Hi, Chris: I know that the Hypervisor toolkit (which with a prebuilt kernel.elf --- note: only for ARMv7) can working, but what I am talking about is that the source code which you pulled from the Git repository will never working. Am I right?
Regards. Charles, Liu ----- Original Message ----- From: "Jenkins, Christipher Derell" <cdj...@sandia.gov> To: <codezero-devel@lists.l4dev.org> Sent: Thursday, May 10, 2012 10:37 PM Subject: Re: [codezero-devel] [EXTERNAL] Re: Codezero for PandaBoard bootloader question > Liu, > > Be sure to use the patched QEMU from the website. I believe the manual gives > the link. I believe you can also download the 0.5.0 toolkit and it have both > insight, gdb, and the patch qemu. Like yourself, I did see issues when I did > a fetch from the Git repo. I believe the codezero team will be releasing some > new code soon (according to emails and mail-serv posts) that should help with > ARMv7 development. > > Best, > Chris J > > -----Original Message----- > From: 起凌 [mailto:chipl...@yeah.net] > Sent: Thursday, May 10, 2012 3:25 AM > To: amit.maha...@b-labs.com; Jenkins, Christipher Derell > Cc: codezero-devel@lists.l4dev.org > Subject: Re: [codezero-devel] [EXTERNAL] Re: Codezero for PandaBoard > bootloader question > Importance: High > > Hello, Dear all: > > In fact, as in my test result, the codezero source code which from Git > repository & the download page that cannot work, even the ARMv5's code. With > the Git version and follow the http://www. l4dev.org/build_from_scratch > document, We can produced a "final.elf" for PB926-Versatile Platform & > ARM926EJ-S processor with an "empty" or "hello_world" baremetal project, but > when you run the final image with qemu & gdb debugger, and after the > following messages: > > ELF Loader: Loader image size: 557KB, placed at physical 0xe00000 - 0xe8b6b8 > Loading the kernel... > Entry point: 0x8000 > Copying to range from 0x8000 to 0x1fdf8 of size: 0x17df8 Clearing memory... > starting from 1fdf8, size: 0 Copying to range from 0x20000 to 0x2d000 of > size: 0xd000 Clearing memory... starting from 2d000, size: 61c8 Copying to > range from 0x34000 to 0x38390 of size: 0x4390 Clearing memory... starting > from 38390, size: 0 > > Loading containers... > > Loading section .cont.0 from top-level elf file. > Loading .img.0 section image... > Entry point: 0xa0000000 > Copying to range from 0x10295c to 0x10297c of size: 0x20 Clearing memory... > starting from 10297c, size: 0 Copying to range from 0x100000 to 0x10297c of > size: 0x297c Clearing memory... starting from 10297c, size: 0 Copying to > range from 0x103000 to 0x1030b4 of size: 0xb4 Clearing memory... starting > from 1030b4, size: 104c > > Total of 1 images in this container. > Total of 1 container images. > elf-loader: kernel entry point is 0x8000 > elf-loader: Starting kernel > > code0: start kernel... > > code0: Init kernel mappings... > code0: Virtual memory enabled. > code0: Kernel area 0xf0008000 - 0xf0039000 remapped as 49 pages > code0: Kernel built on May 10 2012, 15:44:05 > code0: Mapping 0x3000 bytes as RX from 0x100000 physical to 0xa0000000 > virtual for hello_world0 > code0: Mapping 0x2000 bytes as RW from 0x103000 physical to 0xa0003000 > virtual for hello_world0 > > you will see the message as following( when you using "c" or "n" command in > gdb after the init_finalize() ): > > Pager (1) faulted on itself. FSR: 0x1f, FAR: 0xa00040ec, PC: 0xa0001d2c pte: > 0x104ffe CPU0 Exiting. > > this message seems is come from file src/arch/arm/exception-common.c, > function fault_ipc_to_pager(), line 121, I don't know if this is a bug or > not, because in my mind, the correct message should be come from file > conts/baremetal/hello_world/hello,c, function print_hello_world(), line 10, > which as following: > > cont0: Hello world from hello_world0! > > but my test result always failed after init_finalize(). So, that means I do > not believe that someone said that their demos which also come from the > codezero's Git repository can working. Am I wrong? > > By the way, especially for Amit and B-labs R&D team, I found a small bug in > file print-early.c(all of the platform), function printhex8, line 51 & line > 54(this line numbers means in the pb926 platform version, for realview > platform's version is line 58 & 61, beagle's version is line 50 & 53), the > correct code should like this: > > 51 *temp = '\0'; > 52 > 53 if ( !val ) { > 54 *(--temp) = '0'; > > Am I right? > > Regards, > > Charles, Liu > > > ----- Original Message ----- > From: "Amit Mahajan" <amit.maha...@b-labs.com> > To: "Jenkins, Christipher Derell" <cdj...@sandia.gov> > Cc: <codezero-devel@lists.l4dev.org> > Sent: Thursday, May 10, 2012 5:59 AM > Subject: Re: [codezero-devel] [EXTERNAL] Re: Codezero for PandaBoard > bootloader question > > >> On Wed, 2012-05-09 at 21:52 +0000, Jenkins, Christipher Derell wrote: >>> Hello Amit. >>> >>> I'm new to C0 development. I pulled codezero from the Git repository and >>> have the helloworld, mutex, and threading demos working. However, they only >>> work for ARMv5. Also, looking at 'git log' I notice no commits have >>> happened since June 2010. Is the development of codezero still progressing? >>> >> >> Hi Chris, >> >> Yes we are actively working on Codezero and Virtualization. We have not >> released ARM-V7 sources publicly yet. Instead we have release toolkits >> for v7 based platforms like VX and panda board. >> >> Check l4dev.org for the toolkit downloads. >> >> >> -- >> Thanks, >> Amit Mahajan >> >> >> _______________________________________________ >> codezero-devel mailing list >> codezero-devel@lists.l4dev.org >> http://lists.l4dev.org/mailman/listinfo/codezero-devel_lists.l4dev.org > _______________________________________________ > codezero-devel mailing list > codezero-devel@lists.l4dev.org > http://lists.l4dev.org/mailman/listinfo/codezero-devel_lists.l4dev.org > _______________________________________________ codezero-devel mailing list codezero-devel@lists.l4dev.org http://lists.l4dev.org/mailman/listinfo/codezero-devel_lists.l4dev.org