In-flight collision: ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP renaming

In-flight collision: ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP renaming

[Lukas Bulwahn]

Dear Guo Ren, dear Palmer,

on linux-next, I have noticed that you have submitted and added the
commit a3c7d6b642e4 ("riscv: mm: hugetlb: Enable
ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), that adds the select
ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP to the RISCV config. As it
seems, Aneesh Kumar K.V has concurrently renamed this config
throughout the tree with commit 215218dddd25 ("mm/hugetlb_vmemmap:
rename ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), but your commit was
not visible to that renaming. When both commits are merged in a common
tree---as happens with linux-next, and will soon happen with Linus'
tree, your commit a3c7d6b642e4 will do nothing (select no existing
config anymore).

It is easy to fix, adjust this one line to the renaming and 'select
ARCH_WANT_OPTIMIZE_VMEMMAP'. For now, I am just reporting to let you
know; probably, it is best to get a quick fix-up patch with -rc2, once
all the changes landed in -rc1.


Best regards,

Lukas

Re: In-flight collision: ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP renaming

[Matthew Wilcox]

On Mon, Apr 24, 2023 at 11:40:25AM +0200, Lukas Bulwahn wrote:
> It is easy to fix, adjust this one line to the renaming and 'select
> ARCH_WANT_OPTIMIZE_VMEMMAP'. For now, I am just reporting to let you
> know; probably, it is best to get a quick fix-up patch with -rc2, once
> all the changes landed in -rc1.

Actually, it's best that both parties who send a pull request to Linus
note this collision and he can fix it in the merge commit.

Re: In-flight collision: ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP renaming

[Guo Ren]

On Mon, Apr 24, 2023 at 5:40 PM Lukas Bulwahn <lukas.bulwahn@gmail.com> wrote:
>
> Dear Guo Ren, dear Palmer,
>
> on linux-next, I have noticed that you have submitted and added the
> commit a3c7d6b642e4 ("riscv: mm: hugetlb: Enable
> ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), that adds the select
> ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP to the RISCV config. As it
> seems, Aneesh Kumar K.V has concurrently renamed this config
> throughout the tree with commit 215218dddd25 ("mm/hugetlb_vmemmap:
> rename ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), but your commit was
> not visible to that renaming. When both commits are merged in a common
> tree---as happens with linux-next, and will soon happen with Linus'
> tree, your commit a3c7d6b642e4 will do nothing (select no existing
> config anymore).
>
> It is easy to fix, adjust this one line to the renaming and 'select
> ARCH_WANT_OPTIMIZE_VMEMMAP'. For now, I am just reporting to let you
> know; probably, it is best to get a quick fix-up patch with -rc2, once
> all the changes landed in -rc1.
Thx for taking care; please fix that.

>
>
> Best regards,
>
> Lukas



-- 
Best Regards
 Guo Ren

Re: In-flight collision: ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP renaming

[Lukas Bulwahn]

On Fri, May 5, 2023 at 8:16 AM Guo Ren <guoren@kernel.org> wrote:
>
> On Mon, Apr 24, 2023 at 5:40 PM Lukas Bulwahn <lukas.bulwahn@gmail.com> wrote:
> >
> > Dear Guo Ren, dear Palmer,
> >
> > on linux-next, I have noticed that you have submitted and added the
> > commit a3c7d6b642e4 ("riscv: mm: hugetlb: Enable
> > ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), that adds the select
> > ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP to the RISCV config. As it
> > seems, Aneesh Kumar K.V has concurrently renamed this config
> > throughout the tree with commit 215218dddd25 ("mm/hugetlb_vmemmap:
> > rename ARCH_WANT_HUGETLB_PAGE_OPTIMIZE_VMEMMAP"), but your commit was
> > not visible to that renaming. When both commits are merged in a common
> > tree---as happens with linux-next, and will soon happen with Linus'
> > tree, your commit a3c7d6b642e4 will do nothing (select no existing
> > config anymore).
> >
> > It is easy to fix, adjust this one line to the renaming and 'select
> > ARCH_WANT_OPTIMIZE_VMEMMAP'. For now, I am just reporting to let you
> > know; probably, it is best to get a quick fix-up patch with -rc2, once
> > all the changes landed in -rc1.
> Thx for taking care; please fix that.
>

Just for the record:

Conor Dooley sent a patch after the merge of the two trees, addressing
the issue. This patch is included in v6.4-rc1 as commit 26b0812f4cf8
("RISC-V: fixup in-flight collision with ARCH_WANT_OPTIMIZE_VMEMMAP
rename")

So, the issue is resolved.

Lukas