May 2005 - l4-hackers - example.com Site

L4 Linux and compatibility with other L4 servers
by Tiago Jorge 17 May '05

17 May '05

Hello... I'm now trying to boot my own server with the L4Linux 2.4. i'm using this servers: title Tests root (hd0,0) kernel /home/tiago/fiasco_builddir/rmgr -sigma0 modaddr 0x02000000 module /home/tiago/fiasco_builddir/main -nokdb -nowait -serial_esc -comspeed 115200 -comport 1 module /home/tiago/fiasco_builddir/sigma0 module /home/tiago/fiasco_builddir/names module /home/tiago/fiasco_builddir/dm_phys module /home/tiago/fiasco_builddir/l4io module /home/tiago/fiasco_builddir/rtc module /home/tiago/fiasco_builddir/name_server module /home/tiago/fiasco_builddir/flips-lxdrv module /home/tiago/fiasco_builddir/mini_ifconfig eth0 10.10.5.197 255.255.0.0 10.10.255.255 module /home/tiago/fiasco_builddir/WOO 10.10.5.197 eth0 module /home/tiago/fiasco_builddir/vmlinuz.V2 no-scroll no-hlt l4irqack=linux root=/dev/hda1 but as i expected there are many pagefaults and resources allocated. How can i configure this servers (not my own of course :) ) so that they don't interfere with the l4linux work but still do their work(mainly networking with FLIPS)? for example, when put only this servers... kernel /home/tiago/fiasco_builddir/rmgr -sigma0 modaddr 0x02000000 module /home/tiago/fiasco_builddir/main -nokdb -nowait -serial_esc -comspeed 115200 -comport 1 module /home/tiago/fiasco_builddir/sigma0 module /home/tiago/fiasco_builddir/names module /home/tiago/fiasco_builddir/dm_phys the module dm_phys and l4linux start to page fault. thanks for the help Tiago

4 12

Killing L4Linux
by gjoel123＠fastmail.fm 17 May '05

17 May '05

Hi How does one kill L4linux from the DROPS console using the loader program? This is what I am currently doing: 1:I type 'n' at the DROPS console to get task ID's and names. This is what I get: . . F.02 LOADER 12.03 Linux server . . 2. I type 'k' and then '12', and I get: 'Error -11 (not owner) killing task #12'. Any ideas? The aim is ultimately to correctly halt/reboot l4linux from l4linux's own shutdown code. Regards Gordon -- gjoel123(a)fastmail.fm

2 3

tftp issue when starting 3 VMs
by Derick Swanepoel 11 May '05

11 May '05

Hi, I'm trying to start three VMs from grub (one master and two slaves), but the tftp provider fails to load the third kernel. This file is the same as the second VM's kernel, which loads successfully. Here's the error output: ===================================================================== tftp | File not found: "(nd)/fiasco/ds/vmlinuz-slave.l4env" exec | vmlinuz-slave.l4env: Error -8 (item not found) from file provider, exec : path was "(nd)/fiasco/ds". DMphys | [7.0] dm_phys/server/src/close.c:130:if_l4dm_generic_close_component( DMphys : ): DMphys | DMphys: client B.02 does not own dataspace 113 (owner is F.02) exec | [B.2] dm_generic/lib/clientlib/src/close.c:58:l4dm_close(): exec | libdm_generic: close dataspace 113 at 7.00 failed (ret -17, exc 0)! loader | vmlinuz-slave.l4env: Error -8 (item not found) while loading loader | ==> App successfully purged ===================================================================== Here is my grub.conf entry: ===================================================================== title Fiasco + Con + L4Linux [3 VMs] kernel (nd)/fiasco/ds/rmgr -sigma0 task modname "L4Linux task" modaddr 0x02000000 module (nd)/fiasco/ds/fiasco -nowait -nokdb -nofancy -serial -comspeed 115200 -comport 1 module (nd)/fiasco/ds/sigma0 module (nd)/fiasco/ds/names module (nd)/fiasco/ds/log module (nd)/fiasco/ds/dm_phys module (nd)/fiasco/ds/l4io --noirq module (nd)/fiasco/ds/tftp -i -s 192.168.1.253 module (nd)/fiasco/ds/simple_ts module (nd)/fiasco/ds/l4exec module (nd)/fiasco/ds/name_server module (nd)/fiasco/ds/con --l4io module (nd)/fiasco/ds/loader (nd)/fiasco/ds/con.cfg (nd)/fiasco/ds/l4linux-nsrd.cfg (nd)/fiasco/ds/l4linux-slave.cfg (nd)/fiasco/ds/l4linux-slave.cfg vbeset 0x117 ===================================================================== l4linux-slave.cfg looks like this: ===================================================================== verbose 2 modpath "(nd)/fiasco/ds" task "vmlinuz-slave.l4env" "l4irqack=linux mem=64M load_ramdisk=1 ramdisk_size=32768 root=/dev/ram0 ro l4env_rd=(nd)/fiasco/ds/nsrd.gz" priority 0xA0 ===================================================================== Thanks, Derick

