contain the client's thread ID after the system > call returns. > sorry for not explaining it clearly. if I defined two functions in idl file, just like: interface test{ int open(int fd, int flags); int close(fd); }; and I implement them in the server side. I know that when I use open()/close() in the client side, the server will call the function open_component()/close_component() which are in the server side. In the open_component()/close_component() function I want to get thread id of the caller to do some operation flag.but as we know, when we call open()/close() in client side, the open_component()/close_component() function in server side will be executed. so we can not use l4_ipc_wait() or l4_ipc_reply_and_wait() function in the open_component()/close_component() function, or it will be stop and wait for a message in the function. So in order to get the thread id in the open_component()/close_component(), when or where should I use the l4_ipc_wait() or l4_ipc_reply and wait() to get the thread id? thank you! -- fang,

1 0

can server side get the thread id of client side?
by fang zhengshu 28 Aug '09

28 Aug '09

hi all, I want to write a simple example in l4, and there are to task module: client,server in the client side, I use the interface which is inplemented in server side.like: int open(int fd, int flags) I know that the client side can get the server's thread_id throw name service. I want to ask can the server side get the client's task_id or thread_id? or must the client side send the thread_id to server? thank you. -- fang,

2 1

Re: l4-hackers Digest, Vol 76, Issue 10
by fang zhengshu 28 Aug '09

28 Aug '09

hi all, I want to write a simple example in l4, and there are to task module: client,server in the client side, I use the interface which is inplemented in server side.like: int open(int fd, int flags) I know that the client side can get the server's thread_id throw name service. I want to ask can the server side get the client's task_id or thread_id? or must the client side send the thread_id to server? thank you? -- fang,

1 0

one question about l4_thread_ex_regs flags
by Guanghui, Cheng 26 Aug '09

26 Aug '09

Hello: About the l4_thread_ex_regs system call there are 2 flags related with exception ipc. I only know the thread with these flags could generate the exception ipc but i don't know what is the difference between these flags: L4_THREAD_EX_REGS_RAISE_EXCEPTION , L4_THREAD_EX_REGS_ALIEN and L4_THREAD_EX_REGS_ALIEN | L4_THREAD_EX_REGS_ALIEN. So, i write a following program to test the behaviour with different flags : void idle(void *data) { l4_sleep(1000); l4_sleep(1000); l4_sleep(1000); l4_sleep(1000); l4_sleep(1000); } int main() { ... l4_threadid_t me = l4_myself(); l4thread_create(idle, NULL, L4THREAD_CREATE_ASYNC); ... l4_thread_ex_regs_flags(idle_thread, ~0UL, ~0UL, &_preempter, &me, &o, &o, &o, L4_THREAD_EX_REGS_RAISE_EXCEPTION | L4_THREAD_EX_REGS_ALIEN); while(1) { l4_ipc_wait(&src, ...); printf("thread :%d:%d\n", src.id.task, src.id.lthread); } } I tried 3 combination about these 2 flags but the output is the same as this: exceptio| thread :8:3 i don't know why there is only one exception ipc. But the idle function includes 5 system calls. Additionally, it seems the ALIEN state is TEMPORARILY. I set once there is one exception ipc, twice two exception ipc. Can you explain me? Thanks. Cheng Guanghui

3 4

Re: seL4 kernel is formally verified
by gernot＠cse.unsw.EDU.AU 18 Aug '09

18 Aug '09

On Tue, August 18, 2009 10:18 am, John van V. wrote: > Doesn't load. Apologies, that was the link to the internal stage site. The correct link is http://ertos.nicta.com.au/research/l4.verified/ Gernot

1 0

seL4 kernel is formally verified
by Gernot Heiser 18 Aug '09

18 Aug '09

Hi, For those out there who missed the publicity (incl ./): NICTA announced last week that the formal verification of our seL4 kernel has been completed. This makes seL4 the first ever general-purpose OS kernel with a formal proof that the implementation is consistent with the specification. A paper on the project will appear in this year's SOSP. Details at http://ertos.nicta.com.au/news/home.pml#verified Gernot

