CRAY bitfield support in protocol headers: does anyone care?

Egbert Eich eich at
Tue Feb 27 22:02:01 PST 2007

It's not likely that people here don't complain about this here if
noone here has access to such an architecture. And I don't know how
many of these beasts are still around. But I don't think this community
is representative for the community that is into these architectures.

It will certainly take a while before stuff done here has penetrated 
from here down to these architectures but it is not unlikely that
this may eventually happen. Then - when this support is gone -
the surprise for people to get it to run there will be rather
I would think the client side would be more important than the server 
side though.


Josh Triplett writes:
 > Some of the X protocol headers have support for the cray architecture, which
 > does not support data sizes smaller than 64 bits, and thus has to use
 > bitfields in structures to have the right data sizes over the wire.
 > XCB's generated protocol headers do not include this support, and nobody has
 > complained or asked for it.  We want to know if we could reasonably start
 > using our data structures more widely, such as in the server, without needing
 > to add this support.
 > So: does anyone care if we stop supporting machines that can't declare
 > 1-byte, 2-byte, and 4-byte structure fields?
 > - Josh Triplett
 > _______________________________________________
 > xorg mailing list
 > xorg at

More information about the xorg mailing list