users

users@lists.genode.org

2 participants
2342 discussions

More mapping issues.
by Daniel Waddington 17 Jul '13

17 Jul '13

Hi, I am still struggling with memory mapping issues. Unfortunately I cannot give a simple test case. We have a large memory hungry programming running on x86_64 Fiasco.OC. During the initialization of memory (using sub-rm dataspace and attach mappings) we suddenly (mid-way through) get "Ipc error 16 in pagefault from 0". Looking up error code 16, I see its L4_IPC_SEMAPFAILED which suggests that the implicit IPC send mapping of the flex page failed in the kernel. Is this correct? Kernel func map<Mem_space> is failing but I can't find the implementation to track it further. Anyone know where it is? Also, any advice on how to go about diagnosing this? Thanks Daniel

1 1

Re: Porting on i.MX6
by Archie Nidhi 15 Jul '13

15 Jul '13

Again thanks for your reply , the board we are using is Nitrogen 6X ,now the size of our uncompressed file is 4.2MB and compressed is 1.8 MB , the RAM base address of our board is 0x10000000 and RAM size is 0x40000000 , we tried booting by assigning load address as 0x10001000 and the entry point as 10001000 and tftp 0x10800000 . We are stuck at starting kernel Bytes transferred = 1829585 (1bead1 hex) U-Boot > bootm ## Booting kernel from Legacy Image at 10800000 ... Image Name: Image Type: ARM Linux Kernel Image (gzip compressed) Data Size: 1829521 Bytes = 1.7 MiB Load Address: 10001000 Entry Point: 10005d70 Verifying Checksum ... OK Uncompressing Kernel Image ... OK Starting kernel ... Any Idea what the problem is ? On Sat, Jul 13, 2013 at 2:36 AM, Frank Kaiser <frank.kaiser@...24...>wrote: > [image: Boxbe] <https://www.boxbe.com/overview> This message is eligible > for Automatic Cleanup! (frank.kaiser@...24...) Add cleanup rule<https://www.boxbe.com/popup?url=https%3A%2F%2Fwww.boxbe.com%2Fcleanup%3Ftok…>| More > info<http://blog.boxbe.com/general/boxbe-automatic-cleanup?tc_serial=14613466892…> > > Archie Nidhi schrieb: > > Hi thanks for the reply,I am Archie working with Aditya on this . We > use your reprository and then followed the steps given in https://github.com/genodelabs/genode/commit/ce58a63fb661df42e6bf283891589b1… > ,<https://github.com/genodelabs/genode/commit/ce58a63fb661df42e6bf283891589b1…>for creating uImage . > > And then booted using tftp with this. > > # either 'dhcp' or 'setenv ipaddr xxx.xxx.xxx.xxx' > # 'setenv serverip xxx.xxx.xxx.xxx' as long it isn't set via DHCP > # 'tftp 0x70800000 [path_to_uimage]/uImage' > # 'bootm 0x70800000' > > Now the error we are getting in this is > > Using FEC device > TFTP from server 10.6.1.123; our IP address is 10.6.1.122 > Filename 'uImage'. > Load address: 0x10800000 > Loading: ################################################################# > ################################################################# > ################################################################# > ################################################################# > ################################################################# > ################################# > done > Bytes transferred = 1829585 (1bead1 hex) > U-Boot > bootm > ## Booting kernel from Legacy Image at 10800000 ... > Image Name: > Image Type: ARM Linux Kernel Image (gzip compressed) > Data Size: 1829521 Bytes = 1.7 MiB > Load Address: 70140000 > Entry Point: 7018b33c > Verifying Checksum ... OK > Uncompressing Kernel Image ... > > > Any Ideas where we might be going wrong? > > Hello, Archie > > Which type of i.MX6 board are you using? The load address 0x70140000 of > your boot image assumes that your board supports 2 GB of RAM (physical > address range 0x10000000 to 0x8FFFFFFF). AFAIK not all i.MX6 boards have > that much memory, others come with 1 GB instead (physical address range > 0x10000000 to 0x4FFFFFFF). In that case you have to modify the build > scripts so that the löad address is set to 0x40140000 and the entry point > to 0x4018b33c (or other values which fit into the 1 GB range). > > Frank > > > ------------------------------------------------------------------------------ > See everything from the browser to the database with AppDynamics > Get end-to-end visibility with application monitoring from AppDynamics > Isolate bottlenecks and diagnose root cause in seconds. > Start your free trial of AppDynamics Pro today! > http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk > > _______________________________________________ > Genode-main mailing list > Genode-main(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/genode-main > >