5 7

Howto implement a periodic real-time task?
by Andre Puschmann 10 Aug '09

10 Aug '09

Hey guys, I got a L4 real-time application that should run strictly periodic. In each cycle I have to communicate with at least two other L4 threads in different address spaces. The cycle period is in the order of 1000Hz to 5000Hz. As described in [1] I tried to use the CPU reservation server but the sample implementations did not work. Anyway, I figured out that even the CPU reservation server depends on the kernel timer and thus the granularity is 1ms (in PIT mode), which is too coarse for cycle periods larger then 1000Hz. Although I like the idea to use OS primitives to keep the application as portable as possible (ARM port is a goal), I thought about implementing the periodic execution with an extra timer decoupled from the scheduling timer which generates an IRQ at each hit. Any comment/suggestions/ideas about that? Any other experience with periodic real-time tasks on L4? Cheers, Andre [1] http://os.inf.tu-dresden.de/pipermail/l4-hackers/2009/004179.html

3 6

Re: problems running l4linux No.2
by Stefan Kögel 06 Aug '09

06 Aug '09

Hi, I started from a default l4linux26 kernel config. I did not touch anything, except for the things I had to, like acpi, apic, MTRR, MCE .. and the stuff that is missing (olpc, scx200 and these things). Everything else is default. The problem still exists but I was able to gather some info. Executing the l4linux26.cfg prints out: loader | sleeping for 1000 ms loader | vmlinuz: Starting application using libld-l4.s.so loader | vmlinuz,#11: Loading binary loader | vmlinuz,#11: Loading ldso loader | vmlinuz,#11: Starting libld-l4.s.so at 00013940 via 0000cc3c loader | DEBUG file = vmlinuz l4lx | ======> L4Linux 2.6 starting... <======== l4lx | Linux version 2.6.30-l4 (stefan@scarface) (gcc version 4.3.2 (Ubuntu l4lx : 4.3.2-1ubuntu12) ) #10 Wed Aug 5 16:45:57 CEST 2009 l4lx | Binary name: vmlinuz l4lx | Linux kernel command line (5 args): mem=64M load_ramdisk=1 ramdisk_si l4lx : ze=16384 root=/dev/ram l4env_rd=drops-rd.rd l4lx | Image: 00400000 - 00874000 [4560 KiB]. l4lx | Areas: Text: 00400000 - 006fd000 [3060kB] (a bit longer) l4lx | Data: 006fd000 - 007280e8 [172kB] l4lx | Initdata: 0072c000 - 00770000 [272kB] l4lx | BSS: 00772000 - 00853a94 [902kB] l4lx | l4lx_thread_create: Created thread 11.03 (tamer0) l4lx | Tamer0 is 11.03 l4lx | Using tamed mode. ROOT: Task #0c is not allowed to execute cli/sti loader | vmlinuz,#11: WARNING: Can't map I/O space, ROOT denies page (result=0 loader : 0004000) loader | vmlinuz,#11: Not allowed to perform any I/O l4lx | Got 0 out of 65536 I/O ports l4lx | Connecting to l4io server. l4lx | l4env_linux_startup thread 4. l4lx | l4lx_thread_create: Created thread 11.04 (cpu0) l4lx | main thread will be 11.04 l4lx | l4env_register_pointer_section: addr = 0072a000 size = 290816 l4lx | sec-w-init: virt: 0x0072a000 to 0x00770fff [284 KiB] l4lx | Noncontiguous region for sec-w-init l4lx | sec-w-init: Number of physical regions: 2, 290816 Bytes l4lx | sec-w-init: 1: Phys: 0x014f0000 to 0x01500000, Size: 65536 l4lx | sec-w-init: 2: Phys: 0x013b5000 to 0x013ec000, Size: 225280 l4lx | main thread: received startup message. l4lx | Main thread running, waiting... l4lx | memory_setup: mem=64M load_ramdisk=1 ramdisk_size=16384 root=/dev/ram l4lx : l4env_rd=drops-rd.rd l4lx | setup_l4env_memory: Forcing superpages for main memory l4lx | Main memory size: 64MB l4lx | Main memory: virt: 0x00c00000 to 0x04bfffff [65536 KiB] l4lx | Main memory: Number of physical regions: 1, 67108864 Bytes l4lx | Main memory: 1: Phys: 0x07400000 to 0x0b400000, Size: 67108864 l4lx | Filling lower ptabs... l4lx | mainmem = c00000 l4lx | Done (1864 entries). l4lx | l4env_register_pointer_section: addr = 00772000 size = 1056768 l4lx | end: virt: 0x00772000 to 0x00873fff [1032 KiB] l4lx | end: Number of physical regions: 1, 1056768 Bytes l4lx | end: 1: Phys: 0x06499000 to 0x0659b000, Size: 1056768 l4lx | memory_setup done l4lx | l4env_rd_path: drops-rd.rd l4lx | Loading: drops-rd.rd l4lx | INITRD: Size of RAMdisk is 16384KiB l4lx | RAMdisk from 04c00000 to 05c00000 [16384KiB] l4lx | l4env_phys_to_virt: Could not translate phys. address 0x80000 l4lx | l4env_phys_to_virt: Could not translate phys. address 0x80010 l4lx | l4env_phys_to_virt: Could not translate phys. address 0x80020 ... and continues untiel physical address 0x9fff0 and ends with a lock up of the computer. I am wondering about these lines: ROOT: Task #0c is not allowed to execute cli/sti loader | vmlinuz,#11: WARNING: Can't map I/O space, ROOT denies page (result=0 loader : 0004000) loader | vmlinuz,#11: Not allowed to perform any I/O Could that be causing the error? How can I fix this? Thank you for help. Kind Regards, Stefan > Hi, > > On Tue Aug 04, 2009 at 11:11:15 +0200, Stefan K?gel wrote: > >> I am still trying to run l4linux on top of fiasco. I am running it on >> top of a console system and configured L4Linux2.6 without any input >> drivers. Thanks to your suggestion I am loading vmlinuz with a start script. >> >> I can't tell you much, but once booted both serial and console are >> spammed with >> >> ... >> l4lx | l4env_phys_to_virt: Could not translate phys. address 0x94140 >> l4lx | l4env_phys_to_virt: Could not translate phys. address 0x94150 >> l4lx | l4env_phys_to_virt: Could not translate phys. address 0x94160 >> ... >> >> messages. >> >> Did I forget to configure some kind of memory translation unit? >> > > No but I'd guess that there's some driver in the Linux that causes this. > Maybe you could start from the x86-ux or x86-native defconfigs and see > if the same happens? > > > > Adam >

