Hello Ian,
I'll re-jig with the typographical changes (sorry that were so
many ... )
... and re-post - but I'd like to clear up a couple of points first.
On 30 Sep 2011, at 00:00, Ian Lance Taylor wrote:
-#define GNU_SECTION_NAMES "__section_names"
Are we sure it is OK to drop support for __section_names? Won't that
mean that any old LTO files will not work with new gcc? Is that OK?
I was under the impression that an lto built with an earlier revision
of gcc would not interwork anyway.
If that is mistaken, then we could re-enable the option to allow both
(I had that in the previous version, in fact).
+ If we find mach-o sections (with the segment name as specified)
which also
+ contain: a 'sects' wrapper, an index, and a name table, we
expand this into
+ as many sections as are specified in the index. In this case,
there will
+ be a callback for each of these.
+
+ When the omr->segment_name is not specified, then we would call-
back for
+ every mach-o section specified in the file.
+
+ Return 1 if we should continue, 0 if the caller should return.
*/
Is there a reason that you are changing the code to support not
specifying a segment? Should that be a separate change?
I'll separate it out.
I think that it could help us build some small tools to assist in
making GCC output more closely match what the system tool-chain is
expecting (e.g. stripping off LTO sections). But that would require
being able to read all sections as they come and then apply a filter.
Perhaps it's just better to invest the time in BFD, but that does mean
the user having to build a second set of stuff or a combined tree.
thanks for reviewing and sorry about the number of trivial errors.
Iain
