l4-hackers March 2008

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

22 participants
28 discussions

Compiling Fiasco UX problems under Fedora 8
by Olivier.Landemarre＠utbm.fr 17 Mar '08

17 Mar '08

Hello I'm currently trying compil curent Fiasco UX under a Fedora 8 unfortunatly I get this: Linking test mapdb_t g++ -static -Wl,--gc-sections,-T/home/olivier/l4/kernel/fiasco/src/kernel.ux.ld \ -fno-implement-inlines -mpreferred-stack-boundary=2 -m32 -march=i686 -fno-defer-pop -freg-struct-return -g -Wall -W -Wno-parentheses -Wformat=2 -fno-stack-protector -ffunction-sections -gstabs+ -fno-rtti -fno-exceptions -fno-threadsafe-statics -Wno-non-virtual-dtor -O2 --param large-function-insns=10000 -fweb -finline-limit=10000 -frename-registers -fgcse-after-reload \ mapdb_t.o mapdb.o simpleio.o kernel_console.o panic.o warn.o buddy_alloc.o mem_layout.o mem_layout-ux.o per_cpu_data.o globals.o globals-ia32-ux.o loader.o vhw.o kip_init-ia32-amd64-ux.o kip_init-ux.o boot_info.o boot_info-ia32-amd64-ux.o boot_info-ux.o cmdline.o config.o config-ia32-ux.o config-ux.o jdb_symbol.o jdb_util.o tb_entry.o tb_entry-ia32-ux.o perf_cnt.o perf_cnt-ia32-ux.o jdb_tbuf.o x86desc.o emulation.o pic.o pic-ux.o usermode.o cpu-common.o cpu-ia32-ux.o cpu-ux.o trampoline.o cpu_lock.o cpu_lock-generic.o region.o entry_frame.o entry_frame-ia32-ux.o entry_frame-abs-timeout-hack.o kmem-ia32-ux.o kmem-ux.o mapped_alloc.o mem_unit-ux.o ram_quota.o kmem_alloc.o kmem_alloc-ia32-amd64-ux.o per_cpu_data_alloc.o slab_cache_anon.o kmem_slab_simple.o vmem_alloc.o vmem_alloc-ia32-ux.o vmem_alloc-ux.o paging.o paging-ia32-ux.o kmem_slab.o fpu_state.o fpu.o fpu-ia32-ux.o fpu-ux.o sched_context.o switch_lock.o timer.o timer-ia32-amd64-ux.o timer-ux.o timeout.o space_index.o obj_space.o cap_space.o mem_space.o mem_space-user.o mem_space-ia32-ux.o mem_space-ux.o space.o space-caps.o space-ux.o context.o context-ia32-ux.o context-ux.o helping_lock.o cap_space_alloc.o obj_space_alloc.o mapping-ia32-ux.o mapping.o mapping_tree.o libabi.a libjabi.a libdrivers.a libk.a libamm.a libcxx.a sighandler.o -o mapdb_t Running test mapdb_t ... /bin/sh: line 1: 24293 Erreur de segmentation ./mapdb_t --test --quiet > mapdb_t.out make[1]: *** [mapdb_t.ok] Erreur 139 make[1]: quittant le répertoire « /home/olivier/l4/kernel/fiasco/build » make: *** [all] Erreur 2 The Fiasco ux have been compiled but it is the same error I receive in console only an: Segmentation error Nothing more. any idea? Olivier --

3 6

porting L4linux on minix3.
by Pravin 16 Mar '08

16 Mar '08

Hi, I wanted to ask a question regarding the procedure followed in porting linux on L4 (L4Linux). I am exploring the possibility of porting Linux on Minix3 on similar lines. L4 and Minix3 are having similarities. So, in theory, with few modifications, L4Linux should be able to run on Minix also. I wanted some information regarding the approach taken in porting Linux on L4. If there are any documentation/links available about *. how to go in case of these kind of porting of linux on another kernel? *. and how L4linux works internally? *. which architecture dependent code of Linux has been modified and why? then they will be quite helpful to me. -- Thanks and Regards, Pravin Shinde

