modify gcc to handle byteorder issue automatically

[Georgy Yunayev]

Hello Jim,

Friday, August 8, 2003, 8:54:54 AM, you wrote:

JW> This issue has come up before a number of times.  I'm not
JW> convinced that it is a good idea.

Probably. But, if you see the original message, my case is special - I
have a lot of little-endian oriented source code, which doesn't use
any system functions. There is no float/double types, just short/int.
I didn't understand the structure difficult well, probably I just
missed some problems. Currently I don't touch bit-fields ops.

Anyway I have already done it, and have run some tests against it to
see it worked, so the problem is probably closed. I would not say it
was really complicated, it tooks about a week to create a skeleton and
handle ops like *a = 0 and b = *a, in two days I added array ops (both
lvalue and rvalue), then structure pointers, then structure membets.

Have I missed something?

Best regards, Tim Yunaev