2 1

Re: interrupt registration
by eric johnson 05 Aug '09

05 Aug '09

On Tue, Aug 4, 2009 at 4:56 PM, eric johnson <eric.gd.johnson(a)gmail.com>wrote: > > > On Tue, Aug 4, 2009 at 1:48 PM, Adam Lackorzynski < > adam(a)os.inf.tu-dresden.de> wrote: > >> >> On Fri Jul 31, 2009 at 20:36:37 -0700, eric johnson wrote: >> > i m using a pistachio L4 kernel. ( with no operating system on top. ) >> > I am running this on an emulator ( qemu) . So i dont have a way of >> having >> > real interrupts. There might be a way in qemu to do that , but i dont >> know >> > how. >> > >> > My question is : i figure there is an interrupt thread for every >> interrupt. >> > ? Since i dont have any real hardware i would like to have a interrupt >> > thread for a random interrupt no. How do i make sure that this interrupt >> > thread is created ? I am trying to simulate the behaviour of an >> interrupt >> > processing in L4. >> >> Why do you think that Qemu's interrupts are not real from the point of >> view of the software running on it? > > > I dont see any of L4's interrupt code path getting called, like irq_thread > or handle_interrupt. > > In any case when is an irq thread created ? > > > > >> >> >> >> Adam >> -- >> Adam adam(a)os.inf.tu-dresden.de >> Lackorzynski http://os.inf.tu-dresden.de/~adam/<http://os.inf.tu-dresden.de/%7Eadam/> >> >> _______________________________________________ >> l4-hackers mailing list >> l4-hackers(a)os.inf.tu-dresden.de >> http://os.inf.tu-dresden.de/mailman/listinfo/l4-hackers >> > >

1 0