3 5

Re: Failed to create a new task
by ahmad kamal 16 Mar '08

16 Mar '08

here is my latest source codes. unfortunately i got this new error: KERNEL: nothing mapped: (Mem_space) from [4]: 00005000 size: 00001000 to [5] #include <l4/sys/types.h> #include <l4/sys/syscalls.h> #include <l4/sys/ipc.h> #include <l4/rmgr/librmgr.h> #include <l4/sigma0/sigma0.h> #include <stdio.h> #define STACK_SIZE 4096 char stack001[STACK_SIZE]; char stack002[STACK_SIZE]; static void new_task_fn(void) { l4_threadid_t cur_id = l4_myself(); enter_kdebug("new_task_fn\n"); while(1) { printf("Currently running in task id: %d, thread no: %d\n", cur_id.id.task, cur_id.id.lthread); } } static void page_handler(void) { int res; l4_timeout_t timeout = L4_IPC_NEVER; l4_msgdope_t result; /*int i = 0; while(1) { printf("page_handler, taskno: %d, thread id: %d\n", l4_myself().id.task, l4_myself().id.lthread); if (i%10 == 0) { //l4_thread_switch(new_task); l4_threadid_t switch_tid = l4_myself(); switch_tid.id.lthread -= 1; l4_thread_switch(switch_tid); enter_kdebug("i/10\n"); } i++; }*/ l4_addr_t map_area = 0x60000000; l4_addr_t base; l4_umword_t rcv_fpage; l4_umword_t snd_fpage; res = l4_ipc_call(SIGMA0_ID, // thread id L4_IPC_SHORT_MSG, // send descriptor SIGMA0_REQ_FPAGE_ANY, // snd_dword0 snd_fpage, // snd_dword1 L4_IPC_MAPMSG(map_area, L4_WHOLE_ADDRESS_SPACE),// rcv_desc base, // *rcv_dword0 &rcv_fpage, // *rcv_dword1 timeout, &result); if (res) { printf("Error in ipc_call, mapping, err no: %d\n", res); enter_kdebug("Error in ipc_call\n"); } else { enter_kdebug("successfully map fpage\n"); } l4_threadid_t src_id; // a faulting thread l4_umword_t rcv_dword0; // the faulting address l4_umword_t rcv_dword1; // the faulting eip res = l4_ipc_wait(&src_id, L4_IPC_SHORT_MSG, &rcv_dword0, &rcv_dword1, timeout, &result); enter_kdebug("Page handler handle faulting thread\n"); if (res) { //res == 0 ? SUCCESS : FAILED printf("Error occurs in l4_ipc_wait: %d\n", res); enter_kdebug("FAILED in l4_ipc_wait: %d\n"); } l4_umword_t fault_addr = rcv_dword0 &~ 0x00000003; //rcv_dword0 & 0xFFFFFFFC // mask lower bits for (write/grant) flags? l4_umword_t rw_addr = fault_addr | 0x00000002; // mask the 2 nd lower bits to set write flags while(1) { l4_snd_fpage_t fpage; fpage.snd_base = rw_addr; // do wee need to left shift the address??? //we can also use the function l4_fpage(address, size, write, grant) fpage.fpage.fp.grant = 0; fpage.fpage.fp.write = 1; fpage.fpage.fp.page = rw_addr; fpage.fpage.fp.size = L4_LOG2_PAGESIZE; //12 res = l4_ipc_reply_and_wait(src_id, //dest_id L4_IPC_SHORT_FPAGE, //send descriptor fpage.snd_base, //snd_dword0 fpage.fpage.fpage, //snd_dword1 - might change later to .fp &src_id, //src_id L4_IPC_SHORT_MSG, //rcv descriptor &rcv_dword0, //rcv_dword0 &rcv_dword1, //rcv_dword1 L4_IPC_NEVER, //timeout &result); //result if(res) {//res == 0 ? SUCCESS : FAILED printf("FAILED in reply wait: %d\n", res); enter_kdebug("FAILED in reply wait\n"); } } } int main(void) { // 1st thread l4_threadid_t first_tid; l4_umword_t first_eip, first_esp; l4_threadid_t preempter, pager_tid; l4_umword_t old_eflags, old_eip, old_esp; first_tid = l4_myself(); first_eip = first_esp = 0xFFFFFFFF; preempter = pager_tid = L4_INVALID_ID; l4_thread_ex_regs(first_tid, first_eip, first_esp, &preempter, &pager_tid, &old_eflags, &old_eip, &old_esp); // 2nd thread l4_threadid_t sec_tid; l4_umword_t sec_eip, sec_esp; l4_threadid_t sec_pager_tid; sec_tid = first_tid; sec_tid.id.lthread += 1; sec_eip = (l4_umword_t)&page_handler; sec_esp = (l4_umword_t)&stack001[STACK_SIZE-1]; sec_pager_tid = pager_tid; printf("sec_tid, task no: %d, threadid: %d\n eip: %lx esp: %lx\n", sec_tid.id.task, sec_tid.id.lthread, sec_eip, sec_esp); l4_thread_ex_regs(sec_tid, sec_eip, sec_esp, &preempter, &sec_pager_tid, &old_eflags, &old_eip, &old_esp); // new task l4_taskid_t new_task; l4_umword_t mcp; //max controlled priority, min[creator_mcp, specified_mcp] l4_umword_t new_esp; //initial stack pointer for thread000 if new task - active l4_umword_t new_eip; //entry point for thread000 if new task - active l4_threadid_t new_task_pgr; //L4_NIL_ID - inactive task. /* l4_task_new - create/delete task * create inactive/active task * active task - create a new address space with 128 threads and thread000 is started * inactive task - occupies no resources, no address space and no threads * task can be deleted by its chief or higher-level chief */ new_task = first_tid; new_task.id.task += 1; mcp = (l4_umword_t)255; new_esp = (l4_umword_t)&stack002[STACK_SIZE-1]; new_eip = (l4_umword_t)&new_task_fn; new_task_pgr = sec_tid; //printf("new task, task no: %d, threadid: %d\n", new_task.id.task, new_task.id.lthread); printf("new pager, task no: %d, threadid: %d\n eip: %lx esp: %lx\n", new_task_pgr.id.task, new_task_pgr.id.lthread, new_eip, new_esp); if (!rmgr_init()) { // zero - true/error enter_kdebug("Failed in init RMGR\n"); return 0; } if (rmgr_get_task(new_task.id.task)) { enter_kdebug("Failed in get the right of task creation\n"); return 0; } new_task = l4_task_new(new_task, mcp, new_esp, new_eip, new_task_pgr); // l4_umword_t esp2 = 0x40000000 + (new_task.id.task << 20 | new_task.id.lthread << L4_PAGESHIFT) + L4_PAGESIZE; // l4_taskid_t new_task2 = l4_task_new(new_task, 255, esp2, (l4_umword_t)&new_task_fn, new_task_pgr); l4_threadid_t nil_tid = L4_NIL_ID; if (new_task.id.task == nil_tid.id.task) { enter_kdebug("Failed to create new task\n"); return 0; } else { printf("new task, task no: %d, threadid: %d\n", new_task.id.task, new_task.id.lthread); enter_kdebug("new task is succesfully created\n"); } //continue running thread0 int i = 0; while(1) { /*if (i%10 == 0) { //l4_thread_switch(new_task); //l4_thread_switch(sec_tid); //enter_kdebug("i/10\n"); }*/ printf("Continue running thread, task id: %d, thread no: %d\n", first_tid.id.task, first_tid.id.lthread); i++; } return 0; } ----- Original Message ---- From: Adam Lackorzynski <adam(a)os.inf.tu-dresden.de> To: l4-hackers(a)os.inf.tu-dresden.de Sent: Monday, March 10, 2008 7:43:17 AM Subject: Re: Failed to create a new task Hi, On Sat Mar 08, 2008 at 22:54:30 -0800, ahmad kamal wrote: > I've failed to create a new task: > > the l4_task_new in my code return 0 (L4_NIL_ID) which is indicate > either the task was create as inactive or didn't create at all. Good you have posted the source. I think the problem is that you need to get the right for creating new tasks first. Depending whether simple_ts is started or not you need to use l4ts_allocate_task2(0, &tid) or rmgr_get_task(num) before the l4_task_new. Then it should work. Adam -- Adam adam(a)os.inf.tu-dresden.de Lackorzynski http://os.inf.tu-dresden.de/~adam/ _______________________________________________ l4-hackers mailing list l4-hackers(a)os.inf.tu-dresden.de http://os.inf.tu-dresden.de/mailman/listinfo/l4-hackers ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs

2 1

Error handling for reply in dice
by Marc Chaland 13 Mar '08

13 Mar '08

Hi, I implemented an IDL with a allow_reply_only method : start. When client calls this, it is blocked until server decides to send reply. This works fine in general. But when client thread disapears after call, I cannot get error code from reply function on server side. I tried to use error_function, but it is not called. Is there another way to get this info ? Regards Marc

2 5

error running fiasco-ux
by Lucyantie Mazalan 12 Mar '08

12 Mar '08

Hi all, need some help here. I've been trying to use my fiasco-ux to load my module. Thing is, when I try to run the module with: ~/fiasco-ux# ./fiasco -l hello Killed This is what I got. I have copied down the needed binaries ( fiasco, irq0, sigma0, roottask ) in the fiasco-ux directory. Please help. Thanks. Regards, Lucille. ------------------------------------------------------------------ - - DISCLAIMER: This e-mail (including any attachments) may contain confidential information. If you are not the intended recipient, you are hereby notified that any dealing, review, distribution, printing, copying or use of this e-mail is strictly prohibited. If you have received this email in error, please notify the sender or MIMOS Berhad immediately and delete the original message. Opinions, conclusions and other information in this e-mail that do not relate to the official business of MIMOS Berhad and/or its subsidiaries shall be understood as neither given nor endorsed by MIMOS Berhad and/or its subsidiaries and neither MIMOS Berhad nor its subsidiaries accepts responsibility for the same. All liability arising from or in connection with computer viruses and/or corrupted e-mails is excluded to the fullest extent permitted by law.

2 1

run-l4 - to start another task
by ahmad kamal 12 Mar '08

12 Mar '08

hi all, i've been trying to load another l4 task using run-l4 in l4linux, but i've got error the error is -8(cannot a find module) here is my conf file #launch.conf verbose 0 modpath "(hd0,2)/boot/l4app" task "hello" and here is the cmd i run: run-l4 --fprov "BMODFS" "(hd0,2)/boot/l4app/launch.conf" here are few question: do i need to load run-l4 during boot time, in other words do i need to define it in menu.lst? do i need to load the task i would like to call using run-l4 during boot time? Regards, Ahmad Kamal ____________________________________________________________________________________ Never miss a thing. Make Yahoo your home page. http://www.yahoo.com/r/hs

2 1

about L4ka build
by ANPING WANG 12 Mar '08

12 Mar '08

sorry could you give me some help about the L4ka build i have so problem when compile the L4ka user-level application i do step by step from http://l4ka.org/projects/pistachio/ia32/gettingstarted.php when build user-level code there is a problem root@wap-desktop:~/l4ka-pistachio/ia32-user-build# ../user/configure --prefix=../ia32-user-install --with-kerneldir=../ia32-kernel-build configure: error: can not find sources in ../user or .. root@wap-desktop:~/l4ka-pistachio/ia32-user-build# that show there is no source is any thing i should to fix it thinks a lot

