users

users@lists.genode.org

6 participants
2306 discussions

Has Bhyve been ported to Genode?
by Lonnie Cumberland 11 Aug '20

11 Aug '20

Hello, I have been working with VMM's for a while now and it lately my understanding is that Bhyve is much better than VirtualBox. I was wondering if Bhyve has been ported to Genode yet as I would like to see it run with Bhyve+Sculpt+NOVA if that were possible? There is also a minimal port to OS-X called Xhyve ( https://github.com/machyve/xhyve) which is completely self-contained except for just a few calls to the hypervisor from what I read and might be easier to port than the base Bhyve. Cheers, Lonnie

2 2

Genode modified Bender sources?
by Lonnie Cumberland 11 Aug '20

11 Aug '20

Hello All, I am attempting to learn more about Bender (https://github.com/blitz/bender) and the way that it relocates modules so that I can get the NOVA x86_64 bit hypervisor using Bender to relocate the modules. I have been able to read a bit of the code from the bender github repo but in researching in the Genode repositories, I have come across some information that suggests Genode has a modified set of Bender sources: https://github.com/genodelabs/genode/tree/4450b37ff5b22e99133863a6f56d5b715… But, I could not locate the actual modified genode bender source code so that I can look things over. Can you please tell me where I can find these modified sources? Cheers, Lonnie

2 2

How to start my VM with the new Sculpt-CE preview (Problems)
by Lonnie Cumberland 11 Aug '20

11 Aug '20

Greetings All, I was trying to use these steps to get VBox started on Sculpt 20.02 https://genodians.org/m-stein/2019-03-07-vm-with-sculpt-ce-preview But Sculpt seems to error out and the Virtualbox console never starts. Is anyone able to get this to work? I also tried it using the Sculpt OVA imported into a Virtualbox on Ubuntu 20.04 and still the same results. Cheers, Lonnie

2 1

Libc Component Not Logging
by Daly, Connor G. 08 Aug '20

08 Aug '20

Hi all, Genode beginner here. Been working with the platform for a few weeks trying to port a standard Linux application I wrote to Genode. My program uses sockets, both for sending and receiving, and judging from the code in tcp_terminal, the only way to do sockets is with libc rather than native Genode. I've been trying to implement the first few phases of my program in Genode, but when I run them, no log messages seem to appear and I can't really tell if my program is doing anything. When I remove the libc relevant pieces of code and run my program through a base/Component, everything logs just fine. I created a small test program to find my mistake, so this piece of code does not log anything to qemu with a Linux target platform: #include <libc/component.h> #include <base/log.h> void Libc::Component::construct(Libc::Env &) { Genode::log("Hello world"); } For a run script, I'm using: build { core init widget } create_boot_directory install_config { <config verbose="yes"> <parent-provides> <service name="LOG"/> <service name="PD"/> <service name="CPU"/> <service name="ROM"/> <service name="IO_MEM"/> <service name="IO_PORT"/> <service name="RAM"/> </parent-provides> <default-route> <any-service> <parent/> <any-child/> </any-service> </default-route> <default caps="50"/> <start name="widget"> <resource name="RAM" quantum="1M"/> </start> </config>} build_boot_image { core ld.lib.so init widget libc.lib.so vfs.lib.so libm.lib.so } append qemu_args " -nographic " run_genode_until "Hello world" 10 Any help would be greatly appreciated. Thanks! -Connor

