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: Inlining functions with variable-sized structs


On Sun, 28 Nov 2004, Richard Kenner wrote:

> We currently disallow inlining them, but I think the reason is no longer
> correct due to changes I made in the tree walking routines a while ago.
> It's too late to investigate this for 4.0, so I added a comment as a reminded
> to investigate it later.

We have a lot of old comments such as

combine.c:           ??? For 2.5, try to tighten up the MD files in this regard
combine.c:         ??? For 2.5, try to tighten up the MD files in this regard
cse.c:   Perhaps for 2.9.  */
dbxout.c:                   Should make this not happen, after 2.4.  */
expmed.c:     of possible modes in init_expmed?  Save this for GCC 2.7.  */
expmed.c:                  but it does not seem worth the trouble for GCC 2.6.  */
expmed.c:                  above.  For 2.7 we should merge these variants, but for
expmed.c:                  2.6.1 I don't want to touch the code for unsigned since that
sdbout.c:      /* C++ in 2.3 makes nameless symbols.  That will be fixed later.

in the compiler.  It would be useful to file a PR as well, "Should be able 
to inline functions with variable-size structs".  If there turn out to be 
several underlying problems still present, the PR can be a meta-bug with 
dependencies on the underlying problems added.

-- 
Joseph S. Myers               http://www.srcf.ucam.org/~jsm28/gcc/
    jsm@polyomino.org.uk (personal mail)
    joseph@codesourcery.com (CodeSourcery mail)
    jsm28@gcc.gnu.org (Bugzilla assignments and CCs)


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