This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: regrename bug fix


>>>>> Joern Rennecke writes:

Joern> Well, if the patch to find_oldest_value_reg is making a difference to the
Joern> workings of the compiler, that means a non-zero offset is calculated
Joern> somewhere (we got some unrelated latent bug that is triggered by different
Joern> input rtl).
Joern> I've appended a patch that should make your stage1 compiler abort when it
Joern> finds a non-zero offset; you can then analyze if this has introduced the
Joern> AIX problem somehow, or if it fixed a problem you haven't realized you had
Joern> before...
	
	The mainline is not suppose to be left broken for a target that
formerly worked.  This is especially annoying as everyone is focussed on
GCC 3.1.1.  I ask you again, please revert your patch while we debug this.

David


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]