2 1

initial port of golang to genode
by Alexander Tormasov 04 Aug '20

04 Aug '20

I make some very rough port of golang runtime to genode (mostly for nova), https://github.com/tor-m6/genode.git This work is a side result of one of research project made in Innopolis University. It allow compilation and execution of simple go applications/packages (probably without unsupported os features). I use only gccgo version of golang compiler. Things like "go build" in cross mode not supported yet (while we make some techology to produce makefiles from "go build" to compilei go code, not attached yet). Other limitations see below. It build a set of libraries, in form of "libport" and "noux-pkg" each (because we need to run configure which is not supported in libport), as static ones. Also it build a set of standard golang packages (namely fmt, os, syscall, runtime/etc) as files for gccgo compilation. Note: Go runtinme support around 9 systems as backend. I use netbsd as simplest one for the base for new system - "inno". To implement context switching I have to develop it from the scratch; for stack port use alloc_secondary_stack() from Genode. I suspect that I need to use a low-level context switching from kernel (native thread switch), this is not implemented yet. Configuration after build ========================= Golang package (like fmt) is an object file (e.g. fmt.o) which appears in build/x86_64/noux-pkg/libgo/ directory and subdirectories. This path should be included during golang/gccgo compilation (not only linking). Built libraries for Go will be in ${LIB_CACHE_DIR}/libgo/libgobegin.a ${LIB_CACHE_DIR}/libgo/libgolibbegin.a ${LIB_CACHE_DIR}/libgo/libgo.a Current limitations =================== - RAM quota in .run file should be >= 1300M: genode do not support MAP_ANON flag, so, we cant reserve memory (544Gb) for arena, and, therefore, go to Windows-like model with "small" immediate allocation; it force 2GB limitation in code and make direct allocation of ~1.1Gb - can run in VM with 1 CPU only: in general it support checking of CPU number (by backporting patch from current genode code) and trying to run everything, but it hangs somewhere in the middle due to recursive locking (error "deadlock ahead" in acquire() in mutex.cc) May be this is small mistake - don't have time to find and fix (in this place due to slot_jump() gdb do not give normal info related to stack). I found that problem during stack unwind in ... runtime.startm -> runtime.newm -> runtime.newm -> runtime.mcommoninit -> runtime.callers -> runtime_callers() -> backtrace_full() ... here it try to call -> strlcpy() -> _jmp_slot() and make again guard inside this function - it works only on x86_64, tested on nova kernel genode do not support makecontex/setcontext/gencontext family; but it is crucial to support go because of goroutine model; I have to implement it by myself. It is machine-dependent; moreover, context swiching do not work after direct port from libc/etc - I have to make 1 step stack unwind procedure to jump using stack frame to the next function. So, I add everywhere compilation option -fno-omit-frame-pointer to be sure that this approach works. Probably it need only for implementation of the file mycontext.cc - it can work on seL4 as well, while I did not submit additional dirty changes for go runtime (I swithed to nova/gdb) - any go application need to add 3 files into it (may be it worth to place them in another library or libgo.a) - dummy.cc with Genode initialization and some empty functions which are not supported on genode, mycontext.cc and myacontext.s with implementation of x86 [sg]etcontext - during port I make a set of dirty fixes for absent on the platform (genode/nova), in particular "empty" C implementaion of sendfile(), getpgid(), lchown(), mount(), settimeofday(), mlock(), munlock(), mlockall(), munlockall(), sigaltstack(); in go part of runtime port I also "damage" some functions, in particualr route.sysctl(), broke poll.ReadDirent(), partially broke logic in runtime.sysMap/sysReserve because of unsupported mem range reservation, do not use ptrace, broke routing in netbsd version of syscall.toRoutingMessage and some others, see patches for libgo in repos/libports/src/lib/libgo/. so, anything in Golang which use them will fail. Potentially new genode releases will contain something which can replace or fix these and other patched functions - once I compile all code with -O3 option, and it hangs. I did not check why... Questions? tor(a)innopolis.ru, origin in https://github.com/tor-m6/genode.git

