uClibc ARM thumb support

[Joakim Tjernlund]

> -----Original Message-----
> From: uclibc-bounces at uclibc.org 
> [mailto:uclibc-bounces at uclibc.org] On Behalf Of John Bowler
> Sent: den 21 januari 2006 08:02
> To: uclibc at uclibc.org
> Subject: uClibc ARM thumb support
> 
> I updated bug 385 (http://bugs.busybox.net/view.php?id=385) 
> with a new set of patches which apply to the SVN head (as of 
> 20060120) and which apparently work.
> 
> I say 'apparently' because when I tested them with both ARM 
> and Thumb compilation shells apparently fail - either an 
> abort of some kind or a hang.  I can reproduce this both with 
> and without the patches, so I assume it is either my build or 
> the SVN head.
> 
> Running /usr/bin/find as the init did, however, work and 
> /sbin/init partially works (it either crashes itself or the 
> shell it execs to run rcS crashes).
> 
> The patch set still contains the controversial dl-hash.c 
> change, which may be unnecessary because binutils 2.16 may 
> strip out the problematic ARM_STT_TFUNC.  Anyway, the patch 
> set is be fine without the dl-hash.c change (i.e. nothing 
> breaks, although thumb stuff might not work in some cases.)  
> The very limited testing I could do without a shell shows 
> /usr/bin/find working fine but /sbin/init gets a segmentation 
> violation.
> 
> John Bowler <jbowler at acm.org>
> 

Hi John 

Did a little research since our last chat. Found this:
  . If the st_shndx member of the symbol table entry is not SHN_UNDEF, the dynamic linker
  has found a definition for the symbol and uses its st_value member as the symbol's
  address.
  . If the st_shndx member is SHN_UNDEF and the symbol is of type STT_FUNC and the
  st_value member is not zero, the dynamic linker recognizes this entry as special and uses
  the st_value member as the symbol's address.
  . Otherwise, the dynamic linker considers the symbol to be undefined within the executable file
  and continues processing.
in http://refspecs.freestandards.org/elf/elfspec_ppc.pdf, heading Function Addresses

This shows that there are cases when SHN_UNDEF is a valid target.

Don't know much about ARM even less about Thumb but since STT_ARM_TFUNC is
encoded in st_value you might want to try adding to the current dl_hash in SVN:
if (sym->st_value == 1)
	continue;

The dl_hash.c changes does not look sane to me, looks like you have restored the old WEAK
procedure(among some other stuff that I don't understand yet)? Why? 

 Jocke