GNU cpp is predefining a number of symbols, depending on the host and target architecture. This can produce some unexpected results: for example, the expansion of CLIENTHOST if the host name is i386.my.domain. The attached patch creates a new -undef option to xrdb that is passed to cpp. Should it be on by default ?
Created attachment 2781 [details] [review] proposed patch
added patch keyword
Sorry about the phenomenal bug spam, guys. Adding xorg-team@ to the QA contact so bugs don't get lost in future.
I sent this patch to xorg-devel for discussion/review. This bug report is obviously not doing much, so I'm going to close it.
Hey Matt, where did this go? The patch looks fine to me: Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> I think we should leave it off by default as someone might actually be doing something like #ifdef __i386__ in their .Xresources.
(In reply to comment #5) > Hey Matt, where did this go? The patch looks fine to me: > > Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> > > I think we should leave it off by default as someone might actually be doing > something like #ifdef __i386__ in their .Xresources. Yes this is the kind of reasons that made this patch stall. I'll revisit the issue and push the fix disabled. Thanks for reviewing it.
(In reply to comment #5) > Hey Matt, where did this go? The patch looks fine to me: > > Reviewed-by: Jeremy Huddleston <jeremyhu@apple.com> > > I think we should leave it off by default as someone might actually be doing > something like #ifdef __i386__ in their .Xresources. I sent it off: http://lists.x.org/archives/xorg-devel/2010-December/thread.html#16525 but there were corner cases that made people wary of committing it.
I pushed the version that leaves the default the same as current, but lets you specify -undef to pass the -undef flag to cpp.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.