l4-hackers

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

2827 discussions

help l4linux dma usage
by jangwoo 20 Mar '12

20 Mar '12

Hi all. I’m jangwoo. I have some question about l4linux DMA usage. Currently if use dma_alloc_coherent api, it works well. You kwow that “dma_alloc_coherent” get virt addr and phys addr concurrently. But most case of dma usage is to change virt addr into phys addr. Mostly virt addr comes from generic layer of linux kernel. Os it is hard to modify. So, I tried to call virt_to_phys(internal mapping l4x_virt_to_phys), but this function wasn’t direct mapping. I feel like delaying to sync between virt addr and phys addr.(when I check with Trace32) Anyway as “virt_to_phys” have these issues, DMA operation hit miss. Currently l4linux don’t support most dma function. how can I work DMA although it is expedient method? Please. p.s : Not support dma list dma_mmap, ___dma_single_cpu_to_dev, ___dma_page_cpu_to_dev, ___dma_page_dev_to_cpu, dma_map_sg, dma_sync_sg_for_cpu etc thanks all.

1 0

Ankh with l4Linux
by Shashi Sharma 15 Mar '12

15 Mar '12

Hi, I want to enable l4linux to use the Ankh server. But I dont know how to start l4linux so that it can bind to Ankh server. I enabled the l4ankh driver in l4linux. And Following is how I start the Ankh server and l4linux --Ankh----------- -- start ankh loader:startv( {caps = {ankh_service = ankh_clnt:svr(); -- session server vbus = ankh_vbus, -- the virtual bus shm = ankh_shm:m("rws")}, -- shm namespace read/writable log = {"ankh", "g"}, l4re_dbg = L4.Dbg.Warn }, "rom/ankh" ); --L4linux------------------- loader:start( { caps = { log = L4.Env.log:m("rws"), vbus = vbus_l4linux; rtc = rtc_caps.rtc; shm = ankh_shm:m("rws"); ankh = ankh_clnt:create(0,"debug,phys_mac,shm=shm,bufsize=16384"); }, l4re_dbg = L4.Dbg.Warn, log = { "l4linux", "yellow" }, }, "rom/vmlinuz mem=128M earlyprintk=yes showpfexc=1 console=ttyLv0 tcp,nolock ip=192.168.0.200:192.168.0.115:192.168.0.1:255.255.255.0::eth0: l4x_rd=rom/drops-rd.rd root=1:0 ramdisk_size=20000 init=/bin/sh"); Thanks, Shashi

3 7

NFS boot
by Sartakov A. Vasily 15 Mar '12

15 Mar '12

Hello. I try to use NFS in l4linux and have problem. This happens on ARM and on X86: WARN: Page-fault above task size: pfa=bfdfeb40 pc=4007d760 l4linux | 0: 00009228 4017e2d8 0d696914 00008f38 4: 00000000 40095910 bfdff000 bfdfeb40 l4linux | 8: 40094fb0 00000000 40094fb0 0001702c 12: 00000070 bf977ea8 40083e58 4007d760 l4linux | CPSR: 20000010 Err: 00410007 Kernel panic - not syncing: Attempted to kill init! [<0012ebf8>] (unwind_backtrace+0x0/0xf4) from [<00482c70>] (panic+0x58/0x180) [<00482c70>] (panic+0x58/0x180) from [<00150ee0>] (do_exit+0x5b4/0x670) [<00150ee0>] (do_exit+0x5b4/0x670) from [<00150fd8>] (do_group_exit+0x3c/0xb0) [<00150fd8>] (do_group_exit+0x3c/0xb0) from [<0015bf88>] (get_signal_to_deliver+0x190/0x388) [<0015bf88>] (get_signal_to_deliver+0x190/0x388) from [<00128bb8>] (do_signal+0x88/0x5ac) [<00128bb8>] (do_signal+0x88/0x5ac) from [<0012c31c>] (l4x_user_dispatcher+0x1500/0x2574) [<0012c31c>] (l4x_user_dispatcher+0x1500/0x2574) from [<001276c0>] (kernel_execve+0x5c/0xc0) [<001276c0>] (kernel_execve+0x5c/0xc0) from [<00122558>] (init_post+0x98/0xcc) [<00122558>] (init_post+0x98/0xcc) from [<001001f0>] (kernel_init+0xe4/0x114) [<001001f0>] (kernel_init+0xe4/0x114) from [<00126c18>] (kernel_thread_exit+0x0/0x8) panic: going to sleep forever, bye l4linux | panic: going to sleep forever, bye If i boot from other source, I can successful mount NFS root, do one time ls and with next command i will get panic. -- Sartakov A. Vasily

2 1

Re: PowerPC support
by Adam Lackorzynski 05 Mar '12

05 Mar '12

Hi Oliver, On Wed Feb 29, 2012 at 10:11:06 +0100, Oliver Horst wrote: > I've read your reply to last year question about the PowerPC support of > FIASCO [1]. > Is there any progress to be noted on this topic? Unfortunately none. > Further, I'd like to know which actual processor board you targeted the > PowerPC > port to and with that which PowerPC ISA version do you intend to support? One is a 603e one, the other Qemu. At the current stage it probably does not play a big role. Adam -- Adam adam(a)os.inf.tu-dresden.de Lackorzynski http://os.inf.tu-dresden.de/~adam/