1 1

sel4 build broken in 20.05?
by Alexander Tormasov 04 Aug '20

04 Aug '20

I am trying to compile sel4 demo for x86, and hit an error: during compilation of [CC] kernel_final.s in the process which give /usr/local/genode/tool/current/bin/genode-x86-gcc -S -o kernel_final.s kernel_final.c it generates asm error error: impossible constraint in ‘asm’! Any suggestions? Log below [CC] kernel_final.s In file included from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/assert.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/types.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/basic_types.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/types.h:14, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/src/api/faults.c:14: /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/util.h:89:7: warning: conflicting types for built-in function ‘memset’ [-Wbuiltin-declaration-mismatch] void *memset(void *s, unsigned long c, unsigned long n) VISIBLE; ^~~~~~ /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/util.h:91:31: warning: conflicting types for built-in function ‘strncmp’ [-Wbuiltin-declaration-mismatch] int PURE strncmp(const char *s1, const char *s2, int n); ^~~~~~~ In file included from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/api/types.h:23, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/compound_types.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/types.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/src/api/faults.c:14: /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/machine/io.h:35:6: warning: conflicting types for built-in function ‘putchar’ [-Wbuiltin-declaration-mismatch] void putchar(char c); ^~~~~~~ /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/machine/io.h:38:8: warning: conflicting types for built-in function ‘puts’ [-Wbuiltin-declaration-mismatch] word_t puts(const char *s) VISIBLE; ^~~~ /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/src/machine/io.c:154:1: warning: conflicting types for built-in function ‘isdigit’ [-Wbuiltin-declaration-mismatch] isdigit(char c) ^~~~~~~ In file included from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/kernel/x2apic.h:17, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/kernel/apic.h:18, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/plat/pc99/plat/machine/interrupt.h:20, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/plat/pc99/plat/machine/hardware.h:14, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/machine.h:18, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/api/syscall.h:15, from /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/src/api/faults.c:16: /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/machine.h: In function ‘x86_flush_rsb’: /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/machine.h:348:5: warning: asm operand 1 probably doesn’t match constraints asm volatile( ^~~~~~~ /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/include/arch/x86/arch/machine.h:348:5: error: impossible constraint in ‘asm’ /home/tor/gen/20.05-tor/contrib/sel4-7935487f91a31c0cd8aaf09278f6312af56bb935/src/kernel/sel4/Makefile:653: recipe for target 'kernel_final.s' failed make[4]: *** [kernel_final.s] Error 1

2 2

Is Genode good for the Desktop?
by Jason Long 03 Aug '20

03 Aug '20

Hello,Is Genode good for the Desktop and is it based on seL4?What does OS framework mean? Thanks.

1 0

how to find number of cpu (cores) available for application?
by Alexander Tormasov 24 Jul '20

24 Jul '20

I am trying to understand how many cpu or/and cored are available for running application. I check code related to sysctl in libports/src/lib/libc/sysctl.cc<http://sysctl.cc> and found extern "C" int __sysctl(const int *name, u_int namelen, void *oldp, size_t *oldlenp, const void *newp, size_t newlen) { … /* fallback values */ { switch(index_a) { … case CTL_HW: switch(index_b) { case HW_MACHINE: *oldlenp = 0; return 0; case HW_NCPU: *(int*)oldp = 1; *oldlenp = sizeof(int); return 0; } break; … in short, if I used standardised for BSD code (like below in golang) it always return 1! int32 getproccount(void) { int mib[2], out; size_t len; mib[0] = CTL_HW; mib[1] = HW_NCPU; len = sizeof(out); if(sysctl(mib, 2, &out, &len, NULL, 0) >= 0) return (int32)out; else return 0; } Any suggestions? How I can have portable code for Genode to obtain at least NCPU? Sincerely, Alexander

2 1

Genode Community Summer 2020
by Martin Stein 13 Jul '20

13 Jul '20

Last year, for the first time we held the Genode Community Summer for people that would like to work on their Genode project while being able to seek our assistance in person. The event was a success and we would like to give it another try this year. ** For whom? ** The gathering is open for everyone of the Genode community. Hence, it should be regarded as a public event. Each participant should have a concrete project to pursue during the time and be able to work on this project independently most of the time. Unlike a commercial offered workshop, there won't be any lecture or intensive coaching, but simply co-working in an informal and friendly setting. Anything confidential should better be left at home! This way, we can accommodate people with different backgrounds at the same time while the event bears potential for networking effects among different community members. If you are planning to participate as organization, please consider a limit of 2 persons per organization. With this, we aim for preserving versatility and the possibility for interdisciplinary exchange at the Genode Community Summer. ** Where? ** The event will take place in the lower floor of the Genode Labs office, Friedrichstraße 26, 01067 Dresden, Germany. Note, that lodging is up to each participant individually. ** When? ** Over two weeks from 31.8.2020 till 11.9.2020. It is possible to participate only for a part of this time. ** How to register? ** Each participant must register in advance since our space is limited but the event will be at no cost. Please register via mail to martin.stein(a)genode-labs.com. We will hold the event only if at least 4 people register until 17.8.2020. We will inform all participants as soon as we have 4 registrations. Your registration must contain a brief description of the project you're planning to pursue during the gathering. Please also add whether you like to participate for the whole two weeks, and if not, which part of that time you will participate. We've opened an issue in the Genode Github repository that serves for discussions and information sharing around the Genode Community Summer 2020: https://github.com/genodelabs/genode/issues/3818 If you have questions regarding the Genode Community Summer please don't hesitate to ask in this issue or via mail to martin.stein(a)genode-labs.com . Hope to see you there!

