[lto] PATCH: new CALL_EXPR constructors

Sandra Loosemore sandra@codesourcery.com
Mon Aug 7 14:34:00 GMT 2006


Mark Mitchell wrote:
> Paolo Bonzini wrote:

>> It's better not to introduce new TREE_CHAIN usages.  You may put a FIXME comment saying that we may want to remove it. 
> 
> I'm not sure that a FIXME here would be useful.  The only alternative to
> TREE_CHAIN, at the moment, would be a VEC, and that would use more
> memory.  It's a long way from where we are to a TREE_CHAIN-free
> representation; look how hard it's been to make the CALL_EXPR change
> that Sandra's undertaken.  If there were an EXPR_CHAIN, I'd suggest that
> Sandra use that -- but it doesn't look like that's been created yet.
> However, if you feel strongly that there should be a comment, then
> please indicate that, and, Sandra, please do as Paolo requests in this
> respect.

FWIW, I've already fixed this bit to use an obstack, as Paolo suggested.  Here's 
the revised patch (just the tree-ssa-pre.c part).

I also fixed the other bug Paolo found (thanks!) and verified that things still 
build and test properly.

-Sandra

-------------- next part --------------
A non-text attachment was scrubbed...
Name: tree-ssa-pre.diff
Type: text/x-patch
Size: 12882 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20060807/dbf6587a/attachment.bin>


More information about the Gcc-patches mailing list