On Mon, Apr 13, 2009 at 09:53:56PM -0400, Avery Fay wrote: > I fixed this by doing: > > --- conftest.h.old 2009-04-13 19:34:15.856105060 -0400 > +++ conftest.h 2009-04-13 19:25:56.000000000 -0400 > @@ -160,7 +160,7 @@ > /* Only kernels sometime before 2.4.0 don't have this function, and who */ > /* cares anymore */ > #define NV_SMP_CALL_FUNCTION_PRESENT > -#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,29) > +#if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,28) > #define NV_SMP_CALL_FUNCTION_ARGUMENT_COUNT 3 > #else > #define NV_SMP_CALL_FUNCTION_ARGUMENT_COUNT 4 > > Based on the title of this bug it should probably test for 2.6.27, but I > can't confirm that locally. > > I'd point out though that the nvidia conftest.sh way of testing for > functions is way safer. This can easily break if someone is running a > kernel that includes a newer patch. It also seems like a maintenance > nightmare (but of course I'm not the one maintaining it).
Nvidia's way also doesn't work with debian's new way of doing kernel headers in 2.6.29, so it needed replacing. I am trying to come up with a way to do compile tests using kbuild that is clean so we can get the best of both worlds. -- Len Sorensen -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
