World: r3wp

The browser is easiest

Or R3

No, that's only on the live CD

Runs but doesn't display anything.

Look at the end of /var/log/kernel

memmap_instance() RO overlap RW (80000000 + 00001000 -> 80000000)

ERROR : execve(./hello) failed. Too late to recover, will exit

That would be the next thing to work on

that sounds like something alignment-related is broken

Alignment should basically be the same as on Linux

Kaj, what was the link for Syllable's ELF loader source code?

Just want to have a look at the source, I won't debug it right now, 
I'm still preparing slides for tomorrow's presentation.

I'll try this again

Kaj: 
I tested the new section headers on Syllable Desktop.

memmap_instance() RO overlap RW (08048000 + 00001000 -> 08048000)"


besides the entry point address being a problem, this could also 
be due to segment alignment, which we basically ignore, at the moment. 
could you try changing "page-size: 4096" to "page-size: 1" and see 
where that gets us?"

Ok

i think syllable just ignored the alignment setting in the elf header 
last time we tried this

http://syllable.cvs.sourceforge.net/viewvc/syllable/syllable/system/sys/kernel/kernel/elf.c?view=markup

Thanks.

line 1111 originates the error

I've uploaded a new empty with page-size 1

Same error.

hmm, if syllable ignores the alignment i guess we have to properly 
align things

:)

We can load an awful lot of other software, so it seems prudent

in line 183 of formats/ELF.r, change data-ptr: code-ptr + code-size 
to data-ptr: code-ptr + page-size

and change back page-size to 4096 before that

Is that all?

(could you please)

should be a quick way to test this particular problem, yes

ah, crap, no

empty compiles, at least

yeah, but it generates crap :)

Andreas, shouldn't it be rather:
data-ptr: code-ptr + code-size + (page-size // code-size)

yeah, if we want to do proper alignment :)

but it results in the same problem of generating wrong file offsets

(because other code parts re-use data-ptr to calculate the file offset)

give me a few minutes, i'll try to generate a binary that has that 
fixed

I don't think the GCC folks can do debug cycles this fast :-)

you can (it's only C, after all), but you need a quite beefy machine. 
LLVM (C++) on the other hand ...

soo, here we go

http://bolka.at/2011/tmp/hello

trying...

or if you want to compile the binaries on your own:
https://gist.github.com/1bf8d5c54cb630c83d8e

hmm, the browser don't want to save it to disk, it just displays 
it as text...I guess it misses a content-type: application/octet-stream...

Kaj: any workaround?

http://bolka.at/2011/tmp/hello.bin

Andreas: you rules!

Hello is working

nice :)

Displays all the banners, but seg fault on exiting.