On Wed, Apr 03, 2002 at 09:42:24PM -0500, Jack Howarth wrote: > Hello, > Could someone explain to me the point of releasing > Xfree86 4.1.0-15 as is when clearly patch #065 was > going to break builds on most non-intel arches? Actually, the patch was applied to FIX a problem with building xfree86 on ia64. So your assertion is wrong. The bottom line is that it's hit or miss as to whether you can compile XFree86 or not. It depends on what kernel flavor your libc6-dev was compiled against. BenC's last glibc release for i386 was built against a kernel that omitted the SiS ioctl information from drm.h. BenC also says that it's completely up to the guys who build glibc for each architecture what kernel headers they build against. And Herbert Xu told me that I shouldn't ever depend on kernel headers for anything, but encapsulate all information I need into my package. So that's how we got where are today. Somehow it's the application-writer's job to know how the underlying kernel defines its ioctls. How this position is consistent with the very existence of symbolic constants in C library headers is quite beyond me. Apparently, if the Linux kernel driver guys renumber some ioctls, the right thing is for everybody's apps to break instantly. > I simply don't see the logic at play here considering > we're supposed to be closing in on woody's release. I entirely agree. However, I don't have the power to tell Ben, the glibc-architecture builders, or Herbert Xu how to do things. My package is at their mercy. As I understand it, however, the real blame lies with the guys on the linux-kernel list. We wouldn't have this problem if the kernel headers were the same regardless of what modules you decide to compile. -- G. Branden Robinson | You can have my PGP passphrase when Debian GNU/Linux | you pry it from my cold, dead branden@debian.org | brain. http://people.debian.org/~branden/ | -- Adam Thornton
Attachment:
pgpNpTukjeHO9.pgp
Description: PGP signature