[PATCH] Fix PR middle-end/26306

Eric Botcazou ebotcazou@adacore.com
Wed Nov 15 13:03:00 GMT 2006

> I'll not object to such a patch.  In fact, I'll pre-approve it, on one
> condition: fill in the section in the manual about implementation of C
> with respect to "qualifiers implementation" to explain under what
> conditions we generate an access to a volatile lvalue whose value is not
> otherwise required.  If we're going to restore this behavior, and you
> think it ought to be preserved in future, we ought to document it.

Agreed, this will be easier to read than the RTL expander in any cases.  
Moreover I think we should also make the "When is a Volatile Object Accessed" 
section for the C++ compiler reference this new section.  Patch attached.

> I'm not trying to be cute; I really do think those are the reasonable
> choices, if we want to preserve this behavior for the foreseeable future.

The Ada compiler happens to rely on this behavior of the middle-end for proper 
operation, so we certainly are interested in preserving it. :-)

Eric Botcazou
-------------- next part --------------
A non-text attachment was scrubbed...
Name: f712-023-4_fsf.diff
Type: text/x-diff
Size: 7957 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/gcc-patches/attachments/20061115/c3f1bb16/attachment.bin>

More information about the Gcc-patches mailing list