Re: [RFC] PPC: Skip call to __eabi in main()

[Daniel Jacobowitz <drow@false.org>]

On Wed, Aug 06, 2008 at 05:40:37PM -0700, Kevin Buettner wrote:
> With regard to my patch, I'd prefer to commit it in its present form
> and then address improvements to PowerPC instruction decoding at a
> later time.  I considered using my preferred approach when I adjusted
> my patch, but decided against doing so because a different approach
> (that of using explicit masks and comparisons) was already in use.

I think this is reasonable.

> FWIW, this isn't the only approach that I find compelling.  I recently
> worked on a port which utilizes the instruction decoder in opcodes/. 
> This decoder is also used for the disassembler and simulator.  It
> completely decodes the instruction, returning a symbolic (via enums)
> opcodes, and completely decoded instruction offsets, registers, etc.

I wish other ports provided enough detail in libopcodes to do this!
It's an excellent approach.

-- 
Daniel Jacobowitz
CodeSourcery