Get rid of libgcc and nacl_read_tp symbol preservation hacks.

This is no longer needed for libgcc since it is compiled to native code. nacl_read_tp can be preserved correctly with a one-line fix to crt1.x.

This reverses the actions of issue 1520.

BUG= http://code.google.com/p/nativeclient/issues/detail?id=1520

[jvoung - send to chromium..., 2011-11-02 01:21:00]

Do you have a link to the paired nacl_startup change?

[robertm, 2011-11-02 13:55:04]

+1 
please file a bug so we can track these changes together.

Also added krasin who is going to look into the mysteries of 
of our link phases.


On 2011/11/02 01:21:00, jvoung wrote:
> Do you have a link to the paired nacl_startup change?

[pdox, 2011-11-02 16:04:17]

robert,

This is not a bug. I'm cleaning out dead code.

[robertm, 2011-11-02 16:38:13]

can you check the URL for the corresponding CL

On 2011/11/02 16:04:17, pdox wrote:
> robert,
> 
> This is not a bug. I'm cleaning out dead code.

It is not so much about whether this is a bug (call it an issue if you prefer), 
but about linking related CLs so we have some way of making sense of it later
on.

The nacl convention is to have a bug for almost every CL, c.f.
  http://www.chromium.org/nativeclient/day-to-day/codereview

[pdox, 2011-11-02 16:47:04]

Fixed CL link.

Robert, that doc says all "significant" CLs need a bug. This is just cleaning
out dead code that is no longer needed since we switched libgcc to native code.

[pdox, 2011-11-02 17:00:09]

PTAL

[robertm, 2011-11-02 17:27:08]

On 2011/11/02 17:00:09, pdox wrote:
> PTAL

Here is a bug that should be referenced
http://code.google.com/p/nativeclient/issues/detail?id=1520

Please also add this bug to the other CL, i.e.
http://codereview.chromium.org/8437034/

[pdox, 2011-11-02 18:27:51]

added bug

[robertm, 2011-11-02 18:31:46]

LGTM

On 2011/11/02 18:27:51, pdox wrote:
> added bug