3 8

L4Linux Scheduling error with Kernel Module
by Leon Brits 10 May '05

10 May '05

Hi all, My Kernel Module seem to be working fine for most of the time. But I get an intermittent error where the Kernel crashes with the following error: <snip> Scheduling in interrupt KERNEL: 12.9 (tcb=c0484800) killed: Unhandled trap <snip> The Kernel module does not control any interrupts and at this point-in-time is pure virtual. Any ideas why or how to debugging this error is appreciated? Thanks Leon

2 1

Flips tracking of IPC
by Tiago Jorge 10 May '05

10 May '05

hello... I'm having a problem with the FLIPS server. I'm doing some transmission in TCP and i seems that the call from the client to the FLIPS server doesn't arrive. I've putten some printfs along the way in l4vfs_net_io_send_call and he does the call the IPC function... but i've putten also some debug message in l4vfs_net_io_send_component and it doesn't print out... so something is dying around here and i don't know where...:( My questions are... in the FLIPS/L4vfs net_io server side where should i put the debug to see if the call has arrived? (i've tried in net_io-server.c and nothing is shown) i've sent something earlier trought that socket... i mean... it only dies when i try to send something the second time... am i missing something (the socket its the same it stays connected)? is there any problem with FLIPS that prevents me from sending something from a socket that is blocked in a recv in another thread (should't be because TCP is bi-directional)? thanks in advance Tiago

