Strange bug in egcs-1.1.2 suspected

[Martin v. Loewis]

> pommnitz ~/alias>gcc -O2 aliasorig.c -o aliasorig
> pommnitz ~/alias>./aliasorig 
> 0x12

This looks like a bug to me. gcc-2.95 19990614 says
mira% gcc -o a -O2 a.c
mira% ./a             
0x78563412

If you look at the generated assembly code, you'll find that all the
assignments are optimized-aways, and that the compiler *just knows*
that the result is 18 - which, of course, is wrong.

You'll probably find that people don't care that much about egcs 1.1
bugs, anymore, now that gcc 2.95 is soon to be released.

Regards,
Martin