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] |
This patch looks like the right direction, but it could do with some cleaning up:
- The comments need to be rechecked. Some of the old ones are no longer right, and the new ones are not formatted properly. You should never say that something is the 'new way' in a comment, since it won't be new in 10 years time, and the comment will still be around; just say how the existing code works.
- It doesn't remove the FIXME. The FIXME is the underlying cause of these problems. In fact, I think it makes it worse by creating two strings that are modified instead of one. There's no need to allocate the string using GCC if it's only going to be passed directly to printf, you can just use a static variable.
It is gone from rs6000.c from what I say and I tested an objective-c program to make sure- The objective-C stuff should go. I think it's no longer used, but you'd want to test that.
- I bet this breaks x86-darwin.
- What's with the change to darwin_encode_section_info?
That is part of a different change that is need for Ada to build, I submitted it separately already <http://gcc.gnu.org/ml/gcc-patches/2003-07/msg00171.html>, I will not include it in this updated patch though.
Here is an updated patch which is tested on i686-darwin cross compiler (just running make and making sure it does not ICE on a small test program that accesses global variable). Also bootstrapped on powerpc-apple-darwin6.6.
Attachment:
temp2.diff.txt
Description: Text document
Thanks, Andrew Pinski
Index Nav: | [Date Index] [Subject Index] [Author Index] [Thread Index] | |
---|---|---|
Message Nav: | [Date Prev] [Date Next] | [Thread Prev] [Thread Next] |