2 1

Re: Porting on i.MX6
by Archie Nidhi 12 Jul '13

12 Jul '13

Hi thanks for the reply,I am Archie working with Aditya on this . We use your reprository and then followed the steps given in https://github.com/genodelabs/genode/commit/ce58a63fb661df42e6bf283891589b1… ,<https://github.com/genodelabs/genode/commit/ce58a63fb661df42e6bf283891589b1…>for creating uImage . And then booted using tftp with this. # either 'dhcp' or 'setenv ipaddr xxx.xxx.xxx.xxx' # 'setenv serverip xxx.xxx.xxx.xxx' as long it isn't set via DHCP # 'tftp 0x70800000 [path_to_uimage]/uImage' # 'bootm 0x70800000' Now the error we are getting in this is Using FEC device TFTP from server 10.6.1.123; our IP address is 10.6.1.122 Filename 'uImage'. Load address: 0x10800000 Loading: ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################# done Bytes transferred = 1829585 (1bead1 hex) U-Boot > bootm ## Booting kernel from Legacy Image at 10800000 ... Image Name: Image Type: ARM Linux Kernel Image (gzip compressed) Data Size: 1829521 Bytes = 1.7 MiB Load Address: 70140000 Entry Point: 7018b33c Verifying Checksum ... OK Uncompressing Kernel Image ... Any Ideas where we might be going wrong? On Wed, Jul 10, 2013 at 3:52 PM, Nikolay Golikov <nik@...104...> wrote: > [image: Boxbe] <https://www.boxbe.com/overview> This message is eligible > for Automatic Cleanup! (nik@...104...) Add cleanup rule<https://www.boxbe.com/popup?url=https%3A%2F%2Fwww.boxbe.com%2Fcleanup%3Ftok…>| More > info<http://blog.boxbe.com/general/boxbe-automatic-cleanup?tc_serial=14582365019…> > > > > среда, 10 июля 2013 г. пользователь Aditya Kousik писал: > >> Hello >> >> Firstly, thank you for patiently replying to our queries and debugging >> issues so far. >> >> We're trying to port Genode onto an i.MX6, which we've realised is no >> easy task. Our initial approach to this was to tweak existing i.Mx53 files >> and make corresponding changes (arm cortex a9, created imx6 related folders >> in include). As pointed out to us by Stefan, this was completely off track >> as the memory layout may be mixed up (or many more issues). >> >> Here's a brief reboot of our attempt: >> 1. edit create_builddir to add a rule for imx6 board and added >> build.conf.hw_imx6 in builddir folder. >> 2. modify board_base.h in base/include/platform/imx6/drivers and fill >> with addresses by referring to the imx6 reference manual. (We also added >> CORTEX_A9_CLOCK and private memory base) >> 3. read through the Makefile and added spec-hw_imx6 and >> spec-platform_imx6 respectively. >> 4. created imx6 folder with files in os/include/platform/imx6 as well. >> (All with imx53 as reference) >> >> Make: >> Since the drivers for imx6 would take a long time to code, we are NOT >> running demo.run. We tried to build hello.run (without building >> drivers/timer) according the instructions given in your site. >> >> We're getting the following error: >> Program core/imx6/core >> COMPILE _main.o >> ASSEMBLE boot_modules.o >> COMPILE console.o >> COMPILE context_area.o >> COMPILE cpu_session_component.o >> COMPILE cpu_session_support.o >> ASSEMBLE crt0.o >> COMPILE dataspace_component.o >> COMPILE dump_alloc.o >> COMPILE io_mem_session_component.o >> COMPILE io_mem_session_support.o >> COMPILE irq_session_component.o >> COMPILE kernel.o >> COMPILE kernel_support.o >> COMPILE main.o >> make[3]: *** No rule to make target `mode_transition.o', needed by >> `core'. Stop. >> make[2]: *** [core.prg] Error 2 >> make[1]: *** [gen_deps_and_build_targets] Error 2 >> make[1]: Leaving directory `/media/work/assorted/genode/imx' >> Error: Genode build failed >> make: *** [run/hello] Error 255 >> >> Thanks and regards >> Aditya >> > > Hello, four months ago I added basic support for i.MX6 to genode. Results > of this work you can see at https://github.com/decaprox/genode/tree/i.mx6. > > > > ------------------------------------------------------------------------------ > See everything from the browser to the database with AppDynamics > Get end-to-end visibility with application monitoring from AppDynamics > Isolate bottlenecks and diagnose root cause in seconds. > Start your free trial of AppDynamics Pro today! > http://pubads.g.doubleclick.net/gampad/clk?id=48808831&iu=/4140/ostg.clktrk > > _______________________________________________ > Genode-main mailing list > Genode-main(a)lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/genode-main > >