3 9

Pagefaulting in lock instruction
by Tiago Jorge 06 May '05

06 May '05

hello... I'm having a strange page fault in my program. My server has an object called Consensus. This object its responsable in the system to do consensus about something... an has an inner thread that receives some parameters from outside. its header file its the following ################################################## #ifndef CONSENSUS_H #define CONSENSUS_H #include <fcntl.h> #include <stdlib.h> #include <netdb.h> #include <sys/types.h> #include <sys/socket.h> #include <netinet/in.h> #include <arpa/inet.h> #include <l4/thread/thread.h> #include <l4/log/l4log.h> #include <l4/lock/lock.h> #include "WOO_vector.h" #include "WOO_map.h" #include "WOO_string.h" #include "FailureDetector.h" #include "Consensus_Packet.h" #include "WOO_TMO.h" #include "WOO_Groups.h" #include "WOO_common.h" #include "WOO_member.h" #define CONSENSUS_PORTNUMBER 30987 struct estimate{ int eid; vector<string> myEstimate; }; class Consensus{ private: int sock; int trueFlag; socklen_t addrLength; struct sockaddr_in destinationAddr; struct sockaddr_in myAddr; struct sockaddr_in responseAddr; estimate est; phases current_phase; bool mustRunConsensus; //l4lock_t consensus_lock; l4thread_t consTh; int consensus_id; //params int myEid; vector<WOO_member> *eids; l4lock_t *group_lock; map<string, CWOO_TMO> *addr_TMO_running; l4lock_t *lock_TMO_running; map<string, CWOO_TMO> *addr_TMO_decided; l4lock_t *lock_TMO_decided; vector<CWOO_Groups> *addr_WOO_groups_table; l4lock_t *lock_WOO_groups_table; public: l4lock_t consensus_lock; private: void initialize_comm(); bool amICoordinator(vector<WOO_member> v); int coordinator(vector<WOO_member> v); void copyGroup(vector<WOO_member> *toFill); bool deadCoordinator(vector<WOO_member> v, FailureDetector *fd); void cleanFailures(vector<WOO_member> *toClean, FailureDetector *fd); void consensus_broadcast(vector<WOO_member> *toBroadcast, phases phase, packet_type packet, int consensus_round); vector<string> buildEstimate(); bool moreConsensus(); void decide(); public: Consensus(); void start(int eid, vector<WOO_member> *new_eids, l4lock_t *new_group_lock, map<string, CWOO_TMO> *p_addr_TMO_running, l4lock_t *p_lock_TMO_running, map<string, CWOO_TMO> *p_addr_TMO_decided, l4lock_t *p_lock_TMO_decided, vector<CWOO_Groups> *p_addr_WOO_groups_table, l4lock_t *p_lock_WOO_groups_table); void consensus_impl(); void startConsensus(); void stopConsensus(); bool get_mustRunConsensus(); }; void helper_consensus(void *args); #endif // CONSENSUS_H ################################################## when someone wants to start the program we invoke start() that has this code... ################################################## //passar aqui as TMOdecided e as TMOrunning e respectivos locks void Consensus::start (int new_eid, vector < WOO_member > *new_eids, l4lock_t * new_group_lock, map < string, CWOO_TMO > *p_addr_TMO_running, l4lock_t * p_lock_TMO_running, map < string, CWOO_TMO > *p_addr_TMO_decided, l4lock_t * p_lock_TMO_decided, vector < CWOO_Groups > *p_addr_WOO_groups_table, l4lock_t * p_lock_WOO_groups_table) { myEid = new_eid; eids = new_eids; group_lock = new_group_lock; consensus_id = 0; trueFlag = 0x1; consensus_lock = L4LOCK_UNLOCKED; addr_TMO_running = p_addr_TMO_running; lock_TMO_running = p_lock_TMO_running; addr_TMO_decided = p_addr_TMO_decided; lock_TMO_decided = p_lock_TMO_decided; addr_WOO_groups_table = p_addr_WOO_groups_table; lock_WOO_groups_table = p_lock_WOO_groups_table; /*consTh = l4thread_create (helper_consensus, this, L4THREAD_CREATE_ASYNC);*/ this->startConsensus(); this->consensus_impl(); LOG ("estou dps do thread create\n"); } ################################################## for testing porpuses the l4thread_create method its commented because its page faulting also :(. Because it page-faulted when i've called the object's method that its inside helper_consensus... ################################################## void helper_consensus (void *args) { Consensus *cobj = static_cast<Consensus *>(args); cobj->consensus_impl(); } ################################################## i thought it was pagefaulting because of this method... but no... he page-faults event when called normally... but the focus of the problem it's not here yet. when i call the startConsensus() method it page-faults in the l4lock_lock(...) instruction... the method looks like this ################################################## void Consensus::startConsensus () { l4lock_lock(&consensus_lock); <---- it page faults here... LOG("inside\n"); mustRunConsensus = true; est.myEstimate = this->buildEstimate (); LOG("out\n"); l4lock_unlock(&consensus_lock); } ################################################## here it is the dissassembled output... ################################################## <Consensus::startConsensus()>: /home/tiago/l4/pkg/WOO/server/src/Consensus_udp.cc:178 01db6a50 push %ebp 01db6a51 mov %esp,%ebp 01db6a53 push %edi 01db6a54 push %esi 01db6a55 push %ebx 01db6a56 sub $0xc37c,%esp 01db6a5c mov 8(%ebp),%ebx /home/tiago/l4/include/l4/lock/lock.h:97 01db6a5f lea 0xc3c8(%ebx),%eax 01db6a65 mov %eax,0xffff3c84(%ebp) <<--- PAGE FAULT HERE!!!!!!!!!!!! /home/tiago/l4/include/l4/lock/lock.h:98 01db6a6b call 0x1dc0e80 <l4thread_myself> /home/tiago/l4/include/l4/lock/lock.h:101 01db6a70 mov 0xffff3c84(%ebp),%ecx /home/tiago/l4/include/l4/lock/lock.h:98 01db6a76 mov %eax,0xffff3c80(%ebp) /home/tiago/l4/include/l4/lock/lock.h:101 01db6a7c test %ecx,%ecx 01db6a7e je 0x1db6aa6 ################################################## as of my inexperience with Fiasco debug issues... what might be causing this? thanks for your all your help...(and yes... i'm just about to finish my project) Tiago Jorge

