On Wed, Apr 04, 2012 at 08:53:45AM +0200, Helmut Grohne wrote: > On Tue, Apr 03, 2012 at 04:23:17PM +0000, Kamble, Nitin A wrote: > > Thanks for catching the typo. We use "x86_64-linux-gnux32"
> Thanks for the quick reply. > On IRC Steve Langasek pointed out that some part of the difference > resides in the architecture-kernel part. You cannot run a x32 binary on > an arbitrary x86_64 linux kernel, because it needs a (partly) separate > set of syscalls. So ultimately the triplet should differ in the first > two parts (for example x86_64-linux32-gnu). This way of writing things > poses other problems though. It wrongly suggests that a ia32 kernel > could be running x32 binaries. This notion also does not express that > any kernel being able to run x32 binaries will also be able to run amd64 > binaries. To me it seems that this issue is not expressible using the > language of architecture triplets. I therefore suggest sticking with the > x86_64-linux-gnux32 used by yocto. What do you think? I was having a senior moment when I made this comment, and retracted it immediately thereafter. As we are using canonical GNU triplets for multiarch in the final implementation, this of course needs to be a string that the toolchain is also happy with - which means x86_64-linux-$something. -- Steve Langasek Give me a lever long enough and a Free OS Debian Developer to set it on, and I can move the world. Ubuntu Developer http://www.debian.org/ slanga...@ubuntu.com vor...@debian.org
signature.asc
Description: Digital signature