2 1

Dataspace juju
by Daniel Waddington 11 Jul '13

11 Jul '13

Hi, We're having bad juju with dataspaces. As we try larger dataspaces things seem to go wrong (like hanging, bad mappings and exceptions). Can someone check out my test program.. https://github.com/dwaddington/genode/blob/master/testing/src/core-api-1/ma… and see what is going on? Basically if you change the total memory use in the program (line 26) to something ~ > 512MB, then region conflicts happen. I tried this out this test on both fiasco.oc and nova with qemu and real PCs. Same result. The NOVA run reports an "unresolvable exception" by pager:core-api-1. I could of course be doing something wrong with the APIs. Daniel

3 13

env()->rm_session()->attach
by Daniel Waddington 09 Jul '13

09 Jul '13

Hi, Whats the limitation on how much a process can env()->rm_session()->attach(...)? Is this a thread-safe API? Thanks Daniel

2 1

Re: u-boot files - device specific?
by Stefan Kalkowski 09 Jul '13

09 Jul '13

sorry for sending it twice. I intended to send to the mailing list. On 07/08/2013 11:55 AM, Stefan Kalkowski wrote: > Hi Aditya, > > On 07/08/2013 11:03 AM, Aditya Kousik wrote: >> Hi Stefan, >> >> You got us. We tweaked all the iMX53 files to make it into imx6. :D > > so excuse me for thinking something else! > >> (editing create_builddir to add foc_imx6, hw_imx6 etc.) and edited all >> related platform folders and header file register addresses (board_base.h). >> And simply ran make followed by make run/demo. We've got the image.elf >> and uImage. Now stuck on booting these on the i.MX6. > > In that case, it can be really everything: beginning with the wrong > initialization of the serial device, errors in the memory layout, > booting the kernel image to a bad position and so on. It's impossible to > give you a general advice hereby. > > Being in your situation I would start by instrumenting the base-hw > kernel, so it will first print something to serial line. Please have a > look at the kernel function: > > > https://github.com/genodelabs/genode/blob/master/base-hw/src/core/kernel.cc… > > This is the first C-function triggered after a few assembler > instructions in: > > > https://github.com/genodelabs/genode/blob/master/base-hw/src/core/arm/crt0.… > > If it doesn't print anything, you might use a JTAG debugger to have a > look what happens after u-boot jumps to the first kernel instruction. If > you don't have a JTAG debugger available, ensure that your kernel image > doesn't get loaded to memory where u-boot resides in, or to the first > page of physical memory. > > Good luck, and best regards > Stefan > >> >> Thanks and Regards >> Aditya >> >> On 8 Jul 2013 14:24, "Stefan Kalkowski" >> <stefan.kalkowski@...1... >> <mailto:stefan.kalkowski@...1...>> wrote: >> >> Hi Aditya, >> >> On 07/08/2013 08:48 AM, Aditya Kousik wrote: >> > Hello, >> > >> > I'm trying to run Genode on an i.MX6. And I've tried copying >> uImage. But >> > that doesn't seem to be enough. I came across this : >> > http://comments.gmane.org/gmane.os.genode.devel/813 >> >> I hope you're not trying to use the uImage build for i.MX53 on a i.MX6 >> platform? Apart from u-boot, the kernel, as well as Genode's core >> process, and drivers are hardly bound not only to the SoC (i.MX53 or >> i.MX6), but also to the board the SoC is built in. So you can't just run >> the same image built for i.MX53 Quickstart board without modifications >> on some i.MX6 board. >> >> Regards >> Stefan >> >> > >> > Your set of u-boot files, is it pandaboard specific? or can the MLO, >> > u-boot.bin be copied to any device and booted? >> > >> > What's the procedure of booting from using the image.elf? (So far, >> i ran >> > mkimage and created a uImage file. But that's obviously not enough. >> > What's missing?) >> > >> > Thanks and regards, >> > Aditya >> > >> > >> > >> ------------------------------------------------------------------------------ >> > This SF.net email is sponsored by Windows: >> > >> > Build for Windows Store. >> > >> > http://p.sf.net/sfu/windows-dev2dev >> > >> > >> > >> > _______________________________________________ >> > Genode-main mailing list >> > Genode-main(a)lists.sourceforge.net >> <mailto:Genode-main@lists.sourceforge.net> >> > https://lists.sourceforge.net/lists/listinfo/genode-main >> > >> >> -- >> Stefan Kalkowski >> Genode Labs >> >> http://www.genode-labs.com/ · http://genode.org/ >> > -- Stefan Kalkowski Genode Labs http://www.genode-labs.com/ · http://genode.org/