1 0

Re: l4-hackers Digest, Vol 59, Issue 7
by Valery V. Sedletski 12 Mar '08

12 Mar '08

On Mon, 10 Mar 2008 12:00:02 +0100, Adam Lackorzynski <adam(a)os.inf.tu-dresden.de> wrote: >> mnt/data1/src/drops/build/include/x86/l4/util/util.h:84 >> 4019e0: 83 0a 00 orl $0x0,(%edx) >> >> So it is the function: >> >> /** Touch data areas to force mapping read-write */ >> static inline void >> l4_touch_rw(const void*addr, unsigned size) >> { >> const char *bptr, *eptr; >> >> bptr = (const char*)(((unsigned)addr) & L4_PAGEMASK); >> eptr = (const char*)(((unsigned)addr+size-1) & L4_PAGEMASK); >> for(;bptr<=eptr;bptr+=L4_PAGESIZE){ >> asm volatile("or $0,%0 \n" >> : >> : "m" (*(const unsigned*)bptr) >> ); >> } >> } >> >> And instruction "orl" is inside "asm volatile" block. >> >> So, this function touches an area (BSS in this case) to make mapping read-write. But >> why this does not work? > >I have no idea, I've never seen this before. Is there the possibility to >update the source so that we can work on the same basis? > OK, I'll try to update the sources. (First I must update a svn client as my svn client under Linux is too old (I have Red Hat 9.0)). I used an old sources from CVS, not newer SVN sources. Thank you. WBR, Valery

1 0

thread.id.lthread inquiry
by hazwan 11 Mar '08

11 Mar '08

hello Currently im studying on how to build an application using fiasco/L4. Im very new to this subject and it requires me lots of fundemental issues concerning this topic. I need to go deep in the code ie:-the roottask/sigma0/kernel implementation. So far i had try some steps to aquire answers. One of them is by printing out a thread which is initialize as l4_threadid_t. I didnt call any sys calls (l4_myself() to that particular thread. So, what i had done is :- l4_threadid_t thread; l4_threadid_t pager; printf("%x\n", thread.id.lthread); printf("%x\n", pager.id.lthread); when i print this out, the thread will print out its id as 34 in hex. Meanwhile the pager will identify as 0 for its lthread id. The question is why is this happening? If i turn back around the code, it will still print out the thread as 34 for its id and pager is 0. Is it reading some garbage value from the memory or the roottask/sigma0/kernel will identify it as 34 only if we didnt initialize any system calls. I need to study this in depth due to my yearly project which highly involves fiasco/L4 as its platform. If it is applicable, can anyone show me which file(header/*.c etc..) that locates the number of 34 happens. Thanks. p/s: Im also thinking that this is a silly question as well. :-P -- View this message in context: http://www.nabble.com/thread.id.lthread-inquiry-tp15950571p15950571.html Sent from the L4 mailing list archive at Nabble.com.

2 1

Re: Failed to create a new task
by ahmad kamal 10 Mar '08

10 Mar '08

is rmgr a roottask? why we need to rmgr_get_task(int taskno)? ----- Original Message ---- From: Adam Lackorzynski <adam(a)os.inf.tu-dresden.de> To: l4-hackers(a)os.inf.tu-dresden.de Sent: Monday, March 10, 2008 7:43:17 AM Subject: Re: Failed to create a new task Hi, On Sat Mar 08, 2008 at 22:54:30 -0800, ahmad kamal wrote: > I've failed to create a new task: > > the l4_task_new in my code return 0 (L4_NIL_ID) which is indicate > either the task was create as inactive or didn't create at all. Good you have posted the source. I think the problem is that you need to get the right for creating new tasks first. Depending whether simple_ts is started or not you need to use l4ts_allocate_task2(0, &tid) or rmgr_get_task(num) before the l4_task_new. Then it should work. Adam -- Adam adam(a)os.inf.tu-dresden.de Lackorzynski http://os.inf.tu-dresden.de/~adam/ _______________________________________________ l4-hackers mailing list l4-hackers(a)os.inf.tu-dresden.de http://os.inf.tu-dresden.de/mailman/listinfo/l4-hackers ____________________________________________________________________________________ Looking for last minute shopping deals? Find them fast with Yahoo! Search. http://tools.search.yahoo.com/newsearch/category.php?category=shopping

2 1

2025

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 March 2008