2 2

Serial console with qemu/realview-eb
by Mehdi AICHOUCH 05 Mar '12

05 Mar '12

Hi Adam, By adding the option [kernel fiasco -serial_esc] to conf/module.list, the input of the L4Linux console is working great. Thank very much. Mehdi

1 0

page mapping in l4linux
by jangwoo 01 Mar '12

01 Mar '12

Hi all. I have question about mapping memory page in l4linux. I think “dma_map_page” function seem malfunction in below code. struct page *new_page; u32 linux_phys_addr; new_page = alloc_page(GFP_HIGHUSER | __GFP_ZERO | __GFP_NORETRY | __GFP_NOWARN | __GFP_COLD); linux_phys_addr = dma_map_page(NULL, new_page, 0, PAGE_SIZE, DMA_BIDIRECTIONAL); ////// PAGE_SIZE is 4KB _______________________________________________________________ new_page | linux_phys_addr | correct linux_phys_addr | ----------------|----------------------------|------------------------------ | 0xe54b20 0x325d9000 0x48b54b20 0xe54b40 0x325da000 0x48b54b40 0xe54b60 0x325db000 0x48b54b60 as H/W ram-base is 0x40000000, 0x30000000 base address can’t be mapped. Does anybody know a way to work dma_map_page? Best.

2 1

Serial console with qemu/realview-eb
by Mehdi AICHOUCH 01 Mar '12

01 Mar '12

Hi all, I have the same problem with the serial console in QEMU, I tested the same configuration as you. This my test command. [qemu-system-arm -M realview-eb -cpu arm926 -kernel bootstrap_L4Linux_ARM.elf -m 256 -nographic] All output are OK, but no input. I also tried to use the I/O C standard gets(char *) function, in the "hello" example, but I get an error compilation: "undefined reference to gets". Is there any solution or hint since the first post of the problem and after the answer of Julian. thank you in advance. Mehdi

2 1

PowerPC support
by Oliver Horst 29 Feb '12

29 Feb '12

Hi Adam, I've read your reply to last year question about the PowerPC support of FIASCO [1]. Is there any progress to be noted on this topic? Further, I'd like to know which actual processor board you targeted the PowerPC port to and with that which PowerPC ISA version do you intend to support? Best Regards Oliver [1] http://os.inf.tu-dresden.de/pipermail/l4-hackers/2011/004603.html

1 0

IRET returned message
by Julian Vetter 16 Feb '12

16 Feb '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hello Everybody, can someone explain to me what it means when my L4Linux interrupts and jumps into JDB with the message: - ------------------------------------------------------ CPU 0 [f001239c]: IRET returned [ l4lx.cpu0] jdb: g ------------------------------------------------------ I sometimes get this message, and after that my system hangs and I can't jump back to L4Linux by hitting 'g' in the Debugger. Any help would be appreciated Cheers, Julian -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQEcBAEBAgAGBQJPPNeTAAoJEOpGAeJlWp7qw9EH/31w3fA+nAvkbDwF2cjr9r47 qz991ninI6SksDh6OD05b/O/BxxB1fZyaSXAmrnewHVIcuP+bGfgplFnKqjfu4RA qo6PzcpGRCEZpsXYiAhqLxnygxpLTLOXtkFKs9VYfaIkPI4q2DQmSexpZBCTHr27 RY8QO2+eeluR4H8ik3NH9C01pGWmRJJU6Z46wRcIfE+0AZ8NvQBOn4RjYfrfTFbI sW1W7UdQiXc3TurvSzuvaL/tAkT7hpDQOlbEhwDO0SWA8KVtFIeWrKLZWMppglzB TxWGCp5w//LMuG/8DuECDr1C/4u9WwP2Gs/PshtEeLCCX9pqRF/fQfobbtsHlaQ= =7Ifo -----END PGP SIGNATURE-----

2 1

Memory sharing between user and kernel
by 안영호 15 Feb '12

15 Feb '12

Hello everyone, I’m trying to apply an asynchronous batch processing of IPC into l4/fiasco to increase the bandwidth of IPC. I think the main problem is the performance of IPC when virtualization is applied to mobile system. To increase the performance of IPC, I try to use the memory sharing between the address space of user library and that of kernel memory. There is a way to access user memory from kernel, but I couldn’t find any way to read kernel memory in user space. At first, I want to know whether it is meaningful to use an asynchronous batch processing of IPC, or not. Second, when my user library tries to use the virtual address of a kernel global variable by passing it to user address space, “unhandled exception: pc=xxxx” occurs. I think that region mapper makes that exception while it’s checking the range of virtual address. If I want to access kernel memory in user address space, what can I do? Maybe it causes security problems and it’s not proper for virtualization. However, in my opinion, some compromise will be needed for the future mobile virtualization system.

2 1

Jump to page:

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