Message ID | 1507479013-5207-1-git-send-email-yamada.masahiro@socionext.com |
---|---|
Headers | show |
Series | radix-tree: split out struct radix_tree_root out to <linux/radix-tree-root.h> | expand |
On Mon, Oct 09, 2017 at 01:10:01AM +0900, Masahiro Yamada wrote: <...> > > By splitting out the radix_tree_root definition, > we can reduce the header file dependency. > > Reducing the header dependency will help for speeding the kernel > build, suppressing unnecessary recompile of objects during > git-bisect'ing, etc. If we judge by the diffstat of this series, there won't be any visible change in anything mentioned above. <...> > > Masahiro Yamada (12): > radix-tree: replace <linux/spinlock.h> with <linux/spinlock_types.h> > radix-tree: split struct radix_tree_root to <linux/radix-tree-root.h> > irqdomain: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > writeback: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > iocontext.h: replace <linux/radix-tree.h> with > <linux/radix-tree-root.h> > fs: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > blkcg: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > fscache: include <linux-radix-tree.h> > sh: intc: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > net/mlx4: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > net/mlx5: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > drm/i915: replace <linux/radix-tree.h> with <linux/radix-tree-root.h> > > drivers/gpu/drm/i915/i915_gem.c | 1 + > drivers/gpu/drm/i915/i915_gem_context.c | 1 + > drivers/gpu/drm/i915/i915_gem_context.h | 2 +- > drivers/gpu/drm/i915/i915_gem_execbuffer.c | 1 + > drivers/gpu/drm/i915/i915_gem_object.h | 1 + > drivers/net/ethernet/mellanox/mlx4/cq.c | 1 + > drivers/net/ethernet/mellanox/mlx4/mlx4.h | 2 +- > drivers/net/ethernet/mellanox/mlx4/qp.c | 1 + > drivers/net/ethernet/mellanox/mlx4/srq.c | 1 + > drivers/sh/intc/internals.h | 2 +- > include/linux/backing-dev-defs.h | 2 +- > include/linux/blk-cgroup.h | 2 +- > include/linux/fs.h | 2 +- > include/linux/fscache.h | 1 + > include/linux/iocontext.h | 2 +- > include/linux/irqdomain.h | 2 +- > include/linux/mlx4/device.h | 2 +- > include/linux/mlx4/qp.h | 1 + > include/linux/mlx5/driver.h | 2 +- > include/linux/mlx5/qp.h | 1 + > include/linux/radix-tree-root.h | 24 ++++++++++++++++++++++++ > include/linux/radix-tree.h | 8 ++------ > 22 files changed, 46 insertions(+), 16 deletions(-) > create mode 100644 include/linux/radix-tree-root.h > > -- > 2.7.4 >
2017-10-09 3:52 GMT+09:00 Leon Romanovsky <leonro@mellanox.com>: > On Mon, Oct 09, 2017 at 01:10:01AM +0900, Masahiro Yamada wrote: > > <...> >> >> By splitting out the radix_tree_root definition, >> we can reduce the header file dependency. >> >> Reducing the header dependency will help for speeding the kernel >> build, suppressing unnecessary recompile of objects during >> git-bisect'ing, etc. > > If we judge by the diffstat of this series, there won't be any > visible change in anything mentioned above. Of course, judging by the diffstat is wrong. -- Best Regards Masahiro Yamada
On Mon, Oct 09, 2017 at 02:58:58PM +0900, Masahiro Yamada wrote: > 2017-10-09 3:52 GMT+09:00 Leon Romanovsky <leonro@mellanox.com>: > > On Mon, Oct 09, 2017 at 01:10:01AM +0900, Masahiro Yamada wrote: > > > > <...> > >> > >> By splitting out the radix_tree_root definition, > >> we can reduce the header file dependency. > >> > >> Reducing the header dependency will help for speeding the kernel > >> build, suppressing unnecessary recompile of objects during > >> git-bisect'ing, etc. > > > > If we judge by the diffstat of this series, there won't be any > > visible change in anything mentioned above. > > > Of course, judging by the diffstat is wrong. > I'm more than happy to be wrong and you for sure can help me. Can you provide any quantitative support of your claims? Thanks > > > -- > Best Regards > Masahiro Yamada > -- > To unsubscribe from this list: send the line "unsubscribe linux-rdma" in > the body of a message to majordomo@vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html
On Mon, Oct 09, 2017 at 01:10:01AM +0900, Masahiro Yamada wrote: > Reducing the header dependency will help for speeding the kernel > build, suppressing unnecessary recompile of objects during > git-bisect'ing, etc. Well, does it? You could provide measurements showing before/after time to compile, or time to recompile after touching a header file that is included by radix-tree.h and not by radix-tree-root.h. Look at the files included (never mind the transitively included files): #include <linux/bitops.h> #include <linux/bug.h> #include <linux/kernel.h> #include <linux/list.h> #include <linux/preempt.h> #include <linux/rcupdate.h> #include <linux/spinlock.h> #include <linux/types.h> These are not exactly rare files to be included. My guess is that most of the files in the kernel end up depending on these files *anyway*, either directly or through some path that isn't the radix tree.
2017-10-10 21:18 GMT+09:00 Matthew Wilcox <willy@infradead.org>: > On Mon, Oct 09, 2017 at 01:10:01AM +0900, Masahiro Yamada wrote: >> Reducing the header dependency will help for speeding the kernel >> build, suppressing unnecessary recompile of objects during >> git-bisect'ing, etc. > > Well, does it? You could provide measurements showing before/after > time to compile, or time to recompile after touching a header file that > is included by radix-tree.h and not by radix-tree-root.h. > > Look at the files included (never mind the transitively included files): > > #include <linux/bitops.h> > #include <linux/bug.h> > #include <linux/kernel.h> > #include <linux/list.h> > #include <linux/preempt.h> > #include <linux/rcupdate.h> > #include <linux/spinlock.h> > #include <linux/types.h> > > These are not exactly rare files to be included. My guess is that most > of the files in the kernel end up depending on these files *anyway*, either > directly or through some path that isn't the radix tree. Good question. I tested this series, and I confirmed the total number of included headers decreased, but did not decrease as much as I had expected. The statement "most of the files in the kernel end up depending on these files" is true. But, with that excuse, I do not want to conclude this kind of refactoring is pointless. For example, how can we explain commit bc6245e5efd70c41eaf9334b1b5e646745cb0fb3 ? <linux/bug.h> includes the following three. #include <asm/bug.h> #include <linux/compiler.h> #include <linux/build_bug.h> Your statement applies to them too. Actually, I did not see any impact by replacing <linux/bug.h> in my files with <linux/build_bug.h> Generally, people do not pay much attention in decreasing header dependency. One refactoring alone does not produce much benefits, but making continuous efforts will disentangle the knotted threads. Of course, this might be a pipe dream... -- Best Regards Masahiro Yamada
On Tue, Oct 10, 2017 at 09:56:22PM +0900, Masahiro Yamada wrote: > One refactoring alone does not produce much benefits, > but making continuous efforts will disentangle the knotted threads. > Of course, this might be a pipe dream... A lot of people have had that dream, and some of those refactoring efforts have proven worthwhile. But it's not a dream without costs; your refactoring will conflict with other changes. I don't think the benefit here is high enough to pursue this edition of the dream.