1 0

Crash on Windriver Simics : page fault, pd='init -> drivers -> acpi_drv' thread='ep'
by Parfait Tokponnon 09 Jul '20

09 Jul '20

Hello, I want to run Genode on Windriver Simics <https://www.windriver.com/products/simics/>. I've just created a disk image with the demo scenario from a freshly cloned Genode. It crashes with the following error. > page fault, pd='init -> drivers -> acpi_drv' thread='ep' cpu=0 > ip=0x1011062 address=0xbff00000 stack pointer=0x401fbbc0 qualifiers=0x4 > irUwp reason=1 > addr2line -e debug/acpi_drv 0x1011062 > The IP 0x1011062 resolves to repos/base/include/util/mmio.h:54 > /** > * Read '_ACCESS_T' typed from MMIO base + 'offset' > */ > template <typename ACCESS_T> > inline ACCESS_T _read(off_t const &offset) const > { > addr_t const dst = _base + offset; > 54 ACCESS_T const value = *(ACCESS_T volatile *)dst; > return value; > } > When using the Genode::backtrace(), I get this > Error: > PAGE-FAULT IN CORE addr=0x204 ip=0x1b32b8 (read) > > stack pointer 0x403feda0, qualifiers 0x4 irUwp > faulter utcb 0x403ff000, last message item count 0 > #1 0x00000000403feda0 0x00000000001b32b8 > #2 0x00000000403feea8 0x000000000015e7c3 > [ 0] Killed EC:0xffffffff8103b4c0 SC:0xffffffff81148040 V:0xe CS:0x2b > IP:0x1c6ac3(0x2155ca) CR2:0x100000008 ERR:0x4 (Livelock) Pd::root > [ 0] Killed EC:0xffffffff8103bb80 SC:0xffffffff81148040 V:0xe CS:0x2b > IP:0x1b32b8(0x1) CR2:0x204 ERR:0x4 (IPC Abort) Pd::root > [ 0] Killed EC:0xffffffff81139280 SC:0xffffffff81148040 V:0xe CS:0x2b > IP:0x1010ab2(0x0) CR2:0xbff00000 ERR:0x4 (IPC Abort) > [ 0] Killed EC:0xffffffff816cb280 SC:0xffffffff816d9040 V:0xe CS:0x2b > IP:0x1021b10(0x1) CR2:0x1021b10 ERR:0x14 (IPC Timeout) > [ 0] Killed EC:0xffffffff8178d280 SC:0xffffffff8106b040 V:0xe CS:0x2b > IP:0xf11fb(0x1067000) CR2:0x1392fff ERR:0x6 (IPC Timeout) > [ 0] Killed EC:0xffffffff8178d040 SC:0xffffffff8106b040 V:0xe CS:0x2b > IP:0xccd34(0x0) CR2:0xccd34 ERR:0x14 (IPC Timeout) > [init -> drivers -> acpi_drv] Error: Uncaught exception of type > 'Genode::Ipc_error' > [init -> drivers -> acpi_drv] Warning: abort called - thread: main > The IP:0x1010ab2 resolves to /home/parfait/git/genode20.05/repos/os/src/drivers/acpi/acpi.cc:254 where I can read following : Fadt(addr_t a) : Genode::Mmio(a) > { > features = read<Fadt::Feature_flags>(); > reset_type = read<Fadt::Reset_reg_type>(); > 254 reset_addr = read<Fadt::Reset_reg_addr>(); > reset_value = read<Fadt::Reset_value>(); > } > It seems like there is an issue with the parsing of the acpi table, especially the FADT. How can I proceed to resolve this ? -- Parfait T.

3 6

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

users