Re: [openrisc] gcc, data and bss segments

[Chris Ziomkowski<chris@asics.ws>]

> Hi Chris,
> 
> are these buffers uninitialized data? If yes, then this is probably a simple
> bug in or1k gcc backend. I'll look into this.

Yes, they are uninitialized. You can see the same thing by
creating a simple test program:

int foo;

int main(int argc,char** argv)
{
}

You'll notice it puts foo into the data segment.

BTW, a non standard but completely functional hack to save
even a few more bytes (I've never seen this fail on anything)
is to place all zero initialized and unitialized data into
the bss segment. (I control zeroing of the bss segment, and I
will guarantee it is set to 0.) Something to consider if
you're having trouble tracking where to place the data....

Thanks for you're help. Not top priority but something to
look into when you're otherwise bored.

Chris
chris@asics.ws



_______________________________________________________
www.asics.ws - Solutions for your ASIC needs
--
To unsubscribe from openrisc mailing list please visit http://www.opencores.org/mailinglists.shtml