sys-fs/fuse-2.7.1 should have the supported kernel version bumped to ge 2 6 24, see attached patch Reproducible: Always Steps to Reproduce: 1. emerge =sys-fs/fuse-2.7.1 Actual Results: Dies with: * Checking for suitable kernel configuration options... * You need to build the FUSE module from the kernel source, because your kernel is too new Expected Results: Compiles, installs
Created attachment 136741 [details, diff] fuse-2.7.1-fix-2.6.23.patch Short patch for kernel version bump
Why is this not in portage yet, even though 2.6.23 kernel is marked stable?
I'm not a fuse maintainer, but the package you are trying is in the testing tree, and by using that tree you can expect the occasional issue now and then. The stable fuse version should work fine on 2.6.23.
(In reply to comment #3) > I'm not a fuse maintainer, but the package you are trying is in the testing > tree, and by using that tree you can expect the occasional issue now and then. > The stable fuse version should work fine on 2.6.23. > As far as I'm aware, support for 2.6.23 was only finalised in 2.7.1, with 2.7.0 requiring a patch for 2.6.23 support due to changes in kmem_cache_create and some other minor stuff. I can provide the relevant patches against 2.7.0, but it seems more sensible to update and stabalise the 2.7.1(-r1) ebuild. Also, the 2.7.0 ebuild has the 'ge 2 6 23' restriction applied, so no, for multiple reasons the currently stable version of fuse will NOT work with the currently stable kernel sources. Please check your facts before steering users wrong.
(In reply to comment #4) > so no, for multiple > reasons the currently stable version of fuse will NOT work with the currently > stable kernel sources. Please check your facts before steering users wrong. He probably meant that the userland part would work with kernel module from the kernel tree itself and not from the ebuild. It doesn't, or compiling kernel module from ebuild has other advantages?
I assumed 2.7.0 was unaffected due to the resolution of bug #195300. If that is not true, please comment on that bug so that it can be reopened.
(In reply to comment #6) > I assumed 2.7.0 was unaffected due to the resolution of bug #195300. If that is > not true, please comment on that bug so that it can be reopened. > The solution there is to use the in-kernel fuse module. Resolution of this bug will yield a fully function fuse build without the need for kernel recompilation.
Because there was no response from the maintainer on this bug, and even a new versions of fuse was commited yesterday, I just went and commited the change. However, a stabilization of 2.7.1 should be considered, as 2.6.23 is already stable.
thanks :-)