2 3

L4Linux priority
by Leon Brits 02 May '05

02 May '05

Hi all, One is able to set the priority of L4Linux in the loader config file with the "priority" parameter. Q1: How can I change this priority dynamically after the Linux server has been loaded? Q2: Which tasks has the rights to do this - only RMGR? Thanks Leon

2 1

RT issues
by Rene Wittmann 02 May '05

02 May '05

Hi, As I'm trying to implement a real-time application in DROPS, I still have some questions on this topic. Maybe there's someone who can answer them because I didn't find it described it the documentation... 1. The timer granularity seems to be set to 976 microsec. Is there a special background for using this number or can I set it down (maybe even to 10 microsec, which what be a good time resolution for my task)? How can I set it? Or would it be stupid to set it to such a tiny value? 2. When looking at the rt_sched example in the hello-package: the main-thread has a preempter-thread for recognizing timeslice overrun and deadline misses. When does this preempter-thread recognize a TS overrun or a deadline miss? Right after this event (so within the next time resolution) or after the main-thread says that is has finished it's work (meaning after a call to "l4_rt_next_reservation")? 3. How can the preempter-thread tell the main-thread to stop it's work in case of a timeslice overrun or even a deadline miss? 4. Does the system allow a thread's scheduling like this: period-length = 976 microsec, timeslice_mand = 9760 microsec, timeslice_opt = 9760 microsec (which would mean that the system had to reserve timeslices that don't fit into one period...) OR period-length = 9760 microsec, timeslice_mand = 4380 microsec, timeslice_opt = 4380 microsec (which would mean that the system had to reserve all resources for this thread..) (what would happen if another thread comes with a period_length of also 9760 microsec and a timeslice_mand = 6000 microsec? What would be the system's behaviour?) ===================================================================== Other thing: File system support: 1. which filesystems are supported (reiserfs?)? 2. Is there a way to have RT access to disk and accessing files this way? If so, how? Thanks, Rene

3 2

L4 IPC doubt
by Tiago Jorge 02 May '05

02 May '05

hello to you all. Let me describe my cenario... I have my L4 server that receives ipc calls. The code was generated from DICE. And i have a lib and i've implemented a client that uses that lib. My L4 server also uses ipc to comunicate with other servers, mainly with flips that also uses IPC. that problem is the following: My client makes an IPC call and blocks waiting for the answer from my server... while processing that answer, my server has to comunicate with flips using IPC. The problem is here... my server blocks in the ipc_send when calling flips, but when i remove the call from the client it works fine, meaning that the client is not blocked waiting for ipc to return. My questions are... why can't i do IPC while another one that has nothing to do with my server is blocked? there can be only one ipc call in the whole system at a time? they all must be sequential? thanks in advance Tiago

4 8