Re: GCC headers for sanitizers

[christos%astron.com@localhost (Christos Zoulas)]

In article <89bf59b6-1487-2041-3979-e1e84d336140%gmx.com@localhost>,
Kamil Rytarowski  <n54%gmx.com@localhost> wrote:
>-=-=-=-=-=-
>-=-=-=-=-=-
>
>I propose to merge the following patch:
>
>http://netbsd.org/~kamil/patch-00043-gcc-sanitizer-headers.txt
>
>It addresses reports such as:
>"Missing headers in distribution for GCC sanitizers"
>http://gnats.netbsd.org/52265
>
>There is an option to install all the headers available regardless of
>the support quality. This applies for example for lsan_interface.h.
>
>Another option is to install the headers only for those ports that are
>supported. This means that future msan_interface.h will land amd64, but
>not i386. And asan_interface.h could be removed from e.g. atari.
>
>I propose a patch installing all headers for sanitizers that are
>supported on any of the ports. The size of files is relatively small -
>36K (allocator_interface.h, asan_interface.h, common_interface_defs.h
>tsan_interface_atomic.h). This is a cost of preventing an introduction
>of a per-port switch.

Why are you doing that in gcc.old and not gcc?

christos