1 0

Unable to boot
by Archie Nidhi 08 Jul '13

08 Jul '13

Hello I have a few questions We have successfully created an uImage by following the steps given in https://github.com/genodelabs/genode/commit/ce58a63fb661df42e6bf283891589b1… We tried to transfer the uImage by tftp and got the following error U-Boot > setenv serverip 10.6.1.123 U-Boot > setenv ipaddr 10.6.1.122 U-Boot > setenv bootfile uImage U-Boot > tftp Using FEC device TFTP from server 10.6.1.123; our IP address is 10.6.1.122 Filename 'uImage'. Load address: 0x10800000 Loading: ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# ################################################################# #### done Bytes transferred = 2346876 (23cf7c hex) U-Boot > bootm ## Booting kernel from Legacy Image at 10800000 ... Image Name: Image Type: ARM Linux Kernel Image (gzip compressed) Data Size: 2346812 Bytes = 2.2 MiB Load Address: 70140000 Entry Point: 7018b33c Verifying Checksum ... OK Uncompressing Kernel Image ... It is stuck over here , let me know whether there is another way to convert uImage to a form from which we can boot from a sdcard. Thanks and Regards Anubhav

3 4

Re: u-boot files - device specific?
by Aditya Kousik 08 Jul '13

08 Jul '13

Hello, I'm trying to run Genode on an i.MX6. And I've tried copying uImage. But that doesn't seem to be enough. I came across this : http://comments.gmane.org/gmane.os.genode.devel/813 Your set of u-boot files, is it pandaboard specific? or can the MLO, u-boot.bin be copied to any device and booted? What's the procedure of booting from using the image.elf? (So far, i ran mkimage and created a uImage file. But that's obviously not enough. What's missing?) Thanks and regards, Aditya

3 5

Booting Genode on i.MX53
by Archie Nidhi 05 Jul '13

05 Jul '13

Hello, 1) Given that Genode has been ported on iMX53 with u-boot, how do I get it running on either qemu or hardware directly? (cross-compiling for arm, configuring u-boot etc.) 2) I tried creating build directories with ./tool/create_builddir hw_imx53 (and foc_imx53) BUILD_DIR=build and running make run/demo. Got the following errors: 1.1 : foc_imx53 ./dump_tcboffsets tcboffset.bin > tcboffset.h.new mv tcboffset.h.new tcboffset.h ... Making ivt.o ... Making tb_entry_output.o /media/work/genode-13.05/base- foc/contrib/kernel/fiasco/src/kern/tb_entry_output.cc:506:31: error: uninitialized const ‘_ipc_fmt’ [-fpermissive] /media/work/genode-13.05/base-foc/contrib/kernel/fiasco/src/kern/tb_entry_output.cc:141:7: note: ‘const class Tb_entry_ipc_fmt’ has no user-provided default constructor /media/work/genode-13.05/base-foc/contrib/kernel/fiasco/src/kern/tb_entry_output.cc:507:35: error: uninitialized const ‘_ipc_res_fmt’ [-fpermissive] /media/work/genode-13.05/base-foc/contrib/kernel/fiasco/src/kern/tb_entry_output.cc:157:7: note: ‘const class Tb_entry_ipc_res_fmt’ has no user-provided default constructor make[5]: *** [tb_entry_output.o] Error 1 make[4]: *** [all] Error 2 make[3]: *** [/media/work/genode-13.05/foc.imx/kernel/fiasco.oc/fiasco] Error 2 make[2]: *** [fiasco.oc.prg] Error 2 make[1]: *** [gen_deps_and_build_targets] Error 2 make[1]: Leaving directory `/media/work/genode-13.05/foc.imx' Error: Genode build failed make: *** [run/demo] Error 252 1.2: hw_imx53 Program core/imx53/core COMPILE _main.o cc1plus: error: unrecognized command line option ‘-std=gnu++11’ make[3]: *** [_main.o] Error 1 make[2]: *** [core.prg] Error 2 make[1]: *** [gen_deps_and_build_targets] Error 2 make[1]: Leaving directory `/media/work/genode-13.05/hw. imx53' Error: Genode build failed make: *** [run/demo] Error 255 Any ideas? Thanks and Regards Anubhav

6 8

RPC threads on specific cores
by Daniel Waddington 03 Jul '13

03 Jul '13

Hi, Is there a way to have RPC server threads launch on specific cores (as opposed to the default of core 0)? Daniel

2 1

Jump to page:

2026

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

users