Bart Veer wrote:
> This fixes the AT91 SPI driver in the same way as the CortexM STM32
> one.

I object to this change. It requires very recent tools (gcc 4.3.0+). It's 
acceptable for the Cortex HAL to do this as it is a new HAL, but not all 
the AT91 HALs.

Instead I am reinstating the build of spi_at91_init.cxx, but ifdeffing it 
on the presence of CYGBLD_ATTRIB_C_INIT_PRI, which is only defined if GCC 
is recent enough. Patch is attached and checked in.

If there are any other examples of this sort of thing, that I haven't seen 
go past, please fix similarly or let me know.

Jifl
-- 
*See us at Embedded World 2009, Nürnberg, Germany, 3-5 Mar, Stand 11-300*
eCosCentric Limited      http://www.eCosCentric.com/     The eCos experts
Barnwell House, Barnwell Drive, Cambridge, UK.       Tel: +44 1223 245571
Registered in England and Wales: Reg No 4422071.
------["Si fractum non sit, noli id reficere"]------       Opinions==mine