Re: null bytes?

[Erik M. <J.A...@it...>]

On Mon, Jan 28, 2002 at 03:44:24PM -0700, Tim Riker wrote:
> hmm...
> 
> I have not touched that file. Compiling for idr from cvs I see:
> 
> timr@localhost:~/work/vercel/blob/src/blob$ ls -l blob* 
> -rwxr-xr-x    1 timr     timr        46520 Jan 28 14:12 blob
> -rwxr-xr-x    1 timr     timr        46520 Jan 28 14:12 blob-chain
> -rwxr-xr-x    1 timr     timr        45496 Jan 28 14:12 blob-rest
> -rwxr-xr-x    1 timr     timr        77029 Jan 28 14:12 blob-rest-elf32
> -rwxr-xr-x    1 timr     timr          852 Jan 27 01:30 blob-start
> -rwxr-xr-x    1 timr     timr          196 Jan 27 01:30 blob-start-chain
> -rwxr-xr-x    1 timr     timr        34179 Jan 27 01:30 blob-start-chain-elf32
> -rwxr-xr-x    1 timr     timr        35805 Jan 27 01:30 blob-start-elf32
> 
> where blob has 13k of nulls on the end.
> 
> blob-rest-elf32 contains this same block on the end of the code before
> the debug blocks.

I currently can't compile for idr because of compiler errors (SBI_SKCR
undeclared in sa1111_init()), but this is the output from the last time
I compiled it:

erik@arthur:idr/src/blob >pwd
/home/erik/src/sourceforge/build/idr/src/blob
erik@arthur:idr/src/blob >ls -l blob*
-rwxr-xr-x    1 erik     users       32148 Jan 18 23:43 blob
-rwxr-xr-x    1 erik     users       32148 Jan 17 00:22 blob-chain
-rwxr-xr-x    1 erik     users       31124 Jan 17 00:22 blob-rest
-rwxr-xr-x    1 erik     users       76596 Jan 17 00:22 blob-rest-elf32
-rwxr-xr-x    1 erik     users         852 Jan 18 23:43 blob-start
-rwxr-xr-x    1 erik     users         196 Jan  9 00:31 blob-start-chain
-rwxr-xr-x    1 erik     users       34465 Jan  9 00:31 blob-start-chain-elf32
-rwxr-xr-x    1 erik     users       36382 Jan 18 23:43 blob-start-elf32

And this is with today's CVS version for LART (compiled with
--enable-md5):

erik@arthur:lart/src/blob >pwd
/home/erik/src/sourceforge/build/lart/src/blob
erik@arthur:lart/src/blob >ls -l blob*
-rwxr-xr-x    1 erik     users       24632 Jan 21 22:51 blob
-rwxr-xr-x    1 erik     users       24632 Jan 21 22:51 blob-chain
-rwxr-xr-x    1 erik     users       23608 Jan 21 22:51 blob-rest
-rwxr-xr-x    1 erik     users       65826 Jan 21 22:51 blob-rest-elf32
-rwxr-xr-x    1 erik     users         640 Jan 19 00:47 blob-start
-rwxr-xr-x    1 erik     users         196 Jan 19 00:47 blob-start-chain
-rwxr-xr-x    1 erik     users       34466 Jan 19 00:47 blob-start-chain-elf32
-rwxr-xr-x    1 erik     users       36162 Jan 19 00:47 blob-start-elf32

I can't really repeat your problem. I am using both my gcc-2.95.2 
toolchain (available from the LART site) and the gcc-2.95.3 toolchain
(available from LART site and arm-linux FTP site).

Unless something is defining a non-null byte somewhere in the .bss
segment or on the stack, I can't really see what's wrong.


Erik

> Erik Mouw wrote:
> > 
> > On Mon, Jan 28, 2002 at 02:13:45PM -0700, Tim Riker wrote:
> > > Anyone know why blob has close to 14k of nulls on the end?
> > 
> > It doesn't, unless you have put a segment definition *after* the .bss
> > segment in src/blob/rest-ld-script.in .

-- 
J.A.K. (Erik) Mouw, Information and Communication Theory Group, Faculty
of Information Technology and Systems, Delft University of Technology,
PO BOX 5031, 2600 GA Delft, The Netherlands  Phone: +31-15-2783635
Fax: +31-15-2781843  Email: J.A...@it...
WWW: http://www-ict.its.tudelft.nl/~erik/