The new version uses asm/byteorder.h and its function __le16_to_cpu, which do not exist in FreeBSD. I will attach a patch. Please let me know if it is OK to check in the patch.
Created attachment 134833 [details, diff] files/usbutils-0.73-freebsd.patch Patch to use an alternate method in FreeBSD
Created attachment 134834 [details, diff] usbutils-0.73.ebuild The ebuild change to use this patch
usbutils uses autotools ... so should you
Mike, can you elaborate: how would you patch this?
system related defines are not valid in autotool based systems add proper configure checks and key off of those defines
Created attachment 134912 [details, diff] usbutils-0.73.ebuild New ebuild to patch and reconf autotools instead
Created attachment 134913 [details, diff] files/usbutils-0.73-byteorder.patch New patch using autotools to determine how to deal with byteorder. Mike, let me know what you think.
This patch has been submitted to upstream. Please advise on how to proceed for 0.73. Thanks, Joe
i'd indent stuff myself ... #if moo # include ... # if ... # .... but otherwise looks fine did you submit this upstream ?
on second thought, dont use _BYTE_ORDER there is an AC macro: AC_C_BIGENDIAN this will define "WORDS_BIGENDIAN" if targetting a big endian machine
Created attachment 135897 [details, diff] files/usbutils-0.73-byteorder.patch OK, see new patch and let me know how this looks. I did submit to upstream, but I'm sure they will accept this change as well.
Created attachment 135899 [details, diff] files/usbutils-0.73-byteorder.patch Use this one instead.
Created attachment 135900 [details, diff] files/usbutils-0.73-byteorder.patch This one is even more clear I think...
one space, not two ... but otherwise looks fine side note, never use "#" comments in configure.ac files ... use "dnl"
Fixed the indentation (using one space now) and committed. Thanks Mike! Submitted new AC endian macro tweak upstream.
